Point cloud attribute coding by transfer function parameters

Abstract

A point cloud encompassed in a space, points of the point cloud having attributes is compressed and encode. The space is first divided in blocks. For each block comprising at least one point, the geometry of the points is compressed and a transfer function associating coordinates within the block with an attribute value is determined according to the attribute values of the at least one point. Then, compressed geometries and transfer functions are encoded in a data stream.

Description

[0001] POINT CLOUD ATTRIBUTE CODING BY TRANSFER FUNCTION PARAMETERS

[0002] 1. Technical Field

[0003] The present principles generally relate to the domain of encoding, transmitting and decoding point clouds with attributes. In particular, the present principles relate to formatting the attributes of a point cloud independently of how the geometry is encoded or decoded and in a way that allow a parallel encoding and decoding of the geometry and the attributes of a point cloud.

[0004] 2. Background

[0005] The present section is intended to introduce the reader to various aspects of art, which may be related to various aspects of the present principles that are described and / or claimed below. This discussion is believed to be helpful in providing the reader with background information to facilitate a better understanding of the various aspects of the present principles. Accordingly, it should be understood that these statements are to be read in this light, and not as admissions of prior art.

[0006] Advances in 3D capturing and rendering technologies enables new applications and services in the fields of autonomous driving, cultural heritage archival, immersive telepresence and virtual / augmented reality. Point clouds have arisen as one of the main 3D scene representations for such applications. A point cloud frame consists of a set of 3D points, each point being represented by its 3D location and eventually one or more attributes like color, reflectance, normal vector, or transparency. Compression of dense dynamic point clouds with a geometry-based approach requires a huge amount of information. A 3D-to-2D projection may leverage existing 2D video codecs to encode 3D point clouds or 3D meshes. For example, a Geometric Point Cloud Compression (G-PCC) encoder for dense dynamic point clouds combines several tool modules, comprising pruned occupancy tree (octree) plus triangle soups for geometry coding, Region- Adaptive Hierarchical Transform (RAHT) for color attribute coding, motion compensated interframe prediction and context-adaptive arithmetic coding. In such systems, geometry and attributes are sequentially compressed. The geometry is first compressed, then input attributes are transferred onto the reconstructed geometry (after decompression) and compressed. So, at decoder side, the reconstructed geometry must be available (fully decompressed) when decoding the attributes. For real time applications, there is a lack of a parallelisable encoding / decoding scheme.

[0007] 3. Summary

[0008] The following presents a simplified summary of the present principles to provide a basic understanding of some aspects of the present principles. This summary is not an extensive overview of the present principles. It is not intended to identify key or critical elements of the present principles. The following summary merely presents some aspects of the present principles in a simplified form as a prelude to the more detailed description provided below.

[0009] The present principles relate to a method for encoding a point cloud encompassed in a space. The points of the point cloud have attributes. First, the space is divided in blocks. For each block comprising at least one point, the geometry of the points of the block is compressed. At the same time, a transfer function associating coordinates within the block with an attribute value is determined according to the attribute values of the points of the block. Then, the compressed geometries and the transfer functions are encoded in a data stream.

[0010] The present principles also relate to a device comprising a processor and a memory associated with the processor that is configured to implement the method above.

[0011] The present principles also relate to a method for decoding a point cloud encoded according to the principles of the above method. A data stream is received. The data are representative of a point cloud encompassed in a space divided in blocks. Points of the point cloud have attributes. An encoded block comprises at least one point and a transfer function associating coordinates within the block with an attribute value. For each encoded block, the geometry of the points of the block is decompressed and an attribute value is set to the decompressed points by using the transfer function of the block.

[0012] The present principles also relate to a device comprising a processor and a memory associated with the processor that is configured to implement the method above. The present principles also relate to a data stream representative of a point cloud encompassed in a space divided in blocks, points of the point cloud having attributes, an encoded block comprising at least one point and a transfer function associating coordinates within the block with an attribute value.

[0013] 4. Brief Description of Drawings

[0014] The present disclosure will be better understood, and other specific features and advantages will emerge upon reading the following description, the description making reference to the annexed drawings wherein:

[0015] - Figure 1 illustrates a G-PCC scheme wherein geometry and attributes are sequentially compressed;

[0016] - Figure 2 illustrates how the 3D space of a 3D point cloud to encode is divided;

[0017] - Figure 3 shows an example architecture of a device 30 which may be configured to implement encoding and / or decoding methods according to an embodiment of the present principles;

[0018] - Figure 4 shows an example of an embodiment of the syntax of a stream when the data are transmitted over a packet-based transmission protocol;

[0019] - Figure 5 illustrates the scheme for the encoding and decoding of a 3D point cloud block according to the present principles, wherein geometry and attributes are encoded / decoded in parallel by block;

[0020] - Figure 6 shows five examples of predefined graph configurations.

