Point cloud data processing method, device, equipment, medium and product

Abstract

This application discloses a method for processing point cloud data, relating to the technical field of point cloud data. The method, a point cloud data processing device, a point cloud data processing apparatus, a storage medium, and a computer program product are disclosed. The method comprises: dividing point cloud data by scale through upsampling and downsampling to obtain anchor points and non-anchor points distributed at different scales; encoding the distribution probabilities of occupancy code values ​​of anchor points and non-anchor points at different scales to obtain bit streams of occupancy code values ​​of anchor points and non-anchor points at different scales. This application aims to address the technical problem of low point cloud data processing performance.

Description

Technical Field

[0001] The present application relates to the field of point cloud data technology, and in particular to a point cloud data processing method, a point cloud data processing apparatus, a point cloud data processing device, a storage medium, and a computer program product. Background Art

[0002] Point clouds are a powerful representation of three-dimensional data, typically containing a large number of points. To efficiently store and transmit this data, compression encoding is necessary, and the most important aspect of encoding is the probabilistic prediction of the occupancy code for the point cloud data. Currently, two prediction methods have emerged: point prediction based on SparsePCGC (Sparse Tensor-based Multiscale Representation for Point Cloud Geometry Compression), and prediction of eight-bit occupancy codes based on octrees. The former, when processing dense point clouds with tens or even millions of points, poses a significant challenge to memory capacity, resulting in low processing performance. The latter, using 8-bit occupancy codes, is prone to serial encoding issues within the scale, resulting in long encoding times and low performance. Therefore, improving the performance of point cloud data processing is an urgent issue that needs to be addressed. Summary of the Invention

[0003] The main purpose of this application is to provide a point cloud data processing method, a point cloud data processing device, a point cloud data processing equipment, a storage medium and a computer program product, aiming to solve the technical problem of low point cloud data processing performance.

[0004] To achieve the above objectives, the present application proposes a method for processing point cloud data, the method comprising:

[0005] The point cloud data is divided by scale through up and down sampling to obtain anchor points and non-anchor points distributed at different scales;

[0006] The distribution probabilities of the occupied code values ​​of the anchor points and the occupied code values ​​of the non-anchor points are encoded at different scales to obtain bit streams of the occupied code values ​​of the anchor points and the non-anchor points at different scales.

[0007] In one embodiment, the step of encoding the distribution probabilities of the occupancy code values ​​of the anchor points and the occupancy code values ​​of the non-anchor points at different scales to obtain bit streams of the occupancy code values ​​of the anchor points and the non-anchor points at different scales includes:

[0008] If the preset scale is smaller than the second highest scale or the highest scale, a probability prediction is performed on the occupancy code value of the anchor point to obtain a first distribution probability of the occupancy code value of the anchor point being any value between 0 and 15;

[0009] Probabilistically predicting the occupancy code value of the non-anchor point according to the first distribution probability to obtain a second distribution probability of the occupancy code value of the non-anchor point being any value from 0 to 15;

[0010] Encoding the first distribution probability to obtain a bit stream of an occupancy code value of an anchor point;

[0011] The second distribution probability is encoded to obtain a bit stream of occupied code values ​​of non-anchor points.

[0012] In one embodiment, the step of performing probability prediction on the occupancy code value of the anchor point to obtain a first distribution probability that the occupancy code value of the anchor point is any value between 0 and 15 includes:

[0013] Based on the distribution probability of the occupancy code value of the existing anchor point being any value between 0 and 15, a probability prediction is performed on the occupancy code value of the next anchor point to obtain the distribution probability of the occupancy code value of the next anchor point being any value between 0 and 15, until the distribution probability of the occupancy code value of all anchor points being any value between 0 and 15 is predicted.

[0014] In one embodiment, the step of probabilistically predicting the occupancy code value of the non-anchor point based on the first distribution probability to obtain a second distribution probability of the occupancy code value of the non-anchor point being any value between 0 and 15 includes:

[0015] Based on the distribution probability of the occupancy code value of the anchor point being any value between 0 and 15 and the distribution probability of the occupancy code value of the existing non-anchor point being any value between 0 and 15, a probability prediction is performed on the occupancy code value of the next non-anchor point to obtain the distribution probability of the occupancy code value of the next non-anchor point being any value between 0 and 15, until the distribution probability of the occupancy code value of all non-anchor points being any value between 0 and 15 is predicted.

[0016] In one embodiment, the step of encoding the occupancy code value of the anchor point and the occupancy code value of the non-anchor point at different scales to obtain the bit streams of the occupancy code values ​​of the anchor point and the non-anchor point at different scales further includes:

[0017] If the preset scale is the second highest scale or the highest scale, the anchor points and non-anchor points are grouped, and the occupancy code values ​​of the anchor points in the target group are probability predicted to obtain a first distribution probability of the occupancy code values ​​of the anchor points in the target group;

[0018] Probabilistically predicting the occupied code values ​​of the non-anchor points in the target group according to the first distribution probability to obtain a second distribution probability of the occupied code values ​​of the non-anchor points in the target group;

[0019] Encoding the first distribution probability to obtain a bit code stream of an occupancy code value of an anchor point in a target group;

[0020] The second distribution probability is encoded to obtain a bit code stream of occupied code values ​​of non-anchor points in the target group.

