Point cloud data transmission device, point cloud data transmission method, point cloud data reception device, and point cloud data reception method

Abstract

The embodiment discloses a point cloud data sending method, a point cloud data sending device, a point cloud data receiving method and a point cloud data receiving device. The point cloud data transmission method according to an embodiment may comprise the steps of: acquiring point cloud data including points through a lidar device provided with a laser sensor; encoding the point cloud data; and transmitting the signaling data and the encoded point cloud data.

Description

Technical Field

[0001] Embodiments relate to a method and apparatus for processing point cloud content. Background Art

[0002] Point cloud content is content represented by a point cloud, which is a collection of points belonging to a coordinate system representing a three-dimensional space (or volume). Point cloud content can express media configured in three dimensions and is used to provide various services such as virtual reality (VR), augmented reality (AR), mixed reality (MR), XR (extended reality), and self-driving services. However, tens of thousands to hundreds of thousands of point data are required to represent point cloud content. Therefore, a method of efficiently processing large amounts of point data is needed. Summary of the invention

[0003] Technical issues

[0004] An object of the present disclosure, which is intended to solve the above-mentioned problems, is to provide a point cloud data transmitting device, a point cloud data transmitting method, a point cloud data receiving device, and a point cloud data receiving method for efficiently transmitting and receiving a point cloud.

[0005] Another object of the present disclosure is to provide a point cloud data sending device, a point cloud data sending method, a point cloud data receiving device and a point cloud data receiving method for solving time delay and encoding / decoding complexity.

[0006] Another object of the present disclosure is to provide a point cloud data sending device, a point cloud data sending method, a point cloud data receiving device and a point cloud data receiving method that improve the compression performance of point clouds by improving the technology of encoding attribute information of geometry-based point cloud compression (G-PCC).

[0007] Another object of the present disclosure is to provide a point cloud data sending device, a point cloud data sending method, a point cloud data receiving device and a point cloud data receiving method for efficiently compressing and sending point cloud data captured by a LiDAR device, and receiving point cloud data captured by a LiDAR device.

[0008] Another object of the present disclosure is to provide a point cloud data sending device, a point cloud data sending method, a point cloud data receiving device and a point cloud data receiving method for efficiently compressing point cloud data captured by a LiDAR device.

[0009] Another object of the present disclosure is to provide a point cloud data sending device, a point cloud data sending method, a point cloud data receiving device and a point cloud data receiving method for classifying point cloud data into roads and objects to efficiently compress point cloud data captured by a LiDAR device.

[0010] Another object of the present disclosure is to provide a point cloud data sending device, a point cloud data sending method, a point cloud data receiving device and a point cloud data receiving method for classifying point cloud data into roads and objects and applying the classification to a prediction tree to efficiently compress point cloud data captured by a LiDAR device.

[0011] The embodiments are not limited to the above-mentioned purposes, and the scope of the embodiments can be expanded to other purposes that can be inferred by those skilled in the art based on the entire contents of the present disclosure.

[0012] Technical Solution

[0013] To achieve these objectives and other advantages and in accordance with the purposes of the present disclosure, as embodied and broadly described herein, a method for sending point cloud data according to an embodiment may include: acquiring point cloud data containing points through a LiDAR device having a laser sensor; encoding the point cloud data; and sending the encoded point cloud data and signaling data.

[0014] According to an embodiment, the encoding may include: separating the points in the point cloud data into roads and objects; constructing a first prediction tree based on the points separated into the roads, and constructing a second prediction tree based on the points separated into the objects; and compressing the point cloud data based on the prediction by applying the first prediction tree and the second prediction tree.

[0015] According to an embodiment, the signalling data may include information related to the separation of the road and the object.

[0016] According to an embodiment, separating the points into the road and the object may include separating the points in the point cloud data into the road and the object based on a basic height corresponding to a distance from the road to a center of the laser sensor.

[0017] According to an embodiment, the basic height may be estimated based on angle information, radius information, and height information related to a laser sensor having a maximum negative angle among the laser sensors.

[0018] According to an embodiment, separating the points into the roads and the objects may include: separating the points among the points in the point cloud data that are located at a position equal to or lower than the basic height as the roads; and separating the points among the points in the point cloud data that are located at a position higher than the basic height as the objects.

[0019] According to an embodiment, separating the points into the road and the object may include removing points existing within a minimum radius among points acquired by the current laser sensor, wherein the separation may be performed after the removal.

[0020] According to an embodiment, the compressing may include compressing the point cloud data based on intra-frame prediction by applying the first prediction tree and the second prediction tree.

[0021] According to an embodiment, the compressing may include compressing the point cloud data based on inter-frame prediction by applying the first prediction tree, the second prediction tree and at least one reference frame.

[0022] According to an embodiment, the compression may include: based on the current point in the current frame belonging to the road, selecting a point having the same laser ID and similar azimuth as the current point from points in the area separated as the road in the at least one reference frame as a reference point, and setting the reference point as a predicted value of the current point; and based on the current point in the current frame belonging to the object, selecting a point having the same laser ID and similar azimuth as the current point from points in the area separated as the object in the at least one reference frame as a reference point, and setting the reference point as a predicted value of the current point.

[0023] According to an embodiment, a device for sending point cloud data may include: an acquirer, which is configured to acquire point cloud data containing points through a LiDAR device with a laser sensor; an encoder, which is configured to encode the point cloud data; and a transmitter, which is configured to send the encoded point cloud data and signaling data.

[0024] According to an embodiment, the encoder can be configured to: separate the points in the point cloud data into roads and objects; construct a first prediction tree based on the points separated into the roads, and construct a second prediction tree based on the points separated into the objects; and compress the point cloud data based on the prediction by applying the first prediction tree and the second prediction tree.

[0025] According to an embodiment, the signalling data may include information related to the separation of the road and the object.

[0026] According to an embodiment, the encoder may separate the points in the point cloud data into the road and the object based on a basic height corresponding to a distance from the road to a center of the laser sensor.

[0027] According to an embodiment, the basic height may be estimated based on angle information, radius information, and height information related to a laser sensor having a maximum negative angle among the laser sensors.

[0028] According to an embodiment, the encoder may separate the points among the points in the point cloud data that are located at a position equal to or lower than the basic height as the roads; and separate the points among the points in the point cloud data that are located at a position higher than the basic height as the objects.

[0029] According to an embodiment, the encoder may separate the points into the road and the object after removing points existing within a minimum radius among points acquired by a current laser sensor.

[0030] According to an embodiment, the encoder may compress the point cloud data based on intra-frame prediction by applying the first prediction tree and the second prediction tree.

[0031] According to an embodiment, the encoder may compress the point cloud data based on inter-frame prediction by applying the first prediction tree, the second prediction tree, and at least one reference frame.

[0032] According to an embodiment, the encoder may be configured to: based on the current point in the current frame belonging to the road, select a point having the same laser ID and a similar azimuth as the current point from among the points in the area separated as the road in the at least one reference frame as a reference point, and set the reference point as a predicted value of the current point; and based on the current point in the current frame belonging to the object, select a point having the same laser ID and a similar azimuth as the current point from among the points in the area separated as the object in the at least one reference frame as a reference point, and set the reference point as a predicted value of the current point.

[0033] Beneficial Effects

[0034] The point cloud data sending method, point cloud data sending device, point cloud data receiving method and receiving device according to the embodiments can provide high-quality point cloud services.

[0035] According to the point cloud data sending method, point cloud data sending device, point cloud data receiving method and point cloud data receiving device of the implementation mode, various video encoding and decoding methods can be implemented.

[0036] The point cloud data transmitting method, the point cloud data transmitting device, the point cloud data receiving method, and the point cloud data receiving device according to the embodiments can provide general point cloud contents such as autonomous driving services.

[0037] The point cloud data sending method, point cloud data sending device, point cloud data receiving method and point cloud data receiving device according to the embodiments can perform spatial adaptive partitioning of point cloud data for independent encoding and decoding of point cloud data, thereby improving parallel processing and providing scalability.

[0038] According to the implementation manner, the point cloud data sending method, the point cloud data sending device, the point cloud data receiving method and the point cloud data receiving device can perform encoding and decoding by partitioning the point cloud data in units of tiles and / or slices, and signaling necessary data thereof, thereby improving the encoding and decoding performance of the point cloud.

[0039] The point cloud data sending method, point cloud data sending device, point cloud data receiving method and point cloud data receiving device according to the embodiment can support efficient geometric compression of point cloud content captured by a LiDAR device on a mobile vehicle by the following operations: estimating (or predicting) a basic height value, which is the distance from the road to the center of the laser sensor; classifying points in the point cloud data into roads or objects based on the estimated (or predicted) height values; constructing prediction trees for road parts and object parts that are distinguished from each other; and performing compression. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this application. The accompanying drawings illustrate embodiments of the present disclosure and together with the description serve to explain the principles of the present disclosure. In the drawings:

[0041] Figure 1 An exemplary point cloud content providing system according to an embodiment is shown;

[0042] Figure 2 is a block diagram illustrating a point cloud content providing operation according to an embodiment;

[0043] Figure 3 An exemplary point cloud encoder according to an embodiment is illustrated;

[0044] Figure 4 shows examples of octrees and occupancy codes according to an embodiment;

[0045] Figure 5 illustrates an example of point configuration in each LOD according to an embodiment;

[0046] Figure 6 illustrates an example of point configuration in each LOD according to an embodiment;

[0047] Figure 7 A point cloud decoder according to an embodiment is illustrated;

[0048] Figure 8 An exemplary transmitting device according to an embodiment is illustrated;

[0049] Figure 9 An exemplary receiving device according to an embodiment is illustrated;

[0050] Figure 10 An exemplary structure that can be operated in conjunction with a point cloud data transmitting / receiving method / apparatus according to an embodiment is illustrated;

[0051] Figure 11 is a diagram illustrating an example of point cloud data acquired using a LiDAR device according to an embodiment;

[0052] Figure 12 is a diagram illustrating an example of points constituting a road captured by a LiDAR device according to an embodiment;

[0053] Figure 13 (a) is a diagram illustrating an exemplary prediction tree when an object does not exist in point cloud data acquired using a plurality of lasers according to an embodiment;

[0054] Figure 13 (b) is a diagram illustrating an exemplary prediction tree when an object exists in point cloud data acquired using a plurality of lasers according to an embodiment;

[0055] Figure 14 is a diagram illustrating an example of a point of each laser sensor according to an embodiment;

[0056] Figure 15 is a diagram illustrating an example of point cloud data captured by a LiDAR device according to an embodiment;

[0057] Figure 16 is a diagram illustrating an example of a laser sensor and an angle θ according to an embodiment;

[0058] Figure 17 is a diagram illustrating an exemplary position estimation method according to an embodiment;

[0059] Figure 18 (a) is a diagram illustrating an exemplary prediction tree using point cloud data that is not separated into roads and objects according to an embodiment;

[0060] Figure 18 (b) is a diagram illustrating an exemplary case where prediction trees are constructed to have points belonging to a road and points belonging to an object, respectively, according to an embodiment;

[0061] Figure 19(a) is a diagram illustrating an exemplary prediction tree using point cloud data that is not separated into roads and objects according to an embodiment;

[0062] Figure 19 (b) is a diagram illustrating an exemplary case in which prediction trees are respectively constructed to have points belonging to a road and points belonging to an object, and then the prediction trees are connected to each other and then two prediction trees are connected according to an embodiment;

[0063] Figure 20 is a diagram illustrating another exemplary point cloud transmitting device according to an embodiment;

[0064] Figure 21 is a diagram illustrating exemplary operations of a geometry encoder and an attribute encoder according to an embodiment;

[0065] Figure 22 is a diagram illustrating another exemplary point cloud receiving device according to an embodiment;

[0066] Figure 23 is a diagram illustrating exemplary operations of a geometry decoder and an attribute decoder according to an embodiment;

[0067] Figure 24 An exemplary bitstream structure for transmitting / receiving point cloud data according to an embodiment is illustrated;

[0068] Figure 25 An exemplary syntax structure of a sequence parameter set according to an embodiment is illustrated;

[0069] Figure 26 illustrates an exemplary syntax structure of a geometry parameter set according to an embodiment;

[0070] Figure 27 An exemplary syntax structure of a tile parameter set according to an embodiment is illustrated;

[0071] Figure 28 illustrates an exemplary syntax structure of a geometry data unit according to an embodiment;

[0072] Figure 29 An exemplary syntax structure of a geometry data unit header according to an embodiment is illustrated;

[0073] Figure 30 illustrates an exemplary syntax structure of predictive_tree() according to an embodiment;

[0074] Figure 31 An exemplary syntax structure of a geometry PU header according to an embodiment is illustrated;

[0075] Figure 32is a flowchart illustrating an exemplary method of transmitting point cloud data according to an embodiment; and

[0076] Figure 33 is a flow chart illustrating an exemplary method of receiving point cloud data according to an embodiment. DETAILED DESCRIPTION

[0077] Reference will now be made in detail to the preferred embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings. The detailed description given below with reference to the accompanying drawings is intended to illustrate exemplary embodiments of the present disclosure, rather than to show the only embodiments that can be implemented according to the present disclosure. The following detailed description includes specific details in order to provide a thorough understanding of the present disclosure. However, it will be apparent to those skilled in the art that the present disclosure can be practiced without these specific details.

[0078] Although most of the terms used in the present disclosure are selected from common terms widely used in the art, some terms are arbitrarily selected by the applicant and their meanings are explained in detail in the following description as needed. Therefore, the present disclosure should be understood based on the intended meanings of the terms rather than their simple names or meanings.

[0079] Figure 1 An exemplary point cloud content providing system according to an embodiment is shown.

[0080] Figure 1 The point cloud content providing system shown may include a transmitting device 10000 and a receiving device 10004. The transmitting device 10000 and the receiving device 10004 can transmit and receive point cloud data via wired or wireless communication.

[0081] According to an embodiment, the point cloud data sending device 10000 can obtain and process point cloud video (or point cloud content) and send it. According to an embodiment, the sending device 10000 may include a fixed station, a base transceiver system (BTS), a network, an artificial intelligence (AI) device and / or system, a robot, an AR / VR / XR device and / or a server. According to an embodiment, the sending device 10000 may include a device configured to communicate with a base station and / or other wireless devices using a radio access technology (e.g., 5G New RAT (NR), Long Term Evolution (LTE)), a robot, a vehicle, an AR / VR / XR device, a portable device, a home appliance, an Internet of Things (IoT) device, and an AI device / server.

[0082] According to the implementation manner, the sending device 10000 includes a point cloud video acquisition unit 10001, a point cloud video encoder 10002 and / or a transmitter (or communication module) 10003.

[0083] The point cloud video acquisition unit 10001 according to the embodiment acquires the point cloud video through a processing process such as acquisition, synthesis or generation. The point cloud video is a point cloud content represented by a point cloud, which is a set of points located in a 3D space and can be referred to as point cloud video data, point cloud data, etc. The point cloud video according to the embodiment may include one or more frames. A frame represents a still image / picture. Therefore, the point cloud video may include a point cloud image / frame / picture and may be referred to as a point cloud image, frame or picture.

[0084] The point cloud video encoder 10002 according to the embodiment encodes the acquired point cloud video data. The point cloud video encoder 10002 may encode the point cloud video data based on point cloud compression coding. The point cloud compression coding according to the embodiment may include geometry-based point cloud compression (G-PCC) coding and / or video-based point cloud compression (V-PCC) coding or next generation coding. The point cloud compression coding according to the embodiment is not limited to the above-mentioned embodiment. The point cloud video encoder 10002 may output a bit stream containing the encoded point cloud video data. The bit stream may include not only the encoded point cloud video data, but also signaling information related to the encoding of the point cloud video data.

[0085] The transmitter 10003 according to the embodiment transmits a bitstream containing encoded point cloud video data. The bitstream according to the embodiment is encapsulated in a file or a fragment (e.g., a stream fragment) and is transmitted via various networks such as a broadcast network and / or a broadband network. Although not shown in the figure, the transmitting device 10000 may include a packager (or a package module) configured to perform a packaging operation. According to an embodiment, the packager may be included in the transmitter 10003. According to an embodiment, the file or fragment may be sent to the receiving device 10004 via a network, or stored in a digital storage medium (e.g., USB, SD, CD, DVD, Blu-ray, HDD, SSD, etc.). The transmitter 10003 according to the embodiment can communicate with the receiving device 10004 (or the receiver 10005) via a 4G, 5G, 6G or other network by wire / wireless communication. In addition, the transmitter may perform necessary data processing operations according to the network system (e.g., a 4G, 5G or 6G communication network system). The transmitting device 10000 may transmit the encapsulated data in an on-demand manner.

[0086] The receiving device 10004 according to an embodiment includes a receiver 10005, a point cloud video decoder 10006 and / or a renderer 10007. According to an embodiment, the receiving device 10004 may include a device configured to communicate with a base station and / or other wireless devices using a radio access technology (e.g., 5G New RAT (NR), Long Term Evolution (LTE)), a robot, a vehicle, an AR / VR / XR device, a portable device, a home appliance, an Internet of Things (IoT) device, and an AI device / server.

[0087] The receiver 10005 according to the embodiment receives a bitstream containing point cloud video data or a file / segment encapsulated with a bitstream from a network or a storage medium. The receiver 10005 may perform necessary data processing according to a network system (e.g., a communication network system of 4G, 5G, 6G, etc.). The receiver 10005 according to the embodiment may decapsulate the received file / segment and output a bitstream. According to the embodiment, the receiver 10005 may include a decapsulator (or decapsulation module) configured to perform a decapsulation operation. The decapsulator may be implemented as an element (or component) separated from the receiver 10005.

[0088] The point cloud video decoder 10006 decodes the bitstream containing the point cloud video data. The point cloud video decoder 10006 can decode the point cloud video data according to the method in which the point cloud video data is encoded (for example, in the reverse process of the operation of the point cloud video encoder 10002). Therefore, the point cloud video decoder 10006 can decode the point cloud video data by performing point cloud decompression encoding (the reverse process of point cloud compression). The point cloud decompression encoding includes G-PCC encoding.

[0089] The renderer 10007 renders the decoded point cloud video data. In one embodiment, the renderer 10007 may render the decoded point cloud video data according to a viewport, etc. The renderer 10007 may render not only the point cloud video data but also the audio data to output the point cloud content. According to an embodiment, the renderer 10007 may include a display configured to display the point cloud content. According to an embodiment, the display may be implemented as a separate device or component rather than included in the renderer 10007.

[0090] The arrow indicated by the dotted line in the figure represents the transmission path of the feedback information obtained by the receiving device 10004. Feedback information is information that reflects the interactivity with the user who consumes the point cloud content, and includes information about the user (e.g., head orientation information, viewport information, etc.). Specifically, when the point cloud content is content for a service that requires interaction with the user (e.g., self-driving service, etc.), the feedback information may be provided to the content sender (e.g., sending device 10000) and / or the service provider. Depending on the embodiment, the feedback information may be used in the receiving device 10004 and the sending device 10000, or may not be provided.

[0091] The head orientation information according to the embodiment may represent information about the position, orientation, angle and movement of the user's head. The receiving device 10004 according to the embodiment may calculate the viewport information based on the head orientation information. The viewport information is information about the area of ​​the point cloud video that the user is watching (that is, the area that the user is currently watching). That is, the viewport information is information about the area that the user is currently watching in the point cloud video. In other words, the viewport or viewport area may represent the area that the user is watching in the point cloud video. The viewpoint is the point that the user is watching in the point cloud video, and may represent the center point of the viewport area. That is, the viewport is an area centered on the viewpoint, and the size and shape of the area may be determined by the field of view (FOV). Therefore, the receiving device 10004 may extract the viewport information based on the vertical or horizontal FOV supported by the device and the head orientation information. In addition, the receiving device 10004 may perform gaze analysis based on the head orientation information and / or the viewport information to determine the way the user consumes the point cloud video, the area that the user is looking at in the point cloud video, and the gaze time. According to an embodiment, the receiving device 10004 may send feedback information including the gaze analysis result to the sending device 10000. According to an embodiment, a device such as a VR / XR / AR / MR display may extract a viewport area based on the position / orientation of the user's head and the vertical or horizontal FOV supported by the device. According to an embodiment, the head orientation information and the viewport information may be referred to as feedback information, signaling information, or metadata.

[0092] According to an embodiment, feedback information may be obtained during rendering and / or display. The feedback information may be obtained by one or more sensors included in the receiving device 10004. According to an embodiment, the feedback information may be ensured by the renderer 10007 or a separate external element (or device, component, etc.). Figure 1 The dotted line in represents the process of sending feedback information ensured by the renderer 10007. Feedback information can not only be sent to the sending side, but also be consumed on the receiving side. That is, the point cloud content providing system can process (encode / decode / render) point cloud data based on the feedback information. For example, the point cloud video decoder 10006 and the renderer 10007 can prioritize decoding and rendering of point cloud videos only for the area currently viewed by the user based on the feedback information (i.e., head orientation information and / or viewport information).

[0093] In addition, the receiving device 10004 may send feedback information to the sending device 10000. The sending device 10000 (or the point cloud video data encoder 10002) may perform an encoding operation based on the feedback information. Therefore, the point cloud content providing system may efficiently process necessary data (e.g., point cloud data corresponding to the user's head position) based on the feedback information instead of processing (encoding / decoding) the entire point cloud data, and provide the point cloud content to the user.

[0094] According to an implementation, the transmitting device 10000 may be referred to as an encoder, a transmitting device, a transmitter, a transmitting system, etc., and the receiving device 10004 may be referred to as a decoder, a receiving device, a receiver, a receiving system, etc.

[0095] According to the implementation mode Figure 1 The point cloud data processed in the point cloud content providing system (through a series of processes of acquisition / encoding / transmission / decoding / rendering) may be referred to as point cloud content data or point cloud video data. According to an embodiment, point cloud content data may be used as a concept covering metadata or signaling information related to point cloud data.

[0096] Figure 1 The elements of the point cloud content providing system shown may be implemented by hardware, software, a processor and / or a combination thereof.

[0097] Figure 2 is a block diagram illustrating point cloud content providing operations according to an embodiment.

[0098] Figure 2 The block diagram shows Figure 1 The operation of the point cloud content providing system described in . As described above, the point cloud content providing system can process point cloud data based on point cloud compression coding (e.g., G-PCC).

[0099] According to an embodiment, a point cloud content providing system (e.g., a point cloud sending device 10000 or a point cloud video acquisition unit 10001) can acquire a point cloud video (20000). The point cloud video is represented by a point cloud belonging to a coordinate system for expressing a 3D space. According to an embodiment, a point cloud video may include a Ply (Polygon file format or Stanford Triangle format) file. When a point cloud video has one or more frames, the acquired point cloud video may include one or more Ply files. A Ply file contains point cloud data such as point geometry and / or attributes. Geometry includes the position of a point. The position of each point may be represented by a parameter (e.g., the value of the X, Y, and Z axes) representing a three-dimensional coordinate system (e.g., a coordinate system consisting of X, Y, and Z axes). Attributes include attributes of the point (e.g., information about the texture, color (YCbCr or RGB), reflectivity r, transparency, etc. of each point). A point has one or more attributes. For example, a point may have a color attribute or two attributes, color and reflectivity. According to an embodiment, geometry may be referred to as position, geometry information, geometry data, position information, position data, etc., and attribute may be referred to as attribute, attribute information, attribute data, etc. The point cloud content providing system (e.g., the point cloud transmitting device 10000 or the point cloud video acquiring unit 10001) may obtain point cloud data from information related to a point cloud video acquiring process (e.g., depth information, color information, etc.).

[0100] According to an embodiment, a point cloud content providing system (e.g., a transmitting device 10000 or a point cloud video encoder 10002) may encode point cloud data (20001). The point cloud content providing system may encode point cloud data based on point cloud compression coding. As described above, point cloud data may include geometric information and attribute information about points. Therefore, the point cloud content providing system may perform geometric coding for encoding geometry and output a geometric bitstream. The point cloud content providing system may perform attribute coding for encoding attributes and output an attribute bitstream. According to an embodiment, the point cloud content providing system may perform attribute coding based on geometric coding. The geometric bitstream and the attribute bitstream according to the embodiment may be multiplexed and output as one bitstream. The bitstream according to the embodiment may also include signaling information related to geometric coding and attribute coding.

