Point cloud encoding and decoding methods, systems, point cloud encoders and point cloud decoders

By decoupling the prediction mode determination from attribute information reconstruction in point cloud encoding, the process becomes more efficient, enhancing the encoding and decoding performance.

KR103047550B1Active Publication Date: 2026-11-11GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
KR1020237020518
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2020-12-28
Publication Date
2026-11-11
Estimated Expiration
2040-12-28

AI Technical Summary

Technical Problem

The coupling of prediction mode determination and attribute information reconstruction processes in point cloud encoding reduces the effectiveness of point cloud encoding efficiency.

Method used

Decouple the prediction mode determination from the attribute information reconstruction process by determining the prediction mode based on geometric information of a point and its neighboring points, predicting attribute information, obtaining a residual value, and encoding this residual value to generate a point cloud code stream.

Benefits of technology

This approach allows for parallel processing of prediction mode determination and attribute information reconstruction, thereby improving the efficiency of point cloud encoding and decoding.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 112023066609882-PCT00034_ABST
    Figure 112023066609882-PCT00034_ABST
Patent Text Reader

Abstract

The present application provides a point cloud encoding and decoding method, a system, a point cloud encoder, and a point cloud decoder. By determining the prediction mode of attribute information of a current point through geometric information of a point in a point cloud, the process of determining the prediction mode is decoupled from the process of reconstructing attribute information of a point in a point cloud, thereby allowing the two processes to be performed in parallel, thus improving the efficiency of encoding and decoding.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The present application relates to the field of point cloud encoding and decoding technology, and in particular to a point cloud encoding and decoding method, system, point cloud encoder, and point cloud decoder. Background Technology

[0002] Object surfaces are collected using collection devices to form point cloud data, which contains hundreds of thousands or more points. During the video production process, point cloud data is transmitted between point cloud encoding and decoding devices in the form of point cloud media files. However, since such a large number of points poses difficulties for transmission, point cloud encoding devices must compress the point cloud data before transmission.

[0003] Compression of point cloud data primarily involves the compression of geometric information and attribute information. When attribute information is compressed, redundant information in the point cloud data can be reduced or eliminated through prediction. For example, one or more adjacent points of the current point are obtained from the encoded point, and the attribute information of the current point is predicted based on the attribute information of the adjacent points.

[0004] Currently, by determining the prediction mode of the current point based on the reconstructed attribute information of neighboring points, the process of determining the prediction mode and the process of reconstructing attribute information are coupled, thereby reducing the effectiveness of point cloud encoding. The problem to be solved

[0005] Embodiments of the present application provide a point cloud encoding and decoding method, a system, a point cloud encoder, and a point cloud decoder, thereby improving point cloud encoding efficiency. means of solving the problem

[0006] In a first aspect, the present application provides a point cloud encoding method, and said method,

[0007] A step of obtaining geometric and attribute information of the current point in the point cloud;

[0008] A step of determining K neighbor points of the current point based on the geometric information of the current point - where K is a positive integer greater than or equal to 2 - ;

[0009] A step of determining a target prediction mode for attribute information of the current point based on the geometric information of the current point and the geometric information of the K neighboring points;

[0010] A step of predicting attribute information of the current point using the target prediction mode above, and obtaining a predicted value of the attribute information of the current point;

[0011] A step of obtaining a residual value of the attribute information of the current point according to the predicted value of the attribute information of the current point; and

[0012] It includes the step of encoding the residual value of the attribute information of the current point to generate a point cloud code stream.

[0013] In a second aspect, an embodiment of the present application provides a point cloud decoding method, and said method,

[0014] A step of decoding a point cloud code stream to obtain geometric information and attribute information of the current point in the point cloud;

[0015] A step of determining K neighbor points of the current point based on the geometric information of the current point - where K is a positive integer greater than or equal to 2 - ;

[0016] A step of determining a target prediction mode of attribute information of the current point according to the geometric information of the current point and the geometric information of the K neighboring points; and

[0017] It includes the step of predicting attribute information of the current point using the target prediction mode and obtaining a predicted value of the attribute information of the current point.

[0018] In a third aspect, the present application provides a point cloud encoder for executing a method in the first aspect or an implementation of the first aspect. Specifically, the encoder includes a functional unit for executing a method in the first aspect or each implementation of the first aspect.

[0019] In a fourth aspect, the present application provides a point cloud decoder for executing a method in the second aspect or each implementation of the second aspect. Specifically, the decoder includes a functional unit for executing a method in the second aspect or each implementation of the second aspect.

[0020] In a fifth aspect, a point cloud encoder comprising a processor and memory is provided. The memory is for storing a computer program, and the processor calls and operates the computer program stored in the memory to execute the method of the first aspect or each implementation of the first aspect.

[0021] In a sixth aspect, a point cloud decoder comprising a processor and memory is provided. The memory is for storing a computer program, and the processor calls and operates the computer program stored in the memory to execute the method of the second aspect or each implementation of the second aspect.

[0022] In a seventh aspect, a point cloud encoding and decoding system comprising a point cloud encoder and a point cloud decoder is provided. The point cloud encoder is for executing a method in the first aspect or each implementation of the first aspect, and the point cloud decoder is for executing a method in the second aspect or each implementation of the second aspect.

[0023] In the eighth aspect, a chip is provided for implementing a method in any one of the first and second aspects or in each implementation method thereof. Specifically, the chip includes a processor that calls and operates a computer program from memory to enable a device on which the chip is installed to implement a method in any one of the first and second aspects or in each implementation method thereof.

[0024] In the ninth aspect, a computer-readable storage medium for storing a computer program is provided, and said computer program causes a computer to execute any one of the aspects of the first and second aspects or a method in each implementation thereof.

[0025] In the tenth aspect, a computer program product comprising computer program instructions is provided, said computer program instructions cause a computer to execute any one of the aspects in the first and second aspects or a method in each implementation method thereof.

[0026] In the eleventh aspect, a computer program is provided, and when said computer program is operated on a computer, the computer is made to execute any one of the aspects in the first and second aspects or a method in each implementation thereof. Effects of the invention

[0027] Based on the above technical method, by determining the prediction mode of the attribute information of the current point through the geometric information of the point in the point cloud, the process of determining the prediction mode is decoupled from the process of reconstructing the attribute information of the point in the point cloud, so that the two processes can be performed in parallel, thereby improving the efficiency of encoding and decoding. Brief explanation of the drawing

[0028] FIG. 1 is an exemplary block diagram of a point cloud encoding and decoding system (100) mentioned in an embodiment of the present application. FIG. 2 is an exemplary block diagram of a point cloud encoder (200) provided in an embodiment of the present application. FIG. 3 is an exemplary block diagram of a decoder (300) provided in an embodiment of the present application. FIG. 4 is a partial block diagram of the attribute encoding module (400) mentioned in the embodiment of the present application. FIG. 5 is a partial block diagram of the attribute decoding module (500) mentioned in the embodiment of the present application. Figure 6 is a predicted example diagram of an encoding unit of one embodiment. Figure 7 is a predicted example diagram of the decoding unit of one embodiment. FIG. 8 is a flowchart of the point cloud encoding method (600) provided in an embodiment of the present application. FIG. 9 is a flowchart of the point cloud encoding method (600a) provided in an embodiment of the present application. FIG. 10 is a flowchart of the point cloud encoding method (700) provided in an embodiment of the present application. FIG. 11 is a flowchart of a point cloud decoding method (800) provided in an embodiment of the present application. FIG. 12 is a flowchart of a point cloud decoding method (800a) provided in an embodiment of the present application. FIG. 13 is a flowchart of a point cloud decoding method (900) provided in an embodiment of the present application. FIG. 14 is an exemplary block diagram of a point cloud encoder (10) provided in an embodiment of the present application. FIG. 15 is an exemplary block diagram of a point cloud decoder (20) provided in an embodiment of the present application. FIG. 16 is an exemplary block diagram of an electronic device (30) provided in an embodiment of the present application. FIG. 17 is an exemplary block diagram of a point cloud encoding and decoding system (40) provided in an embodiment of the present application. Specific details for implementing the invention

[0029] This application can be applied to the field of point cloud compression technology.

[0030] To facilitate understanding of the embodiments of the present application, the relevant concepts mentioned in the embodiments of the present application are first briefly introduced as follows.

[0031] A point cloud refers to a set of discrete points in a group that are irregularly distributed in space and represent the spatial structure and surface properties of a three-dimensional object or three-dimensional scenario.

[0032] Point cloud data is a specific record form of a point cloud, and a point in the point cloud may include location information of the point and attribute information of the point. For example, the location information of the point may be three-dimensional coordinate information of the point. The location information of the point may also be referred to as geometric information of the point. For example, the attribute information of the point may include at least one of color information and reflectance, etc. For example, the color information may be information in any one color space. For example, the color information may be (RGB). Also, for example, the color information may be information in luminance chromaticity (YcbCr, YUV). For example, Y represents Luma, Cb (U) represents blue color difference, Cr (V) represents red, and U and V represent Chroma to explain color difference information. For example, in a point cloud obtained according to the laser measurement principle, a point in the point cloud may include three-dimensional coordinate information of the point and the laser reflection intensity (reflectance) of the point. For example, in a point cloud obtained according to a shooting measurement principle, the points in the point cloud may include three-dimensional coordinate information of the points and color information of the points. For example, in a point cloud obtained by combining laser measurement and shooting measurement principles, the points in the point cloud may include three-dimensional coordinate information of the points, laser reflection intensity (reflectance) of the points, and color information of the points.

[0033] The acquisition path for point cloud data may include, but is not limited to, at least one of the following acquisition paths: (1) It is generated by a computer device. A computer device can generate point cloud data based on virtual 3D objects and virtual 3D scenarios. (2) It is acquired through 3D laser scanning. Point cloud data of static real-world 3D objects or 3D scenarios can be acquired through 3D laser scanning, and millions of point cloud data can be acquired every second. (3) It is acquired through 3D imaging measurement. Point cloud data of real-world visual scenarios can be acquired by collecting real-world visual scenarios through a 3D imaging device (i.e., a camera device equipped with a group of cameras or multiple lenses and sensors), and point cloud data of dynamic real-world 3D objects or 3D scenarios can be acquired through 3D imaging. (4) Point cloud data of biological tissues and organs is acquired through a medical device. In the medical field, point cloud data of biological tissues and organs can be acquired through medical devices such as Magnetic Resonance Imaging (MRI), Computed Tomography (CT), and electromagnetic positioning information.

[0034] Point clouds can be classified into dense set point clouds and sparse point clouds depending on the acquisition path.

[0035] Point clouds are divided into a first type static point cloud, a second type dynamic point cloud, and a third type dynamic acquired point cloud according to the sequence type of the data.

[0036] In the first type static point cloud, the object is stationary, and the device acquiring the point cloud is also stationary.

[0037] In the second type of dynamic point cloud, the object is in motion, but the device acquiring the point cloud is stationary.

[0038] In the third type dynamic acquisition point cloud, the device acquiring the point cloud is in motion.

[0039] Point clouds are divided into two categories based on their use.

[0040] Category 1 is a machine detection point cloud that can be used in scenarios such as automatic navigation systems, real-time patrol systems, geographic information systems, visual classification robots, and emergency rescue robots.

[0041] Category 2 is a visually detected point cloud that can be used in point cloud application scenarios such as digital cultural heritage, free viewpoint broadcasting, 3D immersive communication, and 3D immersive interaction.

[0042] FIG. 1 is an exemplary block diagram of a point cloud encoding and decoding system (100) described in an embodiment of the present application. It should be noted that FIG. 1 is merely an example, and the point cloud encoding and decoding system of an embodiment of the present application includes, but is not limited to, that shown in FIG. 1. As shown in FIG. 1, the point cloud encoding and decoding system (100) includes an encoding device (110) and a decoding device (120). Here, the encoding device is intended to generate a code stream by encoding (which can be understood as compression) point cloud data and to transmit the code stream to the decoding device. The decoding device decodes the code stream generated by encoding by the encoding device to obtain the decoded point cloud data.

[0043] The encoding device (110) of the embodiment of the present application can be understood as a device equipped with a point cloud encoding function, and the decoding device (120) can be understood as a device equipped with a point cloud decoding function. That is, the encoding device (110) and the decoding device (120) of the embodiment of the present application include a wider range of devices, such as a smartphone, a desktop computer, a mobile computing device, a notebook (e.g., laptop) computer, a tablet computer, a set-top box, a TV, a camera, a display device, a digital media player, a point cloud game console, a vehicle computer, etc.

[0044] In some embodiments, the encoding device (110) may transmit encoded point cloud data (e.g., a code stream) to the decoding device (120) via the channel (130). The channel (130) may include at least one of one or more media and devices capable of transmitting the encoded point cloud data from the encoding device (110) to the decoding device (120).

[0045] In one example, the channel (130) includes one or more communication media that enable the encoding device (110) to transmit the encoded point cloud data directly in real time to the decoding device (120). In this example, the encoding device (110) can modulate the encoded point cloud data according to a communication standard and transmit the modulated point cloud data to the decoding device (120). Here, the communication media includes a wireless communication media such as a radio frequency spectrum, and optionally, the communication media may also include a wired communication media such as one or more physical transmission lines.

