A V-PCC decoding unit division method, system and device
By calculating JND values and improving JND thresholds, the decoding tree unit of point cloud video is divided, and the problem of insufficient perception ability of unoccupied areas in point cloud video compression is solved, encoding efficiency and visual quality are improved, and the depth range is achieved. Statistical and coding mode decisions.
Patent Information
- Application Number
- CN202410867866.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-01
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-07-01
AI Technical Summary
During point cloud video compression, the brightness and depth perception capabilities of existing JND models in unoccupied areas are reduced, resulting in coding efficiency and visual quality problems. Especially in V-PCC technology, traditional JND models are difficult to effectively divide the decoding units in traditional JND models. , affects coding mode decisions.
By calculating the JND values of unoccupied and occupied pixels in attribute frames and geometric frames and improving JND thresholds, they are divided into visually sensitive, medium visually sensitive and non-visually sensitive decoding tree units, and adjusting JND using brightness adaptive and depth masking factors. Threshold, optimized coding unit division to improve depth range statistics and coding mode decisions.
The subjective quality and encoding efficiency of point cloud video are improved, and the visual perception quality of the video is maintained to the greatest extent by reasonably dividing the decoding units, and the optimal depth range is determined to optimize the encoding mode.
Smart Images

Figure CN118450143B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of point cloud compression technology, and in particular to a method, system and device for dividing a V-PCC decoding unit. Background Art
[0002] In order to effectively compress dynamic point clouds, the Motion Picture Experts Group (MPEG) released two point cloud compression coding standards, namely, Geometry-based Point Cloud Compression (G-PCC) and Video-based Point Cloud Compression (V-PCC).
[0003] In the V-PCC process, the dynamic point cloud is often first converted into a 2D video (several frames of 2D images) through a patch projection method; when these patches are packed into 2D images, geometric videos (several frames of geometric images) and attribute videos (several frames of attribute images) are generated respectively; since patches usually have irregular shapes, an occupancy map is generated to indicate whether a pixel in the 2D image belongs to the patch; then the occupancy map, geometric video and attribute video are compressed through a video decoder, such as High Efficiency Video Coding (HEVC) and Versatile Video Coding (VCC).
[0004] In HEVC / VVC, the encoder first divides the input video sequence into coding tree units (CTUs), which can usually be divided into several coding units (CUs) of different sizes.
[0005] At the same time, there is also a coding technology based on the Just Noticeable Difference (JND) model, which only focuses on the details that the human eye can perceive, while ignoring the small changes that are insignificant to the human eye. Therefore, it can effectively reduce the amount of encoded data while maintaining the visual perception quality as much as possible. The JND in the JND algorithm is usually used to represent the threshold of the minimum change in the visual signal that the human eye can perceive.
[0006] However, when the JND model is used in the point cloud video decoding process, the model performance often decreases significantly. The reasons are briefly described as follows:
[0007] For attribute videos, the brightness perception ability of the traditional JND model is reduced due to the use of pixel filling technology in unoccupied areas. Specifically, the JND threshold is used to describe the sensitivity of the human visual system (HVS) to stimulus changes. The three-dimensional points of the dynamic point cloud are spatially projected to generate attribute videos and geometric videos. Due to the irregular shape of the projection patches, there are large areas of unoccupied areas in the attribute video. In order to improve coding efficiency, V-PCC uses the original relevant attribute information in the dynamic point cloud for filling. Although this method improves coding efficiency, it also leads to a decrease in the local contrast of unoccupied areas of the attribute video frame. If the traditional JND model is used directly in the attribute video, it will lead to large errors in the prediction of the JND value;
[0008] For geometric videos, due to the depth filling technology used in unoccupied areas, the depth perception ability of the depth-based JND (JNDD) model in the boundary area is reduced. Specifically, the geometric video represents the depth information of the dynamic point cloud. After the filling operation, there are still large areas of unoccupied areas. The unoccupied areas are filled by the depth filling technology (the residual value between the original block and the predicted block). Since the human eye has different sensitivities to depth changes, HVS also has a visual masking effect on depth information. When using the JNDD model to describe the visual masking effect of the human eye on depth changes, there is a significant depth difference between the texture area and the dark background area in the unfilled geometric video frame. However, the depth filling technology will make the depth changes at the boundary of the geometric video after filling not obvious, which will cause the JNDD model to be unable to recognize the depth changes at the boundary of the geometric video;
[0009] Due to the visual sensitivity and visual masking effects in HVS, the human eye is not able to perceive every type of distortion. Therefore, the minimum visibility threshold of pixel intensity change can also be defined as JND. This is divided into a pixel domain JND model and a sub-band domain JND model. The sub-band domain JND model usually includes three parts: luminance adaptation (LA), contrast masking (CM) and contrast sensitivity function (CSF).