[0021] 5. Detailed description of embodiments

[0022] The present principles will be described more fully hereinafter with reference to the accompanying figures, in which examples of the present principles are shown. The present principles may, however, be embodied in many alternate forms and should not be construed as limited to the examples set forth herein. Accordingly, while the present principles are susceptible to various modifications and alternative forms, specific examples thereof are shown by way of examples in the drawings and will herein be described in detail. It should be understood, however, that there is no intent to limit the present principles to the particular forms disclosed, but on the contrary, the disclosure is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present principles as defined by the claims.

[0023] The terminology used herein is for the purpose of describing particular examples only and is not intended to be limiting of the present principles. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises", "comprising," "includes" and / or "including" when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. Moreover, when an element is referred to as being "responsive" or "connected" to another element, it can be directly responsive or connected to the other element, or intervening elements may be present. In contrast, when an element is referred to as being "directly responsive" or "directly connected" to other element, there are no intervening elements present. As used herein the term "and / or" includes any and all combinations of one or more of the associated listed items and may be abbreviated as" / ".

[0024] It will be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element without departing from the teachings of the present principles.

[0025] Although some of the diagrams include arrows on communication paths to show a primary direction of communication, it is to be understood that communication may occur in the opposite direction to the depicted arrows.

[0026] Some examples are described with regard to block diagrams and operational flowcharts in which each block represents a circuit element, module, or portion of code which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in other implementations, the function(s) noted in the blocks may occur out of the order noted. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending on the functionality involved. Reference herein to “in accordance with an example” or “in an example” means that a particular feature, structure, or characteristic described in connection with the example can be included in at least one implementation of the present principles. The appearances of the phrase in accordance with an example” or “in an example” in various places in the specification are not necessarily all referring to the same example, nor are separate or alternative examples necessarily mutually exclusive of other examples.

[0027] Reference numerals appearing in the claims are by way of illustration only and shall have no limiting effect on the scope of the claims. While not explicitly described, the present examples and variants may be employed in any combination or sub-combination.

[0028] Figure 1 illustrates a G-PCC scheme wherein geometry and attributes are sequentially compressed. In the encoder 11, a point cloud 12 with attributes (for example color or transparency) is obtained from a source. The geometry is first compressed, for example by dividing the 3D space of the 3D point cloud in voxels and representing it as an octree. Then the geometry is decompressed and the attributes are transferred onto the decompressed geometry to be compressed, for example with RAHT techniques. Currently, RAHT is not a parallelisable transform as it encodes the entire frame’s attributes together. In addition, the encoding of the attributes is dependent on the results of the geometry encoder, which introduces errors if the geometry encoding is lossy. Therefore, the encoding of the geometry cannot be executed in parallel with the encoding of the attributes, as one needs to transfer the attributes to the decoded geometry prior to encoding the point cloud attributes. The same problem occurs at the decoder 13 wherein the geometry must be fully decompressed before than the attributes are decompressed on the decompressed geometry to obtain a decoded point cloud 14 with attributes.

[0029] According to the present principles, a scheme to encode / decode the attributes of a point cloud independently of how the geometry is encoded / decoded is proposed.

[0030] Figure 2 illustrates how the 3D space of a 3D point cloud to encode is divided. The point cloud 3D space is divided into 3D-blocks of size Nxx Nyx Nz, as shown in a 2D example in Figure 2, where Nx, Ny, and Nzare predetermined parameters, such as 2, 4, 8, 16, 32, or 64. According to the present principles, the geometry and attribute encoding / decoding is executed for each block individually and only occupied blocks are processed (empty blocks are skipped). The signalling of which block is occupied can be conveyed to the decoder, for example by the octree. In an embodiment, the 3D coordinates of the points in a block are expressed relatively to the block origin.