[0046] In another example, the channel (130) includes a storage medium, and the storage medium can store point cloud data encoded by the encoding device (110). The storage medium includes various local access data storage media such as optical discs, DVDs, flash memory, etc. In the above example, the decoding device (120) can obtain the encoded point cloud data from the storage medium.

[0047] In another example, the channel (130) may include a storage server, said storage server may store point cloud data encoded by the encoding device (110). In this example, the decoding device (120) may download the encoded point cloud data stored from said storage server. Optionally, said storage server may store the encoded point cloud data and send said encoded point cloud data to a decoding device (120), such as a web server (e.g., used on a website), a file transfer protocol (FTP) server, etc.

[0048] In some embodiments, the encoding device (110) includes a point cloud encoder (112) and an output interface (113). Here, the output interface (113) may include at least one of a modulator / demodulator (modem) and an emitter.

[0049] In some embodiments, the encoding device (110) may include a point cloud encoder (112) and an input interface (113), in addition to including a point cloud source (111).

[0050] The point cloud source (111) may include at least one of a point cloud collection device (e.g., a scanner), a point cloud archive, a point cloud input interface, and a computer graphics system, wherein the point cloud input interface is for receiving point cloud data from a point cloud content provider location, and the computer graphics system is for generating point cloud data.

[0051] The point cloud encoder (112) encodes point cloud data from the point cloud source (111) to generate a code stream. The point cloud encoder (112) transmits the encoded point cloud data directly to the decoding device (120) via the output interface (113). The encoded point cloud data can also be stored on a storage medium or storage server and provided to the decoding device (120) for future reading.

[0052] In some embodiments, the decoding device (120) includes an input interface (121) and a point cloud decoder (122).

[0053] In some embodiments, the decoding device (120) may include, in addition to including an input interface (121) and a point cloud decoder (122), also a display device (123).

[0054] Here, the input interface (121) includes at least one of a receiver and a modem. The input interface (121) can receive encoded point cloud data through the channel (130).

[0055] The point cloud decoder (122) is for decoding encoded point cloud data, obtaining decoded point cloud data, and transmitting the decoded point cloud data to a display device (123).

[0056] The display device (123) displays decoded point cloud data. The display device (123) may be connected to the decoding device (120) or may be located outside the decoding device (120). The display device (123) may include various display devices such as a liquid crystal display (LCD), a plasma display, an organic light-emitting diode (OLED) display, or other types of display devices.

[0057] In addition, FIG. 1 is merely an example, and the technical method of the embodiment of the present application is not limited to FIG. 1, and for example, the technology of the present application can also be applied to single-sided point cloud encoding or single-sided point cloud decoding.

[0058] Current point cloud encoders may use the geometry point cloud compression (G-PCC) encoding and decoding framework provided by the Moving Picture Experts Group (MPEG) or the video point cloud compression (V-PCC) encoding and decoding framework, and may also use the AVS-PCC encoding and decoding framework provided by the Audio Video Standard (AVS). Both G-PCC and AVS-PCC are for static sparse point clouds, and their encoding frameworks are largely the same. The G-PCC encoding and decoding framework may be for compressing a first type static point cloud and a third type dynamic acquired point cloud, and the V-PCC encoding and decoding framework may be for compressing a second type dynamic point cloud. The G-PCC encoding and decoding framework may be referred to as the point cloud codec TMC13, and the V-PCC encoding and decoding framework may be referred to as the point cloud codec TMC2.

[0059] Below, using the G-PCC encoding and decoding framework as an example, point cloud encoders and point cloud decoders to which the embodiments of the present application can be applied are described.

[0060] FIG. 2 is an exemplary block diagram of a point cloud encoder (200) provided in an embodiment of the present application.

[0061] As can be seen from the foregoing, since a point in a point cloud may include location information and attribute information, the encoding of a point in a point cloud primarily includes location encoding and attribute encoding. In some examples, the location information of a point in a point cloud is also referred to as geometric information, and the corresponding location encoding of a point in a point cloud may also be referred to as geometric encoding.

[0062] The process of position encoding includes performing preprocessing on points in a point cloud, such as coordinate transformation, quantization, and removal of duplicate points; and then performing geometric encoding on the preprocessed point cloud, such as constructing an octree and performing geometric encoding based on the constructed octree to form a geometric code stream. At the same time, based on the position information output by the constructed octree, the position information of each point in the point cloud data is reconstructed to obtain the reconstructed value of the position information of each point.

[0063] The attribute encoding process includes selecting one of three prediction modes to perform point cloud prediction using the reconstruction information of location information and the raw values ​​of attribute information of a given input point cloud, quantizing the predicted result, and performing arithmetic encoding to form an attribute code stream.

[0064] As shown in FIG. 2, the position encoding is,

[0065] It can be implemented through a Tanmsform coordinates unit (201), a Quantize and remove points unit (202), an Analyze octree unit (203), a Reconstruct geometry unit (204), and a first Arithmetic Encoding unit (205).

[0066] The coordinate conversion unit (201) may be for converting the world coordinates of a point in a point cloud into relative coordinates. For example, to implement converting the coordinates of a point in a point cloud from world coordinates to relative coordinates, the minimum values ​​of the xyz coordinate axes are each subtracted from the geometric coordinates of the point, which corresponds to a DC removal operation.

[0067] The quantization and duplicate point removal unit (202) can reduce the number of coordinates through quantization; after quantization, originally different points can be assigned the same coordinates, and based on this, duplicate points can be deleted through a duplicate removal operation; for example, multiple clouds having the same quantization location and different attribute information can be merged into one cloud through attribute conversion. In some embodiments of the present application, the quantization and duplicate point removal unit (202) is an optional unit module.

[0068] The octree analysis unit (203) can encode location information of quantized points using an octree encoding method. For example, the point cloud is divided according to the shape of the octree, so that the location of the point can correspond to the location of the octree one by one, and geometric encoding is performed by statistically analyzing the location of the point within the octree and marking their flags as 1.

[0069] The geometric reconstruction unit (204) performs location reconstruction based on location information output by the octree analysis unit (203) to obtain location information and reconstruction values ​​for each point in the point cloud data.

[0070] The first arithmetic encoding unit (205) can perform arithmetic encoding on the position information output by the octree analysis unit (203) using an entropy encoding method, that is, generate a geometric code stream using an arithmetic encoding method on the position information output by the octree analysis unit (203); the geometric code stream may also be referred to as a geometry bitstream.

[0071] Attribute encoding is,

[0072] It can be implemented through a color space conversion (Transform colors) unit (210), an attribute transfer unit (211), a Region Adaptive Hierarchical Transform (RAHT) unit (212), a predicting transform unit (213), a lifting transform unit (214), a quantize coefficients unit (215), and a second arithmetic encoding unit (216).

[0073] It should be explained that the point cloud encoder (200) may include more, fewer, or different functional components than in FIG. 2.

[0074] The color space conversion unit (210) may be for converting the RGB color space of a point in the point cloud into a YCbCr format or another format.

[0075] The attribute conversion unit (211) may be intended to convert attribute information of a point in a point cloud to minimize attribute distortion. For example, the attribute conversion unit (211) may be intended to obtain the raw value of the attribute information of the point. For example, the attribute information may be the color information of the point.

[0076] After obtaining the raw value of the attribute information of the point through the attribute conversion unit (211), one of the prediction units can be selected to predict the point in the point cloud. The prediction unit may include a RAHT (212), a predicting transform unit (213), and a lifting transform unit (214).

[0077] In other words, any one of the RAHT (212), the predicting transform unit (213), and the lifting transform unit (214) may be for predicting attribute information of a point in the point cloud and obtaining a predicted value of attribute information of the point, and thereby obtaining a residual value of attribute information of the point based on the predicted value of attribute information of the point. For example, the residual value of attribute information of the point may be a predicted value obtained by subtracting attribute information of the point from the raw value of attribute information of the point.

[0078] In one embodiment of the present application, the predictive transformation unit (213) may also be for generating a level of detail (LOD). The process of generating the LOD includes obtaining the Euclidean distance between points according to the location information of points in the point cloud; and dividing the points into different detail representation layers according to the Euclidean distance. In one embodiment, after arranging the Euclidean distances, different ranges of Euclidean distances may be divided into different detail representation layers. For example, one point may be randomly selected to be the first detail representation layer. Next, the Euclidean distance between the remaining points and the said point is calculated, and the point whose Euclidean distance meets the first threshold requirement is classified into the second detail representation layer. The center of mass of the point in the second detail representation layer is obtained, and the Euclidean distance between the center of mass and points other than the first detail representation layer and the second detail representation layer is calculated, and the point whose Euclidean distance meets the second threshold is classified into the third detail representation layer. In this way, all points are classified into detail representation layers. By adjusting the threshold of the Euclidean distance, the number of points in each LOD layer can be gradually increased. It should be understood that other methods of LOD division may also be used, and this application is not limited thereto.

[0079] It needs to be explained that a point cloud can be directly divided into one or more detail representation layers, or the point cloud can first be divided into multiple point cloud slices, and then each point cloud slice can be divided into one or more LOD layers.

[0080] For example, a point cloud can be divided into multiple point cloud slices, and the number of points in each point cloud slice can be between 550,000 and 1,100,000. Each point cloud slice can be considered as a separate point cloud. Each point cloud slice can also be divided into multiple detail representation layers, and each detail representation layer includes multiple points. In one embodiment, the division of the detail representation layers can be performed according to the Euclidean distance between points.

[0081] The quantization unit (215) may be for quantizing residual values ​​of attribute information of a point. For example, if the quantization unit (215) and the prediction transformation unit (213) are connected to each other, the quantization unit may be for quantizing residual values ​​of attribute information of a point output by the prediction transformation unit (213).

[0082] For example, quantization is performed on the residual value of the attribute information of the point output by the prediction conversion unit (213) using a quantization step size to improve system performance.

[0083] The second arithmetic encoding unit (216) can obtain an attribute code stream by performing entropy encoding on the residual value of the attribute information of the point using zero run length coding. The attribute code stream may be bit stream information.

[0084] FIG. 3 is an exemplary block diagram of a decoder (300) provided in an embodiment of the present application.

[0085] As illustrated in FIG. 3, the decoding framework (300) can obtain a point cloud code stream from an encoding device and obtain location information and attribute information of points in the point cloud by parsing the code. Decoding of the point cloud includes location decoding and attribute decoding.

[0086] The process of position decoding includes performing arithmetic decoding on a geometric code stream; constructing an octree and then performing merging to reconstruct the position information of a point and obtain reconstructed position information of the point; and performing a coordinate transformation on the reconstructed position information of the point to obtain the position information of the point. The position information of the point may also be referred to as the geometric information of the point.

[0087] The attribute decoding process includes obtaining residual values ​​of attribute information of points in a point cloud by parsing an attribute code stream; obtaining inversely quantized residual values ​​of attribute information of points by performing inverse quantization on the residual values ​​of attribute information of points; performing point cloud prediction by selecting one of three prediction modes, RAHT, prediction change, and lifting change, based on the reconstruction information of position information of points obtained in the position decoding process, to obtain a prediction value, and obtaining a reconstruction value of attribute information of points by adding residual values ​​to the prediction value; and obtaining a decoded point cloud by performing color space inversion on the reconstruction value of attribute information of points.

[0088] As shown in FIG. 3, position decoding is,

[0089] It can be implemented through a first arithmetic decoding unit (301), an octree analysis (synthesize octree) unit (302), a geometric reconstruction (Reconstruct geometry) unit (303), and an inverse transform coordinates unit (304).

[0090] Attribute encoding is,

[0091] It can be implemented through a second arithmetic decoding unit (310), an inverse quantize unit (311), a RAHT unit (312), a predicting transform unit (313), a lifting transform unit (314), and an inverse transform colors unit (315).

[0092] It should be explained that decompression is the reverse process of compression, and similarly, the function of each unit in the decoder (300) can refer to the function of the corresponding unit in the encoder (200). Additionally, the point cloud decoder (300) may include more, fewer, or different functional components than in FIG. 3.

[0093] For example, the decoder (300) divides the point cloud into multiple LODs based on the Euclidean distance between points in the point cloud; then, sequentially decodes the attribute information of the points in the LODs; for example, calculates the number of zeros (zero_cnt) in zero-run length encoding techniques and decodes the residuals based on zero_cnt; subsequently, the decoding framework (200) performs inverse quantization based on the decoded residual values, and obtains a reconstructed value of the point cloud by applying a predicted value of the current point based on the inversely quantized residual values ​​until all point clouds are decoded. The current point is set as the closest point to the point in the future LODs, and attribute information of the future point is predicted using the reconstructed value of the current point.

[0094] As can be seen from FIG. 2 above, the point cloud encoder (200) functionally comprises two parts, mainly a position encoding module and an attribute encoding module, wherein the position encoding module is intended to form a geometric code stream by implementing the encoding of position information of the point cloud, and the attribute encoding module is intended to form an attribute code stream by implementing the encoding of attribute information of the point cloud. The present application is mainly related to the encoding of attribute information, and the attribute encoding module in the point cloud encoder mentioned in the present application is introduced below in conjunction with FIG. 4.