[0021] In one embodiment, after the step of obtaining the bit code stream of the occupied code value of the non-anchor point in the target group, the method further includes:

[0022] Probabilistically predicting the occupancy code values ​​of the remaining anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code values ​​of the anchor points and non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining anchor points;

[0023] Constructing, according to the occupancy code values ​​of the anchor points in the target group and the occupancy code values ​​of the remaining anchor points, an occupancy code value of the anchor point when the preset scale is the second highest scale or the highest scale;

[0024] Probabilistically predicting the occupancy code values ​​of the remaining non-anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and the non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code value of the anchor points and the occupancy code values ​​of the non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining non-anchor points;

[0025] According to the occupancy code values ​​of the non-anchor points in the target group and the occupancy code values ​​of the remaining non-anchor points, the occupancy code values ​​of the non-anchor points when the preset scale is the second highest scale or the highest scale are constructed.

[0026] In addition, to achieve the above-mentioned purpose, the present application also proposes a point cloud data processing device, the point cloud data processing device comprising: a partitioning module for partitioning the point cloud data by scale through up and down sampling to obtain anchor points and non-anchor points distributed at different scales;

[0027] The encoding module is used to encode the distribution probabilities of the occupancy code values ​​of the anchor points and the occupancy code values ​​of the non-anchor points at different scales, respectively, to obtain bit code streams of the occupancy code values ​​of the anchor points and the non-anchor points at different scales.

[0028] In addition, to achieve the above-mentioned purpose, the present application also proposes a point cloud data processing device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program is configured to implement the steps of the point cloud data processing method as described above.

[0029] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the point cloud data processing method described above are implemented.

[0030] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the point cloud data processing method as described above.

[0031] Compared with point-based prediction, the present application divides the scale of point cloud data by up and down sampling, and can obtain anchor points and non-anchor points distributed at different scales. The reasonable distribution of anchor points and non-anchor points at different scales can balance the details of the data and the computational complexity, save memory, and is a disguised form of performance improvement. Compared with octree-based prediction, the present application generates a bit stream by encoding the distribution probability of the occupancy code values ​​of anchor points and non-anchor points in the point cloud data at different scales. The number of coding bits used in the present application is more flexible, and it avoids some serial coding problems within the scale of the octree, which greatly reduces the encoding and decoding time and improves the point cloud data processing performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0033] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0034] Figure 1 A schematic diagram of a prediction matrix provided for one embodiment of the method for processing point cloud data of this application;

[0035] Figure 2 A schematic diagram of point cloud data processing provided for one embodiment of the method for processing point cloud data of this application;

[0036] Figure 3 A flowchart of the first embodiment of the method for processing point cloud data of this application is provided;

[0037] Figure 4 This is a schematic diagram of the module structure of the point cloud data processing device of this application;

[0038] Figure 5This is a schematic diagram of the device structure of the hardware operating environment involved in the point cloud data processing method of this application.

[0039] The purpose, features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0040] It should be understood that the specific embodiments described herein are merely used to explain the technical solutions of the present application and are not intended to limit the present application.

[0041] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0042] Because the existing technology has two prediction bases in the lossless coding based on arithmetic coding: the prediction of point occupancy probability and the prediction of octree. The former is based on SparsePCGC, which focuses on point prediction, but point prediction is a huge challenge to memory (the dense point cloud of a frame of a character can have tens or even millions of points), especially for deep learning, the memory in the graphics card is very precious. At a time when graphics card memory is extremely scarce, the use of the above method will continue to challenge the graphics card memory. The latter is based on the octree, and the prediction of the octree is for an eight-bit occupancy code (8 bits) of a supervoxel, but the 8-bit occupancy code is not flexible enough to support more sophisticated models, such as Figure 1 .

[0043] exist Figure 1 In the figure, the left side shows a cube divided into 8 cube sub-blocks of different colors. The points of different colors represent the points of the point cloud data, and the histogram is the probability prediction of each point; the right side shows a cube divided into 8 cube sub-blocks of the same color. The points of the same color represent the 8-bit occupancy code of the point cloud data, and the histogram shows the distribution probability of any value from 0 to 255 obtained by converting the 8-bit occupancy code.

[0044] The main solution of the embodiment of this application is to propose a dual quad occupancy code based on the chessboard division to achieve a balance between performance and memory capacity. The dual quad occupancy code solution models the point cloud lossless compression problem as follows: using up and down sampling, the probability prediction of anchor point occupancy codes and non-anchor point occupancy codes at different scales, such as Figure 2 .

[0045] exist Figure 2 In [1], a cube is divided into two cube sub-blocks with different colors, representing anchor points and non-anchor points respectively. Both anchor points and non-anchor points have 4-bit binary occupancy codes.

[0046] The histogram shows the distribution probability of any value from 0 to 15 obtained by converting the 4-bit occupancy code.