[0101] According to the embodiment, the point cloud content providing system (eg, the transmitting device 10000 or the transmitter 10003) may transmit the encoded point cloud data (20002). Figure 1As shown, the encoded point cloud data can be represented by a geometry bitstream and an attribute bitstream. In addition, the encoded point cloud data can be sent in the form of a bitstream together with signaling information related to the encoding of the point cloud data (e.g., signaling information related to geometry encoding and attribute encoding). The point cloud content providing system can encapsulate the bitstream carrying the encoded point cloud data and send it in the form of a file or a fragment.

[0102] The point cloud content providing system (eg, receiving device 10004 or receiver 10005) according to the embodiment may receive a bit stream containing encoded point cloud data. In addition, the point cloud content providing system (eg, receiving device 10004 or receiver 10005) may demultiplex the bit stream.

[0103] The point cloud content providing system (e.g., the receiving device 10004 or the point cloud video decoder 10005) may decode the encoded point cloud data (e.g., geometry bitstream, attribute bitstream) sent in the bitstream. The point cloud content providing system (e.g., the receiving device 10004 or the point cloud video decoder 10005) may decode the point cloud video data based on the signaling information related to the encoding of the point cloud video data contained in the bitstream. The point cloud content providing system (e.g., the receiving device 10004 or the point cloud video decoder 10005) may decode the geometry bitstream to reconstruct the position (geometry) of the point. The point cloud content providing system may reconstruct the attributes of the point by decoding the attribute bitstream based on the reconstructed geometry. The point cloud content providing system (e.g., the receiving device 10004 or the point cloud video decoder 10005) may reconstruct the point cloud video based on the position according to the reconstructed geometry and the decoded attributes.

[0104] According to an embodiment, a point cloud content providing system (e.g., receiving device 10004 or renderer 10007) may render decoded point cloud data (20004). The point cloud content providing system (e.g., receiving device 10004 or renderer 10007) may use various rendering methods to render the geometry and attributes decoded by the decoding process. Points in the point cloud content may be rendered as vertices with a specific thickness, cubes with a specific minimum size centered at the corresponding vertex position, or circles centered at the corresponding vertex position. All or part of the rendered point cloud content is provided to the user through a display (e.g., a VR / AR display, a general display, etc.).

[0105] The point cloud content providing system (e.g., receiving device 10004) according to the embodiment may obtain feedback information (20005). The point cloud content providing system may encode and / or decode the point cloud data based on the feedback information. Feedback Information and Operations of the Point Cloud Content Providing System According to the Embodiment and Reference Figure 1 The feedback information and operations described are the same, so their detailed description is omitted.

[0106] Figure 3 An exemplary point cloud encoder according to an implementation is shown.

[0107] Figure 3 Show Figure 1 An example of a point cloud video encoder 10002. The point cloud encoder reconstructs and encodes point cloud data (e.g., the location and / or attributes of the points) to adjust the quality of the point cloud content (e.g., lossless, lossy, or near-lossless) according to network conditions or applications. When the total size of the point cloud content is large (e.g., 60 Gbps of point cloud content given for 30 fps), the point cloud content providing system may not be able to stream the content in real time. Therefore, the point cloud content providing system may reconstruct the point cloud content based on the maximum target bit rate to provide the point cloud content according to the network environment, etc.

[0108] As reference Figure 1 and Figure 2 As described, the point cloud encoder can perform geometry encoding and attribute encoding. Geometry encoding is performed before attribute encoding.

[0109] The point cloud video encoder according to the embodiment includes a coordinate transformer (transform coordinates) 30000, a quantizer (quantization and removal of points (voxelization)) 30001, an octree analyzer (analysis of octree) 30002, a surface approximation analyzer (analysis of surface approximation) 30003, an arithmetic encoder (arithmetic coding) 30004, a geometry reconstructor (reconstruction of geometry) 30005, a color transformer (transformation of color) 30006, an attribute transformer (transformation of attributes) 30007, a RAHT transformer 30008, a LOD generator (generated LOD) 30009, a lifting transformer (lifting) 30010, a coefficient quantizer (quantization of coefficients) 30011 and / or an arithmetic encoder (arithmetic coding) 30012. Figure 3 In the point cloud encoder of FIG. 30000, the coordinate transformer 30001, the quantizer 30002, the surface approximation analyzer 30003, the arithmetic encoder 30004 and the geometry reconstructor 30005 can be grouped together and referred to as a geometry encoder. The color transformer 30006, the attribute transformer 30007, the RAHT transformer 30008, the LOD generator 30009, the lifting transformer 30010, the coefficient quantizer 30011 and / or the arithmetic encoder 30012 can be grouped together and referred to as an attribute encoder.

[0110] The coordinate transformer 30000, the quantizer 30001, the octree analyzer 30002, the surface approximation analyzer 30003, the arithmetic encoder 30004 and the geometry reconstructor 30005 may perform geometry coding. The geometry coding according to the embodiment may include octree geometry coding, prediction tree geometry coding, direct coding, triplet geometry coding and entropy coding. Direct coding and triplet geometry coding are applied selectively or in combination. The geometry coding is not limited to the above examples.

[0111] As shown in the figure, the coordinate transformer 30000 according to the embodiment receives the position and transforms it into coordinates. For example, the position can be transformed into position information in a three-dimensional space (e.g., a three-dimensional space represented by an XYZ coordinate system). The position information in the three-dimensional space according to the embodiment can be referred to as geometric information.

[0112] The quantizer 30001 according to the embodiment quantizes the geometry. For example, the quantizer 30001 may quantize the points based on the minimum position value of all points (e.g., the minimum value on each of the X, Y, and Z axes). The quantizer 30001 performs a quantization operation: the difference between the minimum position value and the position value of each point is multiplied by a preset quantization scaling value, and then the nearest integer value is found by rounding the value obtained by the multiplication. Therefore, one or more points may have the same quantized position (or position value). The quantizer 30001 according to the embodiment performs voxelization based on the quantized position to reconstruct the quantized point. As in the case of a pixel (the smallest unit containing 2D image / video information), a point of a point cloud content (or 3D point cloud video) according to the embodiment may be included in one or more voxels. As a composite of volume and pixel, the term voxel refers to a 3D cubic space generated when a 3D space is divided into units (unit = 1.0) based on an axis representing the 3D space (e.g., X-axis, Y-axis, and Z-axis). The quantizer 30001 may match a point group in a 3D space with a voxel. According to an embodiment, a voxel may include only one point. According to an embodiment, a voxel may include one or more points. In order to represent a voxel as a point, the position of the center of the voxel may be set based on the position of one or more points included in the voxel. In this case, the properties of all positions included in a voxel may be combined and assigned to the voxel.

[0113] The octree analyzer 30002 according to the embodiment performs octree geometry encoding (or octree encoding) to present voxels in an octree structure. The octree structure represents points matched to voxels based on an octal tree structure.

[0114] The surface approximation analyzer 30003 according to an embodiment may analyze and approximate an octree. The octree analysis and approximation according to an embodiment is a process of analyzing a region including a plurality of points to efficiently provide an octree and voxelization.

[0115] The arithmetic encoder 30004 according to the embodiment performs entropy coding on the octree and / or the approximate octree. For example, the coding scheme includes arithmetic coding. As a result of the coding, a geometry bitstream is generated.

[0116] The color converter 30006, the attribute converter 30007, the RAHT converter 30008, the LOD generator 30009, the lifting converter 30010, the coefficient quantizer 30011 and / or the arithmetic encoder 30012 perform attribute encoding. As described above, a point may have one or more attributes. The attribute encoding according to the embodiment is also applied to the attributes of a point. However, when the attribute (e.g., color) includes one or more elements, the attribute encoding is applied independently to each element. The attribute encoding according to the embodiment includes color transform encoding, attribute transform encoding, regional adaptive hierarchical transform (RAHT) encoding, interpolation-based hierarchical nearest neighbor prediction (prediction transform) encoding, and interpolation-based hierarchical nearest neighbor prediction (lifting transform) encoding with an update / lifting step. Depending on the point cloud content, the above-mentioned RAHT encoding, prediction transform encoding, and lifting transform encoding may be selectively used, or a combination of one or more encoding schemes may be used. The attribute encoding according to the embodiment is not limited to the above examples.

[0117] The color converter 30006 according to an embodiment performs color conversion encoding to convert the color value (or texture) included in the attribute. For example, the color converter 30006 can convert the format of the color information (e.g., from RGB to YCbCr). Optionally, the operation of the color converter 30006 according to an embodiment can be applied according to the color value included in the attribute.

[0118] The geometry reconstructor 30005 according to an embodiment reconstructs (decompresses) an octree and / or an approximate octree. The geometry reconstructor 30005 reconstructs an octree / voxel based on the result of analyzing the point distribution. The reconstructed octree / voxel may be referred to as a reconstructed geometry (restored geometry).

[0119] The attribute transformer 30007 according to the embodiment performs attribute transformation to transform the attribute based on the reconstructed geometry and / or the position where the geometry encoding is not performed. As described above, since the attribute depends on the geometry, the attribute transformer 30007 can transform the attribute based on the reconstructed geometry information. For example, based on the position value of the point included in the voxel, the attribute transformer 30007 can transform the attribute of the point at the position. As described above, when the center position of the voxel is set based on the position of one or more points included in the voxel, the attribute transformer 30007 transforms the attribute of one or more points. When triplet geometry encoding is performed, the attribute transformer 30007 can transform the attribute based on the triplet geometry encoding.

[0120] The attribute converter 30007 can perform attribute conversion by calculating the average of the attributes or attribute values ​​(e.g., the color or reflectivity of each point) of the neighboring points within a specific position / radius from the center position (or position value) of each voxel. The attribute converter 30007 can apply weights according to the distance from the center to each point when calculating the average. Therefore, each voxel has a position and a calculated attribute (or attribute value).

[0121] The attribute converter 30007 can search for neighboring points within a specific position / radius from the center position of each voxel based on a KD tree or a Morton code. The KD tree is a binary search tree and supports a data structure capable of managing points based on position, so that a nearest neighbor search (NNS) can be performed quickly. The Morton code is generated by presenting the coordinates (e.g., (x, y, z)) representing the 3D positions of all points as bit values ​​and mixing the bits. For example, when the coordinates representing the point position are (5, 9, 1), the bit values ​​of the coordinates are (0101, 1001, 0001). Mixing the bit values ​​in the order of z, y, and x according to the bit index produces 010001000111. This value is represented as a decimal number 1095. That is, the Morton code value of the point with coordinates (5, 9, 1) is 1095. The attribute converter 30007 can sort the points based on the Morton code value and perform NNS through a depth-first traversal process. After the attribute transform operation, KD tree or Morton code is used when NNS is needed in another transform process for attribute encoding.

[0122] As shown, the transformed attributes are input to the RAHT transformer 30008 and / or the LOD generator 30009.

[0123] The RAHT transformer 30008 according to an embodiment performs RAHT encoding for predicting attribute information based on the reconstructed geometric information. For example, the RAHT transformer 30008 may predict attribute information of a higher-level node in the octree based on attribute information associated with a lower-level node in the octree.

[0124] The LOD generator 30009 according to an embodiment generates a level of detail (LOD) to perform predictive transform coding. The LOD according to an embodiment is the degree of detail of the point cloud content. As the LOD value decreases, the details of the point cloud content are degraded. As the LOD value increases, the details of the point cloud content are enhanced. Points can be classified by LOD.

[0125] The lifting transformer 30010 according to an embodiment performs lifting transform coding that transforms point cloud attributes based on weights. As described above, lifting transform coding can be optionally applied.

[0126] The coefficient quantizer 30011 according to the embodiment quantizes the attribute of the attribute encoding based on the coefficient.

[0127] The arithmetic encoder 30012 according to the embodiment encodes quantized properties based on arithmetic coding.

[0128] Although not shown in the figure, Figure 3 The elements of the point cloud encoder may be implemented by hardware, software, firmware, or a combination thereof including one or more processors or integrated circuits configured to communicate with one or more memories included in the point cloud providing device. The one or more processors may execute the above Figure 3 At least one of the operations and / or functions of the elements of the point cloud encoder. In addition, one or more processors may be operable or executable to perform Figure 3 A software program and / or set of instructions for the operation and / or functionality of the elements of the point cloud encoder. One or more memories according to an embodiment may include high-speed random access memory, or include non-volatile memory (e.g., one or more disk storage devices, flash memory devices, or other non-volatile solid-state memory devices).

[0129] Figure 4 Examples of octrees and occupancy codes are shown according to an embodiment.

[0130] As reference Figures 1 to 3 As described, the point cloud content providing system (point cloud video encoder 10002) or the point cloud encoder (e.g., octree analyzer 30002) performs octree geometry encoding (or octree encoding) based on the octree structure to efficiently manage the area and / or position of voxels.

[0131] Figure 4 The upper part of FIG. 1 shows the octree structure. The 3D space of the point cloud content according to the embodiment is represented by the axes (eg, X-axis, Y-axis, and Z-axis) of the coordinate system. d ,2 d ,2 d ) to create an octree structure. Here, 2d can be set to the value of the smallest bounding box that surrounds all points of the point cloud content (or point cloud video). Here, d represents the depth of the octree. The value of d is determined in the following formula. In the following formula, (x int n ,y int n ,z int n ) represents the position (or position value) of the quantized point.

[0132]

[0133] like Figure 4As shown in the middle of the upper part of , the entire 3D space can be divided into eight spaces according to the partition. Each divided space is represented by a cube with six faces. Figure 4 As shown in the upper right part of , each of the eight spaces is divided again based on the axis of the coordinate system (e.g., X-axis, Y-axis, and Z-axis). Therefore, each space is divided into eight smaller spaces. The divided smaller spaces are also represented by cubes with six faces. This division scheme is applied until the leaf nodes of the octree become voxels.

[0134] Figure 4 The lower part shows the octree occupancy code. The occupancy code of the octree is generated to indicate whether each of the eight divided spaces generated by dividing one space contains at least one point. Therefore, a single occupancy code is represented by eight child nodes. Each child node represents the occupancy of the divided space, and the child node has a 1-bit value. Therefore, the occupancy code is represented as an 8-bit code. That is, when the space corresponding to the child node contains at least one point, the node is assigned a value of 1. When the space corresponding to the child node does not contain a point (the space is empty), the node is assigned a value of 0. Since Figure 4 The occupancy code shown is 00100001, thus indicating that the spaces corresponding to the third and eighth subnodes among the eight subnodes each contain at least one point. As shown in the figure, each of the third and eighth subnodes has eight subnodes, and the subnodes are represented by 8-bit occupancy codes. The accompanying drawings show that the occupancy code of the third subnode is 10000111, and the occupancy code of the eighth subnode is 01001111. The point cloud encoder (e.g., arithmetic encoder 30004) according to an embodiment may perform entropy encoding on the occupancy code. In order to increase compression efficiency, the point cloud encoder may perform intra / inter encoding on the occupancy code. The receiving device (e.g., receiving device 10004 or point cloud video decoder 10006) according to an embodiment reconstructs the occupancy code based on the occupancy code.

[0135] According to an embodiment, a point cloud encoder (e.g., Figure 4 The point cloud encoder or octree analyzer 30002) can perform voxelization and octree encoding to store point locations. However, points are not always evenly distributed in the 3D space, so there may be specific areas where fewer points exist. Therefore, performing voxelization on the entire 3D space is inefficient. For example, when a specific area contains few points, voxelization does not need to be performed in the specific area.

[0136] Therefore, for the above-mentioned specific area (or nodes other than the leaf nodes of the octree), the point cloud encoder according to the embodiment may skip voxelization and perform direct encoding to directly encode the point positions included in the specific area. The coordinates of the directly encoded points according to the embodiment are called direct coding mode (DCM). The point cloud encoder according to the embodiment may also perform triplet geometry encoding based on the surface model, which is to reconstruct the point positions in the specific area (or node) based on voxels. Triplet geometry encoding is a geometry encoding that represents an object as a series of triangular meshes. Therefore, the point cloud decoder can generate a point cloud from the mesh surface. Direct encoding and triplet geometry encoding according to the embodiment can be performed selectively. In addition, direct encoding and triplet geometry encoding according to the embodiment can be performed in combination with octree geometry encoding (or octree encoding).

[0137] In order to perform direct encoding, the option of using direct mode to apply direct encoding should be enabled. The node to which direct encoding is to be applied is not a leaf node, and there should be points less than a threshold value within a specific node. In addition, the total number of points to which direct encoding is to be applied should not exceed a preset threshold. When the above conditions are met, the point cloud encoder (or arithmetic encoder 30004) according to the embodiment can perform entropy encoding on the point position (or position value).

[0138] A point cloud encoder (e.g., surface approximation analyzer 30003) according to an embodiment may determine a specific level of the octree (a level less than the depth d of the octree), and may use a surface model starting from this level to perform triplet geometry encoding to reconstruct point positions in the node area based on voxels (triplet mode). A point cloud encoder according to an embodiment may specify a level to which triplet geometry encoding is to be applied. For example, when a specific level is equal to the depth of the octree, the point cloud encoder does not operate in triplet mode. In other words, a point cloud encoder according to an embodiment may operate in triplet mode only when the specified level is less than the depth value of the octree. A 3D cubic area of ​​a node at a specified level according to an embodiment is referred to as a block. A block may include one or more voxels. A block or voxel may correspond to a block. Geometry is represented as a surface within each block. A surface according to an embodiment may intersect each edge of a block at most once.

[0139] A block has 12 edges, so there are at least 12 intersections in a block. Each intersection is called a vertex. When there is at least one occupied voxel adjacent to the edge among all blocks sharing the edge, the vertex existing along the edge is detected. An occupied voxel according to an embodiment refers to a voxel containing a point. The vertex position detected along the edge is the average position of the edge of all voxels adjacent to the edge among all blocks along the shared edge.

[0140] Once the vertex is detected, the point cloud encoder according to the embodiment can perform entropy encoding on the start point (x, y, z) of the edge, the direction vector (Δx, Δy, Δz) of the edge, and the vertex position value (relative position value within the edge). When triplet geometry encoding is applied, the point cloud encoder according to the embodiment (e.g., geometry reconstructor 30005) can generate restored geometry (reconstructed geometry) by performing triangle reconstruction, upsampling, and voxelization processes.

[0141] The vertices located at the edges of the block determine the surface passing through the block. The surface according to an embodiment is a non-planar polygon. In the triangle reconstruction process, the surface represented by the triangle is reconstructed based on the starting point of the edge, the direction vector of the edge and the position value of the vertex. The triangle reconstruction process is performed by i) calculating the centroid value of each vertex, ii) subtracting the center value from each vertex value, and iii) estimating the sum of the squares of the values ​​obtained by subtraction.

[0142] ① ② ③

[0143] Estimate the minimum value of the sum, and perform the projection process according to the axis with the minimum value. For example, when the element x is the smallest, each vertex is projected on the x-axis relative to the center of the block and projected on the (y,z) plane. When the value obtained by the projection on the (y,z) plane is (ai,bi), the value of θ is estimated by atan2(bi,ai), and the vertices are sorted based on the value of θ. The following table shows the vertex combination for creating triangles according to the number of vertices. The vertices are sorted from 1 to n. Table 1 below shows that for four vertices, two triangles can be constructed according to the vertex combination. The first triangle can be composed of vertices 1, 2, and 3 among the sorted vertices, and the second triangle can be composed of vertices 3, 4, and 1 among the sorted vertices.

[0144] [Table 1]

[0145] [Table 1] Triangles formed from vertices sorted in 1,…,n order

[0146] n Triangle 3 (1,2,3) 4 (1,2,3),(3,4,1) 5 (1,2,3),(3,4,5),(5,1,3) 6 (1,2,3),(3,4,5),(5,6,1),(1,3,5) 7 (1,2,3),(3,4,5),(5,6,7),(7,1,3),(3,5,7) 8 (1,2,3),(3,4,5),(5,6,7),(7,8,1),(1,3,5),(5,7,1) 9 (1,2,3),(3,4,5),(5,6,7),(7,8,9),(9,1,3),(3,5,7),(7,9,3) 10 (1,2,3),(3,4,5),(5,6,7),(7,8,9),(9,10,1),(1,3,5),(5,7,9),(9,1,5) 11 (1,2,3),(3,4,5),(5,6,7),(7,8,9),(9,10,11),(11,1,3),(3,5,7),(7,9,11),(11,3,7) 12 (1,2,3),(3,4,5),(5,6,7),(7,8,9),(9,10,11),(11,12,1),(1,3,5),(5,7,9),(9,11,1),(1,5,9)

[0147] An upsampling process is performed to add points in the middle along the edges of the triangle, and voxelization is performed. The added points are generated based on the upsampling factor and the width of the block. The added points are called refined vertices. The point cloud encoder according to an embodiment may voxelize the refined vertices. In addition, the point cloud encoder may perform attribute encoding based on the voxelized position (or position value).

[0148] Figure 5 An example of point configuration in each LOD according to an embodiment is shown.

[0149] As reference Figures 1 to 4 As described, the encoded geometry is reconstructed (decompressed) before performing attribute encoding. When direct encoding is applied, the geometry reconstruction operation may include changing the placement of the directly encoded points (e.g., placing the directly encoded points in front of the point cloud data). When triplet geometry encoding is applied, the geometry reconstruction process is performed by triangle reconstruction, upsampling, and voxelization. Since the attributes depend on the geometry, attribute encoding is performed based on the reconstructed geometry.

[0150] The point cloud encoder (e.g., LOD generator 30009) can classify (or reorganize) the points by LOD. The figure shows the point cloud content corresponding to the LOD. The leftmost screen in the figure shows the original point cloud content. The second screen from the left in the figure shows the distribution of points in the lowest LOD, and the rightmost screen in the figure shows the distribution of points in the highest LOD. That is, the points in the lowest LOD are sparsely distributed, and the points in the highest LOD are densely distributed. That is, as the LOD increases in the direction indicated by the arrow indicated at the bottom of the figure, the space (or distance) between the points becomes narrower.

[0151] Figure 6 An example of point configuration for each LOD is shown according to an embodiment.

[0152] As reference Figures 1 to 5 As described, the point cloud content providing system or point cloud encoder (eg, point cloud video encoder 10002, Figure 3 The point cloud encoder or LOD generator 30009) can generate LOD. LOD is generated by reorganizing points into a set of refinement levels according to the set LOD distance value (or Euclidean distance set). The LOD generation process is performed not only by the point cloud encoder, but also by the point cloud decoder.

[0153] Figure 6 The upper part of shows examples of points (P0 to P9) of point cloud contents distributed in 3D space. Figure 6 In , the original order represents the order of points P0 to P9 before LOD generation. Figure 6 In the LOD-based order, it means the order of points generated according to LOD. Points are reorganized by LOD. In addition, a high LOD contains points belonging to a lower LOD. Figure 6 As shown, LOD0 contains P0, P5, P4, and P2. LOD1 contains the points of LOD0, P1, P6, and P3. LOD2 contains the points of LOD0, the points of LOD1, P9, P8, and P7.

[0154] As reference Figure 3 As described, the point cloud encoder according to the embodiment may selectively or in combination perform prediction transform coding, lifting transform coding, and RAHT transform coding.

[0155] The point cloud encoder according to the embodiment may generate a predictor for a point to perform predictive transform coding for setting a prediction attribute (or a prediction attribute value) of each point. That is, N predictors may be generated for N points. The predictor according to the embodiment may calculate a weight (=1 / distance) based on the LOD value of each point, index information about neighboring points that exist within a set distance of each LOD, and the distance to the neighboring point.