[0095] FIG. 4 is a partial block diagram of an attribute encoding module (400) mentioned in an embodiment of the present application, and the attribute encoding module (400) can be understood as a unit for implementing attribute information encoding among the point cloud encoder (200) shown in FIG. 2. As shown in FIG. 4, the attribute encoding module (400) includes a preprocessing unit (410), a residual unit (420), a quantization unit (430), a prediction unit (440), an inverse quantization unit (450), a reconstruction unit (460), a filtering unit (470), a decoding buffer unit (480), and an encoding unit (490). It should be noted that the attribute encoding module (400) may also include more, fewer, or different functional components.

[0096] In some embodiments, the preprocessing unit (410) may include the color space conversion unit (210) and attribute conversion unit (211) shown in FIG. 2.

[0097] In some embodiments, the quantization unit (430) can be understood as the quantization coefficient unit (215) in FIG. 2, and the encoding unit (490) can be understood as the second arithmetic encoding unit (216) in FIG. 2.

[0098] In some embodiments, the prediction unit (440) may include the RAHT (212), the prediction change unit (213), and the lifting change unit (214) shown in FIG. 2. Specifically, the prediction unit (440) is intended to obtain reconstruction information of the location information of a point in a point cloud, and based on the reconstruction information of the location information of the point, predict attribute information of the point from any one of the point clouds of the RAHT (212), the prediction change unit (213), and the lifting change unit (214) to obtain a predicted value of the attribute information of the point.

[0099] The residual unit (420) can obtain a residual value of the attribute information of a point in the point cloud based on the raw value of the attribute information of a point in the point cloud and the reconstructed value of the attribute information, for example, by subtracting the reconstructed value of the attribute information from the raw value of the attribute information of the point to obtain a residual value of the attribute information of the point.

[0100] The quantization unit (430) can quantize residual values ​​of attribute information, and specifically, the quantization unit (430) can quantize residual values ​​of attribute information of a point based on quantization parameter (QP) values ​​associated with the point cloud. The point cloud encoder can adjust the degree of quantization applied to the point by adjusting the QP values ​​associated with the point cloud.

[0101] The inverse quantization unit (450) can apply inverse quantization to the residual values ​​of the quantized attribute information to reconstruct the residual values ​​of the attribute information from the residual values ​​of the quantized attribute information.

[0102] The reconstruction unit (460) can generate a reconstruction value of the attribute information of a point in the point cloud by adding the residual value of the reconstructed attribute information to the prediction value generated by the prediction unit (440).

[0103] The filtering unit (470) can eliminate or reduce noise in the reconstruction operation.

[0104] The decoding buffer unit (480) can store a reconstructed value of attribute information of a point in the point cloud. The prediction unit (440) can predict attribute information of another point using the reconstructed value of attribute information of the point.

[0105] As can be seen from FIG. 3 above, the point cloud decoder (300) functionally comprises two parts, mainly a position decoding module and an attribute decoding module, wherein the position decoding module is intended to obtain position information of a point by implementing decoding of the geometric code stream of the point cloud, and the attribute decoding module is intended to obtain attribute information of a point by implementing decoding of the attribute code stream of the point cloud. Below, in combination with FIG. 5, the attribute decoding module in the point cloud decoder mentioned in the present application is introduced.

[0106] FIG. 5 is a partial block diagram of an attribute decoding module (500) mentioned in an embodiment of the present application, and the attribute decoding module (500) can be understood as a unit for implementing the decoding of an attribute code stream among the point cloud decoder (300) shown in FIG. 3. As shown in FIG. 5, the attribute decoding module (500) includes a decoding unit (510), a prediction unit (520), an inverse quantization unit (530), a reconstruction unit (540), a filtering unit (550), and a decoding buffer unit (560). It should be noted that the attribute decoding module (500) may include more, fewer, or different functional components.

[0107] The attribute decoding module (500) can receive an attribute code stream. The decoding unit (510) can parse the attribute code stream to extract syntax elements from the attribute code stream. As part of parsing the attribute code stream, the decoding unit (510) can parse syntax elements after encoding in the attribute code stream. The prediction unit (520), the inverse quantization unit (530), the reconstruction unit (540), and the filtering unit (550) can decode attribute information according to the syntax elements extracted from the attribute code stream.

[0108] In some embodiments, the prediction unit (520) determines the prediction mode of a point based on one or more syntax elements parsed from a code stream, and can predict attribute information of the point using the determined prediction mode.

[0109] The inverse quantization unit (530) can obtain the residual value of the attribute information of a point by inversely quantizing (i.e., anti-quantizing) the residual value of the quantized attribute information associated with a point in the point cloud. The inverse quantization unit (530) can determine the degree of quantization using the QP value associated with the point cloud.

[0110] The reconstruction unit (540) reconstructs the attribute information of a point in the point cloud using the residual value of the attribute information of a point in the point cloud and the predicted value of the attribute information of a point in the point cloud. For example, the reconstruction unit (540) can obtain a reconstructed value of the attribute information of a point by adding the residual value of the attribute information of a point in the point cloud to the predicted value of the attribute information of the point.

[0111] The filtering unit (550) can eliminate or reduce noise in the reconstruction operation.

[0112] The attribute decoding module (500) can store a reconstructed value of attribute information of a point in the point cloud in the decoding buffer unit (560). The attribute decoding module (500) can use the reconstructed value of attribute information in the decoding buffer unit (560) as a reference point for future prediction, or transmit the reconstructed value of attribute information to a display device to display it.

[0113] The basic process for encoding and decoding attribute information of a point cloud is as follows. In the encoding stage, preprocessing is performed on the attribute information of the point cloud data to obtain the raw value of the attribute information of the point in the point cloud. The prediction unit (410) selects one of the three prediction methods based on the reconstructed value of the location information of the point in the point cloud to predict the attribute information of the point in the point cloud and obtains the predicted value of the attribute information. The residual unit (420) can calculate the residual value of the attribute information based on the raw value of the attribute information of the point in the point cloud and the predicted value of the attribute information; that is, the difference between the raw value of the attribute information of the point in the point cloud and the predicted value of the attribute information is taken as the residual value of the attribute information of the point in the point cloud. The residual value is quantized by the quantization unit (430) to remove information that is not sensitive to the naked eye, thereby eliminating visual redundancy. The encoding unit (490) receives the residual value of the quantized attribute information output by the quantization unit (430), and can encode the residual value of the quantized attribute information to output an attribute code stream.

[0114] Additionally, the inverse quantization unit (450) may receive the residual value of the quantized attribute information output by the quantization unit (430), and inverse quantize the residual value of the quantized attribute information to obtain the residual value of the attribute information of a point in the point cloud. The reconstruction unit (460) obtains the residual value of the attribute information of a point in the point cloud output by the inverse quantization unit (450) and the predicted value of the attribute information of a point in the point cloud output by the prediction unit (410), and adds the predicted value to the residual value of the attribute information of a point in the point cloud to obtain the reconstruction value of the attribute information of the point. The reconstruction value of the attribute information of the point is filtered through the filtering unit (470) and then buffered in the decoding buffer unit (480) to be used in the prediction process of another point in the future.

[0115] In the decoding unit, the decoding unit (510) parses the attribute code stream to obtain residual values, prediction information, quantization coefficients, etc., of the quantized attribute information of a point in the point cloud, and the prediction unit (520) predicts the attribute information of a point in the point cloud based on the prediction information to generate a predicted value of the attribute information of the point. The inverse quantization unit (530) uses the quantization coefficients obtained from the attribute code stream to inverse quantize the residual values ​​of the quantized attribute information of the point to obtain residual values ​​of the attribute information of the point. The reconstruction unit (440) adds the residual values ​​to the predicted values ​​of the attribute information of the point to obtain a reconstructed value of the attribute information of the point. The filtering unit (550) filters the reconstructed value of the attribute information of the point to obtain decoded attribute information.

[0116] It must be explained that when attribute information of the encoding unit is encoded, mode information or parameter information, such as prediction, quantization, encoding, and filtering, is carried into the attribute code stream if necessary. By parsing the attribute code stream and analyzing it based on existing information, the decoding unit determines the same mode information or parameter information as the encoding unit, such as prediction, quantization, encoding, and filtering, thereby guaranteeing that the reconstructed value of the attribute information obtained by the encoding unit is identical to the reconstructed value of the attribute information obtained by the decoding unit.

[0117] The foregoing describes the basic process of a point cloud codec based on a G-PCC encoding and decoding framework. As technology advances, some modules or steps of the framework or process may be optimized. This application applies to the basic process of a point cloud codec based on the G-PCC encoding and decoding framework, but is not limited to the framework and process.

[0118] FIG. 6 is an example diagram of an encoding unit prediction of one embodiment, and as shown in FIG. 6, it includes the following steps.

[0119] In step S61, a reconstructed value of attribute information of a neighbor point of the current point, such as a color value, is obtained, and based on the reconstructed value of the neighbor point's attribute information, the maximum color difference maxDiff of the neighbor point is calculated and obtained.

[0120] Specifically, it is assumed that the current point has three neighboring points, namely Point 1, Point 2, and Point 3, where the color value of Point 1 is (R1, G1, B1), the color value of Point 2 is (R2, G2, B2), and the color value of Point 3 is (R3, G3, B3). The maximum difference between the three neighboring points in the R, G, and B portions is calculated, and the maximum difference value in the R, G, and B portions is selected and set as maxDiff.

[0121] For example, according to the following formula (1), the maximum color difference maxDiff corresponding to the three neighboring points is obtained.

[0122] (1)

[0123] In step S62, maxDiff is compared with a preset threshold.

[0124] In step S63, if maxDiff is less than the threshold, the target prediction mode of the current point is determined to be a single-pred mode, and the current point is predicted using a weighted average method, for example, the weighted average of the reconstructed values ​​of the attribute information of the three neighboring points is used as the predicted value of the attribute information of the current point.

[0125] In step S64, if maxDiff is greater than or equal to the threshold, the target prediction mode of the current point is determined to be a multi-pred mode. Specifically, three neighbor point prediction variables are obtained by using the reconstructed value of the attribute information of each of the three neighbor points as one prediction variable, and the weighted average value of the reconstructed value of the attribute information of the three neighbor points is used as another prediction variable and denoted as the weighted average prediction variable, thereby obtaining 3+1 prediction variables as shown in the table in FIG. 6, and an index is set for each prediction variable.

[0126] In step S65, a score corresponding to each of the 3+1 predictor variables is calculated, and for example, an RDO value corresponding to each of the 3+1 predictor variables is calculated using a Rate Distortion Optimization (RDO) technique.

[0127] In step S66, the predictor variable with the smallest score is set as the optimal predictor variable for the current point, and entropy encoding is performed on the information of the optimal predictor variable.

[0128] FIG. 7 is an example diagram of a decoding unit prediction of one embodiment, and as shown in FIG. 7, it includes the following steps.

[0129] In step S71, the code stream is decoded to obtain attribute information of the neighbor point of the current point, and the attribute information of the neighbor point is decoded to obtain a reconstructed value of the attribute information of the neighbor point; based on the reconstructed value of the attribute information of the neighbor point, the maximum color difference maxDiff of the neighbor point is calculated and obtained, and for specific details, refer to step 2 in FIG. 6 above.

[0130] In step S72, maxDiff is compared with a preset threshold.

[0131] In step S73, if maxDiff is less than the threshold, the target prediction mode of the current point is determined to be a single-pred mode, and the current point is predicted using a weighted average method, for example, the weighted average of the reconstructed values ​​of the attribute information of the three neighboring points is used as the predicted value of the attribute information of the current point.

[0132] In step S74, if maxDiff is greater than or equal to the threshold, the target prediction mode of the current point is determined to be multi-pred mode.

[0133] In step S75, the code stream is parsed to obtain information on the predictor variable with the smallest score carried in the code stream, and the predictor variable corresponding to the information on the predicted variable is set as the predicted value of the current point.

[0134] As can be seen from Figures 6 and 7 above, when determining the predicted value of the attribute information of the current point, the maximum color difference maxDiff of the neighboring points of the current point must be calculated, and based on maxDiff, it must be determined whether the current point is predicted using a single-pred mode or a multi-pred mode. This determination process depends on the reconstructed value of the attribute information of the neighboring points of the current point, and this dependency prevents the process of determining the prediction mode and the process of reconstructing the attribute information from being performed in parallel, thereby reducing the efficiency of point cloud encoding and decoding.

[0135] To solve the above technical problem, the present application determines the prediction mode of attribute information of the current point through geometric information of the point in the point cloud, thereby allowing the process of determining the prediction mode to be decoupled from the process of reconstructing attribute information of the point in the point cloud, so that the two processes can be performed in parallel, thereby improving the efficiency of encoding and decoding.

[0136] The technical solution provided in one embodiment of the present application is described in detail below by combining specific embodiments.

[0137] Below, the encoding unit is introduced by combining Fig. 8.