[0047] It should be noted that the execution subject of this embodiment may be a processor, or a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, mobile phone, etc., or an electronic device or controller capable of implementing the above functions. The following uses a processor as an example to illustrate this embodiment and the following embodiments.

[0048] Based on this, the embodiment of the present application provides a method for processing point cloud data, referring to Figure 3 , Figure 3 This is a flow chart of the first embodiment of the method for processing point cloud data of the present application.

[0049] In this embodiment, the method for processing point cloud data is applied to the encoding end, and includes steps S10 to S20:

[0050] Step S10: dividing the point cloud data by scale through up and down sampling to obtain anchor points and non-anchor points distributed at different scales;

[0051] It's important to note that point cloud data is a representation of three-dimensional spatial data, consisting of a large number of discrete points, each containing three-dimensional coordinate (X, Y, Z) information. Upsampling and downsampling are techniques used in data processing to change the resolution or sampling rate of a dataset. In point cloud data processing, upsampling typically refers to increasing the number of data points through methods such as interpolation to improve the resolution and detail of the data; downsampling refers to reducing the size of the dataset and lowering computational complexity by removing or merging data points. Upsampling and downsampling techniques play a key role in the compression, transmission, and visualization of point cloud data.

[0052] Scaling is the process of dividing point cloud data into multiple layers based on different levels of detail, or resolution. At each layer (or scale), the density and detail of the data points vary, from coarse-grained global overviews to fine-grained local features. Scaling facilitates multi-scale analysis of data, allowing for simultaneous processing of both overall structure and local details.

[0053] Anchor points and non-anchor points are point cloud data with staggered positions in space, which helps to provide more context information for encoding and decoding, thereby reducing encoding and decoding time and improving encoding and decoding performance.

[0054] Step S20 , encoding the distribution probabilities of the occupied code values ​​of the anchor points and the occupied code values ​​of the non-anchor points at different scales respectively, to obtain bit streams of the occupied code values ​​of the anchor points and the non-anchor points at different scales.

[0055] It should be noted that the occupancy code value of the anchor point refers to a value that indicates whether each point in the point cloud data is occupied. The occupancy code value is a decimal number ranging from 0 to 15, which is obtained by converting the 4-bit binary occupancy code. An occupancy code of 1 indicates that the point is occupied, and 0 indicates that the point is idle. The occupancy code value of the non-anchor point is similar to that of the anchor point.

[0056] In the context of point cloud data encoding, distribution probability refers to the distribution probability of occupying any value between 0 and 15. For point cloud data, encoding can include compression technology to reduce the amount of data and improve storage and transmission efficiency.

[0057] In point cloud data processing, different scales refer to point cloud data of different resolutions or granularities. Encoding point cloud data at different scales means processing the data at different levels of detail.

[0058] For example, the encoder can start at the highest scale and encode anchor points and non-anchor points scale by scale. The decoder can start at the lowest scale and gradually use the code streams sent by the encoder to decode the anchor points and non-anchor points at each scale. Each scale is interdependent; non-anchor points will be decoded in dependence on anchor points. A bit stream is the encoded data sequence used for storage or transmission.

[0059] In one embodiment, step S20 includes steps A10 to A40:

[0060] Step A10: If the preset scale is smaller than the second highest scale or the highest scale, a probability prediction is performed on the occupancy code value of the anchor point to obtain a first distribution probability of the occupancy code value of the anchor point being any value between 0 and 15;

[0061] It should be noted that in multi-scale coding, the preset scale usually refers to a specific scale selected during the encoding process. For example, the point cloud data is divided into 10 scales through up and down sampling technology. The preset scale can be any of the 10 scales. The second highest scale refers to the largest scale among all scales that is smaller than the highest scale, such as the 9th scale; the highest scale refers to the scale with the highest bit rate among all scales, such as the 10th scale.

[0062] The first distribution probability refers to the predicted probability distribution obtained through probability prediction, in which the occupancy code value of the anchor point occurs at any value between 0 and 15.

[0063] Step A20: Probabilistically predict the occupancy code value of the non-anchor point based on the first distribution probability to obtain a second distribution probability of the occupancy code value of the non-anchor point being any value between 0 and 15;

[0064] It should be noted that the first distribution probability refers to the distribution obtained by probabilistically predicting the occupied code value of the anchor point; the second distribution probability refers to the distribution obtained by probabilistically predicting the occupied code value of the non-anchor point based on the first distribution probability, that is, the predicted probability of the occupied code value of the non-anchor point appearing at each value between 0 and 15.

[0065] Step A30: Encode the first distribution probability to obtain a bit stream of an occupancy code value of an anchor point;

[0066] It should be noted that encoding the first distribution probability to generate a bit stream includes entropy coding. In entropy coding, frequently occurring symbols are assigned shorter codes, while less frequently occurring symbols are assigned longer codes. This effectively reduces the length of the data bit stream and improves storage and transmission efficiency.

[0067] Step A40: Encode the second distribution probability to obtain a bit stream of occupied code values ​​of non-anchor points.

[0068] It should be noted that encoding the second distribution probability to generate a bit stream is the same as the encoding method described in step A30.