[0156] The predicted attribute (or attribute value) according to the embodiment is set to the average of the values ​​obtained by multiplying the attribute (or attribute value) (e.g., color, reflectivity, etc.) of the neighboring points set in the predictor of each point by the weight (or weight value) calculated based on the distance to each neighboring point. The point cloud video encoder (e.g., coefficient quantizer 30011) according to the embodiment can quantize and inverse quantize the residual (which can be called residual attribute, residual attribute value, attribute prediction residual, etc.) obtained by subtracting the predicted attribute (or attribute value) of each point from the attribute (attribute value) of each point. Tables 2 and 3 below show the quantization process.

[0157] [Table 2]

[0158] int PCCQuantization(int value, int quantStep) { if (value >= 0) { return floor(value / quantStep + 1.0 / 3.0); } else { return -floor(-value / quantStep + 1.0 / 3.0); } }

[0159] [Table 3]

[0160]

[0161]

[0162] When the predictor of each point has a neighboring point, the point cloud encoder (e.g., the arithmetic encoder 30012) according to the embodiment may perform entropy encoding on the quantized and inverse quantized residual values ​​as described above. When the predictor of each point has no neighboring point, the point cloud encoder (e.g., the arithmetic encoder 30012) according to the embodiment may perform entropy encoding on the attributes of the corresponding point without performing the above operation.

[0163] The point cloud encoder (e.g., lifting transformer 30010) according to an embodiment may generate a predictor for each point, set the calculated LOD and register neighbor points in the predictor, and set weights according to the distance to the neighbor points to perform lifting transform encoding. The lifting transform encoding according to an embodiment is similar to the above-mentioned predictive transform encoding, but is different in that weights are cumulatively applied to attribute values. The process of cumulatively applying weights to attribute values ​​according to an embodiment is configured as follows.

[0164] 1) Create an array quantized weight (QW) for storing weight values ​​of each point. The initial value of all elements of QW is 1.0. Multiply the QW value of the predictor index of the neighbor node registered in the predictor by the weight of the predictor of the current point, and add the values ​​obtained by multiplication.

[0165] 2) Boosting prediction process: The value obtained by multiplying the attribute value of the point by the weight is subtracted from the existing attribute value to calculate the predicted attribute value.

[0166] 3) Create temporary arrays called updateweight and update, and initialize the temporary arrays to zero.

[0167] 4) Accumulate the weight calculated by multiplying the weight calculated for all predictors by the weight stored in the QW corresponding to the predictor index with the updateweight array as the index of the neighbor node. Accumulate the value obtained by multiplying the attribute value of the neighbor node index by the calculated weight with the update array.

[0168] 5) Boosting update process: Divide the attribute values ​​of the update array of all predictors by the weight value of the updateweight array of the predictor index, and add the existing attribute value to the value obtained by the division.

[0169] 6) For all predictors, the predicted attribute is calculated by multiplying the attribute value updated by the lifting update process by the weight (stored in QW) updated by the lifting prediction process. The point cloud encoder (e.g., coefficient quantizer 30011) according to the embodiment quantizes the predicted attribute value. In addition, the point cloud encoder (e.g., arithmetic encoder 30012) performs entropy encoding on the quantized attribute value.

[0170] A point cloud encoder according to an embodiment (e.g., RAHT transformer 30008) may perform RAHT transform coding, in which attributes associated with nodes at a lower level in an octree are used to predict attributes of nodes at a higher level. RAHT transform coding is an example of attribute intra-frame coding by scanning backward through an octree. A point cloud encoder according to an embodiment scans the entire area starting from voxels and repeats a merging process of merging voxels into larger blocks at each step until a root node is reached. The merging process according to an embodiment is performed only on occupied nodes. The merging process is not performed on empty nodes. The merging process is performed on the upper node directly above the empty node.

[0171] The following formula represents the RAHT transformation matrix. In this formula, gl x, y, z represents the average attribute value of the voxels at level l. glx, y, z can be based on and to calculate. and The weight is and

[0172]

[0173] here, Is the low-pass value and is used during the next highest level of merging. represents the high pass coefficient 0000000000. The high pass coefficients of each step are quantized and subjected to entropy coding (e.g., encoded by arithmetic encoder 40012). The weights are calculated as pass and Create the root node as follows.

[0174]

[0175] The values ​​of gDC are also quantized and subjected to entropy decoding like the high-pass coefficients.

[0176] Figure 7 A point cloud decoder according to an embodiment is shown.

[0177] Figure 7 The point cloud decoder shown is an example of a point cloud decoder and can perform a decoding operation, which is Figures 1 to 6 The inverse process of the encoding operation of the point cloud encoder shown.

[0178] As reference Figure 1 and Figure 6 As described, the point cloud decoder can perform geometry decoding and attribute decoding. Geometry decoding is performed before attribute decoding.

[0179] According to an embodiment, a point cloud decoder includes an arithmetic decoder (arithmetic decoding) 7000, an octree synthesizer (synthesized octree) 7001, a surface approximation synthesizer (synthesized surface approximation) 7002 and a geometry reconstructor (reconstructed geometry) 7003, an inverse coordinate transformer (inverse transformed coordinates) 7004, an arithmetic decoder (arithmetic decoding) 7005, an inverse quantizer (inverse quantization) 7006, a RAHT transformer 7007, an LOD generator (generate LOD) 7008, an inverse lifting (inverse lifting) 7009 and / or an inverse color transformer (inverse transformed color) 7010.

[0180] The arithmetic decoder 7000, the octree synthesizer 7001, the surface approximation synthesizer 7002, the geometry reconstructor 7003, and the coordinate inverse transformer 7004 may perform geometry decoding. The geometry decoding according to the embodiment may include direct decoding and triplet geometry decoding. Direct encoding and triplet geometry decoding are selectively applied. The geometry decoding is not limited to the above examples, and as a reference Figures 1 to 6 The inverse process of the geometric encoding described is performed.

[0181] The arithmetic decoder 7000 according to the embodiment decodes the received geometry bitstream based on arithmetic coding. The operation of the arithmetic decoder 7000 corresponds to the inverse process of the arithmetic encoder 30004.

[0182] The octree synthesizer 7001 according to the embodiment may generate an octree by acquiring an occupancy code (or information about the geometry obtained as a result of decoding) from the decoded geometry bitstream. Figures 1 to 6 Describe the configuration in detail.

[0183] When triplet geometry encoding is applied, the surface approximation synthesizer 7002 according to an embodiment may synthesize a surface based on the decoded geometry and / or the generated octree.

[0184] According to an embodiment, the geometry reconstructor 7003 may regenerate the geometry based on the surface and / or the decoded geometry. Figures 1 to 9 As described, direct coding and triplet geometry coding are selectively applied. Therefore, the geometry reconstructor 7003 directly imports the position information about the points to which direct coding is applied and adds them. When triplet geometry coding is applied, the geometry reconstructor 7003 can reconstruct the geometry by performing the reconstruction operations (e.g., triangle reconstruction, upsampling, and voxelization) of the geometry reconstructor 30005. Details and References Figure 6 Those described are the same, so their description is omitted. The reconstructed geometry may include a point cloud screen or frame that does not contain attributes.

[0185] The coordinate inverse transformer 7004 according to an embodiment may acquire a point position by transforming the coordinates based on the reconstructed geometry.

[0186] The arithmetic decoder 7005, the inverse quantizer 7006, the RAHT transformer 7007, the LOD generator 7008, the inverse lifter 7009 and / or the color inverse transformer 7010 may perform attribute decoding. Attribute decoding according to an embodiment includes regional adaptive hierarchical transform (RAHT) decoding, hierarchical nearest neighbor prediction (prediction transform) decoding based on interpolation, and hierarchical nearest neighbor prediction (lifting transform) decoding based on interpolation with an update / lifting step. The above three decoding schemes may be selectively used, or a combination of one or more decoding schemes may be used. Attribute decoding according to an embodiment is not limited to the above examples.

[0187] The arithmetic decoder 7005 according to the embodiment decodes the attribute bit stream through arithmetic coding.

[0188] The inverse quantizer 7006 according to the embodiment inversely quantizes information about the decoded attribute bitstream or the attribute obtained as a result of decoding, and outputs the inversely quantized attribute (or attribute value). The inverse quantization may be selectively applied based on the attribute encoding of the point cloud encoder.

[0189] According to an embodiment, the RAHT transformer 7007, the LOD generator 7008 and / or the inverse lifter 7009 may process the reconstructed geometry and inverse quantized attributes. As described above, the RAHT transformer 7007, the LOD generator 7008 and / or the inverse lifter 7009 may selectively perform a decoding operation corresponding to the encoding of the point cloud encoder.

[0190] The color inverse transformer 7010 according to an embodiment performs inverse transform encoding to inversely transform a color value (or texture) included in the decoded attribute. The operation of the color inverse transformer 7010 may be selectively performed based on the operation of the color transformer 30006 of the point cloud encoder.

[0191] Although not shown in the figure, Figure 7 The elements of the point cloud decoder may be implemented by hardware, software, firmware, or a combination thereof including one or more processors or integrated circuits configured to communicate with one or more memories included in the point cloud providing device. The one or more processors may execute the above Figure 7 At least one or more of the operations and / or functions of the elements of the point cloud decoder. In addition, one or more processors may be operable or executed to perform Figure 7 A software program and / or instruction set for the operation and / or functionality of the elements of a point cloud decoder.

[0192] Figure 8 A transmitting device according to an embodiment is shown.

[0193] Figure 8 The sending device shown is Figure 1 The sending device 10000 (or Figure 3 Example of a point cloud encoder). Figure 8 The sending device shown can be executed with reference to Figures 1 to 6 One or more operations and methods that are the same or similar to those of the described point cloud encoder. A sending device according to an embodiment may include a data input unit 8000, a quantization processor 8001, a voxelization processor 8002, an octree occupancy code generator 8003, a surface model processor 8004, an intra-frame / inter-frame coding processor 8005, an arithmetic encoder 8006, a metadata processor 8007, a color transformation processor 8008, an attribute transformation processor 8009, a prediction / lifting / RAHT transformation processor 8010, an arithmetic encoder 8011 and / or a sending processor 8012.

[0194] The data input unit 8000 receives or acquires point cloud data according to an embodiment. The data input unit 8000 may perform the same operation and / or acquisition method as the point cloud video acquisition unit 10001 (or refer to Figure 2The acquisition process 20000) described above has the same or similar operations and / or acquisition methods.

[0195] The data input unit 8000, the quantization processor 8001, the voxelization processor 8002, the octree occupancy code generator 8003, the surface model processor 8004, the intra / inter coding processor 8005 and the arithmetic encoder 8006 perform geometric coding. Figures 1 to 9 The geometric encodings described are the same or similar, so their detailed description is omitted.

[0196] The quantization processor 8001 according to an embodiment quantizes geometry (e.g., position values ​​of points). The operation and / or quantization of the quantization processor 8001 is similar to the reference Figure 3 The operation and / or quantization of the quantizer 30001 described above are the same or similar. Figures 1 to 9 Same as those described.

[0197] According to an embodiment, the voxelization processor 8002 voxelizes the quantized position value of the point. The voxelization processor 8002 may perform the same as the reference Figure 3 The operation and / or voxelization process of the quantizer 30001 described above is the same or similar to the operation and / or process described above. Figures 1 to 6 Same as those described.

[0198] The octree occupancy code generator 8003 according to the embodiment performs octree encoding on the voxelized position of the point based on the octree structure. The octree occupancy code generator 8003 may generate an occupancy code. The octree occupancy code generator 8003 may perform the same as the reference code. Figure 3 and Figure 4 The operations and / or methods of the point cloud encoder (or octree analyzer 30002) described herein are the same or similar to the operations and / or methods described herein. Figures 1 to 6 Same as those described.

[0199] According to an embodiment, the surface model processor 8004 may perform triplet geometry encoding based on the surface model to reconstruct the point position in a specific area (or node) based on voxels. Figure 3 The operations and / or methods of the point cloud encoder (e.g., surface approximation analyzer 30003) described above are the same or similar to the operations and / or methods. Figures 1 to 6 Same as those described.

[0200] The intra / inter encoding processor 8005 according to an embodiment may perform intra / inter encoding on the point cloud data. The intra / inter encoding processor 8005 may perform encoding that is the same as or similar to intra / inter encoding. According to an embodiment, the intra / inter encoding processor 8005 may be included in the arithmetic encoder 8006.

[0201] According to an embodiment, the arithmetic encoder 8006 performs entropy coding on the octree and / or approximate octree of the point cloud data. For example, the coding scheme includes arithmetic coding. The arithmetic encoder 8006 performs the same or similar operations and / or methods as the operations and / or methods of the arithmetic encoder 30004.

[0202] The metadata processor 8007 according to an embodiment processes metadata (e.g., set values) about the point cloud data and provides it to necessary processing processes such as geometry coding and / or attribute coding. In addition, the metadata processor 8007 according to an embodiment may generate and / or process signaling information related to geometry coding and / or attribute coding. The signaling information according to an embodiment may be encoded separately from the geometry coding and / or attribute coding. The signaling information according to an embodiment may be interleaved.

[0203] The color conversion processor 8008, the attribute conversion processor 8009, the prediction / lifting / RAHT conversion processor 8010 and the arithmetic encoder 8011 perform attribute encoding. Figures 1 to 6 The attribute codes described are the same or similar, so their detailed description is omitted.

[0204] The color transform processor 8008 according to an embodiment performs color transform encoding to transform the color value included in the attribute. The color transform processor 8008 may perform color transform encoding based on the reconstructed geometry. The reconstructed geometry is compared with the reference Figures 1 to 9 In addition, its execution is the same as that of reference Figure 3 The operations and / or methods of the color converter 30006 described above are the same as or similar to the operations and / or methods described above, and detailed descriptions thereof are omitted.

[0205] The attribute transformation processor 8009 according to an embodiment performs attribute transformation to transform attributes based on reconstructed geometry and / or locations where geometry encoding is not performed. The attribute transformation processor 8009 performs the attribute transformation with reference to Figure 3 The operations and / or methods of the attribute converter 30007 described in the foregoing description are the same as or similar to those of the attribute converter 30007. Detailed description thereof is omitted. The prediction / lifting / RAHT transform processor 8010 according to the embodiment may encode the transformed attribute by any one or a combination of RAHT coding, prediction transform coding, and lifting transform coding. The prediction / lifting / RAHT transform processor 8010 performs the same as or similar to the reference. Figure 3The operations of the RAHT transformer 30008, the LOD generator 30009 and the lifting transformer 30010 described above are the same or similar to at least one operation. In addition, the prediction transform coding, the lifting transform coding and the RAHT transform coding are the same as those of the reference Figures 1 to 9 Those described are the same, so their detailed description is omitted.

[0206] The arithmetic encoder 8011 according to the embodiment may encode the encoded attribute based on arithmetic coding. The arithmetic encoder 8011 performs the same or similar operation and / or method as that of the arithmetic encoder 300012.

[0207] The transmitting processor 8012 according to the embodiment may transmit individual bitstreams containing coded geometry and / or coded attributes or metadata information, or transmit a bitstream containing coded geometry and / or coded attributes and metadata information. When the coded geometry and / or coded attributes and metadata information according to the embodiment are configured into a bitstream, the bitstream may include one or more sub-bitstreams. The bitstream according to the embodiment may include signaling information and slice data, and the signaling information includes a sequence parameter set (SPS) for sequence level signaling, a geometry parameter set (GPS) for geometry information coding signaling, an attribute parameter set (APS) for attribute information coding signaling, and a patch parameter set (TPS) for patch level signaling. The slice data may include information about one or more slices. A slice according to the embodiment may include a geometry bitstream Geom00 and one or more attribute bitstreams Attr00 and Attr10.

[0208] A slice is a sequence of syntactic elements that represents the entirety or a portion of an encoded point cloud frame.

[0209] The TPS according to an embodiment may include information about each of the one or more tiles (e.g., coordinate information and height / size information about a bounding box). The geometry bitstream may include a header and a payload. The header of the geometry bitstream according to an embodiment may include a parameter set identifier (geom_parameter_set_id), a tile identifier (geom_tile_id), and a slice identifier (geom_slice_id) included in the GPS, as well as information about the data contained in the payload. As described above, the metadata processor 8007 according to an embodiment may generate and / or process signaling information and send it to the transmission processor 8012. According to an embodiment, an element that performs geometry encoding and an element that performs attribute encoding may share data / information with each other as indicated by the dotted lines. The transmission processor 8012 according to an embodiment may perform operations and / or transmission methods that are the same or similar to those of the transmitter 10003. Details and References Figure 1 andFigure 2 Those described are the same, so their description is omitted.

[0210] Figure 9 A receiving device according to an embodiment is illustrated.

[0211] Figure 9 The receiving device shown is Figure 1 An example of a receiving device 10004. Figure 9 The receiving device shown can execute and refer to Figures 1 to 8 One or more operations and methods that are the same or similar to those of the described point cloud decoder.

[0212] The receiving device according to the embodiment may include a receiver 9000, a receiving processor 9001, an arithmetic decoder 9002, an octtree reconstruction processor based on an occupancy code 9003, a surface model processor (triangle reconstruction, upsampling, voxelization) 9004, an inverse quantization processor 9005, a metadata parser 9006, an arithmetic decoder 9007, an inverse quantization processor 9008, a prediction / lifting / RAHT inverse transform processor 9009, a color inverse transform processor 9010 and / or a renderer 9011. Each decoding element according to the embodiment may perform an inverse process of the operation of the corresponding encoding element according to the embodiment.

[0213] The receiver 9000 according to the embodiment receives point cloud data. The receiver 9000 may perform the same Figure 1 The operation and / or receiving method of the receiver 10005 is the same as or similar to the operation and / or receiving method of the receiver 10005. The detailed description thereof is omitted.

[0214] The reception processor 9001 according to an embodiment may acquire a geometry bitstream and / or an attribute bitstream from the received data. The reception processor 9001 may be included in the receiver 9000 .

[0215] The arithmetic decoder 9002, the octtree reconstruction processor 9003 based on the occupancy code, the surface model processor 9004, and the inverse quantization processor 9005 may perform geometric decoding. Figures 1 to 10 The geometric decoding described is the same or similar, so its detailed description is omitted.

[0216] The arithmetic decoder 9002 according to the embodiment may decode the geometry bitstream based on arithmetic coding. The arithmetic decoder 9002 performs the same or similar operations and / or encoding as those of the arithmetic decoder 7000.

[0217] The octree reconstruction processor 9003 based on the occupancy code according to the embodiment can reconstruct the octree by obtaining the occupancy code from the decoded geometry bitstream (or information about the geometry obtained as a result of decoding). The octree reconstruction processor 9003 based on the occupancy code performs the same or similar operations and / or methods as the operations of the octree synthesizer 7001 and / or the octree generation method. When triplet geometry coding is applied, the surface model processor 9004 according to the embodiment can perform triplet geometry decoding and related geometry reconstruction (e.g., triangle reconstruction, upsampling, voxelization) based on the surface model method. The surface model processor 9004 performs the same or similar operations as the operations of the surface approximation synthesizer 7002 and / or the geometry reconstructor 7003.

[0218] The inverse quantization processor 9005 according to an embodiment may inversely quantize the decoded geometry.

[0219] According to an embodiment, the metadata parser 9006 may parse metadata (eg, setting values) contained in the received point cloud data. The metadata parser 9006 may pass the metadata to the geometry decoding and / or attribute decoding. Figure 8 The metadata described is the same, so a detailed description thereof is omitted.

[0220] The arithmetic decoder 9007, the inverse quantization processor 9008, the prediction / lifting / RAHT inverse transform processor 9009 and the color inverse transform processor 9010 perform attribute decoding. Figures 1 to 8 The attribute decoding described in at least one of them is the same or similar, so its detailed description is omitted.

[0221] The arithmetic decoder 9007 according to the embodiment may decode the attribute bitstream by arithmetic coding. The arithmetic decoder 9007 may decode the attribute bitstream based on the reconstructed geometry. The arithmetic decoder 9007 performs the same or similar operation and / or encoding as that of the arithmetic decoder 7005.

[0222] The inverse quantization processor 9008 according to an embodiment may inversely quantize the decoded attribute bitstream. The inverse quantization processor 9008 performs the same or similar operation and / or method as the inverse quantizer 7006 and / or the inverse quantization method.

[0223] The prediction / lifting / RAHT inverse transform processor 9009 according to an embodiment can process the reconstructed geometric and inverse quantized properties. The prediction / lifting / RAHT inverse transform processor 9009 performs the same Figure 7The color inverse transform processor 9010 according to the embodiment performs inverse transform encoding to inversely transform the color value (or texture) included in the decoded attribute. The color inverse transform processor 9010 performs the same or similar operations and / or decoding as the RAHT transformer 7007, the LOD generator 7008 and / or the inverse lifter 7009 of the embodiment. Figure 7 The operations and / or inverse transform encoding of the color inverse transformer 7010 are the same as or similar to the operations and / or inverse transform encoding of the color inverse transformer 7010. The renderer 9011 according to the embodiment may render point cloud data.

[0224] Figure 10 An exemplary structure that can be operated in conjunction with a point cloud data transmitting / receiving method / apparatus according to an embodiment is shown.

[0225] Figure 10 The structure of 1060 represents a configuration in which at least one of the server 1060, the robot 1010, the self-driving vehicle 1020, the XR device 1030, the smart phone 1040, the home appliance 1050, and / or the head mounted display (HMD) 1070 is connected to the cloud network 1000. The robot 1010, the self-driving vehicle 1020, the XR device 1030, the smart phone 1040, or the home appliance 1050 is referred to as a device. In addition, the XR device 1030 may correspond to a point cloud data (PCC) device according to an embodiment or may be operably connected to a PCC device.

[0226] The cloud network 1000 may represent a network that constitutes a part of a cloud computing infrastructure or exists in a cloud computing infrastructure. Here, the cloud network 1000 may be configured using a 3G network, a 4G or Long Term Evolution (LTE) network, or a 5G network.

[0227] The server 1060 may be connected to at least one of the robot 1010 , the self-driving vehicle 1020 , the XR device 1030 , the smart phone 1040 , the home appliance 1050 , and / or the HMD 1070 via the cloud network 1000 , and may assist in at least a portion of the processing of the connected devices 1010 to 1070 .

[0228] The HMD 1070 represents one of implementation types of an XR device and / or a PCC device according to an embodiment. The HMD type device according to an embodiment includes a communication unit, a control unit, a memory, an I / O unit, a sensor unit, and a power supply unit.

[0229] Hereinafter, various embodiments of devices 1010 to 1050 to which the above-described technology is applied will be described. Figure 10 The illustrated devices 1010 to 1050 may be operably connected / coupled to the point cloud data sending device and receiving device according to the above-described embodiments.

[0230] <PCC + XR>

[0231] The XR / PCC device 1030 may employ PCC technology and / or XR (AR+VR) technology, and may be implemented as an HMD, a head-up display (HUD) disposed in a vehicle, a television, a mobile phone, a smart phone, a computer, a wearable device, a home appliance, a digital signage, a vehicle, a fixed robot, or a mobile robot.

[0232] The XR / PCC device 1030 may analyze 3D point cloud data or image data acquired through various sensors or from an external device and generate position data and attribute data about 3D points. Thus, the XR / PCC device 1030 may acquire information about the surrounding space or real objects, and render and output XR objects. For example, the XR / PCC device 1030 may match an XR object including auxiliary information about the recognized object with the recognized object and output the matched XR object.

[0233] <PCC + XR + Mobile Phone>

[0234] The XR / PCC device 1030 may be implemented as a smart phone 1040 by applying the PCC technology.

[0235] Smart phone 1040 may decode and display point cloud content based on PCC technology.

[0236] <PCC + Self - Driving + XR>

[0237] The self-driving vehicle 1020 may be implemented as a mobile robot, a vehicle, an unmanned aerial vehicle, etc. by applying PCC technology and XR technology.

[0238] The self-driving vehicle 1020 to which the XR / PCC technology is applied may represent a self-driving vehicle provided with a means for providing an XR image, or a self-driving vehicle as a control / interaction target in the XR image. Specifically, as a control / interaction target in the XR image, the self-driving vehicle 1020 may be distinguished from the XR device 1030 and may be operably connected thereto.