[0138] FIG. 8 is a flowchart illustrating a point cloud encoding method (600) provided in an embodiment of the present application, and the embodiment of the present application is applied to the point cloud encoder shown in FIG. 1, FIG. 2 and FIG. 4. As shown in FIG. 8, the method of the embodiment of the present application includes the following steps.

[0139] In step S601, geometric information and attribute information of the current point in the point cloud are obtained.

[0140] A point cloud includes multiple points, and each point may include geometric information of the point and attribute information of the point. Here, the geometric information of the point may be referred to as location information of the point, and the location information of the point may be three-dimensional coordinate information of the point. Here, the attribute information of the point may include at least one of color information and reflectance, etc.

[0141] Here, the current point can be understood as one point in the point cloud currently to be encoded, and in some embodiments, the current point may be referred to as the target point.

[0142] In one example, the attribute information of the current point may be the raw attribute information of the current point.

[0143] In another example, as illustrated in FIG. 2, the point cloud encoder acquires raw attribute information of the current point and then performs color space conversion on the raw attribute information, for example, converting the RGB color space of the current point to a YCbCr format or another format. The attribute information of the current point is acquired by performing attribute conversion on the color space-converted current point to minimize attribute distortion.

[0144] In step S602, K neighbor points of the current point are determined based on the geometric information of the current point, and K is a positive integer greater than or equal to 2.

[0145] What needs to be explained is that the point cloud encoder completes the encoding of geometric information of points in the point cloud and then encodes attribute information of points in the point cloud. Referring to the illustration in FIG. 2, in the process of encoding geometric information, the first arithmetic encoding unit (205) encodes the geometric information of points processed by the octree unit (203) to form a geometric code stream, and in addition, the geometric reconstruction unit (204) reconstructs the geometric information of points processed by the octree unit (203) to obtain a reconstructed value of the geometric information.

[0146] In some embodiments, depending on the reconstructed geometric information of the current point, K neighbor points of the current point can be obtained from already encoded points in the point cloud.

[0147] For example, the geometric information (i.e., geometric coordinates) of a point in a point cloud is converted into a Morton code; specifically, a fixed value (j1, j2, j3) is applied to the geometric information (x, y, z) of the point in the point cloud to generate a Morton code for the point with new coordinates (x+j1, y+j2, z+j3). Points in the point cloud are arranged according to the Morton code of the point to obtain the Morton order of the point cloud. According to the Morton order of the point cloud, K neighbor points of the current point are obtained from the already encoded point in the point cloud, where K is a positive integer greater than or equal to 2.

[0148] In step S603, the target prediction mode of the attribute information of the current point is determined based on the geometric information of the current point and the geometric information of K neighboring points.

[0149] The prediction modes for attribute information of points in a point cloud include multi-pred and single-pred modes.

[0150] Here, the multi-pred mode includes K+1 predictor variables, where K is the number of neighbor points of the current point. K predictor variables are obtained by using the reconstructed value of the attribute information of each of the K neighbor points as one predictor variable, and said predictor variables may also be referred to as neighbor point predictor variables. The weighted average value of the reconstructed value of the attribute information of the K neighbor points is used as the K+1th predictor variable, and said predictor variable may also be referred to as the weighted average predictor variable.

[0151] The single predictor mode includes a single predictor, namely a weighted average predictor formed by the weighted average of the reconstructed attribute information values ​​of K neighboring points.

[0152] The present application determines whether the target prediction mode of the attribute information of the current point is a multi-pred mode or a single-pred mode through the geometric information of the current point and the geometric information of K neighboring points.

[0153] In some embodiments, step S603 includes step S603-A1 and step S603-A2.

[0154] In step S603-A1, a first distance between a set of neighbor points consisting of K neighbor points and the current point is determined based on the geometric information of the current point and the geometric information of K neighbor points.

[0155] In step S603-A2, the target prediction mode of the attribute information of the current point is determined according to the first distance.

[0156] Since the distance between the current point and neighbor points is relatively far, the accuracy is low when predicting attribute information of the current point using a single-pred mode. Therefore, the present application calculates a first distance between the current point and a set of neighbor points consisting of the current point and K neighbor points, and determines a target prediction mode for the attribute information of the current point according to the magnitude of the first distance. By doing so, it implements an accurate determination of the target prediction mode for the attribute information of the current point, thereby enabling accurate prediction of the current point. Furthermore, by determining the prediction mode for the attribute information of the current point through the first distance between the set of neighbor points consisting of K neighbor points and the current point, the process of determining the prediction mode is decoupled from the process of reconstructing the attribute information of the point in the point cloud. This allows the two processes to be performed in parallel, thereby improving encoding efficiency.

[0157] The method for determining the first distance between the current point and a set of neighbor points consisting of K neighbor points in step S603-A1 above includes, but is not limited to, the following methods.

[0158] In method 1, geometric information of the geometric center of mass of a set of neighbor points is determined according to the geometric information of K neighbor points, and a second distance between the current point and the geometric center of mass is determined according to the geometric information of the current point and the geometric information of the geometric center of mass; a first distance is determined according to the second distance, for example, the second distance is set as the first distance, or a corresponding operation (e.g., a round operation) is performed on the second distance and the result of the operation is set as the first distance.

[0159] For example, let K be 3, meaning the current point contains 3 neighbor points, and assume that the 3 neighbor points are denoted as P1, P2, and P3, respectively. The geometric information of P1 is (x1, y1, z1), the geometric information of P2 is (x2, y2, z2), and the geometric information of P3 is (x3, y3, z3). The geometric information of the geometric center of mass C of the set of neighbor points consisting of P1, P2, and P3 is the average value of the geometric information of P1, P2, and P3, that is, the geometric information of the geometric center of mass C according to the following formula (2). Acquires.

[0160] (2)

[0161] Assuming that the geometric information of the current point P0 is (x, y, z), the second distance between the current point P0 and the geometric center of mass C is determined based on the geometric information of the current point P0 and the geometric information of the geometric center of mass C.

[0162] In one example, the second distance between the current point P0 and the geometric center of mass C is the Euclidean distance between the current point P0 and the geometric center of mass C, for example, according to the following formula (3), the second distance between the current point P0 and the geometric center of mass C Determines.

[0163] (3)

[0164] For example, the encoder executes the following program to obtain the Euclidean distance between the current point P0 and the geometric center of mass C.

[0165] int64_t sumDistance = 0;

[0166] if (predictor.neighborCount > 2 && aps.max_num_direct_predictors) {

[0167] auto position = pointCloud.getPosition(indexes[predictorIndex]);

[0168] auto position1 =

[0169] pointCloud.getPosition(indexes[predictor.neighbors[0].predictorIndex]);

[0170] auto position2 =

[0171] pointCloud.getPosition(indexes[predictor.neighbors[1].predictorIndex]);

[0172] auto position3 =

[0173] pointCloud.getPosition(indexes[predictor.neighbors[2].predictorIndex]);

[0174] Vec3<int32_t> neighborCentroid;

[0175] for (int i = 0; i < 3; ++i) {

[0176] neighborCentroid[i] = (position1[i] + position2[i] + position3[i]) / 3;

[0177] }

[0178] for (int i = 0; i < 3; ++i) {

[0179] sumDistance += (position[i] - neighborCentroid[i])

[0180] * (position[i] - neighborCentroid[i]);

[0181] }

[0182] }

[0183] int64_t maxDiff = std::sqrt(sumDistance).

[0184] In one example, the second distance between the current point P0 and the geometric center of mass C is the Manhattan distance between the current point P0 and the geometric center of mass C, for example, according to the following formula (4), the second distance between the current point P0 and the geometric center of mass C Determines.

[0185] (4)

[0186] In method 2, for each neighbor point among K neighbor points, a third distance between the neighbor point and the current point is determined according to the geometric information of the neighbor point and the geometric information of the current point; and the average value of the third distance between each neighbor point and the current point among K neighbor points is set as the first distance.

[0187] Specifically, assuming that the geometric information of the current point P0 is (x, y, z), the first distance between the current point and a set of neighbor points consisting of K neighbor points is given by the following formula (5). Determines.

[0188] (5)

[0189] Here, (xi, yi, zi) is the geometric information of the i-th neighbor point among K neighbor points.

[0190] After determining a first distance between the current point and a set of neighboring points according to the above method, step S603-A2 is executed to determine a target prediction mode of the attribute information of the current point according to the first distance.

[0191] In some embodiments, step S603-A2 includes the following steps S603-A21 and S603-A22.

[0192] In step S603-A21, if the first distance is greater than or equal to the first value, the target prediction mode is determined to be a multi-predictor mode.

[0193] In step S603-A22, if the first distance is smaller than the first value, the target prediction mode is determined to be a single predictor variable mode.

[0194] In this application, when the first distance is greater than or equal to the first value, it is explained that the distance between the current point and the neighboring point is relatively far; in this case, the attribute information of the current point can be predicted using a multi-predictor mode, thereby improving prediction accuracy. When the first distance is less than the first value, it is explained that the distance between the current point and the neighboring point is relatively close; in this case, the attribute information of the current point can be predicted using a single-predictor mode, making the prediction process simple.

[0195] In one possible implementation, the first value is a preset threshold value.

[0196] In one possible implementation, the point cloud encoder determines a first value by executing steps A through C below.

[0197] In Step A, a bounding box of the point cloud is obtained based on the geometric information of the point in the point cloud, and the bounding box is intended to enclose the point cloud. For example, the bounding box can be understood as a rectangular prism that encloses the point cloud, where the length, width, and height of the bounding box are the difference values ​​between the maximum and minimum values ​​of the point in the point cloud on the XYZ coordinate axes, respectively. For example, the length of the bounding box is the difference value between the maximum and minimum values ​​of the point in the point cloud on the X-axis, the width of the bounding box is the difference value between the maximum and minimum values ​​of the point in the point cloud on the Y-axis, and the height of the bounding box is the difference value between the maximum and minimum values ​​of the point in the point cloud on the Z-axis. In other words, the bounding box can be understood as a minimal rectangular prism that encloses the point cloud, and the length, width, and height of the bounding box are denoted as the three sides of the bounding box.

[0198] In step B, the length of the first side of the bounding box of the point cloud is obtained. The first side may be any one side of the bounding box of the point cloud, for example, the first side may be the shortest side of the bounding box or the longest side of the bounding box.

[0199] In step C, a first value is determined according to the length of the first side.

[0200] The method of determining the first value according to the length of the first side in step C above includes, but is not limited to, several methods as follows.

[0201] In method 1, the ratio value of the length of the first side and the preset value is determined as the first value, and the preset value can be determined according to actual demand.

[0202] In method 2, the above step C includes the following steps C1 and C2.

[0203] In step C1, the quantization parameter QP is obtained.

[0204] In step C2, a first value is determined according to QP and the length of the first side.

[0205] Since the quantization parameter QP is associated with the selection of the target prediction mode of the current point, for example, when the QP value is relatively large, if attribute prediction is used for the current point using a multi-predictor mode, the prediction effect is poor, and in such cases, attribute information of the current point is generally predicted using a single-predictor mode. Based on this, the present application determines a first value through QP and the length of a first side, and by making the determined first value associated with QP, the influence of QP can be considered when determining the target prediction mode of the current point based on the first value in the future, thereby improving the accuracy of determining the target prediction mode.

[0206] The method of determining the first value according to QP and the length of the first side in step C2 above includes, but is not limited to, the following methods.

[0207] In the first method, the ratio value of the length of the first side and QP is set as the first value.

[0208] For example, the first value is determined according to the following formula (6).

[0209] (6)

[0210] Here, is the first value, and is the length of the first side, and the above formula (6) indicates that the first side is the shortest side of the bounding box.

[0211] In the second method, a first ratio value between the length of the first side and a first preset value is obtained; a second ratio value between QP and a second preset value is obtained; and a first value is determined according to the first ratio value and the second ratio value. For example, the first ratio value is obtained by shifting the first ratio value to the left by the number of digits rounded by the second ratio value.

[0212] In one example, the first value is determined according to the following formula (7).

[0213] (7)

[0214] Here, T1 is the first preset value and T2 is the second preset value, and is QP divided by the second preset value T2 and rounded up, and “<<” is left shifting, where moving 1 digit to the left corresponds to multiplying by 2.

[0215] In another example, the first value is determined according to the following formula (8).

[0216] (8)

[0217] Here, 2 T3 T3 is equal to the first preset value T1, for example, if T1 is 64, T3 is 6, and “>>” indicates shifting to the right, where shifting 1 digit to the right corresponds to dividing by 2.

[0218] Optionally, the first preset value T1 is greater than the second preset value T2.

[0219] Optionally, the second preset value T2=6.

[0220] Optionally, if the first side is the shortest side of the bounding box, the first preset value is 64.

[0221] Optionally, if the first side is the longest side of the bounding box, the first preset value is 128.

[0222] The present application determines a first value through the above method, and then compares the first value with a first distance between a current point and a set of neighboring points. If the first distance is greater than or equal to the first value, the target prediction mode of the current point is determined to be a multi-predictor mode, and if the first distance is less than the first value, the target prediction mode of the current point is determined to be a single-predictor mode. Subsequently, step S604 as follows is executed.