[0069] In this embodiment, when the preset scale is smaller than the next highest scale or the highest scale, a probability prediction is performed on the occupancy code values ​​of the anchor points to obtain a first distribution probability, which is then encoded to obtain a bit stream of the occupancy code values ​​of the anchor points. Based on the first distribution probability of the anchor points, a probability prediction is performed on the occupancy code values ​​of non-anchor points to obtain a second distribution probability, which is then encoded to obtain a bit stream of the occupancy code values ​​of the non-anchor points. This embodiment embodies a lossless compression process, i.e., when the preset scale is smaller than the next highest scale or the highest scale, the occupancy code values ​​of all anchor points and non-anchor points are transmitted.

[0070] The probability-based coding method helps to reduce the length of the bit code stream and improve the data compression rate, while taking into account the correlation between anchor points and non-anchor points in point cloud data to achieve more efficient data compression and transmission.

[0071] In another embodiment, step S20 further includes:

[0072] If the preset scale is the second highest scale or the highest scale, the anchor points and non-anchor points are grouped, and the occupancy code values ​​of the anchor points in the target group are probability predicted to obtain a first distribution probability of the occupancy code values ​​of the anchor points in the target group;

[0073] Probabilistically predicting the occupied code values ​​of the non-anchor points in the target group according to the first distribution probability to obtain a second distribution probability of the occupied code values ​​of the non-anchor points in the target group;

[0074] Encoding the first distribution probability to obtain a bit code stream of an occupancy code value of an anchor point in a target group;

[0075] The second distribution probability is encoded to obtain a bit code stream of occupied code values ​​of non-anchor points in the target group.

[0076] It should be noted that during the encoding process, anchor points and non-anchor points in the scale with high code rate are grouped and only part of the groups are transmitted, that is, nearly lossless encoding, to optimize coding efficiency and save code rate.

[0077] Exemplarily, the anchor points and non-anchor points are divided into 8 groups each, for a total of 16 groups. Only some of the groups in the 16 groups are transmitted, and the remaining groups are predicted based on known information. The known information includes: the occupancy code values ​​of the anchor points and non-anchor points except for the second highest scale or the highest scale, and the occupancy code values ​​of some groups of the anchor points and non-anchor points at the second highest scale or the highest scale.

[0078] In this embodiment, when the preset scale is the second highest scale or the highest scale, probabilistic prediction is performed on the anchor points and non-anchor points in the target group, that is, the point cloud data with high bit rate is nearly losslessly encoded at a high scale, and only part of the group is encoded and transmitted; this helps to improve coding efficiency, reduce data volume, optimize the storage and transmission of point cloud data, realize a progressive coding process, and at the same time achieve ultra-high fidelity at a high bit rate.

[0079] In this embodiment, the point cloud data is scaled by upsampling and downsampling, so that anchor points and non-anchor points distributed at different scales can be obtained. The reasonable distribution of anchor points and non-anchor points at different scales can balance the details of the data and the computational complexity, save memory, and improve performance in disguise. By encoding and generating a bit stream according to the distribution probability of the occupancy code values ​​of the anchor points and non-anchor points in the point cloud data at different scales, the 4-bit occupancy code used in this embodiment is more flexible and avoids the serial encoding problem within the scale of some octrees, which greatly reduces the encoding and decoding time and improves the point cloud data processing performance.

[0080] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the first embodiment can be referred to the above introduction and will not be described in detail. Step A10, the point cloud data processing method includes:

[0081] Based on the distribution probability of the occupancy code value of the existing anchor point being any value between 0 and 15, a probability prediction is performed on the occupancy code value of the next anchor point to obtain the distribution probability of the occupancy code value of the next anchor point being any value between 0 and 15, until the distribution probability of the occupancy code value of all anchor points being any value between 0 and 15 is predicted.

[0082] It should be noted that in point cloud data encoding, the distribution probability of the occupancy code value of an existing anchor point refers to the distribution probability of the occupancy code value of the already predicted anchor point being any value between 0 and 15. The next anchor point refers to the data point to be encoded. Based on the distribution probability of the occupancy code value of the existing anchor point, the occupancy code value of the next anchor point is probabilistically predicted. This process utilizes the spatial correlation between points in the point cloud data.

[0083] In this embodiment, the distribution probability of the occupancy code value of the next anchor point is predicted based on the occupancy code values ​​of all the current anchor points. This process helps to improve coding efficiency and thus improve coding performance, reduce the amount of data to control memory, and optimize data storage and transmission.

[0084] In one embodiment, in step A10, the method for processing point cloud data further includes:

[0085] Based on the distribution probability of the occupancy code value of the anchor point being any value between 0 and 15 and the distribution probability of the occupancy code value of the existing non-anchor point being any value between 0 and 15, a probability prediction is performed on the occupancy code value of the next non-anchor point to obtain the distribution probability of the occupancy code value of the next non-anchor point being any value between 0 and 15, until the distribution probability of the occupancy code value of all non-anchor points being any value between 0 and 15 is predicted.