[0010] At present, there are many decoding unit division and mode decision algorithms based on the JND model. However, how to efficiently divide the decoding units while maintaining the subjective quality of the video to the greatest extent, so as to facilitate the use of existing technologies to divide depth statistics and determine the optimal depth range, and then complete the coding mode decision, has become a technical problem that needs to be urgently solved in the field of point cloud video compression. Summary of the invention
[0011] The object of the present invention is to provide a method, system and device for dividing a V-PCC decoding unit to solve at least one of the above-mentioned technical problems existing in the prior art.
[0012] In a first aspect, in order to solve the above technical problem, the present invention provides a method for dividing a V-PCC decoding unit, comprising the following steps:
[0013] Step 1: Generate a point cloud video by patch projection based on the input point cloud data; the point cloud video includes a geometry frame, an attribute frame and an occupancy map.
[0014] Step 2: Based on the occupancy map, the brightness adaptive values of the unoccupied pixels and the occupied pixels in the attribute frame in the JND model are calculated, so as to expand the JND threshold difference between the unoccupied pixel area and the occupied pixel area; the JND value of each pixel in the attribute frame is calculated; and the decoding tree unit in the attribute frame is divided by the JND threshold division method.
[0015] In a feasible implementation manner, the step 2 specifically includes:
[0016] Step 211: Based on the occupancy map, calculate the brightness adaptive value of the unoccupied pixels in the attribute frame in the JND model , the specific formula can be:
[0017] ;
[0018] in, Represents pixel The background brightness at Indicates the X-axis coordinate of the pixel; Indicates the Y-axis coordinate of the pixel; Indicates the modulation parameter, which can be 1023 (in 0.9x4096 state); represents the inverse perceptual brightness converter, and the specific formula can be:
[0019] ;
[0020] in, Represents the image brightness value, represents the inverse perceptual quantizer, and the specific calculation formula can be:
[0021] ;
[0022] in, , , , and They represent the parameters of the inverse perceptual quantizer respectively; represents normalized brightness;
[0023] Step 212: Calculate the brightness adaptive value of the occupied pixels in the attribute frame in the JND model , the specific formula can be:
[0024] ;
[0025] Step 213: Calculate the JND value of each pixel in the attribute frame , the specific formula can be:
[0026] ;
[0027] in, represents contrast masking; Indicates a preset constant; represents the joint brightness adaptation function, and the specific formula can be:
[0028] ;
[0029] in, Indicates the occupancy status of the pixel, with a value of 1 for occupied pixels and 0 for unoccupied pixels;
[0030] Step 214: Based on the JND value, the coding tree units (CTUs) in the attribute frame are divided into three categories by using a JND threshold division method: visually sensitive coding tree units (VS CTUs), medium visually sensitive coding tree units (MVS CTUs), and non-visually sensitive coding tree units (NVS CTUs).
[0031] In a feasible implementation manner, the JND threshold includes a first threshold and a second threshold, and the mathematical expression of the specific division method may be:
[0032] ;
[0033] in, represents the first threshold, represents the second threshold value, which can be obtained by statistical analysis of the official test sequence.
[0034] In a feasible implementation manner, the step 2 further includes:
[0035] Step 221: Based on the occupancy map, calculate the improved JND thresholds of unoccupied pixels and occupied pixels in the geometric frame in the JNDD model. , the specific formula can be:
[0036] ;
[0037] in, Represents the visual masking factor, which takes the value of 1 for occupied pixels and 0 for unoccupied pixels; represents the JND threshold in the traditional JNDD model. The specific formula can be:
[0038] ;
[0039] in, Represents pixel Depth level at
[0040] Step 222: Based on the JND value, the decoding tree units in the geometric frame are divided into three categories by using a JND threshold division method: visually sensitive decoding tree units, medium visually sensitive decoding tree units, and non-visually sensitive decoding tree units.