[0223] In step S604, attribute information of the current point is predicted using a target prediction mode, and the predicted value of the attribute information of the current point is obtained.

[0224] In step S605, a residual value of the attribute information of the current point is obtained according to the predicted value of the attribute information of the current point. For example, the difference between the raw value and the predicted value of the attribute information of the current point is used as the residual value of the attribute information of the current point.

[0225] In step S606, the residual value of the attribute information of the current point is encoded to obtain a point cloud code stream.

[0226] The above step S604 includes the following two cases.

[0227] In case 1, if the target prediction mode of the determined current point is a multi-predictor mode, step S604 includes the following steps S604-A1 to S604-A5.

[0228] In step S604-A1, the reconstructed values ​​of the attribute information of K neighboring points are obtained.

[0229] In step S604-A2, the reconstructed values ​​of the attribute information of K neighboring points are used as K predictor variables.

[0230] In step S604-A3, the weighted average of the reconstructed attribute information values ​​of K neighboring points is used as the K+1th predictor variable.

[0231] In step S604-A4, the rate distortion optimization RDO value for each of the K+1 predictor variables is determined.

[0232] In step S604-A5, the predictor variable with the smallest RDO value is set as the predicted value of the attribute information of the current point.

[0233] For example, assuming that K=3, that is, the current point includes three neighbor points P1, P2, and P3 respectively, and that all three neighbor points are already encoded points, the reconstructed value of the attribute information is stored in the decoding buffer unit (480) as shown in FIG. 4, and the point cloud encoder can obtain the reconstructed value of the attribute information of the three neighbor points from the decoding buffer unit (480) and use it to predict the attribute information of the current point. For example, the reconstructed value of the attribute information of each of the three neighbor points is used as one predictor variable for the current point, thereby obtaining three predictor variables. Additionally, if the average value of the reconstructed value of the attribute information of the three neighbor points is used as another predictor variable for the current point, the current point has a total of 3+1 predictor variables as shown in Table 1.

[0234] index Prediction variable (prediction type) 0 weighted average 1 1st neighbor point (1 st neighbor) 2 2nd Neighbor Point (2 nd neighbor) 3 Third neighbor point (3 rd neighbor)

[0235] As can be seen from Table 1, the above 3+1 predictor variables correspond to 3+1 prediction modes, where the first prediction mode is the prediction mode corresponding to the reconstructed value of the attribute information of the first neighbor point, i.e., the first predictor variable, and its index is 1; the second prediction mode is the prediction mode corresponding to the reconstructed value of the attribute information of the second neighbor point, i.e., the second predictor variable, and its index is 2; the third prediction mode is the prediction mode corresponding to the reconstructed value of the attribute information of the third neighbor point, i.e., the third predictor variable, and its index is 3. The fourth prediction mode is the prediction mode corresponding to the average value of the reconstructed values ​​of the attribute information of three neighbor points, i.e., the fourth predictor variable, and its index is 0. Here, the first neighbor point is the neighbor point among the three neighbor points that is closest to the current point in distance, the second neighbor point is the point among the three neighbor points that is second closest to the current point in distance, and the third neighbor point is the point among the three neighbor points that is farthest from the current point in distance. Among the three+1 predictor variables, the Rate Distortion Optimization (RDO) value corresponding to each predictor variable is calculated, and the predictor variable with the smallest Rate Distortion Optimization value is set as the predicted value of the attribute information of the current point, for example, the predictor variable with the smallest RDO is the reconstructed value of the attribute information of point P2. In one example, when determining the RDO value, the reciprocal of the distance between the neighbor point and the current point (e.g., Euclidean distance) can be used as the weight of the neighbor point.

[0236] In case 1, in order for the decoding unit to quickly and accurately determine the predicted value of the current point, the prediction mode information corresponding to the prediction variable with the smallest determined RDO value is carried in the point cloud code stream, so that the decoding unit can directly parse the prediction mode information corresponding to the current point from the point cloud code stream and determine the predicted value of the current point using the prediction variable indicated by the prediction mode information. For example, the prediction variable with the smallest determined RDO value is the reconstructed value of the attribute information of point P2, and the corresponding index is 2, and index 2 can be carried in the point cloud code stream. The decoding unit directly parses index 2 from the point cloud code stream and predicts the attribute information of the current point using the reconstructed value of the attribute information of point P2 corresponding to index 2, for example, the reconstructed value of the attribute information of point P2 is used as the predicted value of the attribute information of the current point.

[0237] In case 2, if the target prediction mode of the current point is a single predictor variable mode, the above step S604 includes the following steps S604-B1 and S604-B2.

[0238] In step S604-B1, the reconstructed values ​​of the attribute information of K neighboring points are obtained.

[0239] In step S604-B2, the weighted average of the reconstructed attribute values ​​of K neighboring points is used as the predicted attribute value of the current point.

[0240] For example, let K=3, meaning that the current point is assumed to include three neighboring points P1, P2, and P3, respectively. The reconstructed values ​​of the attribute information of these three points are obtained, and the weighted average of the reconstructed values ​​of the attribute information of these three neighboring points is taken as the predicted value of the attribute information of the current point.

[0241] In one example, when determining the weighted average of the reconstruction values, the reciprocal of the distance between the neighboring point and the current point (e.g., Euclidean distance) can be used as the weight of the neighboring point.

[0242] When the target prediction mode of the current point is a single predictor variable mode, as shown in Table 2, the predictor variable includes a weighted average value, and the corresponding index may be blank.

[0243] index Prediction variable (prediction type) - weighted average

[0244] In other words, if the target prediction mode of the current point is a single predictor variable mode, the prediction mode information corresponding to the current point in the point cloud code stream may not be carried. Thus, if the decoding unit cannot parse the prediction mode information corresponding to the current point in the point cloud code stream, the target prediction mode in the encoding unit for the current point defaults to a single predictor variable mode, and since the decoding unit also predicts the attribute information of the current point using a single predictor variable mode, consistency between the decoding unit and the encoding unit is guaranteed. In one specific embodiment, as illustrated in FIG. 9, the prediction process (600a) of the encoding unit includes the following steps. In step S600-1, a first distance between a set of neighbor points consisting of three neighbor points and the current point is calculated. It should be explained that although the present application describes the example using three neighbor points, the number of neighbor points of the current point includes but is not limited to three, and may also include two, four, or five, etc., and the present application is not limited thereto.

[0245] In step S600-2, it is determined whether the first distance is greater than or equal to the first value, and if the first distance is greater than or equal to the first value, step S600-3 is executed, and if the first distance is less than the first value, step S600-4 is executed.

[0246] In step S600-3, the target prediction mode of the current point is determined to be the Multi-pred mode.

[0247] In step S600-4, the RDO value corresponding to each predictor variable is calculated. Specifically, the ROD value corresponding to the weighted average predictor variable corresponding to index 0 is calculated first to determine whether the current index is the last index, and if not, the index is cyclic to calculate the ROD value corresponding to the predictor variable corresponding to the next index up to the last index. In this way, the RDO value corresponding to each predictor variable can be calculated through cyclic.

[0248] In step S600-5, the predictor variable with the smallest RDO value is selected as the predicted value of the attribute information of the current point, and step S600-7 is executed.

[0249] In step S600-6, a single predictor mode (Multi-pred) of the current point is determined, the weighted average value of the attribute information of neighboring points is set as the predicted value of the attribute information of the current point, and step S600-7 is executed.

[0250] Step S600-7, terminate.

[0251] Here, when comparing the prediction process of the present application illustrated in FIG. 9 with the prediction process of the prior art illustrated in FIG. 6, the present application determines a first distance between a current point and a set of neighbor points based on the geometric information of the current point and the geometric information of neighbor points. By comparing the first distance and the first value, it is determined whether to use a multi-pred mode or a single-pred mode, and the prediction of the attribute information of the current point is performed. By allowing the selection process of the prediction mode of the current point to be related to the geometric information of the point in the point cloud and to be decoupled from the reconstruction process of the attribute information of neighbor points, the selection process of the prediction mode and the reconstruction process of the attribute information can be performed in parallel, thereby improving the efficiency of encoding.

[0252] FIG. 10 is a flowchart illustrating a point cloud encoding method (700) provided in an embodiment of the present application, and based on the embodiment, as illustrated in FIG. 10, the method of an embodiment of the present application includes the following steps.

[0253] In step S701, geometric information and attribute information of the current point in the point cloud are obtained.

[0254] In step S702, K neighbor points of the current point are determined based on the geometric information of the current point, and K is a positive integer greater than or equal to 2.

[0255] The specific implementation process of steps S701 and S702 above may be described by referring to the descriptions of steps S601 and S602 above, and will not be described again here.

[0256] In step S703, a first distance between a set of neighbor points consisting of K neighbor points and the current point is determined based on the geometric information of the current point and the geometric information of K neighbor points, and the specific implementation process can be described by referring to the relevant description in step S603-A1 above, and is not described again here.

[0257] In step S704, a first value is determined, and for specific details, refer to the description of step S603-A22 above. For example, a bounding box of a point cloud is obtained according to geometric information of a point in a point cloud, a first side length of the bounding box of the point cloud is obtained, and a first value is determined according to the first side length and quantization parameter QP.

[0258] It should be noted that the execution process of the above steps S703 and S704 is not sequential, and step S704 may be executed before step S703, after step S703, or simultaneously with step S703, and the present application is not limited thereto.

[0259] In step S705, it is determined whether the first distance is greater than or equal to the first value; if the first distance is greater than or equal to the first value, step S707 is executed, and if the first distance is less than the first value, step S706 is executed.

[0260] In step S706, if the first distance is smaller than the first value, the attribute information of the current point is predicted using a single predictor variable mode according to the reconstructed value of the attribute information of K neighboring points, and the predicted value of the attribute information of the current point is obtained. For example, the average value of the reconstructed value of the attribute information of K neighboring points is taken as the predicted value of the attribute information of the current point.

[0261] In step S707, if the first distance is greater than or equal to the first value, the attribute information of the current point is predicted using a multi-predictor mode based on the reconstructed attribute information of K neighboring points, and the predicted value of the attribute information of the current point is obtained. For example, RDO values ​​corresponding to K+1 predictors are calculated, and the predictor with the smallest RDO value is selected as the predicted value of the attribute information of the current point.

[0262] In step S708, the residual value of the attribute information of the current point is obtained according to the attribute information and the predicted value of the current point.

[0263] In step S709, the residual value of the attribute information of the current point is encoded to obtain a point cloud code stream, and if the target prediction mode of the current point is a multi-predictor mode, the point cloud code stream includes the residual value of the attribute information of the current point and prediction mode information corresponding to the predictor with the smallest RDO value. Optionally, the prediction mode information is located after the residual value.

[0264] To further explain the technical effects of the present application, the technical solution of the present application was implemented in the G-PCC reference software TMC13 V11.0, and some point cloud test sets (cat1-A and cat1-B) required by the Motion Expert Group (MPEG) were tested under the test conditions of the general test configuration CTC CY, and the test results are as shown in Table 3.

[0265] CY_ai End-to-end BD-AttrRate[%] Luminance Amount (Luma) Chromaticity Amount Cb Chromaticity Amount Cr cat1-A 1.4% 1.4% 1.4% cat1-B 1.6% 1.5% 1.4% medium 1.5% 1.5% 1.4%

[0266] Here, CY_ai represents a test condition that is geometrically lossless and has attributes close to lossless. Points in the cat1-A point cloud test set include color attribute information and other attribute information such as reflectance attribute information, while points in the cat1-B point cloud test set include only color attribute information. BD-AttrRate is one of the key parameters for evaluating the performance of a video encoding algorithm and represents the change in code rate and Peak Signal to Noise Ratio (PSNR) of a video encoded by the new algorithm (i.e., the technical solution of the present application) compared to the original algorithm, specifically the change in code rate when the new algorithm and the original algorithm have the same PSNR. As shown in Table 3, for the cat1-A point cloud test set, the technical solution of the present application showed an increase in code rate of only 1.4% in luminance amount, an increase in code rate of only 1.4% in chromaticity amount Cb, and an increase in code rate of only 1.4% in chromaticity amount Cr compared to the original technology. “Average value” represents the average value of the change in code rate between the cat1-A point cloud test set and the cat1-B point cloud test set. As can be seen from Table 3 above, the technical solution of the present application solves the problem in existing technology where the determination of the prediction mode and the reconstruction of attribute information cannot be decoupled at the cost of relatively small performance distortion. The point cloud encoding method mentioned in the embodiment of the present application was described above, and based on this, the point cloud decoding method mentioned in the present application is described below with respect to the decoding unit.

[0267] FIG. 11 is a flowchart of a point cloud decoding method (800) provided in an embodiment of the present application, and as shown in FIG. 11, the method of an embodiment of the present application includes the following steps.

[0268] In step S801, the point cloud code stream is decoded to obtain geometric information and attribute information of the current point in the point cloud.

[0269] It needs to be explained that after the decoding of the geometric information of the points in the point cloud is completed, the decoding of the attribute information is performed. After the decoding of the geometric code stream is completed, the geometric information of the points in the point cloud can be obtained.