[0086] It should be noted that, based on the distribution probability of the occupied code value of the existing anchor point and the distribution probability of the occupied code value of the existing non-anchor point, the distribution probability of the occupied code value of the next non-anchor point is predicted to be any value between 0 and 15; wherein, the distribution probability of the occupied code value of the existing anchor point refers to the distribution probability of the occupied code value of the anchor point that has been predicted to be any value between 0 and 15, and the distribution probability of the occupied code value of the non-anchor point refers to the distribution probability of the occupied code value of the non-anchor point that has been predicted to be any value between 0 and 15.

[0087] The next non-anchor point refers to a data point beyond the existing non-anchor points. Using the already encoded data point (the occupied code value of the current non-anchor point) to predict the probability distribution of the subsequent data point (the occupied code value of the next non-anchor point) can leverage the correlation between data points to improve prediction accuracy and coding efficiency.

[0088] Exemplarily: based on the first distribution probability, a probability prediction is performed on the occupancy code value of the non-anchor point A to obtain a distribution probability of the occupancy code value of the non-anchor point A;

[0089] Probabilistically predicting the occupancy code value of the non-anchor point B based on the distribution probability of the occupancy code value of the non-anchor point A to obtain the distribution probability of the occupancy code value of the non-anchor point B;

[0090] Probabilistically predicting the occupancy code value of the non-anchor point C based on the distribution probability of the occupancy code value of the non-anchor point A and the distribution probability of the occupancy code value of the non-anchor point B to obtain the distribution probability of the occupancy code value of the non-anchor point C;

[0091] The process repeats until the distribution probability of the occupancy code value of all non-anchor points being any value between 0 and 15 is predicted.

[0092] In this embodiment, in point cloud data encoding, an iterative probability prediction strategy is used to predict the distribution probability of the next non-anchor point occupancy code value by taking the first distribution probability and the occupancy code values ​​of all currently processed non-anchor points as known information, until the distribution probability of the occupancy code values ​​of all non-anchor points between 0 and 15 is predicted. This iterative prediction technology makes full use of the spatial correlation between points in point cloud data and is the key to achieving efficient point cloud data compression.

[0093] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those in the first embodiment can be referred to above and will not be described in detail. After obtaining the bit stream of the occupied code value of the non-anchor preset group, steps F10 to F40 are also included:

[0094] Step F10, based on the occupancy code values ​​of the anchor points and non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code values ​​of the anchor points and non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, probabilistically predicting the occupancy code values ​​of the remaining anchor points in the remaining groups other than the target group, to obtain the occupancy code values ​​of the remaining anchor points;

[0095] It should be noted that, based on the occupancy code values ​​of anchor points and non-anchor points at other scales, as well as the occupancy code values ​​of anchor points and non-anchor points in the target group at the current scale, a probabilistic prediction is made of the occupancy code values ​​of the remaining anchor points in the remaining groups other than the target group to obtain the occupancy code values ​​of the remaining anchor points; wherein, target grouping refers to a method of grouping points in point cloud data according to a rule, which can be based on spatial position. For example, in this embodiment, the number of target groups can be 4 groups, that is, the occupancy code values ​​of 4 groups of anchor points are first encoded and transmitted. The remaining groups other than the target group refer to the point cloud data groups that are not encoded at the current scale during the encoding process. In point cloud data encoding, probabilistic prediction is usually based on existing contextual information.

[0096] Step F20, constructing an occupancy code value of the anchor point when the preset scale is the second highest scale or the highest scale based on the occupancy code value of the anchor point in the target group and the occupancy code values ​​of the remaining anchor points;

[0097] It should be noted that constructing means concatenating the occupancy code values ​​of the anchor points in the target group and the occupancy code values ​​of the remaining anchor points to construct the occupancy code value of the anchor point when the preset scale is the second highest scale or the highest scale.

[0098] Step F30, based on the occupancy code values ​​of the anchor points and non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code values ​​of the anchor points and the occupancy code values ​​of the non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, probabilistically predicting the occupancy code values ​​of the remaining non-anchor points in the remaining groups other than the target group, to obtain the occupancy code values ​​of the remaining non-anchor points;

[0099] It should be noted that the non-anchor points that are not encoded at the current scale are referred to as the remaining non-anchor points in the remaining groups other than the target group. Based on the occupied code values ​​of the anchor points and non-anchor points when the preset scale is less than the next highest or highest scale, the occupied code values ​​of the anchor points when the preset scale is the next highest or highest scale, and the occupied code values ​​of the non-anchor points in the target group, the occupied code values ​​of the remaining non-anchor points are probabilistically predicted to obtain the occupied code values ​​of the remaining non-anchor points.

[0100] Step F40 : constructing the occupancy code value of the non-anchor point when the preset scale is the second highest scale or the highest scale according to the occupancy code value of the non-anchor point in the target group and the occupancy code value of the remaining non-anchor points.

[0101] It should be noted that constructing means concatenating the occupied code values ​​of the non-anchor points in the target group and the occupied code values ​​of the non-anchor points outside the target group to obtain the occupied code values ​​of the non-anchor points when the preset scale is the second highest scale or the highest scale.