[0239] The self-driving vehicle 1020 having means for providing XR / PCC images may acquire sensor information from sensors including cameras, and output generated XR / PCC images based on the acquired sensor information. For example, the self-driving vehicle 1020 may have a HUD and output XR / PCC images thereto, thereby providing passengers with XR / PCC objects corresponding to real objects or objects presented on the screen.

[0240] When the XR / PCC object is output to the HUD, at least a portion of the XR / PCC object may be output to overlap with the real object to which the passenger's eyes are directed. On the other hand, when the XR / PCC object is output on a display provided in the self-driving vehicle, at least a portion of the XR / PCC object may be output to overlap with the object on the screen. For example, the self-driving vehicle 1220 may output XR / PCC objects corresponding to objects such as a road, another vehicle, a traffic light, a traffic sign, a two-wheeled vehicle, a pedestrian, and a building.

[0241] Virtual reality (VR) technology, augmented reality (AR) technology, mixed reality (MR) technology, and / or point cloud compression (PCC) technology according to an embodiment are applicable to various devices.

[0242] In other words, VR technology is a display technology that only provides CG images of real-world objects, backgrounds, etc. On the other hand, AR technology refers to a technology that displays a virtually created CG image on an image of a real object. MR technology is similar to the above-mentioned AR technology in that the virtual objects to be displayed are mixed and combined with the real world. However, MR technology is different from AR technology in that AR technology clearly distinguishes between real objects and virtual objects created as CG images and uses virtual objects as supplementary objects to real objects, while MR technology treats virtual objects as objects with properties equivalent to real objects. More specifically, an example of the application of MR technology is a holographic service.

[0243] Recently, VR, AR, and MR technologies are often referred to as extended reality (XR) technologies, rather than being clearly distinguished from each other. Therefore, embodiments of the present disclosure are applicable to any of VR, AR, MR, and XR technologies. Encoding / decoding based on PCC, V-PCC, and G-PCC technologies are applicable to such technologies.

[0244] The PCC method / apparatus according to the embodiment may be applied to a vehicle that provides a self-driving service.

[0245] Vehicles providing self-driving services are connected to the PCC device for wired / wireless communication.

[0246] When a point cloud data (PCC) transmitting / receiving device according to an embodiment is connected to a vehicle for wired / wireless communication, the device may receive / process content data related to an AR / VR / PCC service (which may be provided together with a self-driving service) and send it to the vehicle. In the case where the PCC transmitting / receiving device is installed on the vehicle, the PCC transmitting / receiving device may receive / process content data related to the AR / VR / PCC service and provide it to the user based on a user input signal input through a user interface device. The vehicle or user interface device according to the embodiment may receive a user input signal. The user input signal according to the embodiment may include a signal indicating a self-driving service.

[0247] As described above, the point cloud content providing system can use one or more cameras (for example, an infrared camera capable of protecting depth information, an RGB camera capable of extracting color information corresponding to the depth information, etc.), a projector (for example, an infrared pattern projector configured to protect depth information, etc.), LiDAR, etc. to generate point cloud content (or point cloud data).

[0248] LiDAR refers to a device configured to measure the distance by measuring the time it takes for the irradiated light to reflect from an object and return. It provides accurate three-dimensional information about the real world as point cloud data over a wide area and long distance. Such a large amount of point cloud data can be widely used in various fields that use computer vision technology, such as autonomous vehicles, robots, and 3D map production. That is, the LiDAR device uses a radar system that is configured to measure the coordinates of the position of the reflector by emitting a laser pulse and measuring the time it takes for the laser pulse to reflect on an object (i.e., a reflector) to generate point cloud content. According to an embodiment, depth information can be extracted by the LiDAR device. The point cloud content generated by the LiDAR device can be composed of multiple frames, and the multiple frames can be integrated into one content.

[0249] LiDAR can be generated by i＝1,...,N The lasers are composed of N lasers (or laser sensors) at positions (where N = 16, 32, 64, etc.), and the lasers can be at an azimuth angle relative to the Y axis. Capture point cloud data while rotating. This type of LiDAR model is called a rotational LiDAR model, and the point cloud content captured and generated by the rotational LiDAR model has angle characteristics.

[0250] The acquired (or captured) point cloud (or point cloud data) may consist of a collection of points, each of which may have geometric information and attribute information. As described above, the geometric information is 3D position (XYZ) information, and the attribute information is color (RGB, YUV, etc.) and / or reflectivity value.

[0251] According to an embodiment, for data captured by a rotating LiDAR device, the compression efficiency can be improved by applying an angle mode in the geometric encoding / decoding process. The angle mode is based on (r, i) instead of the compression method of (x, y, z), where r represents the radius, represents an azimuth or an azimuth angle, and i represents the i-th laser of the LiDAR (e.g., a laser index). In other words, since the frames of the point cloud content generated using the LiDAR device are not stitched together but are individual frames, each frame can have an origin of 0, 0, 0, and therefore the angle mode can be used by changing the coordinate system to a spherical coordinate system.

[0252] During G-PCC encoding according to an embodiment, point cloud data may be split into tiles according to regions, and each tile may be split into slices for parallel processing. G-PCC encoding may include compressing geometry on a slice-by-slice basis and compressing attribute information based on reconstructed geometry (decoded geometry) obtained based on position information changed by compression. Geometric information may be compressed using an octree-based, prediction tree-based, or triplet-based compression method.

[0253] G-PCC decoding according to an embodiment may include decoding slice-by-slice geometry encoded and transmitted by a transmitting device, and decoding attribute information based on the geometry reconstructed through the decoding operation.

[0254] Figure 11 is a diagram illustrating an example of point cloud data acquired using a LiDAR device according to an embodiment.

[0255] like Figure 11 As shown, the point cloud content captured by the LiDAR device on the moving vehicle may include both roads and objects. In other words, a street may include not only roads, but also many objects such as trees, buildings, cars, and people. In the present disclosure, the point cloud content may be referred to as point cloud data or point cloud. There may be one or more objects. Multiple objects may be referred to as objects or object groups or object blocks.

[0256] When configuring a point cloud by capturing consecutive frames, the motion characteristics presented in the consecutive frames of points constituting a road may be different from the motion characteristics presented in the consecutive frames of points constituting an object (or multiple objects). Specifically, in the case of a road captured by a LiDAR capture device (or LiDAR device), the relative height from the LiDAR capture device may be constant. In this case, the points may be generated in a circular shape around a center point that is the location of the sensor, such as Figure 12 shown.

[0257] Figure 12 is a diagram illustrating an example of points constituting a road captured by a LiDAR device according to an embodiment.

[0258] According to an embodiment, when there is no object other than the road in the point cloud frame captured by the LiDAR device on the moving vehicle, or when the relative height between the road and the center position of the sensor is constant, the point cloud frame is always captured as follows: Figure 12 The shapes shown are captured in frames. For this reason, it is difficult to identify the motion represented by road points across consecutive frames. Furthermore, applying the motion generated by the object to the road may be meaningless and may have an adverse effect, increasing the bitstream size and thus reducing compression efficiency. In contrast, applying the motion predicted from the road to the object may result in inaccurate motion prediction, as the predicted motion is unlikely to be accurate.

[0259] Therefore, a method may be needed to separate (or distinguish or classify) roads and objects from point cloud content continuously captured by a LiDAR device on a moving vehicle.

[0260] The present disclosure provides embodiments related to a method for separating roads and objects from point cloud content to efficiently support compression of point cloud content captured by a LiDAR device on a moving vehicle. The object may include one or more objects.

[0261] Specifically, the present disclosure describes embodiments of a method for separating roads and objects from point cloud content to efficiently support intra-prediction / inter-prediction of point cloud content captured by a LiDAR device on a moving vehicle.

[0262] The present disclosure provides an embodiment of a method for separating roads and objects from point cloud content based on the characteristics of a LiDAR device. For a LiDAR device such as a rotating LiDAR, the number of laser sensors in the LiDAR device, the angle and height of each laser sensor, etc. are provided as input information. Roads and objects are separated from point cloud content based on the above input information.

[0263] The present disclosure provides embodiments of a method of applying geometric information (also referred to as geometry or geometric data) related to separated roads and objects to a prediction tree.

[0264] According to an embodiment, the geometric information may be compressed using a method such as octree-based geometric compression, prediction tree-based geometric compression, or triplet-based geometric compression.

[0265] In other words, the most efficient way to compress point cloud content captured by a rotating LiDAR device on a moving vehicle is to apply geometric information to a prediction tree.

[0266] According to an embodiment, prediction tree-based geometric compression (or encoding) includes generating a prediction tree, performing tree-based prediction, selecting an optimal mode, and performing encoding.

[0267] Specifically, for point cloud content captured by rotating LiDAR devices, angle mode can be enabled, which can provide better compression.

[0268] Prediction-based compression of geometric information according to an embodiment is performed by defining a prediction structure for point cloud data. The structure is represented by a prediction tree having vertices associated with each point in the point cloud data. The prediction tree may include a root vertex (referred to as a root point) and leaf vertices (referred to as leaf points), where a point below a root point may have at least one child, and the depth increases toward the leaf point.

[0269] According to an embodiment, a prediction tree may be generated as follows.

[0270] For each node, the nearest node with the same laser index can be set as the parent node. If the parent node with the same laser index cannot be set in the prediction tree, the node with the smallest prediction residual among the nodes with the previous radar index can be set as the parent node.

[0271] Figure 13 (a) is a diagram illustrating an exemplary prediction tree when there is no object in the point cloud data acquired using four lasers according to an embodiment. That is, when there is no object but there is a road in the point cloud frame captured by the LiDAR device on the moving vehicle, a prediction tree may be generated. Figure 13 The prediction tree is shown in (a). This is because all lasers can reach the road.

[0272] As an example, in a prediction tree for a frame acquired using a laser with laser index 3, node 50011 (or point) is a child of node 50012. That is, node 50012 is a parent of node 50011. Node 50013 is a grandparent of node 50011 and is a parent of node 50012. When a parent node with the same laser index (i.e., laser index 3) cannot be set in the prediction tree, as in the case of node 50014, a node with a minimum prediction residual among nodes with a different laser index (e.g., a previous laser index) can be set as a parent node. Figure 13 In (a), node 50015 is one of the nodes in the prediction tree obtained using a laser with a laser index of 2, and is the node with the smallest prediction residual relative to node 50014.

[0273] Figure 13(b) is a diagram illustrating an exemplary prediction tree when an object exists in point cloud data acquired using four lasers according to an embodiment.

[0274] In other words, there are many objects on the street, such as roads, trees, buildings, cars, and people. Therefore, when the laser encounters an object before reaching the road, the laser hits the object, so it can generate Figure 13 A prediction tree of the form shown in (b).

[0275] That is to say, Figure 13 (a) and Figure 13 (b) shows an example of a prediction tree structure that can be generated when angle mode is enabled. Figure 13 (a) and Figure 13 In (b), each node has the latest node associated with the same laser index i as a parent. If the parent node is not available, the current node cannot generate a prediction. If the parent node is not available, the current node will use the node with the smallest prediction residual in the previous laser index j as a parent.

[0276] Figure 14 is a diagram illustrating an example of a point for each laser sensor according to an embodiment. That is, Figure 14 The shapes of the spots with the same laser are shown.

[0277] According to an embodiment, the geometric prediction tree encoding may include the following prediction mode specific prediction calculation method based on the generated prediction tree: For the current vertex in the prediction tree, P0 represents the position of the parent, p1 represents the position of the grandparent, and p2 represents the position of the great-grandparent.

[0278] In other words, each point (or node) may be predicted by applying one of various prediction modes based on the point positions of its parent, grandparent, great-grandparent, etc.

[0279] For example, the prediction mode corresponding to 0 can be defined as no prediction, the prediction mode corresponding to 1 can be defined as incremental prediction (p0), the prediction mode corresponding to 2 can be defined as linear prediction (2p0-p1), and the prediction mode corresponding to 3 can be defined as parallelogram prediction (2p+p1-p2).

[0280] The prediction method can be configured such that if the relative distance difference between the parent / grandfather (parent's parent) / great-grandfather (parent's parent's parent) is regular based on the generated prediction tree, the prediction is well performed. That is, the residual error between the predicted position and the current position can be reduced.

[0281] However, if Figure 13As shown in (b), the laser sensor can generate a zigzag prediction tree by moving back and forth between the road and the object. In this case, the relative distance difference between the parent / grandfather / great-grandfather may not be regular, which may increase the size of the compressed geometry bitstream to be sent.

[0282] Therefore, the present disclosure proposes a method for separating roads and objects to efficiently support compression of point cloud content captured by a LiDAR device on a mobile vehicle, and a method for efficiently performing geometric prediction tree encoding based on the separated roads and objects. In other words, in order to reduce the jagged shape of the prediction tree structure, the present disclosure proposes a method for separating roads and objects for geometric prediction tree encoding.

[0283] According to an embodiment, the separation of roads and objects in the point cloud content (i.e., point cloud data) may be performed by an encoder on the transmitting side or a decoder on the receiving side. For example, in the case where the separation of roads and objects in the point cloud data is performed by an encoder on the transmitting side, the separation may be skipped on the receiving side. On the other hand, in the case where the separation of roads and objects in the point cloud data is performed by a decoder on the receiving side, the separation may be skipped on the transmitting side. According to an embodiment, whether the separation of point cloud data from roads and objects is to be performed by an encoder on the transmitting side or a decoder on the receiving side may be indicated using signaling information, or may be determined based on signaling information.

[0284] According to an embodiment, the partitioning of roads and objects may be performed by an encoder on the transmission side, and in a decoding process, a reconstruction operation may be performed differently for each of the roads and objects by a decoder on the reception side.

[0285] According to an embodiment, it is assumed that the LiDAR device has multiple laser sensors, and the laser ID of the laser sensor closest to the road has a maximum value, and the laser ID and the radius of the captured road may decrease as the distance (ie, height) from the road increases. In other words, for content captured by the LiDAR device on a moving vehicle, as the value of the laser ID increases (ie, the laser sensor position decreases), the radius of the captured road may increase (or decrease).

[0286] According to an embodiment, the method of separating roads and objects may include estimating (or predicting) the height of the LiDAR sensor from the road (i.e., the distance from the road to the center of the sensor) and performing separation based on the estimated (or predicted) height. According to an embodiment, the present disclosure may classify points captured at the current laser ID as roads when they exist at a position at or below the estimated height value, and classify them as objects when they exist at a higher position. Here, an object may include one or more objects and may be referred to as an object group.

[0287] According to an embodiment, when roads and objects are separated from point cloud data, a prediction unit having points separated as roads may be constructed, and a prediction unit having points separated as objects (or object groups) may be constructed. Then, inter-frame prediction or intra-frame prediction may be performed for each prediction unit. Here, the prediction unit may be a maximum prediction unit (LPU) and / or a prediction unit (PU). For simplicity, the LPU may be referred to as a first prediction unit and the PU may be referred to as a second prediction unit.

[0288] According to an embodiment, when roads and objects are separated from point cloud data, a first prediction unit having points separated into roads may be constructed, and a first prediction unit having points separated into objects (or object groups) may be constructed. Then, inter-frame prediction or intra-frame prediction may be performed for each of the first prediction units.

[0289] According to an embodiment, a first prediction unit constructed as a point having an object may be further split into a second prediction unit. In other words, when the objects have different characteristics, a first prediction unit consisting of a plurality of objects may be split into a second prediction unit.

[0290] According to an embodiment, global motion may be applied to the first prediction unit for inter-frame prediction, and local motion may be applied to the second prediction unit for inter-frame prediction.

[0291] According to an embodiment, the global motion may be applied to the first prediction unit consisting of objects, and may not be applied to the first prediction unit consisting of roads.

[0292] Regarding intra prediction / inter prediction according to an embodiment, definitions of the following terms are described in the present disclosure.

[0293] 1) I (Intra) frame; P (Predicted) frame; B (Bidirectional) frame.

[0294] Frames to be encoded / decoded may be divided into I frames, P frames, and B frames. A frame may be referred to as a picture or the like.

[0295] For example, the frames may be transmitted in the order of I frame→P frame→(B frame)→(I frame|P frame)→... The B frame may be omitted.

[0296] 2) Reference Frame

[0297] The reference frame may be a frame referred to in encoding / decoding the current frame.

[0298] The immediately preceding I frame or P frame referred to for encoding / decoding the current P frame may be referred to as a reference frame. The immediately preceding I frame or P frame and the immediately following I frame or P frame referred to for encoding / decoding the current B frame may be referred to as reference frames.

[0299] 3) Frame and intra-frame prediction coding / inter-frame prediction coding

[0300] Intra-frame prediction encoding may be performed for an I frame, and inter-frame prediction encoding may be performed for a P frame and a B frame.

[0301] When a change rate of a P frame relative to a previous reference frame is greater than a certain threshold, intra-frame prediction encoding may be performed on the P frame as in the case of an I frame.

[0302] 4) Criteria for determining I (Intra) frames

[0303] Among the plurality of frames, every k-th frame may be designated as an I frame. Alternatively, a score related to the correlation between frames may be set, and a frame with a higher score may be configured as an I frame.

[0304] 5) I frame encoding / decoding

[0305] When encoding / decoding point cloud data having multiple frames, the geometry of the I frame may be encoded / decoded based on an octree or a prediction tree. Next, attribute information about the I frame may be encoded / decoded based on a prediction / lifting transform scheme or a RAHT scheme based on the reconstructed geometry information. In the present disclosure, the geometry of the I frame may be encoded / decoded based on a prediction tree.

[0306] 6) P frame encoding / decoding

[0307] When encoding / decoding point cloud data having a plurality of frames according to an embodiment, a P frame may be encoded / decoded based on a reference frame.

[0308] In this case, the coding unit for inter-frame prediction of a P frame may be a frame, a tile, a slice, an LPU (also referred to as a first prediction unit), or a PU (also referred to as a second prediction unit).

[0309] In this article, roads and objects can be separated from point cloud data (or point cloud content) captured (or acquired) by a LiDAR device, an LPU with separated points of roads can be constructed, and a PU with separated points of objects can be constructed. Point cloud data (or point cloud content) can be configured by frame, tile, or slice. In other words, point cloud content, frames, tiles, slices, etc. can be referred to as point cloud data. For example, roads and objects can be separated from slice-level point cloud data, an LPU with points of roads can be constructed, and an LPU with points of objects can be constructed.

[0310] Hereinafter, embodiments related to separating roads and objects from point cloud data at a frame level, a tile level, or a slice level will be described in detail.

[0311] Removing Noise Based on the Minimum Radius

[0312] According to an embodiment, noise may be removed before separating roads and objects from the point cloud data.

[0313] For example, in the case of capturing point cloud content via a LiDAR device on a moving vehicle, a portion of the vehicle may be captured by the sensor depending on the mounting location.

[0314] Figure 15 is a diagram illustrating an example of point cloud data captured by a LiDAR device according to an embodiment.

[0315] like Figure 15 As shown, the center of the LiDAR device is captured as the origin (0, 0, 0), and within a certain radius there are points 50020. These points can be objects, or primarily the part of the vehicle that carries the LiDAR device.

[0316] In one embodiment, points 50020 within a certain radius may be considered noise and removed.

[0317] In another embodiment, points 50020 within a certain radius may be considered as roads. When separating roads and objects, points 50020 may be separated as roads. This is because, similar to roads, motion may always appear uniform across consecutive frames.

[0318] In order to identify points 50020 located within a specific radius, a minimum radius may be input and set, or a minimum radius may be calculated and set. Then, points within the set radius may be considered as noise or road. For example, when considered as noise, the points are removed before separating roads and objects from the point cloud data.

[0319] According to an embodiment, the method of separating roads and objects can be based on predicting the position of a laser sensor using a specific laser ID (laserID). Here, laserID is identification information for identifying each laser (or laser sensor) and can represent a laser index. The laser sensor can be called a capture sensor.

[0320] In other words, for point cloud content captured by a LiDAR device on a moving vehicle, when the height of the capturing sensor (i.e., laser sensor) from the road and the laserID of the selected sensor are given, the road and the object can be separated from the point cloud data based on the given information. When the information is not given, the following method can be used to predict the height of the sensor from the road, and the road and the object can be separated based on the predicted information.

[0321] According to an embodiment, the LiDAR device may include multiple sensors (referred to as laser sensors or capture sensors), and may provide angle (θ) information and height (z) information about each sensor. Here, the angle θ may be an angle at which the corresponding laser sensor faces downward (e.g., toward the ground).

[0322] According to an embodiment, for point cloud content captured by a LiDAR device on a moving vehicle, as the laserID increases (i.e., the laser sensor position decreases, such as Figure 16 As shown), the radius of the road captured by the corresponding laser sensor can be increased.

[0323] In this case, the sensor with the smallest angle θ (i.e., the sensor with the most negative angle) can be selected, such as Figure 16 As shown (the sensor with the smallest θ is selected). This is because the nearest road will be captured at the smallest angle θ and is likely to be free of objects.

[0324] Then, the radius of the point captured using the selected sensor can be found (finding the radius of the point captured using the sensor). In this operation, at least one of the maximum radius, the minimum radius, or the average radius can be used. In some embodiments, the radius can be calculated or obtained by removing noise based on the minimum radius.

[0325] Once the radius is found, the height from the road to the center of the sensor can be estimated based on the radius, the angle of the selected sensor, and the height from the center of the sensor to the selected sensor.

[0326] Figure 16 : is a diagram illustrating an example of a laser sensor and an angle θ according to an embodiment.

[0327] exist Figure 16 , laser sensor 50031 with laser index equal to 0 is an example of a sensor with a maximum negative angle.

[0328] Figure 17 is a diagram illustrating an exemplary position estimation method according to an embodiment.

[0329] exist Figure 17 In , θ represents the angle of the sensor in the LiDAR device, and z represents the height of the sensor in the LiDAR device. For example, θ[i] represents the angle of the i-th laser sensor in the LiDAR device, and z[i] represents the height from the center of the sensor in the LiDAR device to the i-th laser sensor.

[0330] For example, refer to Figure 16 and Figure 17, z represents the height from the center of the sensor (0, 0, 0) 50033 to the laser sensor 50031 in the LiDAR device. In other words, Figure 17 , 50031 represents the laser sensor having the largest negative angle among the laser sensors, and 50033 represents the center of the laser sensor. In addition, θ represents the angle of the laser sensor 50031 at an angle of θ degrees, and z represents the height of the laser sensor 50031 as an angle z.

[0331] exist Figure 17 In the example, the height can be called the basic height.

[0332] In some embodiments, the relative distance of the center position of the LiDAR device from the road can be input as the basic height. Alternatively, the relative distance (height) can be calculated based on the sensor with the smallest angle θ (ie, the sensor with the largest negative angle) 50031.

[0333] In the present disclosure, the basic height (estimated_height) may be predicted using Equation 1 given below.

[0334] [Formula 1]

[0335] estimated_height=-(radius*tan(sensor.θ)-sensor.z)+α

[0336] In the above formula, α may be a constant used for error correction.

[0337] In the present disclosure, estimated_height may be referred to as predicted height, estimated height, or basic height. In other words, in the present disclosure, the above formula 1 may be used to estimate the height of the LiDAR sensor from the road to separate the road from the object. In the present disclosure, estimated_height refers to the basic height, which is the distance from the road to the center (0, 0, 0) of the laser sensor.

[0338] According to an embodiment, when the position of a point is at or below the estimated_height, the point can be considered (or separated) as a road. Moreover, when the position of a point is above the estimated_height, the point can be considered (or separated) as an object. That is, if the z position of the point is below the estimated height, the point can be classified as a road. Otherwise, it can be classified as an object (or object area).

[0339] Next, a method of generating a prediction tree based on separated roads and objects will be described. The prediction tree can be used for intra prediction or inter prediction of geometric information (ie, points).