[0270] The point cloud code stream includes an attribute code stream and a geometric code stream, and the geometric code stream can be decoded to obtain geometric information of the current point, and the attribute code stream can be decoded to obtain attribute information of the current point.

[0271] In step S802, K neighbor points of the current point are determined based on the geometric information of the current point, and K is a positive integer greater than or equal to 2.

[0272] Specifically, based on the geometric information of the current point, K neighbor points are obtained that are closest to the current point in distance from points in the point cloud whose attribute information has already been decoded. The specific implementation process of step S802 can be described by referring to the description of step S602, and is not described again here.

[0273] In step S803, the target prediction mode of the attribute information of the current point is determined based on the geometric information of the current point and the geometric information of K neighboring points.

[0274] In some embodiments, the step S803 includes the following steps S803-A1 and S803-A2.

[0275] In step S803-A1, a first distance between a set of neighbor points consisting of K neighbor points and the current point is determined based on the geometric information of the current point and the geometric information of K neighbor points.

[0276] In step S803-A2, the target prediction mode of the attribute information of the current point is determined according to the first distance.

[0277] The method for determining the first distance between the current point and the set of neighbor points consisting of K neighbor points according to the geometric information of the current point and the geometric information of K neighbor points in the above S803-A1 includes, but is not limited to, the method as described below.

[0278] In method 1, the above S803-A1 includes steps S803-A11 and S803-A12.

[0279] In step S803-A11, the geometric information of the geometric center of mass of the set of neighboring points is determined based on the geometric information of K neighboring points.

[0280] In step S803-A12, a second distance between the current point and the geometric center of mass is determined based on the geometric information of the current point and the geometric information of the geometric center of mass.

[0281] For example, the second distance between the current point and the geometric center of mass is the Euclidean distance between the current point and the geometric center of mass.

[0282] For example, the second distance between the current point and the geometric center of mass is the Manhattan distance between the current point and the geometric center of mass.

[0283] In step S803-A13, the first distance is determined according to the second distance. For example, the second distance is set as the first distance.

[0284] In method 2, the above step S803-A1 includes steps S803-A13 and S803-A14.

[0285] In step S803-A13, for each neighbor point among K neighbor points, a third distance between the neighbor point and the current point is determined based on the geometric information of the neighbor point and the geometric information of the current point.

[0286] In step S803-A14, the average value of the third distance between each neighbor point and the current point among K neighbor points is set as the first distance.

[0287] In some embodiments, step S803-A2 includes step S803-A21 and step S803-A22.

[0288] In step S803-A21, if the first distance is greater than or equal to the first value, the target prediction mode is determined to be a multi-predictor mode.

[0289] In step S803-A21, if the first distance is smaller than the first value, the target prediction mode is determined to be a single predictor variable mode.

[0290] In one example, the first value is a preset value.

[0291] In another example, the point cloud decoder determines a first value by executing steps D1 through D4 as follows.

[0292] In step D1, a point cloud code stream is decoded to obtain geometric information of a point in the point cloud, and specifically, a geometric code stream of the point cloud is decoded to obtain geometric information of a point in the point cloud.

[0293] In step D2, a bounding box of the point cloud is obtained according to the geometric information of the point in the point cloud, and the bounding box is intended to enclose the point cloud, and for example, the bounding box can be understood as a minimum rectangular prism that encloses the point cloud.

[0294] In step D3, the length of the first side of the bounding box of the point cloud is obtained, where the first side is any one side of the bounding box, for example, the shortest side of the bounding box or the longest side of the bounding box.

[0295] In step D4, a first value is determined according to the length of the first side.

[0296] In some embodiments, the step of determining a first value according to the length of the first side in step D4 includes the following steps D41 and D42.

[0297] In step D41, a point cloud code stream is decoded to obtain quantization parameters (QP); to be explained, the encoding unit encodes the QP into a set of attribute parameters, said parameter set is encoded into a point cloud code stream, and the decoding unit obtains the QP by decoding the set of attribute parameters.

[0298] In step D42, a first value is determined according to QP and the length of the first side.

[0299] An implementation method for determining the first value according to QP and the length of the first side in step D42 above includes, but is not limited to, the following methods.

[0300] In method 1, the ratio value of the length of the first side and QP is set as the first value.

[0301] In method 2, a first ratio value between the length of a first side and a first preset value is obtained; a second ratio value between QP and a second preset value is obtained; and a first value is determined according to the first ratio value and the second ratio value. For example, the first ratio value is shifted to the left by the digit rounded by the second ratio value to obtain the first value.

[0302] In one example, the first value is determined according to the following formula.

[0303] ,

[0304] Here, geometryThreshold is the first value, minimumbBoxLength is the length of the first side, T1 is the first preset value, T2 is the second preset value, and This is QP divided by the second pre-set value T2 and rounded up.

[0305] In one example, the first value is determined according to the following formula.

[0306] ,

[0307] Here, 2 T3 is equal to the first preset value T1.

[0308] Optionally, the first preset value T1 is greater than the second preset value T2.

[0309] Optionally, the second preset value T2=6.

[0310] Optionally, if the first side is the shortest side of the bounding box, the first preset value is 64.

[0311] Optionally, if the first side is the longest side of the bounding box, the first preset value is 128.

[0312] In step S804, attribute information of the current point is predicted using a target prediction mode, and the predicted value of the attribute information of the current point is obtained.

[0313] The present application parses prediction mode information corresponding to the current point from the point cloud code stream when it is determined that the target prediction mode of the current point is a multi-predictor mode. When it is determined that the target prediction mode of the current point is a single-predictor mode, it defaults to predicting the predicted value of the attribute information of the current point using the single-predictor mode without parsing prediction mode information corresponding to the current point from the point cloud code stream. Specifically, it includes the following two cases.

[0314] In case 1, if the target prediction mode is a multi-predictor mode, the above step S804 includes the following steps S804-A1 to S804-A3.

[0315] In step S804-A1, the point cloud code stream is decoded to obtain prediction mode information of the attribute information of the current point.

[0316] In step S804-A2, the reconstructed value of the attribute information of the target neighbor point corresponding to the prediction mode information among K neighbor points is obtained.

[0317] In step S804-A3, the reconstructed value of the attribute information of the target neighbor point is used as the predicted value of the attribute information of the current point.

[0318] As can be seen from the technical method of the encoding unit above, if it is determined that the target prediction mode of the current point is a multi-predictor mode, the prediction mode information corresponding to the prediction variable with the smallest ROD value is carried in the attribute code stream and transmitted to the decoding unit.

[0319] The decoding unit first parses the geometric code stream of the point cloud to obtain geometric information of each point in the point cloud, and obtains K neighbor points that are closest in distance to the current point based on the geometric information of the point in the point cloud. Based on the geometric information of the K neighbor points and the geometric information of the current point, a first distance is obtained between the current point and a set of neighbor points consisting of K neighbor points. If the first distance is greater than or equal to the first value, the target prediction mode of the current point is determined to be a multi-predictor mode. At this time, the decoding unit decodes the attribute code stream of the point cloud to obtain prediction mode information of the attribute information of the current point carried in the attribute code stream, and the prediction mode information may be an index value in Table 1 above. For example, if the index value included in the prediction mode information is 2, it indicates that the encoding unit has set the reconstructed value of the attribute information of the second neighbor point P2 as the predicted value of the attribute information of the current point. For ease of explanation, the second neighbor point P2 is denoted as the target neighbor point here. Based on this, the decoding unit obtains a reconstructed value of the attribute information of a target neighbor point P2 corresponding to the prediction mode information from the decoded point cloud, and sets the reconstructed value of the attribute information of the target neighbor point as the predicted value of the attribute information of the current point.

[0320] What needs to be explained is that the prediction mode information of the attribute information of the current point and the residual value of the attribute information of the current point are combined and encoded, for example, by having the prediction mode information located after the residual value, when the residual value of the current point is obtained by decoding the code stream, the prediction mode information of the current point can be obtained.

[0321] In case 2, if the target prediction mode is a single predictor variable mode, the above step S804 includes the following steps S804-B1 to S804-B2.

[0322] In step S804-B1, the reconstructed values ​​of the attribute information of K neighboring points are obtained.

[0323] In step S804-B2, the weighted average of the reconstructed attribute values ​​of K neighboring points is used as the predicted attribute value of the target point.

[0324] The above K neighbor points are points that have already been decoded in the point cloud, and as shown in FIG. 5, the reconstructed value of the attribute information of the decoded points is stored in the decoding buffer unit (560), and the decoder can obtain the reconstructed value of the attribute information of the K neighbor points from the decoding buffer unit (560). The weighted average value of the reconstructed value of the attribute information of the K neighbor points is used as the predicted value of the attribute information of the target point.

[0325] In one example, when determining the weighted average of the reconstructed values, the reciprocal of the distance between the neighboring point and the current point (e.g., Euclidean distance) is used as the weight of the neighboring point.

[0326] In one specific embodiment, as illustrated in FIG. 12, the prediction process (800a) of the decoding unit includes the following steps.

[0327] In step S800-1, a first distance is obtained between a set of neighbor points consisting of three neighbor points and the current point.

[0328] In step S800-2, it is determined whether the first distance is greater than or equal to the first value; if the first distance is greater than or equal to the first value, step S800-3 is executed, and if the first distance is less than the first value, step S800-5 is executed.

[0329] In step S800-3, the target prediction mode of the current point is determined to be the Multi-pred mode.

[0330] In step S800-4, the point cloud code stream is decoded to obtain prediction mode information of the attribute information of the current point, and the reconstructed value of the attribute information of the target neighbor point corresponding to the prediction mode information is used as the reconstructed value of the attribute information of the current point.

[0331] In step S800-5, the target prediction mode of the current point is determined to be a single-pred mode, and the weighted predictor (i.e., the weighted average of the reconstructed values ​​of the attribute information of K neighboring points) is set as the reconstructed value of the attribute information of the current point.

[0332] Here, when comparing the prediction process of the present application illustrated in FIG. 12 with the prediction process of the prior art illustrated in FIG. 7, the present application determines a first distance between the current point and a set of neighbor points based on the geometric information of the current point and the geometric information of neighbor points, compares the first distance with a first value to determine whether to use a multi-pred mode or a single-pred mode, and performs prediction of the attribute information of the current point. If it is determined that a multi-pred mode is used, prediction mode information is obtained from the point cloud code stream, and if it is determined that a single-pred mode is used, prediction mode information is not obtained from the point cloud code stream. By doing so, the process of selecting the prediction mode of the current point is associated with the geometric information of the point in the point cloud and is decoupled from the process of reconstructing the attribute information of neighbor points, thereby allowing the process of selecting the prediction mode and the process of reconstructing the attribute information to be performed in parallel, and thus improving the efficiency of decoding.

[0333] FIG. 13 is a flowchart of a point cloud decoding method (900) provided in an embodiment of the present application, and as shown in FIG. 13, the method of an embodiment of the present application includes the following steps.

[0334] In step S901, the point cloud code stream is decoded to obtain geometric information and attribute information of the current point in the point cloud.

[0335] In step S902, K neighbor points of the current point are determined based on the geometric information of the current point, and K is a positive integer greater than or equal to 2.

[0336] In step S903, a first distance between the current point and a set of neighbor points consisting of K neighbor points is determined based on the geometric information of the current point and the geometric information of K neighbor points.

[0337] In step S904, a first value is determined. Specifically, the description of step S803-A21 above may be referenced. For example, a bounding box of a point cloud is obtained according to geometric information of a point in a point cloud, a first side length of the bounding box of the point cloud is obtained, and a first value is determined according to the first side length and quantization parameter QP.

[0338] It should be noted that the execution process of the above steps S903 and S904 is not in any sequential order, and step S904 may be executed before step S903, after step S903, or simultaneously with step S903, and the present application is not limited thereto.

[0339] In step S905, it is determined whether the first distance is greater than or equal to the first value; if the first distance is greater than or equal to the first value, step S907 is executed, and if the first distance is less than the first value, step S906 is executed.

[0340] In step S906, if the first distance is smaller than the first value, the attribute information of the current point is predicted using a single predictor mode based on the reconstructed value of the attribute information of K neighboring points, and the predicted value of the attribute information of the current point is obtained. For example, the average value of the reconstructed value of the attribute information of K neighboring points is taken as the predicted value of the attribute information of the current point.

[0341] In step S907, if the first distance is greater than or equal to the first value, the point cloud code stream is decoded to obtain the prediction mode information of the attribute information of the current point carried in the point cloud code stream.

[0342] In step S908, a reconstructed value of the attribute information of a target neighbor point corresponding to the prediction mode information among K neighbor points is obtained, and the reconstructed value of the attribute information of the target neighbor point is used as the predicted value of the attribute information of the current point.

[0343] In step S909, the point cloud code stream is decoded to obtain residual values ​​of attribute information of the current point, and based on the residual values ​​and predicted values ​​of attribute information of the current point, reconstructed values ​​of attribute information of the current point are obtained. For example, residual values ​​are applied to predicted values ​​of attribute information of the current point to obtain reconstructed values ​​of attribute information of the current point.

[0344] It should be understood that FIGS. 8 through 13 are merely examples of the present application and should not be understood as a limitation to the present application.