[0102] In this embodiment, first, the occupancy code values ​​of the remaining anchor points are predicted using the known occupancy code value information of the anchor points and non-anchor points, as well as the occupancy code values ​​of the anchor points and non-anchor points in the target group at the next highest scale or the highest scale. Then, the complete occupancy code values ​​of the anchor points at the next highest scale or the highest scale are constructed by integrating the occupancy code values ​​of the anchor points in the target group and the occupancy code values ​​of the remaining anchor points.

[0103] Then, the known occupancy code value information is used to predict the uncoded non-anchor point data to obtain the occupancy code values ​​of the non-anchor points outside the target group; finally, by integrating the occupancy code values ​​of the non-anchor points in the target group and the occupancy code values ​​of the remaining non-anchor points, the complete occupancy code values ​​of the non-anchor points at the second highest or highest scale are constructed.

[0104] Probabilistic prediction achieves a balance between memory and performance in point cloud data encoding, reducing data redundancy. By processing data at different scales, we optimize the allocation of computing resources, balancing memory usage and computing performance, thereby maintaining data accuracy while reducing storage and transmission resource consumption.

[0105] Exemplarily, the method for processing point cloud data, applied to a decoding end, includes:

[0106] receiving a bit stream of occupancy code values ​​of anchor points and non-anchor points at different scales;

[0107] The bit streams of the occupied code values ​​of the anchor points and non-anchor points are decoded at different scales to obtain the occupied code values ​​of the anchor points and non-anchor points at different scales.

[0108] The step of decoding the bit streams of the occupied code values ​​of the anchor points and non-anchor points at different scales comprises:

[0109] If the preset scale is smaller than the second highest scale or the highest scale, predicting a first distribution probability of the anchor point occupancy code value being any value between 0 and 15 based on the bit stream of the anchor point occupancy code value, and predicting the anchor point occupancy code value based on the first distribution probability;

[0110] According to the bit code stream of the occupied code value of the non-anchor point, a second distribution probability of the occupied code value of the non-anchor point being any value from 0 to 15 is predicted, and the occupied code value of the non-anchor point is predicted according to the second distribution probability.

[0111] The step of decoding the bit streams of the occupied code values ​​of the anchor points and non-anchor points at different scales further comprises:

[0112] Probabilistically predict the occupancy code values ​​of the remaining anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code values ​​of the anchor points and non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining anchor points;

[0113] Constructing, based on the occupancy code values ​​of the anchor points in the target group and the occupancy code values ​​of the remaining anchor points, the occupancy code value of the anchor point when the preset scale is the second highest scale or the highest scale;

[0114] Probabilistically predicting the occupancy code values ​​of the remaining non-anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and the non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code value of the anchor points and the occupancy code values ​​of the non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining non-anchor points;

[0115] According to the occupancy code values ​​of the non-anchor points in the target group and the occupancy code values ​​of the remaining non-anchor points, the occupancy code values ​​of the non-anchor points when the preset scale is the second highest scale or the highest scale are constructed.

[0116] The above method for processing the point cloud data applied at the decoding end is the inverse process of the method for processing the point cloud data applied at the encoding end. The specific implementation steps and details are not repeated here.

[0117] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the method for processing point cloud data of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0118] This application also provides a point cloud data processing device, please refer to Figure 4 , the point cloud data processing device includes:

[0119] A partitioning module 10 is used to partition the point cloud data by scale through up and down sampling to obtain anchor points and non-anchor points distributed at different scales;

[0120] The encoding module 20 is configured to encode the distribution probabilities of the occupancy code values ​​of the anchor points and the occupancy code values ​​of the non-anchor points at different scales, respectively, to obtain bit streams of the occupancy code values ​​of the anchor points and the non-anchor points at different scales.

[0121] Optionally, the encoding module 20 is further configured to, if the preset scale is smaller than the second highest scale or the highest scale, perform probability prediction on the occupancy code value of the anchor point to obtain a first distribution probability that the occupancy code value of the anchor point is any value between 0 and 15;

[0122] Probabilistically predicting the occupancy code value of the non-anchor point according to the first distribution probability to obtain a second distribution probability of the occupancy code value of the non-anchor point being any value from 0 to 15;

[0123] Encoding the first distribution probability to obtain a bit stream of an occupancy code value of an anchor point;

[0124] The second distribution probability is encoded to obtain a bit stream of occupied code values ​​of non-anchor points.

[0125] Optionally, the encoding module 20 is further configured to perform a probability prediction on the occupancy code value of the next anchor point based on the distribution probability that the occupancy code value of the existing anchor point is any value between 0 and 15, and obtain the distribution probability that the occupancy code value of the next anchor point is any value between 0 and 15, until the distribution probability that the occupancy code value of all anchor points is any value between 0 and 15 is obtained.

[0126] Optionally, the encoding module 20 is further configured to perform a probability prediction on the occupancy code value of the next non-anchor point based on the distribution probability that the occupancy code value of the anchor point is any value between 0 and 15 and the distribution probability that the occupancy code value of the existing non-anchor point is any value between 0 and 15, thereby obtaining the distribution probability that the occupancy code value of the next non-anchor point is any value between 0 and 15, until the distribution probability that the occupancy code value of all non-anchor points is any value between 0 and 15 is obtained.