[0340] According to an embodiment, the point cloud encoder of the transmitting device may generate a prediction tree for prediction tree encoding of geometric information. The prediction tree may be configured as a single tree based on the maximum number of points. However, in the case where roads and objects are connected as parent / child, they may be connected in a zigzag manner, which may result in a loss of consistency.

[0341] Therefore, in the present disclosure, a prediction tree can be constructed by dividing a single tree into a road portion and an object portion. In other words, a first prediction tree can be constructed using points in a single prediction tree that are individually classified as roads, and a second prediction tree can be constructed using points that are individually classified as objects.

[0342] Figure 18 (a) is a diagram illustrating an exemplary prediction tree using point cloud data that is not separated into roads and objects.

[0343] Figure 18 (b) is a diagram illustrating an exemplary case where points in point cloud data are separated into roads and objects and then prediction trees are constructed using points belonging to roads and points belonging to objects, respectively.

[0344] Figure 19 (a) is a diagram illustrating an exemplary prediction tree using point cloud data that has not been separated into roads and objects. In other words, Figure 18 (a) and Figure 19 (a) illustrates the same tree.

[0345] Figure 19 (b) is a diagram illustrating an exemplary case in which points in point cloud data are separated into roads and objects and prediction trees are constructed using points belonging to the roads and points belonging to the objects, respectively, and then the prediction trees are connected to each other.

[0346] That is, the prediction tree 50041 for the road portion and the prediction tree 50042 for the object portion may not be connected, such as Figure 18 As shown in (b), or can be connected, such as Figure 19 as shown in (b).

[0347] In an embodiment, the prediction tree 50041 for the road portion and the prediction tree 50042 for the object portion may be connected by finding and connecting a point 50043 in the prediction tree 50041 for the road portion that is closest to the first point in the prediction tree 50042 for the object portion.

[0348] Next is a method of searching for an inter-prediction reference point based on a separated road and an object when performing inter-prediction, that is, a method of searching for a reference point for inter-prediction.

[0349] In the geometric prediction tree encoding method, a point with a similar azimuth angle can be selected from points in the reference frame that have the same laser ID as the current point as a reference point for inter-frame prediction and used as a prediction value.

[0350] According to an embodiment, when points in point cloud data are separated into roads and objects, reference frames may also be separated into roads and objects when searching for reference points for inter-frame prediction.

[0351] For example, when the current point is a road, a point with a similar azimuth value can be selected from points with the same laser ID in the road reference point area in the reference frame as a reference point and used as a prediction value. In other words, when the point in the current frame is located in the road area, the reference point can be searched in the road area in the reference frame.

[0352] For example, when the current point belongs to an object (or an object area), a point with a similar azimuth value can be selected from points with the same laser ID among points in the object reference point area in the reference frame as a reference point and used as a prediction value. In other words, when a point in the current frame is classified as an object, a reference point can be searched among points in the object area in the reference frame.

[0353] Figure 20 is a diagram illustrating another exemplary point cloud transmitting device according to an embodiment.

[0354] The point cloud transmitting apparatus according to an embodiment may include a data input unit 51001, a coordinate transformer 51002, a quantization processor 51003, a space partitioner 51004, a signaling processor 51005, a geometry encoder 51006, an attribute encoder 51007, and a transmitting processor 51008. According to an embodiment, the coordinate transformer 51002, the quantization processor 51003, the space partitioner 51004, the geometry encoder 51006, and the attribute encoder 51007 may be referred to as a point cloud video encoder.

[0355] Figure 20 The point cloud data sending device may correspond to Figure 1 The sending device 10000, Figure 1 Point cloud video encoder 10002, Figure 1 Transmitter 10003, Figure 2 Acquisition 20000 / Encoding 20001 / Transmission 20002, Figure 3 Point cloud video encoder, Figure 8 The sending device Figure 10 devices, etc. Figure 20 Each of the components in the corresponding figures may correspond to software, hardware, a processor connected to a memory, and / or a combination thereof.

[0356] The data input unit 51001 can execute Figure 1 Some or all operations of the point cloud video acquisition unit 10001, or can be performed Figure 8 The coordinate converter 51002 may perform some or all operations of the data input unit 8000. Figure 3 In addition, the quantization processor 51003 may perform some or all operations of the coordinate transformer 30000. Figure 3 Some or all of the operations of the quantizer 30001, or may perform Figure 8 Some or all operations of the quantization processor 8001. That is, the data input unit 51001 can receive data to encode the point cloud data. The data may include geometric data (which may be referred to as geometry, geometric information, etc.), attribute data (which may be referred to as attributes, attribute information, etc.), and parameter information indicating encoding related settings.

[0357] The coordinate transformer 51002 can support coordinate transformation of point cloud data, such as changing the xyz axis or transforming the Cartesian coordinate system (x, y, z) into a spherical coordinate system (r, i).

[0358] The quantization processor 51003 can quantize the point cloud data. For example, the quantization processor 51003 can adjust the compression rate of the point cloud data by multiplying the x, y, and z position values ​​of the point cloud data by a scaling factor set according to a scaling factor (scaling factor = geometric quantization value). The scaling factor can conform to a set value or can be included in the bitstream as parameter information and delivered to the receiving side.

[0359] The spatial partitioner 51004 can spatially partition the point cloud data quantized and output by the quantization processor 51003 into one or more 3D blocks based on the bounding box and / or sub-bounding box. For example, the spatial partitioner 51004 can partition the quantized point cloud data into tiles or slices for accessing content region by region or parallel processing. In one embodiment, the signaling information for the spatial partitioning is entropy encoded by the signaling processor 51005 and then sent in the form of a bit stream through the sending processor 51008.

[0360] In one embodiment, the point cloud content may be a person, multiple persons, an object, or multiple objects such as an actor. In a broad sense, the point cloud content may be a map for self-driving or a map for indoor navigation of a robot. In this case, the point cloud content may be a large amount of locally connected data. In this case, the point cloud content cannot be encoded / decoded at once, so tile partitioning may be performed before compressing the point cloud content. For example, room #101 in a building may be partitioned into one tile, and room #102 in the building may be partitioned into another tile. In order to apply parallelization to the partitioned tiles to support fast encoding / decoding, the tiles may be partitioned (or split) into slices again. This operation may be referred to as slice partitioning (or splitting).

[0361] That is, according to an embodiment, a tile may represent a partial area of ​​a 3D space occupied by point cloud data (e.g., a rectangular cube). According to an embodiment, a tile may include one or more slices. A tile according to an embodiment may be partitioned into one or more slices, so that a point cloud video encoder may encode point cloud data in parallel.

[0362] A slice may represent a data unit (or bitstream) that can be independently encoded by a point cloud video encoder according to an embodiment and / or a data unit (or bitstream) that can be independently decoded by a point cloud video decoder. A slice may be a collection of data in a 3D space occupied by point cloud data, or a collection of some data among point cloud data. A slice according to an embodiment may represent an area or a collection of points included in a tile according to an embodiment. According to an embodiment, a tile may be partitioned into one or more slices based on the number of points included in a tile. For example, a tile may be a collection of points partitioned according to the number of points. According to an embodiment, a tile may be partitioned into one or more slices based on the number of points, and some data may be split or merged during the partitioning process. That is, a slice may be a unit that can be independently encoded in a tile. The tiles obtained by spatial partitioning in this way may be partitioned into one or more slices for fast and efficient processing.

[0363] The point cloud video encoder according to the embodiment may encode the point cloud data slice by slice, tile by tile, or frame by frame, wherein a tile includes one or more slices and a frame includes one or more tiles. In addition, the point cloud video encoder according to the embodiment may perform different quantization and / or transformations tile by tile or slice by slice.

[0364] The positions of one or more 3D blocks (e.g., slices) spatially partitioned by the spatial partitioner 51004 are output to the geometry encoder 51006, and attribute information (or attributes) are output to the attribute encoder 51007. The positions may be position information about a point contained in a partitioned unit (a box, a block, a tile, a tile group, or a slice), and may be referred to as geometry information.

[0365] The geometry encoder 51006 performs inter-frame prediction or intra-frame prediction-based encoding for the position output from the spatial partitioner 51004 to output a geometry bitstream. In this case, for prediction-based encoding of P frames, the geometry encoder 51006 can separate roads and objects from point cloud data at the frame level, tile level, or slice level by applying the above-mentioned road and object separation method to point cloud data at the frame level, tile level, or slice level. Then, after constructing a prediction tree for each of the road portion and the object portion, the encoder can compress the geometric information using intra-frame prediction or inter-frame prediction. When using inter-frame prediction, the geometry encoder 51006 can further perform the following operations before constructing the prediction tree: construct LPUs of points with separated roads and points of objects, respectively. In some embodiments, predicted motion can be applied to LPUs including points classified as roads. Predicted motion (e.g., global motion) can be applied only to LPUs including points classified as objects. In other embodiments, it may be determined by rate-distortion optimization (RDO) whether to apply predicted motion to an LPU including a point of a road and / or an LPU including a point of an object, and the motion may be applied to a reference frame based on the determination. In other words, it may be predicted by RDO whether it is beneficial to apply a motion vector within an LPU including a point of a road and / or an LPU including a point of an object, and the predicted result may be signaled. In one embodiment, the motion vector may be a global motion vector. The motion vector may be a local motion vector. In addition, the motion vector may be both a global motion vector and a local motion vector. In addition, it may be determined by comparing the bitstream size given when applying the motion vector, etc., whether the application of the motion vector is beneficial. In addition, the global motion vector may be obtained by estimating the overall motion across the frame. According to an embodiment, based on the LPU including a point of an object, a PU may be configured for the corresponding object. The local motion vector may then be applied to the PU. The geometry encoder 51006 may reconstruct the encoded geometry information and output the reconstructed information to the attribute encoder 51007.

[0366] The attribute encoder 51007 encodes (ie, compresses) the attributes (eg, partitioned attribute source data) output from the spatial partitioner 51004 based on the reconstructed geometry output from the geometry encoder 51006, and outputs an attribute bitstream.

[0367] Figure 21 is a diagram illustrating exemplary operations of the geometry encoder 51006 and the attribute encoder 51007 according to an embodiment.

[0368] In one embodiment, a quantization processor may be further provided between the spatial partitioner 51004 and the voxelization processor 53001. The quantization processor quantizes the positions of one or more 3D blocks (e.g., slices) spatially partitioned by the spatial partitioner 51004. In this case, the quantization processor may perform Figure 3 Some or all of the operations of quantizer 30001, or performing Figure 8 When a quantization processor is further provided between the spatial partitioner 51004 and the voxelization processor 53001, some or all operations of the quantization processor 8001 may be omitted or may not be omitted. Figure 20 Quantization processor 51003.

[0369] The voxelization processor 53001 according to an embodiment performs voxelization based on the position of one or more spatially partitioned 3D blocks (e.g., slices) or their quantized positions. Voxelization refers to the smallest unit representing position information in 3D space. That is, the voxelization processor 53001 can support the process of rounding the geometric position value of the scaled point to an integer. According to an embodiment, the points of the point cloud content (or 3D point cloud video) may be included in one or more voxels. According to an embodiment, a voxel may include one or more points. In one embodiment, when quantization is performed before voxelization, multiple points may belong to one voxel.

[0370] In the present disclosure, when two or more points are included in one voxel, the two or more points are referred to as repeated points. That is, in the geometric encoding process, repeated points can be generated by geometric quantization and voxelization.

[0371] The voxelization processor 53001 may output duplicate points belonging to one voxel without merging the points, or may merge the duplicate points into one point to be output.

[0372] According to an embodiment, the points partitioned into tiles, slices or frames by the spatial partitioner 51004 are voxelized by the voxelization processor 53001, and the voxelized points are separated into roads or objects by the road / object splitter 53002.

[0373] For example, Figure 16As shown, assuming that the LiDAR device has multiple laser sensors, and the laser ID of the laser sensor closest to the road has the maximum value, and the laser ID and the radius of the captured road can decrease as the distance from the road (i.e., height) increases. In this case, for the content captured by the LiDAR device on the moving vehicle, the radius of the captured road can increase as the value of the laser ID increases (i.e., the laser sensor position decreases).

[0374] According to an embodiment, the road / object splitter 53002 may estimate (or predict) the height of the LiDAR sensor from the road based on Formula 1, and classify each of the voxelized points as a road or an object based on the estimated (or predicted) height (estimated_height). According to an embodiment, the road / object splitter 53002 may classify the point as a road when the position of the voxelized point captured by the current laserID is lower than or equal to the estimated_height, and classify the point as an object when the position is higher than the estimated_height. In order to find the estimated_height for road / object separation by applying Formula 1, information indicating whether the road / object is separated, a height error value, an angle θ of the sensor, a z value of the sensor, a radius of the sensor, etc. may be provided to the road / object splitter 53002. Then, the estimated_height for road / object separation found by applying Formula 1 may be sent to the receiving side. Here, the object may include one or more objects, which may be referred to as an object group. The z value represents the height from the center of the sensor in the LiDAR device to the corresponding laser sensor. In another embodiment of the present disclosure, the height of the LiDAR sensor from the road may be directly input.

[0375] According to an embodiment, in order to separate the road and the object as described above, the road / object segmentation part 53002 may receive the road / object splitting and prediction related option information as input. According to an embodiment, the road / object splitting and prediction related option information may include information indicating whether to split the road / object (road_object_split_flag), estimated height information (estimated_height) for road / object separation, or information for estimating basic height information (e.g., the angle θ of the sensor, the z value of the sensor, the radius of the sensor). According to an embodiment, the road / object splitting and prediction related option information may also include information for identifying each object (object_id). The road / object splitting and prediction related option information may be included in at least one of a geometry parameter set, a tile parameter set (also called a tile list), or a geometry slice header and sent to a point cloud decoder on the receiving side. In addition, the road / object splitting and prediction related option information may be included in at least one of a geometry LPU header or a geometry PU header and sent to a point cloud decoder on the receiving side.

[0376] For example, when the value of the information indicating whether to split roads / objects is true, the road / object splitter 53002 can split points in the point cloud data at the frame level, tile level, or slice level into roads and objects by applying the above-mentioned capture sensor position prediction method.

[0377] When the frame of the input point cloud data (i.e., the frame to which the input point belongs) is an I frame, the geometric information intra-frame predictor 53004 according to an embodiment may apply geometric intra-frame prediction encoding to the geometric information about the I frame. According to an embodiment, when the points have been classified as roads and objects by the road / object splitter 53002, the geometric information intra-frame predictor 53004 may divide a single prediction tree into a road portion and an object portion to construct prediction trees for these portions. For example, it may first construct a first prediction tree with points classified as roads, and then construct a second prediction tree with points classified as objects. In this case, the prediction tree for the road portion and the prediction tree for the object portion may not be connected, such as Figure 18 As shown in (b), or can be connected, such as Figure 19 According to an embodiment, in order to connect the prediction tree for the road portion and the prediction tree for the object portion, a point closest to the first point in the prediction tree for the object portion may be found in the prediction tree for the road portion and connected to the first point.

[0378] According to an embodiment, once the geometric information intra-frame predictor 53004 generates Figure 18 (b) or Figure 19The prediction tree shown in (b) can predict each point based on the position of the point's parent, grandparent, great-grandparent, etc. by applying one of four intra prediction modes (e.g., intra prediction mode 0 to intra prediction mode 3).

[0379] According to an embodiment, intra prediction mode 0 is a mode in which prediction is not performed and is called no prediction. In other words, when intra prediction mode 0 is selected, the position value (ie, coordinate value) of the current point is transmitted without compression.

[0380] Intra-frame prediction mode 1 is a mode in which the position value p0 (i.e., coordinate value) of the parent node (or parent point) of the current point is set as the prediction value. This mode is called incremental prediction. The residual (or residual position value) of intra-frame prediction mode 1 is obtained by subtracting the prediction value of intra-frame prediction mode 1 from the position value of the current point.

[0381] Intra-frame prediction mode 2 is a mode in which the position value p0 of the parent node (or parent point) of the current point is multiplied by 2 minus the position value p1 of the grandparent node (or grandparent point) of the current point (=2p0-p1) as the prediction value. This mode is called linear prediction. Then, the residual (or residual position value) of intra-frame prediction mode 2 is obtained by subtracting the prediction value of intra-frame prediction mode 2 from the position value of the current point.

[0382] Intra-frame prediction mode 3 is a mode in which the position value p0 of the parent node (or parent point) plus the position value p1 of the grandparent node (or grandparent point) minus the position value p2 of the great-grandparent node (or great-grandparent point) (= p0 + p1 - p2) is set as the prediction value. This mode is called parallelogram prediction. Then, the residual (or residual position value) of intra-frame prediction mode 3 is obtained by subtracting the prediction value of intra-frame prediction mode 3 from the position value of the current point.

[0383] In other words, each point may be predicted based on the point positions of its parent, grandparent, great-grandparent, etc. by applying one of various intra prediction modes (eg, intra prediction mode 0 to intra prediction mode 3).

[0384] To this end, the element indicated by 53003 (also referred to as a determination unit) checks whether the points output from the voxelization processor 53001 or the points separated into roads and objects by the road / object splitter 53002 belong to an I frame or a P frame.

[0385] When the frame identified by the determination unit 53003 is a P frame, and the points in the point cloud data at the frame level, tile level, or slice level have been separated into roads and objects by the road / object splitter 53002, the LPU splitter 53005 constructs an LPU with points belonging to roads and an LPU with points belonging to objects to support inter-frame prediction. Then, for each LPU, information indicating whether the LPU is a road or an object (or a group of objects) can be signaled in the road / object splitting and prediction related option information to be sent to the decoder on the receiving side. In other words, when the receiving side can identify whether the LPU sent from the sending side is composed of points of a road or points of an object, the road / object splitting operation can be skipped on the receiving side. In addition, since the points in the point cloud data are split into LPUs based on whether they are roads or objects, the bitstream to be sent can be split into roads and objects.

[0386] That is, once the points in the point cloud data are separated into roads and objects by the road / object splitter 53002, the LPU splitter 53005 constructs an LPU having points separately classified as roads, and constructs an LPU having points separately classified as objects (object groups). Inter-frame prediction can then be performed by applying a global motion vector to each LPU or by using a previous frame. Here, using a previous frame means not using a motion vector.

[0387] According to an embodiment, the motion predictor 53006 may predict global motion for each LPU, and the region (i.e., LPU) that needs to be split may be split into PUs by the PU splitter 53007 to predict local motion. This process may be performed until no further PU splitting is required or only until a specified level. The LPU / PU split state and predicted motion for each LPU / PU may be included in the road / object splitting and prediction related option information and sent to the decoder at the receiving side.

[0388] Then, the motion compensation application unit 53008 can determine through RDO whether to apply the predicted motion to the generated LPU / PU, and can apply the motion to the reference frame according to the determination. For example, when it is determined through RDO that the gain of inter-frame prediction is large, it can perform motion compensation by applying the global motion vector and / or local motion vector obtained through motion prediction to the previous frame (i.e., the reference frame). In this case, whether to apply motion to each LPU / PU can be included in the road / object splitting and prediction related option information and sent to the decoder on the receiving side. Once motion compensation is performed, the geometric information can be compressed and signaled by the geometric information inter-frame predictor 53009 based on the motion compensation point. That is, the geometric information inter-frame predictor 53009 performs inter-frame prediction based on the current LPU / PU and the previous frame that is motion compensated or the previous frame that is not motion compensated.

[0389] In other embodiments, the motion predictor 53006 may use global motion estimation to obtain a global motion vector for an LPU consisting of points of an object, and skip global motion estimation for an LPU consisting of points of a road. In addition, the PU splitter 53007 may split an LPU consisting of points of an object into as many PUs as the object, and may not split an LPU consisting of points of a road into PUs. For example, when the objects included in an LPU consisting of points of an object (or an object group) have characteristics different from each other, the LPU may be split into as many PUs as the object, and each PU may be assigned an object ID (e.g., object_id). In addition, for each PU, a local motion vector may be obtained by motion prediction. Then, the object ID information and PU information assigned to each PU may be included in the road / object splitting and prediction related option information and sent to the decoder on the receiving side. In this case, the motion compensation application unit 53008 applies motion compensation to the LPU and / or PU consisting of points of the object, and does not apply motion compensation to the LPU consisting of points of the road. The geometry information inter-frame predictor 53009 performs inter-frame prediction based on the current LPU / PU and a motion-compensated previous frame or a non-motion-compensated previous frame.

[0390] In other embodiments, when the frame identified by the determination unit 53003 is a P frame and the road / object splitter 53002 has not yet split the road and the object, the LPU splitter 53005 can split the points of the point cloud data at the frame level, tile level or slice level voxelized by the voxelization processor 53001 into LPUs based on one or a combination of two or more of the height, radius and / or azimuth. In addition, the LPU can be further split into PUs based on a combination of one or more of the height, radius and / or azimuth. The motion compensation application unit 53008 can then determine through RDO whether to apply the predicted motion to the split LPU / PU, and can apply the motion to the reference frame based on the determination. For example, when it is determined through RDO that the gain of inter-frame prediction is large, it can perform motion compensation by applying the global motion vector and / or the local motion vector obtained by motion prediction to the previous frame (ie, the reference frame). In this case, whether the motion is applied to each LPU / PU can be included in the road / object splitting and prediction related option information and sent to the decoder on the receiving side. Once motion compensation is performed, the geometry information may be compressed and signaled based on the motion compensated points by the geometry information interframe predictor 53009. That is, the geometry information interframe predictor 53009 performs interframe prediction based on the current LPU / PU and a motion compensated previous frame or a non-motion compensated previous frame.

[0391] According to an embodiment, in order for the geometry information inter-frame predictor 53009 to perform inter-frame prediction, a reference frame is stored in a reference frame buffer 53011 and provided to the geometry information inter-frame predictor 53009 when needed.

[0392] In this way, the LPU / PU splitter 53005 can construct (or generate) one LPU according to the points of the object, construct (or generate) another LPU according to the points of the road, and the PU splitter 53007 can split the LPU composed of the points of the object into PUs. Then, the motion predictor 53006 and the motion compensation application unit 53008 apply the predicted motion to the LPU and / or PU composed of the points of the object, and do not apply the motion to the LPU composed of the points of the road.

[0393] The geometry information inter-frame predictor 53009 according to an embodiment may perform prediction tree-based inter-frame encoding based on a difference in a geometry prediction value between a current frame and a motion-compensated reference frame or a non-motion-compensated previous frame.

[0394] According to an embodiment, when the frame of the input point cloud data (i.e., the frame to which the input point belongs) is a P frame, the geometric information inter-frame predictor 53009 may apply geometric inter-frame prediction encoding to the geometric information about the P frame. According to an embodiment, when the points have been classified as roads and objects by the road / object splitter 53002, the geometric information inter-frame predictor 53009 may divide a single prediction tree into a road portion and an object portion to construct prediction trees for these portions. For example, it may first construct a first prediction tree with points classified as roads, and then construct a second prediction tree with points classified as objects. In this case, the prediction tree for the road portion and the prediction tree for the object portion may not be connected, such as Figure 18 As shown in (b), or can be connected, such as Figure 19 According to an embodiment, in order to connect the prediction tree for the road portion and the prediction tree for the object portion, a point closest to the first point in the prediction tree for the object portion may be found in the prediction tree for the road portion and connected to the first point.

[0395] According to an embodiment, once the geometric information inter-frame predictor 53009 generates Figure 18 (b) or Figure 19 The prediction tree shown in (b) can perform inter-frame prediction based on the positions of reference points found in one or more reference frames.

[0396] In this case, the geometry information inter-frame predictor 53009 may select a point with a similar azimuth angle from among points in the reference frame having the same laser ID as the current point as a reference point for inter-frame prediction and use it as a prediction value.

[0397] According to an embodiment, when points in point cloud data are separated into roads and objects, reference frames may also be separated into roads and objects when searching for reference points for inter-frame prediction.