[0041] In a feasible implementation manner, the step 2 further includes: after executing step 213, firstly dividing the CTU in the attribute frame into a near layer CTU and a far layer CTU, and then executing step 214 respectively based on the near layer CTU and / or the far layer CTU; in this way, the characteristic that there is a significant difference in the division depth of the coding unit between the far layer frame and the near layer frame in the V-PCC can be utilized, and the division depths of the far layer CTU and the near layer CTU in the attribute frame can be counted respectively, so as to obtain a better depth range.
[0042] In a feasible implementation, step 2 also includes: after executing step 221, first dividing the CTU in the geometric frame into a near-layer CTU and a far-layer CTU, and then executing step 222 respectively based on the near-layer CTU and / or the far-layer CTU; in this way, the characteristic that there is a significant difference in the division depth of the coding unit between the far-layer frame and the near-layer frame in V-PCC can be utilized, and the division depths of the far-layer CTU and the near-layer CTU in the geometric frame can be counted respectively, so as to obtain the optimal depth range.
[0043] Step 3: Based on the decoding tree unit division result, the division depth statistics are performed respectively to determine the corresponding optimal depth range so as to make a subsequent coding mode decision for the CTU.
[0044] In the second aspect, based on the same inventive concept, the present application also provides a V-PCC decoding unit division system using the above method, including a data receiving module, a data processing module and a result generating module:
[0045] The data receiving module is used to receive input point cloud data;
[0046] The data processing module includes a point cloud video unit and an attribute frame division unit;
[0047] The point cloud video unit generates a point cloud video by patch projection according to the input point cloud data; the point cloud video includes a geometry frame, an attribute frame and an occupancy map;
[0048] The attribute frame division unit calculates brightness adaptive values of unoccupied pixels and occupied pixels in the attribute frame in the JND model based on the occupancy map; calculates the JND value of each pixel in the attribute frame; and divides the decoding tree unit in the attribute frame by a JND threshold division method;
[0049] The result generation module performs division depth statistics based on the decoding tree unit division results.
[0050] In a feasible implementation, the data processing module also includes a geometric frame division unit, which calculates the improved JND thresholds of unoccupied pixels and occupied pixels in the geometric frame in the JNDD model based on the occupancy map; and divides the decoding tree units in the geometric frame by a JND threshold division method based on the JND value.
[0051] In the third aspect, based on the same inventive concept, the present application also provides a V-PCC decoding unit division device, including a processor, a memory and a bus, the memory stores instructions and data that can be read by the processor, the processor is used to call the instructions and data in the memory to execute the V-PCC decoding unit division method as described above, and the bus connects the functional components for transmitting information.
[0052] By adopting the above technical solution, the present invention has the following beneficial effects:
[0053] The present invention provides a method, system and device for dividing a decoding unit of V-PCC, which helps to efficiently divide the CTU in the attribute frame according to the JND value by obtaining the JND value and the JND threshold for occupied pixels and unoccupied pixels respectively, so as to reasonably apply HVS to V-PCC, which can maintain the subjective quality of the video to the greatest extent, and is conducive to the subsequent determination of the optimal depth range and coding mode decision; the present application also uses the JND threshold of the geometric frame to efficiently divide the CTU in the geometric frame, further improving the application effect of HVS in V-PCC, and is more conducive to the subsequent determination of the optimal depth range and coding mode decision. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0055] Figure 1 A flow chart of a method for dividing a V-PCC decoding unit provided by an embodiment of the present invention;
[0056] Figure 2 for Figure 1 Specific flow chart of step 2 in ;
[0057] Figure 3 for Figure 1 Additional flow chart for step 2 in ;
[0058] Figure 4 A system diagram of a V-PCC decoding unit division method provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0059] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0060] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.