[0127] Optionally, the encoding module 20 is further configured to, if the preset scale is the second highest scale or the highest scale, group the anchor points and non-anchor points according to a preset number of groups, perform probability prediction on the occupancy code values ​​of the anchor points in the target group, and obtain a first distribution probability of the occupancy code values ​​of the anchor points in the target group;

[0128] Probabilistically predicting the occupied code values ​​of the non-anchor points in the target group according to the first distribution probability to obtain a second distribution probability of the occupied code values ​​of the non-anchor points in the target group;

[0129] Encoding the first distribution probability to obtain a bit code stream of an occupancy code value of an anchor point in a target group;

[0130] The second distribution probability is encoded to obtain a bit code stream of occupied code values ​​of non-anchor points in the target group.

[0131] Optionally, the encoding module 20 is further configured to probabilistically predict the occupancy code values ​​of the remaining anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code values ​​of the anchor points and non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining anchor points;

[0132] Constructing, according to the occupancy code values ​​of the anchor points in the target group and the occupancy code values ​​of the remaining anchor points, an occupancy code value of the anchor point when the preset scale is the second highest scale or the highest scale;

[0133] Probabilistically predicting the occupancy code values ​​of the remaining non-anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and the non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code value of the anchor points and the occupancy code values ​​of the non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining non-anchor points;

[0134] According to the occupancy code values ​​of the non-anchor points in the target group and the occupancy code values ​​of the remaining non-anchor points, the occupancy code values ​​of the non-anchor points when the preset scale is the second highest scale or the highest scale are constructed.

[0135] The point cloud data processing device provided in this application utilizes the point cloud data processing method of the aforementioned embodiment to resolve the technical issue of low point cloud data processing performance. Compared to the prior art, the point cloud data processing device provided in this application achieves the same beneficial effects as the point cloud data processing method of the aforementioned embodiment. Other technical features of the point cloud data processing device are the same as those disclosed in the aforementioned embodiment and are not further elaborated upon here.

[0136] The present application provides a point cloud data processing device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the point cloud data processing method in the above-mentioned first embodiment.

[0137] Reference below Figure 5 , which shows a schematic diagram of the structure of a point cloud data processing device suitable for implementing the embodiments of the present application. The point cloud data processing device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (such as in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 5 The point cloud data processing device shown is merely an example and should not limit the functions and scope of use of the embodiments of the present application.

[0138] like Figure 5As shown, the point cloud data processing device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 1002 or programs loaded from a storage device 1003 into a random access memory (RAM) 1004. RAM 1004 also stores various programs and data required for the operation of the point cloud data processing device. Processing device 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage device 1003 including, for example, a magnetic tape, hard disk, etc.; and communication devices 1009. The communication device 1009 can allow the point cloud data processing device to communicate wirelessly or wired with other devices to exchange data. Although the figure shows a point cloud data processing device with various systems, it should be understood that it is not required to implement or have all of the systems shown. More or fewer systems may be implemented or have alternatively.

[0139] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0140] The point cloud data processing device provided in this application, employing the point cloud data processing method of the aforementioned embodiment, can resolve the technical issue of low point cloud data processing performance. Compared to the prior art, the beneficial effects of the point cloud data processing device provided in this application are the same as those of the point cloud data processing method provided in the aforementioned embodiment. Other technical features of the point cloud data processing device are the same as those disclosed in the aforementioned embodiment and are not further elaborated here.

[0141] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0142] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0143] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, a computer program) stored thereon, and the computer-readable program instructions are used to execute the point cloud data processing method in the above-mentioned embodiment.

[0144] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0145] The computer-readable storage medium may be included in the point cloud data processing device; or may exist independently without being incorporated into the point cloud data processing device.

[0146] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by a point cloud data processing device, the point cloud data processing device is caused to: divide the point cloud data by scale through upsampling and downsampling to obtain anchor points and non-anchor points distributed at different scales;

[0147] The distribution probabilities of the occupied code values ​​of the anchor points and the occupied code values ​​of the non-anchor points are encoded at different scales to obtain bit streams of the occupied code values ​​of the anchor points and the non-anchor points at different scales.

[0148] Computer program code for performing the operations of the present application may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0149] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, program segment or a part of code, and the module, program segment or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented by a dedicated hardware-based system that performs the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions.

[0150] The modules described in the embodiments of the present application may be implemented in software or hardware, wherein the name of a module does not necessarily limit the unit itself.

[0151] The computer-readable storage medium provided in this application stores computer-readable program instructions (i.e., a computer program) for executing the aforementioned point cloud data processing method, thereby resolving the technical issue of low point cloud data processing performance. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the point cloud data processing method provided in the aforementioned embodiment, and are not further elaborated here.

[0152] The present application also provides a computer program product, comprising a computer program, which implements the steps of the above-mentioned point cloud data processing method when executed by a processor.

[0153] The computer program product provided in this application can solve the technical problem of low performance in processing point cloud data. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the point cloud data processing method provided in the above embodiment, and will not be repeated here.