[0031] Figure 5 illustrates the scheme for the encoding and decoding of a 3D point cloud block according to the present principles, wherein geometry and attributes are encoded / decoded in parallel by block. Let f (Pa, x, y, z) be a transfer function that returns the attribute a corresponding to a point located at ( , y, z) relative to the block origin. The value returned by the function is controlled by the parameter vector Pa. In a point cloud, the attributes can be the YUV color, RGB color, YCbCr color, reflectance, etc, and a may assume the values Y, U, V, R, G, B, and so on, and there is one parameter for each attribute, for example, PY, Pv, and Pv. From the input point cloud block, the parameters that describe how the attributes change with position are obtained. The parameters can be a lossless or lossy representation of the attributes, i.e., the values of f(Pa, x, y, z) for the points (x, y, z) of the point cloud can be the exact values of the corresponding attribute at that position, or an approximation. The parameters Paare then quantized and entropy coded to be added to the bitstream. Additionally, the parameters may be down-sampled to further reduce the amount of data to encode. At the decoder side, the parameters are entropy decoded, dequantized, and they may be up-sampled to get the reconstructed parameters Paper block. The geometry decoder generates a list of reconstructed points. For the k-th reconstructed point is (x^, y^, z^), the reconstructed attributes is determined by

[0032] In a first embodiment, from the input point cloud block, an NxX NyX Nzimage, referred to as la, is constructed and represents the attribute a for every possible position inside the block. The value of / a(x, y, z) is equal to the value of the attribute a of the point at position (x, y, z) if there is only one single point of the input point cloud block at this position. It vaults the average value of the attribute a of all points at position (x, y, z) if there is more than one point of the input point cloud block at this position. It occurs when the input point cloud block has collocated points. The value of / a(x, y, z) is equal to an interpolated attribute a based on the average of the closest points if there are no points at position (x, y, z) in other cases. The parameters Paare the transformed coefficients of a 3D transform, such as 3D DCT (discrete cosine transform), 3D DST (discrete sine transform), or 3D DFT / FFT (discrete Fourier transform, fast Fourier transform). The transform to be used can be the same for all blocks or determined for each block and signalled to the decoder. The parameters Pacan be down-sampled independently for each attribute channel. The down-sampling can be processed directly on the image la, by reducing its resolution, or at the transformed coefficients by discarding high frequency coefficients. The transformed coefficients are quantized by dividing them by a quantization matrix and the coefficients are rounded to the nearest integer. The coefficients are rasterized into a vector using a 3D zig-zag path, where low- pass coefficients are at the start of the vector and high-pass coefficients are at the end.

[0033] In this first embodiment, at the decoder, after entropy decoding the parameters, the values are dequantized by multiplying them by the same quantization matrix that was used at the encoder. If the parameters have been down-sampled at the encoder, they are up-sampled at the decoder. If the down-sampling has been applied on the transformed coefficients, the up-sampling can be performed by zero-padding the locations of the discarded high frequency coefficients. If the downsampling has been applied on the 3D image / , the upsampling can be performed by reconstructing the image I from the transformed dequantized coefficients and interpolating the image to its original resolution. Finally, with the reconstructed geometry of the block, function f inverse transforms the coefficients given by Paand picks the reconstructed attributes from the corresponding positions at the reconstructed Nxx Nyx Nzimage.

[0034] In a second embodiment, the parameter vector Pais the coefficients of the GFT (graph Fourier transform). At a first step, one graph configuration of connected nodes is selected from a list of predefined graph configurations inside the block where the nodes are sampling positions. Figure 6 shows five examples of predefined graph configurations. The average attribute value of the closest points is assigned to each node to the graph configuration. The graph configuration can be conveyed to the decoder by the index of the predefined configurations. The parameters Paare the transformed coefficients of the GFT. They are quantized and entropy coded.

[0035] The decoder, after entropy decoding and dequantizing the coefficients, applies the inverse transform to obtain the reconstructed attributes of each node. Finally, with the reconstructed geometry of the block, function f interpolates the attribute by averaging the attribute values of the nodes that are close to the reconstructed points’ 3D location. In a third embodiment, the function f is a neural network and the parameters P are learned. So, f can be a pretrained network with a large point cloud dataset, a network that is trained per- scene (overfitting) or a finetuned network initialized with a pretrained network. In variants, f can also be a generic network independently from point (x,y, z) or it can be locally trained for each block. In the second case, a set of smaller neural networks are trained, and the corresponding parameters are quantized and entropy coded in parallel. This accelerates the processing.