[0061] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0062] In the description of the present invention, the terms "component", "module", "system", etc. are intended to refer to computer-related entities, such as but not limited to hardware, firmware, a combination of hardware and software, software, or software in execution. For example, a component can be but not limited to: a process running on a processor, a processor, an object, an executable, an execution thread, a program, and / or a computer. For example, an application running on a computing device and the computing device can both be components. One or more components can be located within an execution process and / or an execution thread, and a component can be located on one computer and / or distributed on two or more computers. In addition, these components can be executed from various computer-readable media with various data structures stored thereon. Components can communicate by means of local and / or remote processes, such as according to signals having one or more data packets, such as data from a component that interacts with another component in a local system, a distributed system, and / or interacts with other systems by means of signals on a network such as the Internet.
[0063] In the description of the present invention, “CTU partition” and “CU partition” may be used interchangeably, “coding block”, “CU”, and “region” may be used interchangeably, and “image”, “picture”, and “frame” may be used interchangeably.
[0064] In the description of the present invention, "decoding" may mean "encoding" or "decoding", and "decoder" may mean "encoder" or "decoder" or "codec", which is determined according to the complementarity of the video encoding and decoding process.
[0065] In the description of the present invention, the terms "video" and "frame" are used interchangeably because the present application does not involve the essential difference between "video" and "frame" but mainly focuses on the frames in the video.
[0066] The present invention is further explained below in conjunction with specific implementation modes.
[0067] It should also be noted that the following specific embodiments or specific implementations are a series of optimized settings listed in the present invention to further explain the specific content of the invention, and these settings can be used in combination or in association with each other.
[0068] Embodiment 1:
[0069] like Figure 1 As shown, this embodiment provides a V-PCC decoding unit division method, including the following steps:
[0070] Step 1: Generate a point cloud video by patch projection based on the input point cloud data; the point cloud video includes a geometry frame, an attribute frame and an occupancy map.
[0071] Step 2: Based on the occupancy map, the brightness adaptive values of the unoccupied pixels and the occupied pixels in the attribute frame in the JND model are calculated, so as to expand the JND threshold difference between the unoccupied pixel area and the occupied pixel area; the JND value of each pixel in the attribute frame is calculated; and the decoding tree unit in the attribute frame is divided by the JND threshold division method.
[0072] Furthermore, if Figure 2 As shown, the step 2 specifically includes:
[0073] Step 211: Based on the occupancy map, calculate the brightness adaptive value of the unoccupied pixels in the attribute frame in the JND model , the specific formula can be:
[0074] ;
[0075] in, Represents pixel The background brightness at Indicates the X-axis coordinate of the pixel; Indicates the Y-axis coordinate of the pixel; Indicates the modulation parameter, which can be 1023 (in 0.9x4096 state); represents the inverse perceptual brightness converter, and the specific formula can be:
[0076] ;
[0077] in, Represents the image brightness value, represents the inverse perceptual quantizer, and the specific calculation formula can be:
[0078] ;
[0079] in, , , , and They represent the parameters of the inverse perceptual quantizer respectively; represents normalized brightness;
[0080] Step 212: Calculate the brightness adaptive value of the occupied pixels in the attribute frame in the JND model , the specific formula can be:
[0081] ;
[0082] Step 213: Calculate the JND value of each pixel in the attribute frame , the specific formula can be:
[0083] ;
[0084] in, represents contrast masking; Indicates a preset constant; represents the joint brightness adaptation function, and the specific formula can be:
[0085] ;
[0086] in, Indicates the occupancy status of the pixel, with a value of 1 for occupied pixels and 0 for unoccupied pixels;
[0087] Step 214: Based on the JND value, the coding tree units (CTUs) in the attribute frame are divided into three categories by using a JND threshold division method: visually sensitive coding tree units (VS CTUs), medium visually sensitive coding tree units (MVS CTUs), and non-visually sensitive coding tree units (NVS CTUs).
[0088] Furthermore, the JND threshold includes a first threshold and a second threshold, and the mathematical expression of the specific division method can be:
[0089] ;
[0090] in, represents the first threshold, Indicates the second threshold.