[0398] For example, when the current point is a road, a point having the same laserID and similar azimuth as the current point can be selected from the points in the area separated as the road in the reference frame as a reference point and used as a prediction value. In other words, when the point in the current frame is located in the road area, the reference point can be searched in the road area in the reference frame. According to an embodiment, the residual (or residual position value) is obtained by subtracting the prediction value found by searching in the road area in the reference frame from the position value of the current point. In this case, the prediction mode may be referred to as an inter-frame prediction mode in this article.

[0399] For example, when the current point belongs to an object (or an object area), a point having the same laser ID and a similar azimuth as the current point can be selected from the points in the object reference point area in the reference frame as a reference point and used as a prediction value. In other words, when the point in the current frame is classified as an object, the reference point can be searched among the points in the object area in the reference frame. According to an embodiment, the residual (or residual position value) is obtained by subtracting the prediction value found by searching in the object area in the reference frame from the position value of the current point. In this case, the prediction mode may be referred to as an inter-frame prediction mode in this article.

[0400] The geometry information inter-frame predictor 53009 can perform prediction by applying one of various inter-frame prediction modes in the prediction tree, such as Figure 18 (b) or Figure 19 The inter prediction mode may vary depending on the number of reference frames, the position of the point used for prediction, etc.

[0401] Here, the prediction value may be referred to as a prediction value of the current point position or a prediction position value. In addition, the reference frame may be a motion compensated frame or a non-motion compensated frame.

[0402] The geometry information entropy encoder 53010 according to the embodiment performs entropy encoding on the geometry information encoded based on the intra prediction of the geometry information intra predictor 53004 or the inter prediction of the geometry information inter predictor 53009, and outputs a geometry bit stream (also referred to as a geometry information bit stream). In this case, if the points in the point cloud data have been separated into roads and objects and processed, the geometry bit stream can be split into roads and objects to be output.

[0403] The geometry reconstructor according to the embodiment reconstructs (or restores) the geometry information based on the position changed by the intra-frame prediction-based encoding or the inter-frame prediction-based encoding, and outputs the reconstructed geometry information (referred to as reconstructed geometry) to the attribute encoder 51007. This is because the reconstructed (or restored) geometry information is required to compress the attribute information because the attribute information depends on the geometry information (position). In addition, the reconstructed geometry information is stored in the reference frame buffer 53011 so as to be provided as a reference frame for the inter-frame prediction encoding of the P frame. The reference frame buffer 53011 also stores the attribute information reconstructed by the attribute encoder 51007. That is, the reconstructed geometry information and the reconstructed attribute information stored in the reference frame buffer 53011 can be used as a previous reference frame for the geometry information inter-frame prediction encoding performed by the geometry information inter-frame predictor 53009 of the geometry encoder 51006 and the attribute information inter-frame prediction encoding performed by the attribute information inter-frame predictor 55005 of the attribute encoder 51007.

[0404] The color transformation processor 55001 of the attribute encoder 51007 corresponds to Figure 3 Color Changer 30006 or Figure 8 The color transform processor 8008 of the embodiment. The color transform processor 55001 according to an embodiment performs color transform encoding of the transformed color values ​​(or textures) included in the attributes provided from the data input unit 51001 and / or the spatial partitioner 51004. For example, the color transform processor 55001 can transform the format of the color information (e.g., from RGB to YCbCr). The operation of the color transform processor 55001 according to an embodiment can be optionally applied according to the color values ​​included in the attributes. In another embodiment, the color transform processor 55001 can perform color transform encoding based on the reconstructed geometry.

[0405] According to an embodiment, the attribute encoder 51007 may perform recoloring according to whether lossy encoding is applied to the geometry information. To this end, the element assigned with reference numeral 55002 (or referred to as a determination unit) checks whether the geometry encoder 51006 has applied lossy encoding to the geometry information.

[0406] For example, when the determination unit 55002 determines that lossy encoding has been applied to the geometric information, the recolorer 55003 performs color readjustment (or recoloring) due to the lost points to reconfigure the attribute (color). That is, the recolorer 55003 can find and set the attribute value suitable for the position of the lost point in the source point cloud data. In other words, when the position information value changes due to the scaling factor applied to the geometric information, the recolorer 55003 can predict the attribute value suitable for the changed position.

[0407] According to an embodiment, the operation of the reshader 53003 may be optionally applied, depending on whether the duplicate points are merged. According to an embodiment, the duplicate points may be merged by the voxelization processor 53001 of the geometry encoder 51006.

[0408] In an embodiment of the present disclosure, when the voxelization processor 53001 merges points belonging to a voxel into one point, the recolorizer 55003 may perform color readjustment (ie, recoloring).

[0409] Reshader 55003 performs the same Figure 3 Attribute transformer 30007 or Figure 8 The operations and / or methods of the attribute transformation processor 8009 are the same as or similar to the operations and / or methods.

[0410] When the determination unit 55002 determines that lossy encoding has not been applied to the geometric information, it is checked by the element assigned with reference numeral 55004 (or referred to as the determination unit) whether encoding based on inter-frame prediction has been applied to the geometric information.

[0411] When the determination unit 55004 determines that the inter-prediction-based encoding has not been applied to the geometric information, the attribute information intra-frame predictor 55006 performs intra-frame prediction encoding on the input attribute information. According to an embodiment, the intra-frame prediction encoding method performed by the attribute information intra-frame predictor 55006 may include prediction transform encoding, lifting transform encoding, and RAHT encoding.

[0412] When the determination unit 55004 determines that inter-frame prediction-based encoding has been applied to the geometric information, the attribute information inter-frame predictor 55005 performs inter-frame prediction encoding on the input attribute information. According to an embodiment, the attribute information inter-frame predictor 55005 may encode the residual based on the difference in attribute prediction values ​​between the current frame and the motion compensation reference frame.

[0413] According to the implementation scheme, the attribute information entropy encoder 55008 performs entropy encoding on the attribute information encoded by the attribute information intra-frame predictor 55006 based on intra-frame prediction, or performs entropy encoding on the attribute information encoded by the attribute information inter-frame predictor 55005 based on inter-frame prediction, and outputs an attribute bit stream (or called attribute information bit stream).

[0414] The attribute reconstructor according to the embodiment restores (or reconstructs) attribute information based on the attribute changed by intra-frame prediction encoding or inter-frame prediction encoding, and stores the reconstructed attribute information (or referred to as reconstructed attribute) in the reference frame buffer 53009. That is, the reconstructed geometry information and the reconstructed attribute information stored in the reference frame buffer 53009 can be used as a previous reference frame for inter-frame prediction encoding of geometry information by the geometry information inter-frame predictor 53007 and inter-frame prediction encoding of attribute information by the attribute information inter-frame predictor 55005 of the attribute encoder 51007.

[0415] The geometry bit stream compressed and output by the geometry encoder 51006 based on intra-frame prediction or inter-frame prediction and the attribute bit stream compressed and output by the attribute encoder 51007 based on intra-frame prediction or inter-frame prediction are output to the transmission processor 51008.

[0416] According to an embodiment, the sending processor 51008 may execute Figure 8 The operations and / or sending methods of the sending processor 8012 are the same as or similar to those of the sending processor 8012, or can perform the same as those of the sending processor 8012. Figure 1 The operation and / or transmission method of the transmitter 10003 is the same as or similar to the operation and / or transmission method of the transmitter 10003. For further details that will not be described below, refer to Figure 1 or Figure 8 Description.

[0417] The transmission processor 51008 can separately transmit the geometry bit stream output from the geometry encoder 51006, the attribute bit stream output from the attribute encoder 51007, and the signaling bit stream output from the signaling processor 51005, or can multiplex the bit streams into a single bit stream to be transmitted.

[0418] According to an embodiment, the sending processor 51008 may encapsulate the bitstream into files or segments (eg, streaming segments) and send the files or segments through various networks (such as a broadcast network and / or a broadband network).

[0419] The signaling processor 51005 may generate and / or process signaling information and output it in the form of a bit stream to the transmission processor 51008. The signaling information generated and / or processed by the signaling processor 51005 may be provided to the geometry encoder 51006, the attribute encoder 51007, and / or the transmission processor 51008 for geometry encoding, attribute encoding, and transmission processing. Alternatively, the signaling processor 51005 may be provided with signaling information generated by the geometry encoder 51006, the attribute encoder 51007, and / or the transmission processor 51008.

[0420] As used herein, signaling information may be signaled and sent in a parameter set such as a sequence parameter set (SPS), a geometry parameter set (GPS), an attribute parameter set (APS), or a tile parameter set (TPS) (also referred to as a tile list). Signaling information may be signaled and sent per coding unit (such as a slice or tile) of each video. According to the present disclosure, the signaling information may include metadata related to the point cloud data (e.g., set values, etc.) and may be provided to a geometry encoder 51006, an attribute encoder 51007, and / or a transmission processor 51008 for geometry encoding, attribute encoding, and transmission processing. Depending on the application, the signaling information may also be defined at a system level, such as a file format, HTTP dynamic adaptive streaming (DASH), or MPEG media transport (MMT), or at a wired interface level, such as a high-definition multimedia interface (HDMI), a display port, a video electronics standard association (VESA), or a CTA.

[0421] The method / apparatus according to the embodiment may signal relevant information to add / perform the operation of the embodiment. The signaling information may be used by the sending device and / or the receiving device.

[0422] In one embodiment, the road / object splitting and prediction related option information to be used for prediction of the road / object splitting and geometry information of the point may be signaled in at least one of the geometry parameter set, the tile parameter set, or the geometry slice header. Alternatively, the road / object splitting and prediction related option information may be signaled in a separate LPU header (referred to as geom_lpu_header) and / or PU header (referred to as geom_pu_header).

[0423] According to an embodiment, road / object separation and LPU / PU generation based thereon may be performed by a geometry encoder on the sending side or a geometry decoder on the receiving side. For example, when road / object separation and LPU / PU generation based thereon are performed by a geometry encoder on the sending side, they may be skipped on the receiving side. When skipped on the sending side, road / object separation and LPU / PU generation based thereon may be performed by a geometry decoder on the receiving side. According to an embodiment of the present disclosure, whether to split a road / object may be determined based on the value of information indicating whether to split a road / object. For example, when the value of the information indicating whether to split a road / object input to the geometry encoder on the sending side is true, the geometry encoder on the sending side may perform road / object separation and LPU / PU generation (or splitting) based thereon. When the value of the information indicating whether to split a road / object input to the geometry decoder on the receiving side is true, the geometry decoder on the receiving side may perform road / object separation and LPU / PU generation based thereon. In the former case, the value of the information indicating whether to split a road / object may be signaled as false in the road / object splitting and prediction-related option information and sent to the geometry decoder on the receiving side. In this case, the geometry decoder on the receiving side skips road / object separation and LPU / PU generation based thereon.

[0424] In other words, all signaling information may be sent to the decoder on the receiving side. In this case, road / object separation may be performed only by the geometry encoder on the transmitting side, and information indicating whether the road / object is separated and object ID information may be sent to the geometry decoder on the receiving side through the LPU / PU so that the geometry decoder may reconstruct the configuration of the bitstream. Therefore, only when the geometry encoder separates the road / object and the road / object separation is sent through the configuration of the bitstream, the value of the information indicating whether the road / object is separated may be signaled as false when sent to the geometry decoder.

[0425] Figure 22 is a diagram illustrating another exemplary point cloud receiving apparatus according to an embodiment.

[0426] The point cloud receiving device according to the embodiment may include a receiving processor 61001, a signaling processor 61002, a geometry decoder 61003, an attribute decoder 61004, and a post-processor 61005. According to the embodiment, the geometry decoder 61003 and the attribute decoder 61004 may be referred to as a point cloud video decoder. According to the embodiment, the point cloud video decoder may be referred to as a PCC decoder, a PCC decoding unit, a point cloud decoder, a point cloud decoding unit, etc.

[0427] Figure 22 The point cloud receiving device can correspond to Figure 1The receiving device 10004, the receiver 10005, the point cloud video decoder 10006, Figure 2 Transmission 20002 / decoding 20003 / rendering 20004, Figure 7 Point cloud video decoder, Figure 9 The receiving device Figure 10 devices, etc. Figure 22 Each of the components in the corresponding figures may correspond to software, hardware, a processor connected to a memory, and / or a combination thereof.

[0428] The receiving processor 61001 according to the embodiment may receive a single bit stream, or may receive a geometry bit stream (also referred to as a geometry information bit stream), an attribute bit stream (also referred to as an attribute information bit stream), and a signaling bit stream respectively. In this case, the bit stream (or geometry bit stream) may be split into roads and objects and received. When a file and / or a fragment is received, the receiving processor 61001 may decapsulate the received file and / or fragment and output a bit stream.

[0429] When receiving (or decapsulating) a single bitstream, the receiving processor 61001 can demultiplex the geometry bitstream, the attribute bitstream and / or the signaling bitstream from the single bitstream. The receiving processor 61001 can output the demultiplexed signaling bitstream to the signaling processor 61002, output the demultiplexed geometry bitstream to the geometry decoder 61003, and output the demultiplexed attribute bitstream to the attribute decoder 61004.

[0430] Upon receiving (or decapsulating) the geometry bitstream, attribute bitstream and / or signaling bitstream respectively, the receiving processor 61001 may deliver the signaling bitstream to the signaling processor 61002, the geometry bitstream to the geometry decoder 61003, and the attribute bitstream to the attribute decoder 61004.

[0431] The signaling processor 61002 may parse and process signaling information (e.g., information contained in SPS, GPS, APS, TPS, metadata, etc.) from the input signaling bitstream, and provide the processed information to the geometry decoder 61003, the attribute decoder 61004, and the post-processor 61005. In another embodiment, the signaling information contained in the geometry slice header and / or the attribute slice header may also be parsed by the signaling processor 61002 before decoding the corresponding slice data. That is, when the point cloud data is partitioned into tiles and / or slices at the transmitting side, the TPS includes the number of slices included in each tile, so the point cloud video decoder according to the embodiment may check the number of slices and quickly parse the information for parallel decoding.

[0432] Therefore, the point cloud video decoder according to the present disclosure can receive an SPS with a reduced amount of data, thereby quickly parsing the bitstream containing point cloud data. The receiving device can decode the tiles when they are received, and can decode each slice based on the GPS and APS included in each tile, thereby maximizing the decoding efficiency. Alternatively, the receiving device can perform intra-frame prediction decoding or inter-frame prediction decoding for the point cloud data of each PU based on the road / object splitting and prediction related option information signaled in the GPS, TPS, geometric slice header, LPU header and / or PU header, thereby maximizing the decoding efficiency. In this case, it is assumed that the points have been separated into roads and objects on the sending side.

[0433] That is, the geometry decoder 61003 can perform the same operation as described above for the compressed geometry bitstream based on the signaling information (e.g., geometry-related parameters). Figure 20 The geometry is reconstructed by reversing the operation of the geometry encoder 51006. The geometry restored (or reconstructed) by the geometry decoder 61003 is provided to the attribute decoder 61004. Here, the geometry-related parameters may include road / object splitting and prediction-related option information to be used for intra-prediction reconstruction or inter-prediction reconstruction of the geometry information.

[0434] In addition, the geometry decoder 61003 can perform the above-mentioned road / object splitting according to the road / object splitting and prediction related option information, and perform intra-frame prediction and / or inter-frame prediction of the geometry information based on the splitting. For example, when the value of the information indicating whether to split the road / object (e.g., road_object_split_flag) included in the road / object splitting and prediction related option information is false, the process of road / object splitting, LPU / PU splitting and prediction can be performed based on the road / object splitting and prediction related option information. As described above, when performing road / object splitting and LPU / PU splitting, motion compensation can be applied on the sending side for inter-frame prediction reconstruction of geometry information. On the other hand, when the value of the information indicating whether to split the road / object included in the road / object splitting and prediction related option information is true, the process of road / object splitting and LPU / PU splitting (or generation) can be skipped.

[0435] The attribute decoder 61004 can perform the same decoding as described above on the compressed attribute bitstream based on the signaling information (e.g., attribute related parameters) and the reconstructed geometry. Figure 20 The attribute encoder 51007 performs the reverse process to reconstruct the attributes. According to an embodiment, when the point cloud data has been split into tiles and / or slices at the sending side, the geometry decoder 61003 and the attribute decoder 61004 can perform geometry decoding and attribute decoding on a per-tile and / or per-slice basis.

[0436] Figure 23 is a diagram illustrating exemplary operations of the geometry decoder 61003 and the attribute decoder 61004 according to an embodiment.

[0437] Included Figure 23 The geometric information entropy encoder 63001, the inverse quantization processor 63008 and the inverse coordinate transformer 63009 in the geometric decoder 61003 can perform Figure 7 Some or all of the operations of the arithmetic decoder 7000 and the inverse coordinate transformer 7004, or performing Figure 9 The position reconstructed by the geometry decoder 61003 is output to the post-processor 61005.

[0438] According to an embodiment, when road / object splitting and prediction-related option information for predictive reconstruction of geometric information is signaled via at least one of a geometry parameter set (GPS), a tile parameter set (TPS) (or referred to as a tile list), a geometry slice header, a geometry LPU header, or a geometry PU header, it may be acquired by the signaling processor 61002 and provided to the geometry decoder 61003, or may be directly acquired by the geometry decoder 61003.

[0439] According to an embodiment, the road / object splitting and prediction related option information may include information indicating whether to split the road / object (road_object_split_flag), estimated height information (estimated_height) for road / object separation, or information for estimating basic height information (e.g., the angle θ of the sensor, the z value of the sensor, the radius of the sensor). According to an embodiment, the road / object splitting and prediction related option information may also include information for identifying each object (object_id). The road / object splitting and prediction related option information may be included in at least one of a geometry parameter set, a tile parameter set (also referred to as a tile list), or a geometry slice header, and sent to a point cloud decoder on the receiving side. In addition, the road / object splitting and prediction related option information may be included in at least one of a geometry LPU header or a geometry PU header, and sent to a point cloud decoder on the receiving side. According to an embodiment, the road / object splitting and prediction related option information may also include information for identifying each object (object_id). The road / object splitting and prediction related option information may be received in at least one of a geometry parameter set, a tile parameter set (also referred to as a tile list), or a geometry slice header. In addition, road / object splitting and prediction related option information may be received in at least one of the geometry LPU header or the geometry PU header. In the present disclosure, information to be included in the road / object splitting and prediction related option information may be added, deleted, or modified by a person skilled in the art, and thus the implementation is not limited to the above examples.

[0440] In other words, the geometry information entropy decoder 63001 performs entropy decoding on the input geometry bit stream and outputs the decoded bit stream to the road / object splitter 63002.

[0441] When the value of the information indicating whether to split the road / object included in the road / object splitting and prediction related option information is true, the road / object splitter 63002 can split the roads and objects from the entropy-decoded geometric data based on the road / object splitting and prediction related option information.

[0442] According to an embodiment, as described above, a road and an object may be separated based on estimated height information (estimated_height) for road / object separation.

[0443] According to an embodiment, the road / object splitter 63002 may classify the current point in the current frame as a road when it is at a position equal to or lower than the estimated height (estimated_height) of the road / object separation, and classify it as an object when the current point in the current frame is at a higher position.

[0444] According to an embodiment, in order to separate the above-mentioned roads and objects, road / object splitting and prediction related option information may be provided to the road / object splitter 63002. According to an embodiment, the road / object splitting and prediction related option information may include information indicating whether to split the road / object and predicted height information for road / object separation. According to an embodiment, the road / object splitting and prediction related option information may also include object ID information assigned to each PU.

[0445] When the value of the information indicating whether to split the road / object included in the road / object splitting and prediction related option information is false, the road / object splitter 63002 skips the road / object splitting. Alternatively, in the case where a bit stream split into an object and a road is received and the LPU can be identified as an LPU composed of points of the road or an LPU composed of points of the object through inter-frame prediction related information, the road / object splitting can be skipped.

[0446] According to an embodiment, when intra-prediction-based encoding is applied to the geometry information at the transmitting side, the geometry decoder 61003 performs intra-prediction-based reconstruction for the geometry information. On the other hand, when inter-prediction-based encoding is applied to the geometry information at the transmitting side, the geometry decoder 61003 performs inter-prediction-based reconstruction for the geometry information. In the case of receiving a bit stream split into roads and objects, the geometry information can be reconstructed by performing decoding for each of the roads and objects.

[0447] To this end, the element assigned with reference numeral 63002 (or referred to as a determination unit) checks whether encoding based on intra-frame prediction has been applied to the geometric information or encoding based on inter-frame prediction has been applied to the geometric information.

[0448] When the determination unit 63003 determines that intra-prediction-based encoding has been applied to the geometry information, the entropy-decoded geometry information is provided to the geometry information intra-prediction reconstructor 63004. On the other hand, when the determination unit 63003 determines that inter-prediction-based encoding has been applied to the geometry information, the entropy-decoded geometry information or the geometry information split into roads and objects by the road / object splitter 63002 is output to the LPU / PU splitter 63002.

[0449] The geometry information intra prediction reconstructor 63004 decodes and reconstructs the geometry information based on the intra prediction method. That is, the geometry information intra prediction reconstructor 63004 can separate the road portion and the object portion delivered by the road / object identification information in the prediction tree. When searching for a reference point in inter prediction reconstruction, the separated points can be applied to identify whether the current point is part of the road or part of the object.

[0450] When the frame of geometric information to be decoded is a P frame, the LPU / PU splitter 63005 splits the reference frame into LPU / PU based on road / object splitting and prediction related option information, and the road / object splitting and prediction related option information is signaled to support inter-frame prediction based reconstruction and indicate LPU / PU splitting.

[0451] In other words, the LPU / PU splitter 63005 generates LPUs using the points classified by the road / object splitter 63002. In this case, each LPU can be configured as a road and an object (or an object group) based on the road / object splitting and prediction related option information. To this end, information for identifying whether each LPU is a road or an object (or an object group) (i.e., whether it is a road / object) can be included in the road / object splitting and prediction related option information.

[0452] In addition, the LPU split into object groups may be composed of multiple objects. Therefore, when the objects have different characteristics, the LPU / PU splitter 63005 may split the LPU composed of multiple objects into PUs, and may reconstruct the LPU using the road / object splitting and prediction-related option information including the object ID information assigned to each PU.

[0453] The motion compensation application unit 63006 according to the embodiment may apply a motion vector (e.g., a global motion vector and / or a local motion vector) to the LPU / PU split from the reference frame to generate predicted geometric information. Here, the motion vector may be received via signaling information. In one embodiment, the motion compensation application unit 63006 may determine whether to apply the motion vector to the LPU / PU by referring to the road / object splitting and prediction-related option information. In another embodiment, the motion compensation application unit 63006 may perform motion compensation by applying a global motion vector to the LPU split into an object (or object group) and applying a local motion vector to the PU. In addition, the motion compensation application unit 63006 may skip motion compensation for the LPU split into roads.

[0454] The geometry information inter prediction reconstructor 63007 according to the embodiment decodes and reconstructs the geometry information based on the inter prediction method. That is, the geometry information encoded by the geometry inter prediction can be reconstructed based on the geometry information about the motion compensated reference frame (or non-motion compensated reference frame).

[0455] In this case, the geometric information inter-frame prediction reconstructor 63007 may be provided with a predicted height value for road / object separation, so that the reference frame can be separated into roads and objects. It can search for a reference point for the current point. In this operation, when the current point belongs to a road, a point with the same laser ID and a similar azimuth can be searched as a reference point in the area separated as a road in the reference frame. In addition, when the current point belongs to an object, a point with the same laser ID and a similar azimuth can be searched as a reference point in the area separated as an object in the reference frame. This process can be the same for the encoder on the sending side and the decoder on the receiving side.

[0456] The geometric information reconstructed by the geometric information intra-frame prediction reconstructor 63004 or the geometric information reconstructed by the geometric information inter-frame prediction reconstructor 63007 is input to the geometric information inverse transform / inverse quantization processor 63008.