[0036] In the three embodiments, blocks can be encoded and decoded in parallel. The decoding of the attributes is performed by block and, so, there is no need to wait for the decoding of the entire geometry to decode the attributes. As the parameters describe how the attributes change according to point locations within a 3D block, the parameters also englobe the attribute transfer stage. The parameters can be estimated directly from the input point cloud, in parallel with the geometry encoding. The division of the point cloud into blocks allows for the blocks to be processed in parallel, allowing acceleration in the encoding / decoding process for real time applications

[0037] Figure 3 shows an example architecture of a device 30 which may be configured to implement encoding and / or decoding methods according to an embodiment of the present principles. The device is linked with other devices via their bus 31 and / or via I / O interface 36.

[0038] Device 30 comprises following elements that are linked together by a data and address bus 31 :D

[0039] - a processor 32 (or CPU), which is, for example, a DSP (or Digital Signal Processor);

[0040] - a ROM (or Read Only Memory) 33;

[0041] - a RAM (or Random Access Memory) 34;

[0042] - a storage interface 35;

[0043] - an I / O interface 36 for reception of data to transmit, from an application; and

[0044] - a power supply (not represented in Figure 2), e.g. a battery.

[0045] In accordance with an example, the power supply is external to the device. In each of mentioned memory, the word « register » used in the specification may correspond to area of small capacity (some bits) or to very large area (e.g. a whole program or large amount of received or decoded data). The ROM 33 comprises at least a program and parameters. The ROM 33 may store algorithms and instructions to perform techniques in accordance with present principles. When switched on, the CPU 32 uploads the program in the RAM and executes the corresponding instructions.

[0046] The RAM 34 comprises, in a register, the program executed by the CPU 32 and uploaded after switch-on of the device 30, input data in a register, intermediate data in different states of the method in a register, and other variables used for the execution of the method in a register.

[0047] The implementations described herein may be implemented in, for example, a method or a process, an apparatus, a computer program product, a data stream, or a signal. Even if only discussed in the context of a single form of implementation (for example, discussed only as a method or a device), the implementation of features discussed may also be implemented in other forms (for example a program). An apparatus may be implemented in, for example, appropriate hardware, software, and firmware. The methods may be implemented in, for example, an apparatus such as, for example, a processor, which refers to processing devices in general, including, for example, a computer, a microprocessor, an integrated circuit, or a programmable logic device. Processors also include communication devices, such as, for example, computers, cell phones, portable / personal digital assistants ("PDAs"), and other devices that facilitate communication of information between end-users.

[0048] Device 30 is linked, for example via bus 31 to a set of sensors 37 and to a set of rendering devices 38. Sensors 37 may be, for example, cameras, microphones, temperature sensors, Inertial Measurement Units, GPS, hygrometry sensors, IR or UV light sensors or wind sensors. Rendering devices 38 may be, for example, displays, speakers, vibrators, heat, fan, etc.

[0049] In accordance with examples, the device 30 is configured to implement a method according to the present principles of encoding, decoding and rendering a 3D point clouds with attributes, and belongs to a set comprising:

[0050] - a mobile device;

[0051] - a communication device;

[0052] - a game device;

[0053] - a tablet (or tablet computer);

[0054] - a laptop; - a still picture camera;

[0055] - a video camera.

[0056] Figure 4 shows an example of an embodiment of the syntax of a stream when the data are transmitted over a packet-based transmission protocol. Figure 4 shows an example structure 4 of a stream encoding point clouds according to the present principle. The structure consists in a container which organizes the stream in independent elements of syntax. The structure may comprise a header part 41 which is a set of data common to every syntax element of the stream. For example, the header part comprises some of metadata about syntax elements, describing the nature and the role of each of them. The structure comprises a payload comprising an element of syntax 42 and at least one element of syntax 43 (there may be an element of syntax 43 for each type of attribute data, for instance one for the color, one for the reflectance, one for the normal vectors, etc.). Syntax element 42 comprises data representative of the geometry of the point cloud, that is, for example, a series of bits representative of the 3D blocks, for example represented as octrees.

[0057] The implementations described herein may be implemented in, for example, a method or a process, an apparatus, a computer program product, a data stream, or a signal. Even if only discussed in the context of a single form of implementation (for example, discussed only as a method or a device), the implementation of features discussed may also be implemented in other forms (for example a program). An apparatus may be implemented in, for example, appropriate hardware, software, and firmware. The methods may be implemented in, for example, an apparatus such as, for example, a processor, which refers to processing devices in general, including, for example, a computer, a microprocessor, an integrated circuit, or a programmable logic device. Processors also include communication devices, such as, for example, Smartphones, tablets, computers, mobile phones, portable / personal digital assistants ("PDAs"), and other devices that facilitate communication of information between end-users.