[0091] Furthermore, if Figure 3 As shown, the step 2 also includes:
[0092] Step 221: Based on the occupancy map, calculate the improved JND thresholds of unoccupied pixels and occupied pixels in the geometric frame in the JNDD model. , the specific formula can be:
[0093] ;
[0094] in, Represents the visual masking factor, which takes the value of 1 for occupied pixels and 0 for unoccupied pixels; represents the JND threshold in the traditional JNDD model. The specific formula can be:
[0095] ;
[0096] in, Represents pixel Depth level at
[0097] Step 222: Based on the JND value, the decoding tree units in the geometric frame are divided into three categories by using a JND threshold division method: visually sensitive decoding tree units, medium visually sensitive decoding tree units, and non-visually sensitive decoding tree units.
[0098] Furthermore, step 2 also includes: after executing step 213, first dividing the CTU in the attribute frame into a near layer CTU and a far layer CTU, and then executing step 214 respectively based on the near layer CTU and / or the far layer CTU; in this way, the characteristic that there is a significant difference in the division depth of the coding unit between the far layer frame and the near layer frame in the V-PCC can be utilized, and the division depths of the far layer CTU and the near layer CTU in the attribute frame can be counted respectively, so as to obtain a better depth range.
[0099] Furthermore, the step 2 also includes: after executing step 221, first dividing the CTU in the geometric frame into a near-layer CTU and a far-layer CTU, and then executing step 222 respectively based on the near-layer CTU and / or the far-layer CTU; in this way, the characteristic that there is a significant difference in the division depth of the coding unit between the far-layer frame and the near-layer frame in V-PCC can be utilized, and the division depths of the far-layer CTU and the near-layer CTU in the geometric frame can be counted respectively, so as to obtain the optimal depth range.
[0100] Step 3: Based on the decoding tree unit division result, the division depth statistics are respectively performed to determine the corresponding optimal depth range so as to make a subsequent coding mode decision for the CTU; the division depth statistics can utilize the division depth statistics method in the prior art.
[0101] Embodiment 2:
[0102] like Figure 4 As shown, this embodiment provides a V-PCC decoding unit division system using the above method, including a data receiving module, a data processing module and a result generating module:
[0103] The data receiving module is used to receive input point cloud data;
[0104] The data processing module includes a point cloud video unit and an attribute frame division unit;
[0105] The point cloud video unit generates a point cloud video by patch projection according to the input point cloud data; the point cloud video includes a geometry frame, an attribute frame and an occupancy map;
[0106] The attribute frame division unit calculates brightness adaptive values of unoccupied pixels and occupied pixels in the attribute frame in the JND model based on the occupancy map; calculates the JND value of each pixel in the attribute frame; and divides the decoding tree unit in the attribute frame by a JND threshold division method;
[0107] The result generation module performs division depth statistics based on the decoding tree unit division results.
[0108] Furthermore, the data processing module also includes a geometric frame division unit, which calculates the improved JND thresholds of unoccupied pixels and occupied pixels in the geometric frame in the JNDD model based on the occupancy map; and divides the decoding tree units in the geometric frame by a JND threshold division method based on the JND value.
[0109] Embodiment three:
[0110] This embodiment provides a decoding unit division device for V-PCC, including a processor, a memory and a bus, wherein the memory stores instructions and data readable by the processor, the processor is used to call the instructions and data in the memory to execute the decoding unit division method of V-PCC as described above, and the bus connects the functional components to transmit information.
[0111] In another embodiment, the present solution can be implemented by a device, which may include a corresponding module for performing each or several steps in the above-mentioned embodiments. The module may be one or more hardware modules specially configured to perform the corresponding steps, or implemented by a processor configured to perform the corresponding steps, or stored in a computer-readable medium for implementation by a processor, or implemented by some combination.
[0112] The processor performs the various methods and processes described above. For example, the method implementation in the present solution can be implemented as a software program, which is tangibly contained in a machine-readable medium, such as a memory. In some embodiments, part or all of the software program can be loaded and / or installed via a memory and / or a communication interface. When the software program is loaded into the memory and executed by the processor, one or more steps in the method described above can be executed. Alternatively, in other embodiments, the processor can be configured to perform one of the above methods by any other appropriate means (for example, by means of firmware).