[0154] The above description is only part of the embodiments of the present application and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for processing point cloud data, characterized in that: The method comprises: The point cloud data is divided by scale through up and down sampling to obtain anchor points and non-anchor points distributed at different scales; The distribution probabilities of the occupied code values ​​of the anchor points and the occupied code values ​​of the non-anchor points are encoded at different scales to obtain bit streams of the occupied code values ​​of the anchor points and the non-anchor points at different scales.

2. The method according to claim 1, wherein The step of encoding the distribution probabilities of the occupancy code values ​​of the anchor points and the occupancy code values ​​of the non-anchor points at different scales to obtain bit streams of the occupancy code values ​​of the anchor points and the non-anchor points at different scales comprises: If the preset scale is smaller than the second highest scale or the highest scale, a probability prediction is performed on the occupancy code value of the anchor point to obtain a first distribution probability of the occupancy code value of the anchor point being any value between 0 and 15; Probabilistically predicting the occupancy code value of the non-anchor point according to the first distribution probability to obtain a second distribution probability of the occupancy code value of the non-anchor point being any value from 0 to 15; Encoding the first distribution probability to obtain a bit stream of an occupancy code value of an anchor point; The second distribution probability is encoded to obtain a bit stream of occupied code values ​​of non-anchor points.

3. The method according to claim 2, wherein The step of probabilistically predicting the occupancy code value of the anchor point to obtain a first distribution probability of the occupancy code value of the anchor point being any value from 0 to 15 comprises: Based on the distribution probability of the occupancy code value of the existing anchor point being any value between 0 and 15, a probability prediction is performed on the occupancy code value of the next anchor point to obtain the distribution probability of the occupancy code value of the next anchor point being any value between 0 and 15, until the distribution probability of the occupancy code value of all anchor points being any value between 0 and 15 is predicted.

4. The method according to claim 2, wherein The step of probabilistically predicting the occupancy code value of the non-anchor point according to the first distribution probability to obtain a second distribution probability of the occupancy code value of the non-anchor point being any value from 0 to 15 comprises: Based on the distribution probability of the occupancy code value of the anchor point being any value between 0 and 15 and the distribution probability of the occupancy code value of the existing non-anchor point being any value between 0 and 15, a probability prediction is performed on the occupancy code value of the next non-anchor point to obtain the distribution probability of the occupancy code value of the next non-anchor point being any value between 0 and 15, until the distribution probability of the occupancy code value of all non-anchor points being any value between 0 and 15 is predicted.

5. The method according to claim 1, wherein The step of encoding the occupancy code value of the anchor point and the occupancy code value of the non-anchor point at different scales to obtain the bit code streams of the occupancy code values ​​of the anchor point and the non-anchor point at different scales further comprises: If the preset scale is the second highest scale or the highest scale, the anchor points and non-anchor points are grouped, and the occupancy code values ​​of the anchor points in the target group are probability predicted to obtain a first distribution probability of the occupancy code values ​​of the anchor points in the target group; Probabilistically predicting the occupied code values ​​of the non-anchor points in the target group according to the first distribution probability to obtain a second distribution probability of the occupied code values ​​of the non-anchor points in the target group; Encoding the first distribution probability to obtain a bit code stream of an occupancy code value of an anchor point in a target group; The second distribution probability is encoded to obtain a bit code stream of occupied code values ​​of non-anchor points in the target group.

6. The method according to claim 5, wherein After the step of obtaining the bit code stream of the occupied code value of the non-anchor point in the target group, the method includes: Probabilistically predicting the occupancy code values ​​of the remaining anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code values ​​of the anchor points and non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining anchor points; Constructing, according to the occupancy code values ​​of the anchor points in the target group and the occupancy code values ​​of the remaining anchor points, an occupancy code value of the anchor point when the preset scale is the second highest scale or the highest scale; Probabilistically predicting the occupancy code values ​​of the remaining non-anchor points in the remaining groups other than the target group based on the occupancy code values ​​of the anchor points and the non-anchor points when the preset scale is smaller than the second highest scale or the highest scale, and the occupancy code value of the anchor points and the occupancy code values ​​of the non-anchor points in the target group when the preset scale is the second highest scale or the highest scale, to obtain the occupancy code values ​​of the remaining non-anchor points; According to the occupancy code values ​​of the non-anchor points in the target group and the occupancy code values ​​of the remaining non-anchor points, the occupancy code values ​​of the non-anchor points when the preset scale is the second highest scale or the highest scale are constructed.

7. A point cloud data processing device, characterized in that: The device comprises: The partitioning module is used to partition the point cloud data by scale through up and down sampling to obtain anchor points and non-anchor points distributed at different scales; The encoding module is used to encode the distribution probabilities of the occupancy code values ​​of the anchor points and the occupancy code values ​​of the non-anchor points at different scales, respectively, to obtain bit code streams of the occupancy code values ​​of the anchor points and the non-anchor points at different scales.

8. A point cloud data processing device, characterized in that: The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for processing point cloud data according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the point cloud data processing method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the method for processing point cloud data according to any one of claims 1 to 6 are implemented.

Images