[0058] Implementations of the various processes and features described herein may be embodied in a variety of different equipment or applications, particularly, for example, equipment or applications associated with data encoding, data decoding, view generation, texture processing, and other processing of images and related texture information and / or depth information. Examples of such equipment include an encoder, a decoder, a post-processor processing output from a decoder, a pre-processor providing input to an encoder, a video coder, a video decoder, a video codec, a web server, a set-top box, a laptop, a personal computer, a cell phone, a PDA, and other communication devices. As should be clear, the equipment may be mobile and even installed in a mobile vehicle.

[0059] Additionally, the methods may be implemented by instructions being performed by a processor, and such instructions (and / or data values produced by an implementation) may be stored on a processor-readable medium such as, for example, an integrated circuit, a software carrier or other storage device such as, for example, a hard disk, a compact diskette (“CD”), an optical disc (such as, for example, a DVD, often referred to as a digital versatile disc or a digital video disc), a random access memory (“RAM”), or a read-only memory (“ROM”). The instructions may form an application program tangibly embodied on a processor-readable medium. Instructions may be, for example, in hardware, firmware, software, or a combination. Instructions may be found in, for example, an operating system, a separate application, or a combination of the two. A processor may be characterized, therefore, as, for example, both a device configured to carry out a process and a device that includes a processor-readable medium (such as a storage device) having instructions for carrying out a process. Further, a processor-readable medium may store, in addition to or in lieu of instructions, data values produced by an implementation.

[0060] As will be evident to one of skill in the art, implementations may produce a variety of signals formatted to carry information that may be, for example, stored or transmitted. The information may include, for example, instructions for performing a method, or data produced by one of the described implementations. For example, a signal may be formatted to carry as data the rules for writing or reading the syntax of a described embodiment, or to carry as data the actual syntax-values written by a described embodiment. Such a signal may be formatted, for example, as an electromagnetic wave (for example, using a radio frequency portion of spectrum) or as a baseband signal. The formatting may include, for example, encoding a data stream and modulating a carrier with the encoded data stream. The information that the signal carries may be, for example, analog or digital information. The signal may be transmitted over a variety of different wired or wireless links, as is known. The signal may be stored on a processor-readable medium. A number of implementations have been described. Nevertheless, it will be understood that various modifications may be made. For example, elements of different implementations may be combined, supplemented, modified, or removed to produce other implementations. Additionally, one of ordinary skill will understand that other structures and processes may be substituted for those disclosed and the resulting implementations will perform at least substantially the same function(s), in at least substantially the same way(s), to achieve at least substantially the same result(s) as the implementations disclosed. Accordingly, these and other implementations are contemplated by this application.

Claims

CLAIMS1. A method for encoding a point cloud encompassed in a space, points of the point cloud having attributes, the method comprising:- dividing the space in blocks;- for each block comprising at least one point:• compressing a geometry of the at least one point;• determining a transfer function associating coordinates within the block with an attribute value according to the attribute values of the at least one point; and- encoding compressed geometries and transfer functions in a data stream.

2. A device for encoding a point cloud encompassed in a space, points of the point cloud having attributes, the device comprising a memory associated with a processor configured for:- dividing the space in blocks;- for each block comprising at least one point:• compressing a geometry of the at least one point;• determining a transfer function associating coordinates within the block with an attribute value according to the attribute values of the at least one point; and- encoding compressed geometries and transfer functions in a data stream.

3. A method comprising:- obtaining, from a data stream, data representative of a point cloud encompassed in a space divided in blocks, points of the point cloud having attributes, an encoded block comprising at least one point and a transfer function associating coordinates within the block with an attribute value;- for each encoded block:• decompressing a geometry of the at least one point;• for each of the at least one point, setting its attribute value by using the transfer function of the block.

4. A device comprising a memory associated with a processor configured for:- obtaining, from a data stream representative of a point cloud encompassed in a space divided in blocks, points of the point cloud having attributes, an encoded block comprising at least one point and a transfer function associating coordinates within the block with an attribute value; - for each encoded block:• decompressing a geometry of the at least one point;• for each of the at least one point, setting its attribute value by using the transfer function of the block.

5. A data stream representative of a point cloud encompassed in a space divided in blocks, points of the point cloud having attributes, an encoded block comprising at least one point and a transfer function associating coordinates within the block with an attribute value.

Images