[0113] The device can be implemented using a bus architecture. The bus architecture can include any number of interconnecting buses and bridges, depending on the specific application and overall design constraints of the hardware. The bus connects various circuits including one or more processors, memories, and / or hardware modules together. The bus can also connect various other circuits such as peripherals, voltage regulators, power management circuits, external antennas, etc.
[0114] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Component (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.
[0115] The video encoder may be implemented or executed using a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, or any combination thereof designed to perform the functions described herein. The general purpose processor may be a microprocessor, but alternatively, the processor may also be any conventional processor, controller, microcontroller, or state machine.
[0116] The processor may also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a combination of multiple microprocessors, a combination of one or more microprocessors and a DSP core, or any other such structure. In addition, at least one processor may include one or more modules operable to perform one or more steps and / or operations described above.
[0117] When a video encoder is implemented using hardware circuits such as ASIC, FPGA, etc., it may include various circuit blocks configured to perform various functions. Those skilled in the art may design and implement these circuits in various ways according to various constraints imposed on the entire system to implement the various functions disclosed in the present invention.
[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for dividing a V-PCC decoding unit, characterized in that: include: Step 1: Generate point cloud video by patch projection based on input point cloud data; The point cloud video includes a geometry frame, an attribute frame and an occupancy map; Step 2: Based on the occupancy map, calculate the brightness adaptation values of the unoccupied pixels and the occupied pixels in the attribute frame in the JND model; Calculate the JND value of each pixel in the attribute frame; divide the decoding tree unit in the attribute frame by the JND threshold division method; the JND model represents the minimum perceptible distortion model; the JND value represents the threshold value at which the human eye can perceive the minimum change of the visual signal; the JND threshold value is used to describe the sensitivity of the human visual system to stimulus changes; Step 3: Based on the decoding tree unit division results, the division depth statistics are performed respectively; The step 2 specifically includes: Step 211: Based on the occupancy map, calculate the brightness adaptive value of the unoccupied pixels in the attribute frame in the JND model , the specific formula is: ; in, Represents pixel The background brightness at Indicates the X-axis coordinate of the pixel; Indicates the Y-axis coordinate of the pixel; represents the modulation parameters; represents the inverse perceptual brightness converter, and the specific formula is: ; in, Represents the image brightness value, represents the inverse perceptual quantizer, and the specific calculation formula is: ; in, , , , and They represent the parameters of the inverse perceptual quantizer respectively; represents normalized brightness; Step 212: Calculate the brightness adaptive value of the occupied pixels in the attribute frame in the JND model , the specific formula is: ; Step 213: Calculate the JND value of each pixel in the attribute frame , the specific formula is: ; in, represents contrast masking; Indicates a preset constant; represents the joint brightness adaptation function, and the specific formula is: ; in, Indicates the occupancy status of the pixel, with a value of 1 for occupied pixels and 0 for unoccupied pixels; Step 214: Based on the JND value, the decoding tree units in the attribute frame are divided into three categories by using a JND threshold division method: visually sensitive decoding tree units, medium visually sensitive decoding tree units, and non-visually sensitive decoding tree units.
2. The method according to claim 1, characterized in that The modulation parameter value in step 211 is 1023.
3. The method according to claim 1, characterized in that The JND threshold includes a first threshold and a second threshold. The mathematical expression of the specific division method is: ; in, represents the first threshold, represents the second threshold, represents the non-visually sensitive decoding tree unit, represents a medium visually sensitive decoding tree unit, Represents a visually sensitive decoding tree unit.
4. The method according to claim 1, characterized in that: The step 2 also includes: Step 221: Based on the occupancy map, calculate the improved JND thresholds of unoccupied pixels and occupied pixels in the geometric frame in the JNDD model. , the JNDD model represents a depth-based JND model; the specific formula is: ; in, Represents the visual masking factor, which takes the value of 1 for occupied pixels and 0 for unoccupied pixels; It represents the JND threshold in the traditional JNDD model. The specific formula is: ; in, Represents pixel Depth level at Step 222: Based on the JND value, the decoding tree units in the geometric frame are divided into three categories by using a JND threshold division method: visually sensitive decoding tree units, medium visually sensitive decoding tree units, and non-visually sensitive decoding tree units.