[0457] The geometric information inverse transform / inverse quantization processor 63008 performs an inverse operation of the transform performed by the geometric information transform / quantization processor 51003 of the transmitting device on the reconstructed geometric information, and performs inverse quantization by multiplying the result of the inverse operation by a scaling factor (=geometric quantization value) to generate reconstructed geometric information. In other words, the geometric information inverse transform / inverse quantization processor 63008 can inversely quantize the geometric information by applying the scaling factor (scaling factor=geometric quantization value) included in the signaling information to the x value, y value, and z value of the geometric position of the reconstructed point.

[0458] The inverse coordinate transformer 63009 may perform an inverse operation of the coordinate transformation performed by the coordinate transformer 51002 of the transmitting device on the inverse quantized geometric information. For example, the inverse coordinate transformer 63009 may reconstruct the xyz axis changed on the transmitting side or inversely transform the transformed coordinates into xyz Cartesian coordinates.

[0459] According to an embodiment, the geometric information inversely quantized by the geometric information inverse transform / inverse quantization processor 63008 is processed in the geometric reconstruction process and stored in the reference frame buffer 63010, and is also output to the attribute decoder 61004 for attribute decoding.

[0460] According to an embodiment, the attribute residual information entropy decoder 65001 of the attribute decoder 61004 can perform entropy decoding on the input attribute bitstream.

[0461] According to an embodiment, when encoding based on intra-frame prediction has been applied to attribute information at the transmission side, the attribute decoder 61004 performs reconstruction based on intra-frame prediction for the attribute information. On the other hand, when encoding based on inter-frame prediction has been applied to attribute information at the transmission side, the attribute decoder 61004 performs reconstruction based on inter-frame prediction for the attribute information.

[0462] To this end, the element assigned with reference numeral 65002 (or referred to as a determination unit) checks whether encoding based on intra-frame prediction or encoding based on inter-frame prediction has been applied to the attribute information.

[0463] When the determination unit 65002 determines that encoding based on intra-frame prediction has been applied to the attribute information, the entropy-decoded attribute information is provided to the attribute information intra-frame prediction reconstructor 65003. On the other hand, when the determination unit 65002 determines that encoding based on inter-frame prediction has been applied to the attribute information, the entropy-decoded attribute information is provided to the attribute information inter-frame prediction reconstructor 65004.

[0464] The attribute information inter-frame prediction reconstructor 65004 decodes and reconstructs the attribute information based on the inter-frame prediction method. That is, the attribute information predicted by the inter-frame prediction encoding is reconstructed.

[0465] The attribute information intra prediction reconstructor 65003 decodes and reconstructs the attribute information based on the intra prediction method. That is, the attribute information predicted by the intra prediction coding is reconstructed. The intra coding method may include predictive transform coding, lifting transform coding and RAHT coding.

[0466] According to an embodiment, the reconstructed attribute information may be stored in the reference frame buffer 63010. The geometry information and attribute information stored in the reference frame buffer 63010 may be provided to the geometry information inter-frame prediction reconstructor 63007 and the attribute information inter-frame prediction reconstructor 65004 as a previous reference frame.

[0467] According to an embodiment, the reconstructed attribute information may be provided to the color inverse transform processor 65005 to reconstruct the RGB color. In other words, the color inverse transform processor 65005 performs inverse transform encoding to inversely transform the color value (or texture) included in the reconstructed attribute information, and then outputs the attribute to the post-processor 61005. The color inverse transform processor 65005 performs the same as Figure 7 The color inverse converter 7010 or Figure 9 The operation and / or inverse transform encoding of the color inverse transform processor 9010 is the same as or similar to the operation and / or inverse transform encoding.

[0468] The post-processor 61005 can reconstruct the point cloud data by matching the geometric information (i.e., position) reconstructed and output by the geometry decoder 61003 with the attribute information reconstructed and output by the attribute decoder 61004. In addition, when the reconstructed point cloud data is in units of tiles and / or slices, the post-processor 61005 can perform a process opposite to the spatial partitioning on the transmitting side based on the signaling information.

[0469] Figure 24 Another exemplary bitstream structure for transmitting / receiving point cloud data according to an embodiment is illustrated.

[0470] According to an embodiment, from Figure 1 , Figure 2 , Figure 3 , Figure 8 , Figure 20 and Figure 22 The bitstream output by any point cloud video encoder can be Figure 24 The form shown.

[0471] According to an embodiment, Figure 24 The term "slice" in may be referred to as "data unit".

[0472] In addition, Figure 24 In the description, each abbreviation has the following meaning. Each abbreviation may be referred to by another term within the scope of equivalent meaning. SPS: sequence parameter set; GPS: geometry parameter set; APS: attribute parameter set; TPS: tile parameter set; Geom: geometry bitstream = geometry slice header + [geometry PU header + geometry PU data] | geometry slice data); Attr: attribute bitstream = attribute data unit header + [attribute PU header + attribute PU data] | attribute data unit data; LPU: large prediction unit; PU: prediction unit.

[0473] In the present disclosure, relevant information may be signaled in order to add / perform the embodiments described so far. The signaling information according to the embodiments may be used by a point cloud video encoder on the transmitting side or a point cloud video decoder on the receiving side.

[0474] The point cloud video encoder according to the embodiment may generate a point cloud video by encoding the geometric information and the attribute information as described above. Figure 24 In addition, signaling information related to the point cloud data may be generated and processed by at least one of a geometry encoder, an attribute encoder, or a signaling processor of the point cloud video encoder, and may be included in the bitstream.

[0475] As an example, a point cloud video encoder configured to perform geometry encoding and / or attribute encoding may generate Figure 24In addition, the signaling information related to the point cloud data can be generated and processed by the metadata processor of the point cloud data sending device and included in the Figure 24 The point cloud shown.

[0476] The signaling information may be received / obtained by at least one of a geometry decoder, an attribute decoder, or a signaling processor of a point cloud video decoder.

[0477] A bitstream according to an embodiment may be divided into a geometry bitstream, an attribute bitstream, and a signaling bitstream and transmitted / received, or one combined bitstream may be transmitted / received.

[0478] When the geometry bitstream, attribute bitstream and signaling bitstream according to the embodiment are configured in one bitstream, the bitstream may include one or more sub-bitstreams. The bitstream according to the embodiment may include a sequence parameter set (SPS) for sequence level signaling, a geometry parameter set (GPS) for signaling geometry information encoding, one or more attribute parameter sets (APS) (APS0, APS1) for signaling attribute information encoding, a tile parameter set (TPS) (also called a tile list) for signaling at the tile level, and one or more slices (slice 0 to slice n). That is, the bitstream of the point cloud data according to the embodiment may include one or more tiles, and each of the tiles may be a slice group including one or more slices (slice 0 to slice n). The TPS according to the embodiment may contain information about each of the one or more tiles (e.g., coordinate value information and height / size information about the bounding box). Each slice may include a geometry bitstream (Geom0) and one or more attribute bitstreams (Attr0 and Attr1).

[0479] The geometry bitstream in each slice (or referred to as geometry slice) may include a geometry slice header and one or more geometry PUs (Geom PU0, Geom PU1). Each geometry LPU may include a geom LPU header and geom LPU data. Each geometry PU may include a geom PU header and geom PU data.

[0480] Each attribute bitstream (or attribute slice) in each slice may include an attribute slice header and one or more attribute PUs (Attr PU0, Attr PU1). Each attribute LPU may include an Attr LPU header and Attr LPU data. Each attribute PU may include an attr PU header and attr PU data.

[0481] A portion of the road / object splitting and prediction related option information according to an embodiment may be added to and signaled in the GPS and / or TPS (also referred to as a tile list).

[0482] A portion of the road / object splitting and prediction related option information may be added to and signaled in the geometry slice header for each slice.

[0483] Part of the road / object splitting and prediction related option information may be signaled in the geom LPU header.

[0484] Part of the road / object splitting and prediction related option information may be signaled in the geom PU header.

[0485] As used herein, road / object segmentation and prediction related option information is used interchangeably with motion option information for road / object segmentation and geometry prediction tree encoding.

[0486] According to an embodiment, parameters required for encoding and / or decoding of point cloud data may be newly defined in a parameter set of point cloud data (e.g., SPS, GPS, APS, and TPS (or referred to as a tile list), etc.) and / or a header of a corresponding slice. For example, the parameters may be added to the GPS when encoding and / or decoding geometric information, and may be added to the tile (TPS) and / or slice header when performing tile-based encoding and / or decoding. When performing LPU-based encoding and / or decoding, the parameters may be added to the Geom LPU header and / or the Attr LPU header. In addition, when performing PU-based encoding and / or decoding, the parameters may be added to the geometry PU header and / or the attribute PU header.

[0487] like Figure 24 As shown, the bit stream of the point cloud data is divided into tiles, slices, LPUs and / or PUs so that the point cloud data can be divided into areas to be processed. The areas of the bit stream may have different levels of importance. Therefore, when the point cloud data is divided into tiles, different filters (encoding methods) or different filter units may be applied to the tiles, respectively. When the point cloud data is divided into slices, different filters or different filter units may be applied to the slices, respectively. When the point cloud data is divided into LPUs, different filters or different filter units may be applied to the LPUs, respectively. In addition, when the point cloud data is divided into PUs, different filters or different filter units may be applied to the PUs, respectively.

[0488] By following Figure 24The bitstream structure shown transmits point cloud data, and the transmitting device according to the embodiment can allow the encoding operation to be applied differently according to the importance level, and allow the use of a good quality encoding method in an important area. In addition, efficient encoding and transmission can be supported according to the characteristics of the point cloud data, and attribute values ​​can be provided according to user needs.

[0489] Since the receiving device according to the embodiment is based on Figure 24 The bitstream structure shown receives point cloud data, so it can apply different filtering (decoding methods) to corresponding areas (areas divided into tiles or slices) according to the processing capability of the receiving device, rather than using complex decoding (filtering) methods for the entire point cloud data. Therefore, better image quality can be provided for areas important to the user, and appropriate latency can be ensured in the system.

[0490] As described above, tiles or slices are provided to process point cloud data by dividing the point cloud data into regions. When dividing the point cloud data into regions, an option can be set to generate a different set of neighbor points for each region. Thus, a selection method with low complexity and low reliability, or a selection method with high complexity and high reliability can be provided.

[0491] According to an embodiment, at least one of the GPS, TPS, geometry slice header, geometry LPU header, or geometry PU header may include road / object splitting and prediction related option information.

[0492] According to an embodiment, the road / object splitting and prediction related option information may include information indicating whether to split the road / object (road_object_split_flag), estimated height information (estimated_height) for road / object separation, or information for estimating basic height information (e.g., angle θ of the sensor, z value of the sensor, radius of the sensor). According to an embodiment, the road / object splitting and prediction related option information may also include identification information for identifying the LPU, information for identifying whether the LPU is composed of points classified into the road or points classified into the object, information indicating whether the global motion is applied to the LPU, information for identifying the PU, and information for identifying each object (object_id). A portion of the information included in the road / object splitting and prediction related option information may be signaled in the GPS, and another portion thereof (or repeated information) may be signaled in the TPS or geometry slice header. In addition, a portion of the information included in the road / object splitting and prediction related option information may be signaled in the geometry LPU header, and another portion thereof (or repeated information) may be signaled in the geometry PU header.

[0493] A field (which is a term used in syntax to be described later in this disclosure) may have the same meaning as a parameter or a syntax element.

[0494] Figure 25 2 is a diagram illustrating an embodiment of a syntax structure (seq_parameter_set_rbsp()) of a sequence parameter set (SPS) according to the present disclosure. The SPS may include sequence information about a point cloud data bitstream.

[0495] The SPS according to an embodiment may include a main_profile_compatibility_flag field, a unique_point_positions_constraint_flag field, a level_idc field, a sps_seq_parameter_set_id field, a sps_bounding_box_present_flag field, a sps_source_scale_factor_numerator_minus1 field, a sps_source_scale_factor_denominator_minus1 field, a sps_num_attribute_sets field, a log2_max_frame_idx field, an axis_coding_order field, a sps_bypass_stream_enabled_flag field, and a sps_extension_flag field.

[0496] The main_profile_compatibility_flag field may indicate whether the bitstream conforms to the main profile. For example, main_profile_compatibility_flag equal to 1 may indicate that the bitstream conforms to the main profile. For example, main_profile_compatibility_flag equal to 0 may indicate that the bitstream conforms to a profile different from the main profile.

[0497] When unique_point_positions_constraint_flag is equal to 1, all output points can have unique positions in each point cloud frame referenced by the current SPS. When unique_point_positions_constraint_flag is equal to 0, two or more output points can have the same position in any point cloud frame referenced by the current SPS. For example, even when all points are unique in the corresponding slice, slices and other points in the frame may overlap. In this case, unique_point_positions_constraint_flag is set to 0.

[0498] level_idc indicates the level to which the bitstream conforms.

[0499] sps_seq_parameter_set_id provides an identifier of the SPS for reference by other syntax elements.

[0500] The sps_bounding_box_present_flag field indicates whether there is a bounding box in the SPS. For example, sps_bounding_box_present_flag equal to 1 indicates that there is a bounding box in the SPS, and sps_bounding_box_present_flag equal to 0 indicates that the size of the bounding box is not defined.

[0501] According to an embodiment, when sps_bounding_box_present_flag is equal to 1, the SPS may further include a sps_bounding_box_offset_x field, a sps_bounding_box_offset_y field, a sps_bounding_box_offset_z field, a sps_bounding_box_offset_log2_scale field, a sps_bounding_box_size_width field, a sps_bounding_box_size_height field, and a sps_bounding_box_size_depth field.

[0502] sps_bounding_box_offset_x indicates the x offset of the source bounding box in Cartesian coordinates. When there is no x offset of the source bounding box, the value of sps_bounding_box_offset_x is 0.

[0503] sps_bounding_box_offset_y indicates the y offset of the source bounding box in Cartesian coordinates. When there is no y offset of the source bounding box, the value of sps_bounding_box_offset_y is 0.

[0504] sps_bounding_box_offset_z indicates the z offset of the source bounding box in Cartesian coordinates. When there is no z offset of the source bounding box, the value of sps_bounding_box_offset_z is 0.

[0505] sps_bounding_box_offset_log2_scale indicates the scaling factor used to scale the quantized x, y, and z source bounding box offsets.

[0506] sps_bounding_box_size_width indicates the width of the source bounding box in Cartesian coordinates. The value of sps_bounding_box_size_width can be 1 when there is no width of the source bounding box.

[0507] sps_bounding_box_size_height indicates the height of the source bounding box in Cartesian coordinates. The value of sps_bounding_box_size_height can be 1 when there is no source bounding box height.

[0508] sps_bounding_box_size_depth indicates the depth of the source bounding box in Cartesian coordinates. The value of sps_bounding_box_size_depth may be 1 when there is no depth of the source bounding box.

[0509] sps_source_scale_factor_numerator_minus1 plus 1 indicates the scale factor numerator of the source point cloud.

[0510] sps_source_scale_factor_denominator_minus1 plus 1 indicates the scale factor denominator of the source point cloud.

[0511] sps_num_attribute_sets indicates the number of coded attributes in the bitstream.

[0512] The SPS according to an embodiment includes an iteration statement that is repeated as many times as the value of the sps_num_attribute_sets field. In an embodiment, i is initialized to 0 and incremented by 1 each time the iteration statement is executed. The iteration statement is repeated until the value of i becomes equal to the value of the sps_num_attribute_sets field. The iteration statement may include an attribute_dimension_minus1[i] field and an attribute_instance_id[i] field. attribute_dimension_minus1[i] plus 1 indicates the number of components of the i-th attribute.

[0513] attribute_instance_id[i] specifies the instance ID of the i-th attribute.

[0514] According to an embodiment, when the value of the attribute_dimension_minus1[i] field is greater than 1, the iteration statement may further include an attribute_secondary_bitdepth_minus1[i] field, an attribute_cicp_colour_primaries[i] field, an attribute_cicp_transfer_characteristics[i] field, an attribute_cicp_matrix_coeffs[i] field, and an attribute_cicp_video_full_range_flag[i] field.

[0515] attribute_secondary_bitdepth_minus1[i] plus 1 specifies the bit depth of the secondary component of the i-th attribute signal.

[0516] attribute_cicp_colour_primaries[i] indicates the chromaticity coordinates of the colour attribute source primaries of the i-th attribute.

[0517] attribute_cicp_transfer_characteristics[i] indicates a reference electro-optical transfer characteristic function of a color attribute as a function of source input linear light intensity, with a nominal real value range of 0 to 1, or indicates the inverse of a reference electro-optical transfer characteristic function as a function of output linear light intensity.

[0518] attribute_cicp_matrix_coeffs[i] describes the matrix coefficients used when deriving the luma signal and the chroma signal from the green, blue and red or Y, Z and X primaries of the i-th attribute.

[0519] attribute_cicp_video_full_range_flag[i] specifies the black level and range of the luma signal and chroma signal derived from the E'Y, E'PB, and E'PR or E'R, E'G, and E'B real-valued component signals of the i-th attribute.

[0520] The known_attribute_label_flag[i] field indicates whether the known_attribute_label[i] field or the attribute_label_four_bytes[i] field is signaled for the i-th attribute. For example, when known_attribute_label_flag[i] is equal to 0, it indicates that the known_attribute_label[i] field is signaled for the i-th attribute. When known_attribute_label_flag[i] is equal to 1, it indicates that the attribute_label_four_bytes[i] field is signaled for the i-th attribute.

[0521] known_attribute_label[i] specifies the type of the i-th attribute. For example, a known_attribute_label[i] of 0 specifies that the i-th attribute is color. A known_attribute_label[i] of 1 specifies that the i-th attribute is reflectivity. A known_attribute_label[i] of 2 specifies that the i-th attribute is a frame index. Additionally, a known_attribute_label[i] of 4 specifies that the i-th attribute is transparency. A known_attribute_label[i] of 5 specifies that the i-th attribute is a normal.

[0522] attribute_label_four_bytes[i] indicates a known attribute type with a 4-byte code.

[0523] According to an embodiment, attribute_label_four_bytes[i] equal to 0 may indicate that the i-th attribute is color. attribute_label_four_bytes[i] equal to 1 may indicate that the i-th attribute is reflectivity. attribute_label_four_bytes[i] equal to 2 may indicate that the i-th attribute is frame index. attribute_label_four_bytes[i] equal to 4 may indicate that the i-th attribute is transparency. attribute_label_four_bytes[i] equal to 5 may indicate that the i-th attribute is normal.

[0524] log2_max_frame_idx indicates the number of bits used to signal the syntax variable frame_idx.

[0525] axis_coding_order specifies the correspondence between the X output axis label, the Y output axis label, and the Z output axis label and the three position components in the reconstructed point cloud RecPic[pointidx][axis], where axis = 0..2.

[0526] sps_bypass_stream_enabled_flag equal to 1 specifies that the bypass coding mode can be used for reading the bitstream. As another example, sps_bypass_stream_enabled_flag equal to 0 specifies that the bypass coding mode is not used when reading the bitstream.

[0527] sps_extension_flag indicates whether there is an sps_extension_data syntax structure in the SPS syntax structure. For example, sps_extension_present_flag equal to 1 indicates that there is an sps_extension_data syntax structure in the SPS syntax structure. sps_extension_present_flag equal to 0 indicates that there is no such syntax structure.

[0528] When a value of the sps_extension_flag field is 1, the SPS according to an embodiment may further include a sps_extension_data_flag field.

[0529] sps_extension_data_flag can have any value.

[0530] Figure 26 An exemplary syntax structure of a geometry parameter set (geometry_parameter_set()) (GPS) including road / object splitting and prediction related option information according to an embodiment is illustrated. GPS may include information on a method for encoding geometry information related to point cloud data contained in one or more slices.

[0531] According to an embodiment, road / object splitting and prediction related option information (i.e., motion option information for road / object splitting and geometry prediction tree encoding) may be added to a geometry parameter set (GPS) and signaled for encoding / decoding geometry information frame by frame. In other words, by combining one or more items of road / object splitting and prediction related option information included in the GPS, inter-frame prediction of geometry may be efficiently supported. As used herein, the name of the signaling information may be understood within the scope of the meaning and function of the signaling information.

[0532] The GPS according to the embodiment may include at least a gps_geom_parameter_set_id field, a gps_seq_parameter_set_id field, a geom_tree_type field, and a road_object_split_flag field.

[0533] gps_geom_parameter_set_id provides an identifier specific to GPS for reference by other syntax elements.

[0534] gps_seq_parameter_set_id specifies the value of sps_seq_parameter_set_id for the active SPS.

[0535] geom_tree_type indicates the encoding type of geometric information. For example, geom_tree_type equal to 0 may indicate that the geometric information (ie, position information) is encoded using an octree, and geom_tree_type equal to 1 may indicate that the information is encoded using a prediction tree.

[0536] road_object_split_flag specifies whether road / object splitting has been applied to the frame. For example, when the value of road_object_split_flag is true, it may indicate that road / object splitting has been applied to the frame.

[0537] When the value of road_object_split_flag is true, the GPS according to an embodiment may further include an estimated_height field and a road_object_connection_flag field.

[0538] estimated_height indicates the basic height of the sensor applied to the frame. In other words, estimated_height specifies the road / object separation height used to identify whether the point belongs to a road or an object. For example, when the point is at or below the road / object separation height, the point can be classified as a road. In the opposite case, the point can be classified as an object. In another embodiment, the basic height can be obtained by calculation.

[0539] road_object_connection_flag specifies whether to connect a separated road applied to a frame with a separated object.

[0540] Figure 27 An exemplary syntax structure of a tile parameter set (tile_parameter_set()) (TPS) including road / object splitting and prediction related option information according to an embodiment is illustrated. In some embodiments, a TPS may also be referred to as a tile list. A TPS includes information related to each tile on a per-tile basis.

[0541] According to an embodiment, road / object splitting and prediction related option information (i.e., motion option information for road / object splitting and geometry prediction tree encoding) may be added to TPS and signaled for encoding / decoding geometry information on a per-tile basis. In other words, by combining one or more road / object splitting and prediction related option information included in TPS, inter-frame prediction of geometry may be efficiently supported. As used herein, the name of the signaling information may be understood within the scope of the meaning and function of the signaling information.

[0542] The TPS according to an embodiment includes a num_tiles field.

[0543] num_tiles indicates the number of tiles signaled for the bitstream. When not present, num_tiles is inferred to be 0.

[0544] A TPS according to an embodiment includes an iteration statement that iterates as many times as the value of the num_tiles field. In an embodiment, i is initialized to 0 and incremented by 1 each time the iteration statement is executed. The iteration statement iterates until the value of i reaches the value of num_tiles. The iteration statement may include a tile_bounding_box_offset_x[i] field, a tile_bounding_box_offset_y[i] field, a tile_bounding_box_offset_z[i] field, a tile_bounding_box_size_width[i] field, a tile_bounding_box_size_height[i] field, a tile_bounding_box_size_depth[i] field, and a road_object_split_flag[i] field. When the value of road_object_split_flag[i] is true, the iteration statement may also include an estimated_height[i] field and a road_object_connection_flag[i] field.

[0545] tile_bounding_box_offset_x[i] indicates the x offset of the i-th tile in Cartesian coordinates.

[0546] tile_bounding_box_offset_y[i] indicates the y offset of the i-th tile in Cartesian coordinates.

[0547] tile_bounding_box_offset_z[i] indicates the z offset of the i-th tile in Cartesian coordinates.

[0548] tile_bounding_box_size_width[i] indicates the width of the i-th tile in Cartesian coordinates.

[0549] The tile_bounding_box_size_height[i] field indicates the height of the i-th tile in Cartesian coordinates.

[0550] tile_bounding_box_size_depth[i] indicates the depth of the i-th tile in Cartesian coordinates.

[0551] road_object_split_flag[i] specifies whether road / object splitting has been applied to the frame. For example, when the value of road_object_split_flag is true, it may indicate that road / object splitting has been applied to the i-th tile.

[0552] estimated_height[i] indicates the base height of the sensor applied to the i-th tile. In other words, estimated_height[i] specifies the road / object separation height used to identify whether a point belongs to a road or an object. For example, when a point is at or below the road / object separation height, the point can be classified as a road. In the opposite case, the point can be classified as an object. In another embodiment, the base height can be obtained by calculation.