[0345] Although preferred embodiments of the present application have been described in detail above by combining the drawings, the present application is not limited to the specific details of the above embodiments, and various simple modifications to the technical methods of the present application may be made within the scope of the technical concept of the present application, and all such simple modifications fall within the scope of protection of the present application. For example, each specific technical feature described in the above specific embodiments may be combined in any suitable manner provided thereis not to contradict, and to avoid unnecessary duplication, the present application does not describe each possible combination method separately. Furthermore, for example, different embodiments of the present application may be combined arbitrarily, and such combinations should be deemed to be as disclosed in the present application, provided they do not depart from the spirit of the present application.

[0346] It should be further understood that in each method embodiment of the present application, the number of each process does not imply a sequence of execution, and the execution order of each process is determined according to its function and internal logic, and does not impose any limitations on the implementation process of the embodiments of the present application. Furthermore, in the embodiments of the present application, the term “and / or” is used merely to describe the association of associated objects and indicates that three types of relationships may exist. Specifically, A and / or B may represent three cases: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character symbol “ / ” in this text generally indicates that the preceding and succeeding associated objects are in an “or” relationship.

[0347] Above, an embodiment of the method of the present application has been described in detail by combining FIGS. 8 to 13, and below, an embodiment of the device of the present application has been described in detail by combining FIGS. 14 to 16.

[0348] FIG. 14 is an exemplary block diagram of a point cloud encoder (10) provided in an embodiment of the present application.

[0349] As illustrated in FIG. 14, the point cloud encoder (10) is,

[0350] Acquisition unit (11) for acquiring geometric information and attribute information of the current point in the point cloud;

[0351] A neighbor point determination unit (12) for determining K neighbor points of the current point according to the geometric information of the current point - where K is a positive integer greater than or equal to 2 - ;

[0352] A prediction mode determination unit (13) for determining a target prediction mode of attribute information of the current point according to the geometric information of the current point and the geometric information of the K neighboring points; and

[0353] It includes an encoding unit (14) for generating a point cloud code stream by using the target prediction mode to predict attribute information of the current point, obtaining a predicted value of attribute information of the current point, obtaining a residual value of attribute information of the current point according to the predicted value of attribute information of the current point, and encoding the residual value of attribute information of the current point.

[0354] In some embodiments, the prediction mode determination unit (13) specifically determines a first distance between a set of neighbor points consisting of K neighbor points and the current point according to geometric information of the current point and geometric information of the K neighbor points; and determines a target prediction mode of attribute information of the current point according to the first distance.

[0355] In some embodiments, the prediction mode determination unit (13) specifically determines that if the first distance is greater than or equal to the first value, the target prediction mode is a multi-predictor mode; and if the first distance is less than the first value, the target prediction mode is a single-predictor mode.

[0356] In some embodiments, the encoding unit (14) specifically obtains a reconstructed value of attribute information of K neighbor points when the target prediction mode is a multi-predictor mode; sets the reconstructed value of attribute information of K neighbor points as K predictors; sets the weighted average value of the reconstructed value of attribute information of K neighbor points as the K+1th predictor; determines the rate distortion optimization RDO value of each predictor among the K+1 predictors; and sets the predictor with the smallest RDO value as the predicted value of attribute information of the current point.

[0357] In some embodiments, the point cloud code stream includes prediction mode information corresponding to the predictor variable with the smallest RDO value.

[0358] In some embodiments, the prediction mode determination unit (13) specifically obtains a reconstructed value of the attribute information of the K neighbor points when the target prediction mode is a single prediction variable mode; and sets the weighted average value of the reconstructed value of the attribute information of the K neighbor points as the predicted value of the attribute information of the current point.

[0359] In some embodiments, the encoding unit (14) also obtains a bounding box of the point cloud according to geometric information of a point in the point cloud—the bounding box is intended to enclose the point cloud—; obtains the length of a first side of the bounding box of the point cloud; and determines the first value according to the length of the first side.

[0360] In some embodiments, the encoding unit (14) specifically obtains a quantization parameter (QP); and determines the first value according to the QP and the length of the first side.

[0361] In some embodiments, the encoding unit (14) is specifically for making the ratio value of the length of the first side and the QP the first value.

[0362] In some embodiments, the encoding unit (14) specifically obtains a first ratio value of the length of the first side and a first preset value; obtains a second ratio value of the QP and a second preset value; and determines the first value according to the first ratio value and the second ratio value.

[0363] In some embodiments, the encoding unit (14) specifically shifts the first ratio value to the left by a digit rounded to the second ratio value to obtain the first value.

[0364] In some embodiments, the encoding unit (14) is specifically for determining the first value according to the following formula;

[0365] ,

[0366] Here, geometryThreshold is the first value, minimumbBoxLength is the length of the first side, and 2 T3 is equal to the above-mentioned first preset value T1, and the It is rounded up by dividing the above QP by the above second preset value T2.

[0367] Optionally, the first preset value T1 is greater than the second preset value T2.

[0368] Optionally, the second preset value T2 is 6.

[0369] Optionally, if the first variation is the shortest variation of the bounding box, the first preset value is 64.

[0370] Optionally, if the first variation is the longest variation of the bounding box, the first preset value is 128.

[0371] In some embodiments, the prediction mode determination unit (13) specifically determines the geometric information of the geometric center of mass of the set of neighbor points according to the geometric information of the K neighbor points; determines a second distance between the current point and the geometric center of mass according to the geometric information of the current point and the geometric information of the geometric center of mass; and determines the first distance according to the second distance.

[0372] In some embodiments, the prediction mode determination unit (13) is specifically for making the second distance the first distance.

[0373] In one example, the second distance between the current point and the geometric center of mass is the Euclidean distance between the current point and the geometric center of mass.

[0374] In another example, the second distance between the current point and the geometric center of mass is the Manhattan distance between the current point and the geometric center of mass.

[0375] In some embodiments, the prediction mode determination unit (13) specifically determines a third distance between the neighbor point and the current point for each neighbor point among the K neighbor points, according to the geometric information of the neighbor point and the geometric information of the current point; and sets the average value of the third distance between each neighbor point and the current point among the K neighbor points as the first distance.

[0376] It should be understood that device embodiments and method embodiments may correspond to each other, and similar descriptions may refer to method embodiments. To avoid duplication, they are not described again here. Specifically, the point cloud encoder (10) illustrated in FIG. 14 can execute the method of the embodiment of the present application, and at least one of the other operations and functions of each unit in the point cloud encoder (10) is intended to implement the corresponding process in each method, such as method (600) and method (700), respectively, and for brevity, is not described again here.

[0377] FIG. 15 is an exemplary block diagram of a point cloud decoder (20) provided in an embodiment of the present application.

[0378] As illustrated in FIG. 15, the point cloud decoder (20) is,

[0379] A decoding unit (21) for decoding a point cloud code stream to obtain geometric information and attribute information of the current point in the point cloud;

[0380] A neighbor point determination unit (22) for determining K neighbor points of the current point according to the geometric information of the current point - where K is a positive integer greater than or equal to 2 - ;

[0381] A prediction mode determination unit (23) for determining a target prediction mode of attribute information of the current point according to the geometric information of the current point and the geometric information of the K neighboring points; and

[0382] It may include a parsing unit (24) for obtaining a predicted value of the attribute information of the current point by predicting the attribute information of the current point using the target prediction mode above.

[0383] In some embodiments, the prediction mode determination unit (23) specifically determines a first distance between a set of neighbor points consisting of K neighbor points and the current point according to the geometric information of the current point and the geometric information of the K neighbor points; and determines a target prediction mode of attribute information of the current point according to the first distance.

[0384] In some embodiments, the prediction mode determination unit (23) specifically determines that if the first distance is greater than or equal to the first value, the target prediction mode is a multi-predictor mode; and if the first distance is less than the first value, the target prediction mode is a single-predictor mode.

[0385] In some embodiments, the parsing unit (24) specifically decodes the point cloud code stream if the target prediction mode is a multi-predictor mode to obtain prediction mode information of the attribute information of the current point; obtains a reconstructed value of the attribute information of the target neighbor point corresponding to the prediction mode information among the K neighbor points; and sets the reconstructed value of the attribute information of the target neighbor point as the predicted value of the attribute information of the current point.

[0386] In some embodiments, the parsing unit (24) specifically obtains a reconstructed value of the attribute information of the K neighbor points when the target prediction mode is a single prediction variable mode; and sets the weighted average value of the reconstructed value of the attribute information of the K neighbor points as the predicted value of the attribute information of the target point.

[0387] In some embodiments, the parsing unit (24) also decodes the point cloud code stream to obtain geometric information of a point in the point cloud; obtains a bounding box of the point cloud according to the geometric information of the point in the point cloud—the bounding box is intended to enclose the point cloud—; obtains the length of a first side of the bounding box of the point cloud; and determines the first value according to the length of the first side.

[0388] In some embodiments, the parsing unit (24) specifically decodes the point cloud code stream to obtain a quantization parameter QP; and determines the first value according to the QP and the length of the first side.

[0389] In some embodiments, the parsing unit (24) specifically makes the ratio value of the length of the first side and the QP the first value.

[0390] In some embodiments, the parsing unit (24) specifically obtains a first ratio value of the length of the first side and a first preset value; obtains a second ratio value of the QP and a second preset value; and determines the first value according to the first ratio value and the second ratio value.

[0391] In some embodiments, the parsing unit (24) specifically shifts the first ratio value to the left by a digit rounded to the second ratio value to obtain the first value.

[0392] In some embodiments, the parsing unit (24) is specifically for determining the first value according to the following formula;

[0393] ,

[0394] Here, geometryThreshold is the first value, minimumbBoxLength is the length of the first side, and 2 T3 is equal to the above-mentioned first preset value T1, and the It is rounded up by dividing the above QP by the above second preset value T2.

[0395] Optionally, the first preset value T1 is greater than the second preset value T2.

[0396] Optionally, the second preset value T2 is 6.

[0397] Optionally, if the first variation is the shortest variation of the bounding box, the first preset value is 64.

[0398] Optionally, if the first variation is the longest variation of the bounding box, the first preset value is 128.

[0399] In some embodiments, the prediction mode determination unit (23) specifically determines the geometric information of the geometric center of mass of the set of neighbor points according to the geometric information of the K neighbor points; determines a second distance between the current point and the geometric center of mass according to the geometric information of the current point and the geometric information of the geometric center of mass; and determines the first distance according to the second distance.

[0400] In some embodiments, the prediction mode determination unit (23) is specifically for making the second distance the first distance.

[0401] In one example, the second distance between the current point and the geometric center of mass is the Euclidean distance between the current point and the geometric center of mass.

[0402] In another example, the second distance between the current point and the geometric center of mass is the Manhattan distance between the current point and the geometric center of mass.

[0403] In some embodiments, the prediction mode determination unit (23) specifically determines a third distance between the neighbor point and the current point for each neighbor point among the K neighbor points, according to the geometric information of the neighbor point and the geometric information of the current point; and sets the average value of the third distance between each neighbor point and the current point among the K neighbor points as the first distance.

[0404] It should be understood that device embodiments and method embodiments may correspond to each other, and similar descriptions may refer to method embodiments. To avoid duplication, they are not described again here. Specifically, the point cloud decoder (20) illustrated in FIG. 15 may correspond to a corresponding subject in at least one of the method (800) and method (900) of the embodiment of the present application, and at least one of the other operations and functions of each unit in the point cloud decoder (20) is intended to implement a corresponding process in each method, such as at least one of the method (800) and method (900), and for brevity, they are not described again here.

[0405] The apparatus and system of the embodiment of the present application have been described above from the angle of the functional unit by combining the drawings. It should be understood that the functional unit may be implemented in a hardware form, may be implemented in a software form, or may be implemented in a combination of hardware and software units. Specifically, each step of the method embodiment in the embodiment of the present application may be completed by at least one of a hardware integrated logic circuit and a software instruction in a processor, and the step of the method disclosed by combining the embodiment of the present application may be directly reflected as being executed and completed by a hardware decoding processor, or may be executed and completed by a combination of hardware and software units in a decoding processor. Optionally, the software unit may be located in a storage medium widely known in the art, such as random memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in memory, and the processor reads information from memory and completes the step in the method embodiment by combining with hardware.

[0406] FIG. 16 is an exemplary block diagram of an electronic device (30) provided in an embodiment of the present application.

[0407] As shown in FIG. 16, the electronic device (30) may be a point cloud encoder or a point cloud decoder according to an embodiment of the present application, and the electronic device (30) is,

[0408] It may include a memory (31) and a processor (32), wherein the memory (31) is for storing a computer program (34) and transmitting the program code to the processor (32). In other words, the processor (32) can call and operate the computer program (34) from the memory (31) to implement the method of the embodiment of the present application.

[0409] For example, the processor (32) may be for executing a step in the method (200) according to instructions in the computer program (34).

[0410] In some embodiments of the present application, the processor (32) is,

[0411] It may include, but is not limited to, general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic elements, individual gate or transistor logic elements, individual hardware components, etc.