5. The method according to claim 1, characterized in that: The step 2 further includes: after executing step 213, firstly dividing the CTU in the attribute frame into a near layer CTU and a far layer CTU, and then executing step 214 respectively based on the near layer CTU and / or the far layer CTU; the CTU represents a decoding tree unit.
6. The method according to claim 4, characterized in that The step 2 also includes: after executing step 221, firstly dividing the CTU in the geometric frame into a near layer CTU and a far layer CTU, and then executing step 222 respectively based on the near layer CTU and / or the far layer CTU; the CTU represents a decoding tree unit.
7. A V-PCC decoding unit division system, characterized in that: Including data receiving module, data processing module and result generation module: The data receiving module is used to receive input point cloud data; The data processing module includes a point cloud video unit and an attribute frame division unit; The point cloud video unit generates a point cloud video by patch projection based on the input point cloud data; The point cloud video includes a geometry frame, an attribute frame and an occupancy map; The attribute frame division unit calculates brightness adaptive values of unoccupied pixels and occupied pixels in the attribute frame in the JND model based on the occupancy map; Calculate the JND value of each pixel in the attribute frame; The decoding tree units in the attribute frame are divided by a JND threshold division method; the JND model represents a minimum perceptible distortion model; the JND value represents a threshold value at which the human eye can perceive a minimum change in a visual signal; The JND threshold is used to describe the sensitivity of the human visual system to stimulus changes; specifically, it includes: Step 211: Based on the occupancy map, calculate the brightness adaptive value of the unoccupied pixels in the attribute frame in the JND model , the specific formula is: ; in, Represents pixel The background brightness at Indicates the X-axis coordinate of the pixel; Indicates the Y-axis coordinate of the pixel; represents the modulation parameters; represents the inverse perceptual brightness converter, and the specific formula is: ; in, Represents the image brightness value, represents the inverse perceptual quantizer, and the specific calculation formula is: ; in, , , , and They represent the parameters of the inverse perceptual quantizer respectively; represents normalized brightness; Step 212: Calculate the brightness adaptive value of the occupied pixels in the attribute frame in the JND model , the specific formula is: ; Step 213: Calculate the JND value of each pixel in the attribute frame , the specific formula is: ; in, represents contrast masking; Indicates a preset constant; represents the joint brightness adaptation function, and the specific formula is: ; in, Indicates the occupancy status of the pixel, with a value of 1 for occupied pixels and 0 for unoccupied pixels; Step 214: Based on the JND value, the decoding tree units in the attribute frame are divided into three categories by using a JND threshold division method: visually sensitive decoding tree units, medium visually sensitive decoding tree units, and non-visually sensitive decoding tree units; The result generation module performs division depth statistics based on the decoding tree unit division results.
8. The system according to claim 7, characterized in that The data processing module further comprises a geometric frame partitioning unit, which calculates improved JND thresholds of unoccupied pixels and occupied pixels in the geometric frame in a JNDD model based on the occupancy map; Based on the JND value, the decoding tree unit in the geometric frame is divided by the JND threshold division method; the JNDD model represents a depth-based JND model; specifically includes: Step 221: Based on the occupancy map, calculate the improved JND thresholds of unoccupied pixels and occupied pixels in the geometric frame in the JNDD model. , the JNDD model represents a depth-based JND model; the specific formula is: ; in, Represents the visual masking factor, which takes the value of 1 for occupied pixels and 0 for unoccupied pixels; It represents the JND threshold in the traditional JNDD model. The specific formula is: ; in, Represents pixel Depth level at Step 222: Based on the JND value, the decoding tree units in the geometric frame are divided into three categories by using a JND threshold division method: visually sensitive decoding tree units, medium visually sensitive decoding tree units, and non-visually sensitive decoding tree units.
9. A decoding unit division device for V-PCC, characterized in that: It includes a processor, a memory and a bus, wherein the memory stores instructions and data read by the processor, the processor is used to call the instructions and data in the memory to execute any method as claimed in claim 1-6, and the bus connects the functional components to transmit information.