[0553] road_object_connection_flag[i] specifies whether to connect the separated road applied to the i-th tile with the separated object.

[0554] According to an embodiment, the geometry data unit (ie, geometry_slice_bitstream()) may include a geometry data unit header (ie, geometry_slice_header()) and geometry data unit data.

[0555] Figure 28 An exemplary syntax structure of a geometry data unit according to an embodiment is illustrated.

[0556] According to an embodiment, when a value of the geom_tree_type field is 0, the geometry data unit data includes occupancy_tree( ), and when a value of the geom_tree_type field is 1, the geometry data unit data includes predictive_tree( ).

[0557] geom_tree_type indicates the encoding type of the geometric information. For example, geom_tree_type equal to 0 may indicate that the geometric information (ie, position information) has been encoded using an octree, and geom_tree_type equal to 1 may indicate that the geometric information has been encoded using a prediction tree.

[0558] Figure 29 An exemplary syntax structure of a geometry data unit header (ie, a geometry slice header (geometry_slice_header())) according to an embodiment of the present disclosure is illustrated.

[0559] According to an embodiment, road / object splitting and prediction related option information (i.e., motion option information for road / object splitting and geometry prediction tree encoding) may be added to a geometry data unit header (i.e., geometry slice header) and signaled to encode / decode geometry information slice by slice. In other words, by combining one or more road / object splitting and prediction related option information included in the geometry data unit header, geometric road / object splitting and intra / inter prediction may be efficiently supported. As used herein, the name of the signaling information may be understood within the scope of the meaning and function of the signaling information.

[0560] The geometry data unit header according to an embodiment may include at least a gsh_geometry_parameter_set_id field, a gsh_tile_id field, a gsh_slice_id field, or a road_object_split_flag field.

[0561] gsh_geometry_parameter_set_id specifies the value of gps_geom_parameter_set_id for the active GPS.

[0562] gsh_tile_id specifies the identifier of the tile referenced by the geometry data unit header (ie, geometry slice header (GSH)).

[0563] gsh_slice_id specifies the identifier of the slice referenced by other syntax elements.

[0564] road_object_split_flag specifies whether road / object splitting has been applied to geometry slicing. For example, when the value of the road_object_split_flag field is true, it may indicate that road / object splitting has been applied to geometry slicing.

[0565] When the value of road_object_split_flag is true, the geometry data unit header according to an embodiment may further include an estimated_height field and a road_object_connection_flag field. The geometry data unit header according to an embodiment may further include an is_road_flag field.

[0566] estimated_height indicates the basic height of the sensor applied to the slice. In other words, estimated_height specifies the road / object separation height used to identify whether the point belongs to a road or an object. For example, when the point is at or below the road / object separation height, the point can be classified as a road. In the opposite case, the point can be classified as an object. In another embodiment, the basic height can be obtained by calculation.

[0567] road_object_connection_flag specifies whether to connect separated roads applied to slicing with separated objects.

[0568] is_road_flag may indicate whether a point included in a slice belongs to a road or an object.

[0569] Figure 30 An exemplary syntax structure of predictive_tree() according to an embodiment is illustrated. That is, when the value of the geom_tree_type field is 1, predictive_tree() is included in the geometry data unit.

[0570] According to an embodiment, road / object splitting and prediction related option information (i.e., motion option information for road / object splitting and geometry prediction tree encoding) may be added to a prediction tree node and signaled for encoding / decoding geometry information. As used herein, the name of the signaling information may be understood within the scope of the meaning and function of the signaling information.

[0571] According to an embodiment, predictive_tree() may include a roadPtnCnt field and an objPtnCnt field.

[0572] roadPtnCnt specifies the number of points belonging to the road.

[0573] objPtnCnt specifies the number of points belonging to the object.

[0574] According to an embodiment, predictive_tree() includes an iteration statement that iterates until the value of ptree_end_of_slice is false, wherein the iteration statement may include predictive_tree_node(0,roadPtnCnt), ptree_end_of_road, predictive_tree_node(0,objPtnCnt), and ptree_end_of_slice.

[0575] ptree_end_of_road indicates a separator (or indicator information) which specifies the end of a prediction tree composed of points belonging to a road.

[0576] ptree_end_of_slice indicates a delimiter (or indicator information) which specifies the end of a prediction tree consisting of points belonging to a slice.

[0577] predictive_tree_node(0,roadPtnCnt) and predictive_tree_node(0,objPtnCnt) are configured to carry points belonging to roads and points belonging to objects.

[0578] According to an embodiment, a frame, tile or slice may be divided into one or more LPUs based on a road / object separation method. For example, a geometry slice may include a geometry slice header and one or more geometry LPUs. One or more geometry LPUs may include an LPU consisting of points of a road and an LPU consisting of points of an object (or multiple objects). Each LPU may include a geometry LPU header and geometry LPU data.

[0579] According to an embodiment, an LPU consisting of points of an object group may be split into as many PUs as the objects in the object group. Each PU may include a geometry PU header and geometry PU data.

[0580] Figure 31 An exemplary syntax structure of a geometry PU header (geom_pu_header()) including road / object splitting and prediction related option information according to an embodiment is illustrated. In other words, the road / object splitting and prediction related option information (also called object splitting and geometry prediction tree encoding information about intra / inter prediction related LPU / PU) can be signaled by generating a geometry PU header.

[0581] In other words, by combining one or more road / object splitting and prediction related option information contained in the geometry PU header, geometric intra prediction / inter prediction can be efficiently supported. As used herein, the name of the signaling information can be understood within the scope of the meaning and function of the signaling information.

[0582] The geometry PU header according to an embodiment may include a pu_tile_id field, a pu_slice_id field, a pu_cnt field, and an is_road_flag field.

[0583] pu_tile_id specifies the ID of the tile to which the PU belongs.

[0584] pu_slice_id specifies the ID of the slice to which the PU belongs.

[0585] pu_cnt specifies the number of PUs in the slice.

[0586] is_road_flag may indicate whether a point contained in an LPU block belongs to a road or an object.

[0587] The Geometry PU header includes the following fields according to the number of PUs.

[0588] pu_id[puidx] specifies the ID of the PU.

[0589] pu_split_flag specifies whether the PU block has been further split.

[0590] When the value of is_road_flag is false, that is, the points included in the PU are points separated into objects (or object groups), the geometry PU header may further include an object_id field.

[0591] object_id specifies an object ID assigned to a PU. In other words, object_id indicates object identification information for identifying an object to which a point included in a PU block belongs.

[0592] pu_has_motion_vector_flag specifies whether the PU block has a motion vector. For example, pu_has_motion_vector_flag equal to 1 may indicate that the PU has a motion vector, and pu_has_motion_vector_flag equal to 0 may indicate that the PU does not have a motion vector.

[0593] When pu_has_motion_vector_flag is equal to 1, the geometry PU header may also include motion vector related information according to the value of motion_desc_type.

[0594] When motion_desc_type is equal to 0, the geometry PU header may also include a pu_motion_mat[pu_id][k][l] field.

[0595] pu_motion_mat[pu_id][k][l] specifies the motion matrix applied to the PU block identified by pu_id.

[0596] When motion_desc_type is equal to 1, the geometry PU header may further include a pu_motion_rot_vector[pu_id][k] field and a pu_motion_trans[pu_id][k] field.

[0597] pu_motion_rot_vector[pu_id][k] specifies the motion rotation vector applied to the PU block identified by pu_id.

[0598] pu_motion_trans[pu_id][k] specifies the motion translation vector applied to the PU block identified by pu_id.

[0599] When motion_desc_type is equal to 2, the geometry PU header may further include a pu_motion_rot_type[pu_id] field, a pu_motion_rot[pu_id] field, and a pu_motion_trans[pu_id][k] field.

[0600] pu_motion_rot_type[pu_id] specifies the type of motion rotation applied to the PU block identified by pu_id. For example, pu_motion_rot_type[pu_id] may indicate a radius when equal to 0, and an angle (in degrees) when equal to 1.

[0601] pu_motion_rot[pu_id] specifies the motion rotation value applied to the PU block identified by pu_id. According to an embodiment, the rotation axis rotates around the road normal vector.

[0602] Figure 32 is a flowchart illustrating a method of transmitting point cloud data according to an embodiment.

[0603] The method of transmitting point cloud data according to an embodiment may include an operation of acquiring point cloud data 71001, an operation of encoding the point cloud data 71002, and an operation of transmitting the encoded point cloud data and signaling information 71003. In this case, a bit stream containing the encoded point cloud data and signaling information may be encapsulated in a file and transmitted.

[0604] In operation 71001 of acquiring point cloud data, you can perform Figure 1 Some or all of the operations of the point cloud video acquisition unit 10001, or may be performed Figure 8 Some or all operations of the data input unit 8000.

[0605] In operation 71002 of encoding point cloud data, it is possible to perform Figure 1 Point cloud video encoder 10002, Figure 2 Code 20001, Figure 3 Point cloud video encoder, Figure 8 Point cloud video encoder, Figure 20The geometry encoder and attribute encoder of Figure 21 Some or all of the operations of the geometry encoder and attribute encoder to encode the geometry information and attribute information.

[0606] Operation 71002 of encoding point cloud data according to an embodiment may include compressing geometric information associated with the input point cloud data and compressing attribute information associated with the data.

[0607] According to an embodiment, in the operation of compressing geometric information, inter-frame prediction-based or intra-frame prediction-based encoding is performed for the position of point cloud data (i.e., geometric information) to output a geometric bitstream. In this case, when the frame of the point cloud data is a P frame, by applying the above-mentioned road and object separation method to the frame-level, tile-level, or slice-level point cloud data for prediction-based encoding of the P frame, roads and objects can be separated from the frame-level, tile-level, or slice-level point cloud data. Then, a prediction tree can be constructed for each of the road portion and the object portion, and the geometric information can be compressed using intra-frame prediction or inter-frame prediction.

[0608] For details of separating the points of the point cloud data into roads and objects and constructing corresponding prediction trees to perform compression (which will not be described below), refer to the above given Figures 10 to 21 Description.

[0609] The geometric information about each compressed point is entropy encoded and output in the form of a geometry bitstream.

[0610] According to an embodiment, the operation of compressing the attribute information includes compressing the attribute information based on the position and / or reconstructed geometry information for which geometry coding has not been performed. In one embodiment, the attribute information may be encoded using any one or a combination of one or more of RAHT coding, LOC-based predictive transform coding, and lifting transform coding.

[0611] The compressed attribute information is entropy encoded and output in the form of an attribute bitstream.

[0612] In the present disclosure, the signaling information may include road / object splitting and prediction related option information for encoding of the road / object splitting and the geometry prediction tree.

[0613] The road / object splitting and prediction related option information may include information indicating whether the road / object is separated, predicted height information for road / object separation, LPU / PU related information, motion vector related information, and information for identifying each object. The road / object splitting and prediction related option information may be included in at least one of a geometry parameter set, a tile parameter set (also called a tile list), or a geometry slice header and sent to the receiving side. In addition, the road / object splitting and prediction related option information may be included in at least one of a geometry LPU header or a geometry PU header and sent to the receiving side.

[0614] Figure 33 is a flow chart illustrating an exemplary method of receiving point cloud data according to an embodiment.

[0615] The method of receiving point cloud data according to an embodiment may include an operation 81001 of receiving encoded point cloud data and signaling information, an operation 81002 of decoding the point cloud data based on the signaling information, and an operation 81003 of rendering the decoded point cloud data.

[0616] According to an embodiment, the operation 81001 of receiving point cloud data and signaling information can be performed by Figure 1 Receiver 10005, Figure 2 Transmission 20002 or decoding 20003 or Figure 9 The receiver 9000 or the receiving processor 9001 is used for execution.

[0617] In operation 81002 of decoding point cloud data according to an embodiment, it is possible to perform Figure 1 Point cloud video decoder 10006, Figure 2 Decoding 20003, Figure 8 Point cloud video decoder, Figure 9 Point cloud video decoder, Figure 22 Geometry decoder and attribute decoder or Figure 23 Some or all of the operations of the geometry decoder and attribute decoder to decode the geometry information and attribute information.

[0618] Operation 81002 of decoding point cloud data according to an embodiment includes decoding geometric information and decoding attribute information.

[0619] The operation of decoding the geometric information may include decoding (ie, reconstructing) the geometric information based on the road / object splitting and prediction related option information included in the signaling information. For further details, refer to Figures 10 to 23 Description.

[0620] The operation of decoding the attribute information may include decoding (ie, decompressing) the attribute information based on the reconstructed geometric information. In one embodiment, the attribute information may be decoded using any one or a combination of one or more of RAHT coding, LOC-based predictive transform coding, and lifting transform coding.

[0621] The rendering operation 81003 according to the embodiment may include reconstructing the point cloud data based on the restored (or reconstructed) geometric information and attribute information, and rendering the data according to various rendering schemes. For example, the points in the point cloud content may be rendered as vertices with a certain thickness, cubes with a certain minimum size centered at the corresponding vertex position, circles centered at the vertex position, etc. All or part of the rendered point cloud content may be presented to the user through a display (e.g., a VR / AR display, a general display, etc.). The rendering operation 81003 according to the embodiment may be performed by Figure 1 Renderer 10007, Figure 2 Rendering 20004 or Figure 9 The renderer 9011 is used to execute.

[0622] As described above, in a scene where each frame is captured and stored by a LiDAR device on a moving vehicle, there may be continuity between frames, so inter-frame prediction technology can be used to efficiently perform compression. Specifically, since the captured road and object point clouds have different motion characteristics, the roads and objects can be split to efficiently apply inter-frame prediction technology. As a result, motion prediction can be performed quickly and accurately for each of the roads and objects, thereby reducing compression time. In addition, by improving the efficiency of the inter-frame prediction technology through accurate prediction, the bitstream size can be reduced. Inaccurate motion prediction can significantly increase the bitstream size, thereby reducing compression efficiency.

[0623] In these embodiments, a road / object splitting method is proposed to support efficient geometric compression of content captured by a LiDAR device on a moving vehicle. In addition, prediction trees for roads and objects can be constructed separately and then compression can be performed. Therefore, a point cloud content stream with improved geometric compression efficiency can be provided in a G-PCC encoder / decoder.

[0624] In this way, the sending method / apparatus can efficiently compress the point cloud data to send the data, and can deliver signaling information for the data. Therefore, the receiving method / apparatus can efficiently decode / reconstruct the point cloud data.

[0625] Each of the above components, modules or units can be software, processors or hardware components that execute the continuous process stored in the memory (or storage unit). Each of the steps described in the above embodiments can be performed by a processor, a software component or a hardware component. Each module / block / unit described in the above embodiments can be operated as a processor, software or hardware. In addition, the method presented by the embodiment can be executed as a code. The code can be written on a processor-readable storage medium, so it is read by a processor provided by the device.

[0626] In this specification, when a component "includes" or "comprises" an element, unless otherwise specified, it means that the component also includes or comprises another element. In addition, the term "...module (or unit)" disclosed in the specification refers to a unit for processing at least one function or operation, and can be implemented by hardware, software, or a combination of hardware and software.

[0627] Although the embodiments have been explained with reference to each of the accompanying drawings for simplicity, new embodiments may be designed by combining the embodiments illustrated in the accompanying drawings. If a computer-readable recording medium having a program for executing the embodiments mentioned in the foregoing description is designed by a person skilled in the art, it may fall within the scope of the appended claims and their equivalents.

[0628] The device and method may not be limited to the configuration and method of the above-mentioned embodiments. The above-mentioned embodiments may be configured by selectively combining with each other in whole or in part to achieve various modifications.

[0629] Although preferred embodiments of the embodiments have been shown and described, the embodiments are not limited to the specific embodiments described above, and various modifications may be made by a person of ordinary skill in the art without departing from the spirit of the embodiments claimed in the claims, and these modifications should not be understood separately from the technical concept or view of the embodiments.

[0630] Various elements of the device of the embodiment can be implemented by hardware, software, firmware or a combination thereof. Various elements in the embodiment can be implemented by a single chip (e.g., a single hardware circuit). According to the embodiment, the components according to the embodiment can be implemented as separate chips respectively. According to the embodiment, at least one or more components of the device according to the embodiment may include one or more processors capable of executing one or more programs. One or more programs may execute any one or more operations / methods according to the embodiment or include instructions for executing them. The executable instructions for executing the method / operation of the device according to the embodiment may be stored in a non-transitory CRM or other computer program product configured to be executed by one or more processors, or may be stored in a temporary CRM or other computer program product configured to be executed by one or more processors. In addition, the memory according to the embodiment can be used as not only covering volatile memory (e.g., RAM), but also covering the concepts of non-volatile memory, flash memory and PROM. In addition, it can also be implemented in the form of a carrier (e.g., transmission via the Internet). In addition, the processor-readable recording medium can be distributed to a computer system connected via a network so that the processor-readable code can be stored and executed in a distributed manner.

[0631] In this document, the terms " / " and "," should be interpreted as indicating "and / or". For example, the expression "A / B" may mean "A and / or B". In addition, "A, B" may mean "A and / or B". In addition, "A / B / C" may mean "at least one of A, B, and / or C". In addition, "A / B / C" may mean "at least one of A, B, and / or C". In addition, in this document, the term "or" should be interpreted as indicating "and / or". For example, the expression "A or B" may include 1) only A, 2) only B, and / or 3) both A and B. In other words, the term "or" in this document should be interpreted as indicating "in addition or alternatively".

[0632] The various elements of the embodiments can be implemented by hardware, software, firmware or a combination thereof. The various elements of the embodiments can be performed by a single chip such as a single hardware circuit. According to the embodiments, the elements can be selectively performed by separate chips respectively. According to the embodiments, at least one of the elements of the embodiments can be performed in one or more processors including instructions for performing operations according to the embodiments.

[0633] The operation according to the embodiment described in this specification can be performed by a sending / receiving device including one or more memories and / or one or more processors according to the embodiment. One or more memories can store programs for processing / controlling the operation according to the embodiment, and one or more processors can control the various operations described in this specification. One or more processors can be referred to as controllers, etc. In an embodiment, the operation can be performed by firmware, software, and / or a combination thereof. Firmware, software, and / or a combination thereof can be stored in a processor or a memory.

[0634] Terms such as first and second can be used to describe various elements of the embodiments. However, various components according to the embodiments should not be limited by the above terms. These terms are only used to distinguish one element from another element. For example, a first user input signal can be referred to as a second user input signal. Similarly, a second user input signal can be referred to as a first user input signal. The use of these terms should not be interpreted without departing from the scope of various embodiments. Both the first user input signal and the second user input signal are user input signals, but unless the context clearly stipulates otherwise, they do not mean the same user input signal.

[0635] The terms used to describe the embodiments are used only to describe specific embodiments and are not intended to limit the embodiments. As used in the description of the embodiments and the claims, unless the context clearly stipulates otherwise, the singular form includes plural referents. The expression "and / or" is used to include all possible combinations of terms. Terms such as "including" or "having" are intended to indicate the presence of graphics, quantities, steps, elements and / or components, and should be understood as not excluding the possibility of additional presence of graphics, quantities, steps, elements and / or components. As used herein, conditional expressions such as "if" and "when" are not limited to optional situations, and are intended to be interpreted as performing related operations or explaining related definitions according to specific conditions when specific conditions are met. The embodiments may include changes / modifications within the scope of the claims and their equivalents.

[0636] Invention Mode

[0637] As described above, the relevant contents are described in the best mode for carrying out the embodiment.

[0638] Industrial Applicability

[0639] As described above, the embodiments may be applied in whole or in part to point cloud data sending / receiving devices and systems. It is obvious to those skilled in the art that various changes or modifications may be made to the embodiments within the scope of the embodiments. Therefore, the embodiments are intended to cover modifications and variations as long as they fall within the scope of the attached claims and their equivalents.

Claims

1. A method for sending point cloud data, the method comprising: Acquire point cloud data including points by using a LiDAR device having a laser sensor; Encoding the point cloud data; as well as Sending encoded point cloud data and signaling data, The encoding includes: Separating the points in the point cloud data into roads and objects; constructing a first prediction tree based on the points separated into the roads, and constructing a second prediction tree based on the points separated into the objects; and The point cloud data is compressed based on predictions by applying the first prediction tree and the second prediction tree, wherein the signaling data includes information related to separation of the road and the object.

2. The method according to claim 1, wherein: Separating the points into the roads and the objects comprises: The points in the point cloud data are separated into the road and the object based on a base height corresponding to a distance from the road to a center of the laser sensor.

3. The method according to claim 2, wherein: The basic height is estimated based on angle information, radius information, and height information about a laser sensor having a largest negative angle among the laser sensors.

4. The method according to claim 2, wherein: Separating the points into the roads and the objects comprises: separating points existing at positions equal to or lower than the basic height among the points in the point cloud data as the roads; and Points existing at positions higher than the basic height among the points in the point cloud data are separated as the objects.

5. The method according to claim 1, wherein: Separating the points into the roads and the objects comprises: Remove the points within the minimum radius from the points acquired by the current laser sensor. wherein said separating is performed after said removing.

6. The method according to claim 1, wherein: The compression includes: The point cloud data is compressed based on intra-frame prediction by applying the first prediction tree and the second prediction tree.

7. The method according to claim 1, wherein: The compression includes: The point cloud data is compressed based on inter-frame prediction by applying the first prediction tree, the second prediction tree and at least one reference frame.

8. The method according to claim 7, wherein: The compression includes: Based on the current point in the current frame belonging to the road, selecting a point having the same laser ID and a similar azimuth as the current point from among the points in the area separated as the road in the at least one reference frame as a reference point, and setting the reference point as a predicted value of the current point; and Based on the fact that the current point in the current frame belongs to the object, a point having the same laser ID and similar azimuth as the current point is selected from the points in the area separated as the object in the at least one reference frame as a reference point, and the reference point is set as the predicted value of the current point.

9. A device for sending point cloud data, the device comprising: An acquirer configured to acquire point cloud data containing points through a LiDAR device having a laser sensor; an encoder configured to encode the point cloud data; and a transmitter configured to transmit the encoded point cloud data and signaling data, Wherein, the encoder is configured as: Separating the points in the point cloud data into roads and objects; constructing a first prediction tree based on the points separated into the roads, and constructing a second prediction tree based on the points separated into the objects; and The point cloud data is compressed based on predictions by applying the first prediction tree and the second prediction tree, wherein the signaling data includes information related to separation of the road and the object.

10. The device according to claim 9, wherein: The encoder separates the points in the point cloud data into the road and the object based on a base height corresponding to a distance from the road to a center of the laser sensor.

11. The device according to claim 10, wherein: The basic height is estimated based on angle information, radius information, and height information about a laser sensor having a largest negative angle among the laser sensors.

12. The device according to claim 10, wherein: The encoder separates the points among the points in the point cloud data that exist at a position equal to or lower than the basic height as the roads; and separates the points among the points in the point cloud data that exist at a position higher than the basic height as the objects.

13. The device according to claim 9, wherein: The encoder separates the points into the road and the object after removing points existing within a minimum radius among points acquired by a current laser sensor.

14. The device according to claim 9, wherein: The encoder compresses the point cloud data based on intra-frame prediction by applying the first prediction tree and the second prediction tree.

15. The device according to claim 9, wherein: The encoder compresses the point cloud data based on inter-frame prediction by applying the first prediction tree, the second prediction tree, and at least one reference frame.

16. The device according to claim 15, wherein: The encoder is configured as: Based on the current point in the current frame belonging to the road, selecting a point having the same laser ID and a similar azimuth as the current point from among the points in the area separated as the road in the at least one reference frame as a reference point, and setting the reference point as a predicted value of the current point; and Based on the fact that the current point in the current frame belongs to the object, a point having the same laser ID and similar azimuth as the current point is selected from the points in the area separated as the object in the at least one reference frame as a reference point, and the reference point is set as the predicted value of the current point.

Images