[0412] In some embodiments of the present application, the memory (31) is,

[0413] It may include, but is not limited to, at least one of volatile memory and non-volatile memory. Here, the non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), eraseable programmable read-only memory (EPROM), electrically eraseable programmable read-only memory (EEPROM), or flash memory. The volatile memory may be random access memory (RAM) that serves as an external cache. Through exemplary rather than limited descriptions, many forms of RAM can be used, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory (Direct Rambus RAM, DR RAM).

[0414] In some embodiments of the present application, the computer program (34) may be divided into one or more units, said one or more units stored in the memory (31) and executed by the processor (32) to complete the method provided in the present application. said one or more units may be a series of computer program instruction segments capable of completing a specific function, said instruction segments are intended to describe the process of execution of the computer program (34) in the electronic device (30).

[0415] As shown in FIG. 16, the electronic device (30) also,

[0416] It may include a transceiver (33), and the transceiver (33) may be connected to the processor (32) or memory (31).

[0417] Here, the processor (32) can control the transceiver (33) to communicate with another device, specifically, to transmit information or data to another device or to receive information or data transmitted by another device. The transceiver (33) may include a transmitter and a receiver. The transceiver (33) may also additionally include an antenna, and the number of antennas may be one or multiple.

[0418] It should be understood that each component of the electronic device (30) is connected to one another through a bus system, wherein, in addition to including a data bus, the bus system further includes a power bus, a control bus, and a status signal bus.

[0419] FIG. 17 is an exemplary block diagram of a point cloud encoding and decoding system (40) provided in an embodiment of the present application.

[0420] As illustrated in FIG. 17, the point cloud encoding and decoding system (40) may include a point cloud encoder (41) and a point cloud decoder (42), wherein the point cloud encoder (41) is for executing the point cloud encoding method mentioned in the embodiment of the present application, and the point cloud decoder (42) is for executing the point cloud decoding method mentioned in the embodiment of the present application.

[0421] The present application further provides a computer storage medium, said computer storage medium has a computer program stored therein, and when said computer program is executed by a computer, said computer is made to execute the method of the method embodiment. Alternatively, an embodiment of the present application further provides a computer program product including instructions, said instructions being executed by a computer, said computer is made to execute the method of the method embodiment.

[0422] When implemented using software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, a process or function according to an embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a dedicated computer, a computer network, or other programmable device. The computer instructions may be stored on a computer-readable storage medium or transferred from one computer-readable storage medium to another computer-readable storage medium; for example, the computer instructions may be transferred from one website, computer, server, or data center to another website, computer, server, or data center via a wired method (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or a wireless method (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium accessible to a computer or a data storage device such as a server or data center comprising one or more available media aggregations. The above-mentioned usable media may be magnetic media (e.g., floppy disk, hard disk, tape), optical media (e.g., digital point cloud optical disc (digital video disc, DVD)), or semiconductor media (e.g., solid state disk (SSD)).

[0423] A person skilled in the art will understand that each exemplary unit and algorithm step described in the embodiments disclosed herein may be combined and implemented in electronic hardware or in a combination of computer software and electronic hardware. Whether such functions are implemented in a hardware manner or in a software manner is determined by the specific application and design constraints of the technical method. A person skilled in the art may implement the described functions using different methods for each specific application, but such implementations should not be construed as being outside the scope of this application.

[0424] It should be understood that in some of the embodiments provided in this application, the disclosed systems, devices, and methods may be implemented in other ways. For example, the device embodiments described above are merely exemplary, and for example, the division of the units is merely a logical functional division, and there may be other division methods in actual implementation, for example, multiple units or components may be combined or integrated into different systems, or some features may be ignored or not implemented. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be implemented through some interfaces, and the indirect coupling or communication connection of the devices or units may be electrical, mechanical, or of other forms.

[0425] The unit described as a separating member may be physically separated or not physically separated, and the member represented as a unit may be a physical unit or not a physical unit; that is, it may be located in one place or distributed across multiple network units. Depending on actual needs, some or all of the units may be selected to realize the purpose of the present embodiment. For example, each functional unit in each embodiment of the present application may be integrated into a single processing unit, each unit may be an independent physical entity, or two or more units may be integrated into a single unit.

[0426] The above description is merely a specific embodiment of the present application, and the scope of protection of the present application is not limited thereto. A person skilled in the art to which the present application pertains will readily understand that any change or substitution within the technical scope disclosed in the present application falls within the scope of protection of the present application. Accordingly, the scope of protection of the present application must be based on the scope of protection of the claims.

Claims

Claim 1 A point cloud encoding method comprises: a step of obtaining geometric information and attribute information of a current point in a point cloud; a step of determining K neighbor points of the current point according to the geometric information of the current point, wherein K is a positive integer greater than or equal to 2; a step of determining a target prediction mode for attribute information of the current point according to the geometric information of the current point and the geometric information of the K neighbor points; a step of predicting attribute information of the current point using the target prediction mode to obtain a predicted value for attribute information of the current point; a step of obtaining a residual value for attribute information of the current point according to the predicted value for attribute information of the current point; and a step of encoding the residual value for attribute information of the current point to generate a point cloud code stream; wherein the step of determining a target prediction mode for attribute information of the current point according to the geometric information of the current point and the geometric information of the K neighbor points comprises a step of determining a first distance between a set of neighbor points composed of the K neighbor points and the current point according to the geometric information of the current point and the geometric information of the K neighbor points. A point cloud encoding method comprising: a step of determining a target prediction mode of attribute information of the current point according to the first distance; and further comprising: a step of obtaining a bounding box of the point cloud according to geometric information of a point in the point cloud - the bounding box is intended to enclose the point cloud -; a step of obtaining a first side length of the bounding box of the point cloud; and a step of determining a first value for determining a target prediction mode according to the first side length. Claim 2 delete Claim 3 A point cloud encoding method according to claim 1, wherein if the target prediction mode is a multi-predictor mode, the step of predicting attribute information of the current point using the target prediction mode to obtain a predicted value of attribute information of the current point comprises: obtaining a reconstructed value of attribute information of K neighbor points; setting the reconstructed value of attribute information of K neighbor points as K predictors; setting the weighted average value of the reconstructed value of attribute information of K neighbor points as the K+1th predictor; determining the Rate Distortion Optimization (RDO) value of each predictor among the K+1 predictors; and setting the predictor with the smallest RDO value as the predicted value of attribute information of the current point. Claim 4 delete Claim 5 A point cloud decoding method comprises: a step of obtaining geometric information and attribute information of a current point in a point cloud; a step of determining K neighbor points of the current point according to the geometric information of the current point, wherein K is a positive integer greater than or equal to 2; a step of determining a target prediction mode for attribute information of the current point according to the geometric information of the current point and the geometric information of the K neighbor points; and a step of predicting attribute information of the current point using the target prediction mode to obtain a predicted value for attribute information of the current point; wherein the step of determining a target prediction mode for attribute information of the current point according to the geometric information of the current point and the geometric information of the K neighbor points comprises: a step of determining a first distance between a set of neighbor points composed of the K neighbor points and the current point according to the geometric information of the current point and the geometric information of the K neighbor points; and a step of determining a target prediction mode for attribute information of the current point according to the first distance; and wherein the point cloud decoding method comprises a step of decoding a point cloud code stream to obtain geometric information of a point in a point cloud. A point cloud decoding method further comprising: a step of obtaining a bounding box of the point cloud according to geometric information of a point in the point cloud, wherein the bounding box is intended to enclose the point cloud; a step of obtaining the length of a first side of the bounding box of the point cloud; and a step of determining a first value for determining a target prediction mode according to the length of the first side. Claim 6 delete Claim 7 A point cloud decoding method according to claim 5, wherein if the target prediction mode is a multi-predictor mode, the step of predicting attribute information of the current point using the target prediction mode to obtain a predicted value of attribute information of the current point comprises: a step of decoding the point cloud code stream to obtain prediction mode information of attribute information of the current point; a step of obtaining a reconstructed value of attribute information of a target neighbor point corresponding to the prediction mode information among the K neighbor points; and a step of setting the reconstructed value of attribute information of the target neighbor point as the predicted value of attribute information of the current point. Claim 8 A point cloud decoding method according to claim 5, wherein if the target prediction mode is a single prediction variable mode, the step of predicting attribute information of the current point using the target prediction mode to obtain a predicted value of attribute information of the current point comprises: a step of obtaining a reconstructed value of attribute information of K neighboring points; and a step of setting the weighted average value of the reconstructed value of attribute information of K neighboring points as the predicted value of attribute information of the current point. Claim 9 delete Claim 10 A point cloud decoding method according to claim 5, wherein the step of determining the first value according to the length of the first side comprises: decoding the point cloud code stream to obtain a quantization parameter (QP); and determining the first value according to the QP and the length of the first side. Claim 11 A point cloud decoding method according to claim 10, wherein the step of determining the first value according to the QP and the first side length includes the step of making the ratio value of the first side length and the QP the first value. Claim 12 A point cloud decoding method according to claim 11, wherein the step of determining the first value according to the QP and the first side length comprises: a step of obtaining a first ratio value between the first side length and a first preset value; a step of obtaining a second ratio value between the QP and a second preset value; and a step of determining the first value according to the first ratio value and the second ratio value. Claim 13 A point cloud decoding method according to claim 12, wherein the step of determining the first value according to the first ratio value and the second ratio value includes the step of obtaining the first value by shifting the first ratio value to the left by the number of digits rounded by the second ratio value. Claim 14 In paragraph 12, the step of determining the first value according to the first ratio value and the second ratio value includes the step of determining the first value according to the following formula, and , geometryThreshold is the first value, and the minimumbBoxLength is the length of the first side, and 2 T3 is equal to the above first preset value (T1), and the A point cloud decoding method characterized by dividing the above QP by the above second preset value (T2) and rounding up. Claim 15 A point cloud decoding method according to claim 12, characterized in that the first preset value is greater than the second preset value. Claim 16 A point cloud decoding method according to claim 5, wherein the step of determining a first distance between the current point and the set of neighbor points according to the geometric information of the current point and the geometric information of the K neighbor points comprises: a step of determining the geometric information of the center of mass of the set of neighbor points according to the geometric information of the K neighbor points; a step of determining a second distance between the current point and the center of mass according to the geometric information of the current point and the center of mass; and a step of determining the first distance according to the second distance. Claim 17 A point cloud decoding method according to claim 16, wherein the step of determining the first distance according to the second distance includes the step of making the second distance the first distance. Claim 18 A point cloud decoding method according to claim 16, characterized in that the second distance between the current point and the geometric center of mass is the Euclidean distance or Manhattan distance between the current point and the geometric center of mass. Claim 19 A point cloud decoding method according to claim 5, wherein the step of determining a first distance between the current point and the set of neighbor points according to the geometric information of the current point and the geometric information of the K neighbor points comprises: for each neighbor point among the K neighbor points, a step of determining a third distance between the neighbor point and the current point according to the geometric information of the neighbor point and the geometric information of the current point; and a step of setting the average value of the third distance between each neighbor point and the current point among the K neighbor points as the first distance. Claim 20 A point cloud decoder comprising a processor and memory; said memory is for storing a computer program; said processor calls and operates the computer program stored in said memory to perform decoding on a point cloud and obtain geometric information and attribute information of a current point in the point cloud; said K determines K neighbor points of the current point according to the geometric information of the current point, wherein K is a positive integer greater than or equal to 2; said K determines a target prediction mode for attribute information of the current point according to the geometric information of the current point and the geometric information of said K neighbor points; and said K predicts attribute information of the current point using the target prediction mode to obtain a predicted value for attribute information of the current point; and when the target prediction mode for attribute information of the current point is determined according to the geometric information of the current point and the geometric information of said K neighbor points, said processor calls and operates the computer program stored in said memory to determine a first distance between a set of neighbor points consisting of said K neighbor points and the current point according to the geometric information of the current point and the geometric information of said K neighbor points. A point cloud decoder characterized by executing a step of determining a target prediction mode of attribute information of the current point according to the first distance; wherein the processor also calls and operates a computer program stored in the memory to obtain a bounding box of the point cloud according to geometric information of a point in the point cloud - the bounding box is intended to encircle the point cloud -; a step of obtaining a first side length of the bounding box of the point cloud; and a step of determining a first value for determining a target prediction mode according to the first side length. Claim 21 delete Claim 22 delete Claim 23 delete Claim 24 delete Claim 25 delete Claim 26 delete Claim 27 delete Claim 28 delete Claim 29 delete Claim 30 delete Claim 31 delete Claim 32 delete Claim 33 delete Claim 34 delete Claim 35 delete Claim 36 delete Claim 37 delete Claim 38 delete Claim 39 delete Claim 40 delete Claim 41 delete Claim 42 delete Claim 43 delete Claim 44 delete Claim 45 delete Claim 46 delete Claim 47 delete

Citation Information

Patent Citations

  • Method and apparatus for adaptive point cloud attribute coding

    WO2020198180A1

  • Point cloud attribute encoding method, decoding method, encoding equipment and decoding equipment

    CN111145090A

  • Three-dimensional data encoding method, three-dimensional data decoding method, three-dimensional data encoding device, and three-dimensional data decoding device

    WO2019240215A1