Method and apparatus of video coding in 4:4:4 chroma format
Patent Information
- Application Number
- JP2024135648
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2020-01-25
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2041-01-05
Smart Images

Figure 2024161497000001_ABST
Abstract
Description
[Technical field]
[0001] (Related Applications) This application is a continuation of U.S. Provisional Application No. 62 / 957,273, filed January 5, 2020, entitled “METHODS AND APPARATUS ON 4:4:4 VIDEO CODING,” and U.S. Provisional Application No. 62 / 957,273, filed January 25, 2020, entitled “METHODS AND APPARATUS ON 4:4:4 VIDEO CODING,” "IDEO CODING (Method and Apparatus for 4:4:4 Video Coding)" This application claims priority to U.S. Provisional Application No. 62 / 965,859, entitled "Compounding a Polymer for Use in a Medical Device," both of which are incorporated by reference in their entireties.
[0002] (Technical field) This application relates generally to encoding and compression of video data, and more particularly to a method and system for performing a clipping operation before an adaptive color-space transform (ACT). [Background technology]
[0003] Digital video is supported by a variety of electronic devices, such as digital televisions, laptop or desktop computers, tablet computers, digital cameras, digital recording devices, digital media players, video game consoles, smart phones, video teleconferencing devices, video streaming devices, etc. Such electronic devices include video encoding and decoding technologies such as MPEG-4, ITU-T H.263, ITU-T H.264 / MPEG-4 Part 10 AVC (Advanced Video Coding), and other standards. ng), HEVC (High Efficiency Video C) Video compression is typically implemented in a variety of video applications, including video coding (VCR) and video encoding (VVC) standards, to transmit, receive, encode, decode, and / or store digital video data. Video compression generally involves performing spatial (intra-frame) prediction and / or temporal (inter-frame) prediction to reduce or remove redundancy inherent in video data. For block-based video coding, a video frame is divided into one or more slices, with each slice having multiple video blocks, which may also be referred to as coding tree units (CTUs). Each CTU may contain one coding unit (CU), and / or a coding tree unit (CU) may be used to encode, decode, and / or store video data. A CU may be recursively split into smaller CUs until a certain minimum CU size is reached. Each CU (also named leaf CU) contains one or more transform units (TUs) and one or more predictive units (PUs). Each CU can be in intra mode, inter mode or intra block mode. Video blocks in an intra-coded (I) slice in a video frame may be coded using spatial prediction with respect to reference samples in neighboring blocks within the same video frame, or in one of two inter-coded (IBC) modes. Video blocks in an intra-coded (I) slice in a video frame are coded using spatial prediction with respect to reference samples in neighboring blocks within the same video frame. Video blocks in an inter-coded (P (forward predicted image) or B (bidirectionally predicted image)) slice in a video frame may use spatial prediction with respect to reference samples in neighboring blocks within the same video frame, or may use temporal prediction with respect to reference samples in other previous and / or future reference video frames.
[0004] Spatial or temporal prediction based on previously coded reference blocks, e.g., neighboring blocks, results in a predicted block for the current video block being coded. The process of finding the reference block may be accomplished by a block matching algorithm. The residual data representing pixel differences between the current block being coded and the predictive block is called the residual block or prediction error. Inter-coded blocks are coded according to a motion vector and a residual block that point to a reference block in a reference frame forming the predictive block. The process of determining a motion vector is commonly referred to as motion prediction. Intra-coded blocks are coded according to an intra-prediction mode and a residual block. For further compression, the residual block may be transformed from the pixel domain to a transform domain, e.g., the frequency domain, resulting in residual transform coefficients, which are then quantized. The quantized transform coefficients, initially arranged in a two-dimensional array, may be scanned to generate a one-dimensional vector of transform coefficients, which are then entropy coded into a video bitstream to achieve even further compression.
[0005] The encoded video bitstream is then stored on a computer-readable recording medium (e.g., flash memory) that is accessed by another electronic device with digital video capabilities, or transmitted directly to the electronic device via wired or wireless connection. The electronic device then performs video decompression (the inverse process of video compression discussed above), e.g., by parsing the encoded video bitstream to obtain syntax elements from the bitstream, reconstructing digital video data from the encoded video bitstream into an original format based at least in part on the syntax elements obtained from the bitstream, and rendering the reconstructed digital video data on a display of the electronic device.
[0006] Digital video quality has increased from High Definition to 4Kx. As we move to 2K or 8K×4K, the amount of video data to be encoded / decoded increases exponentially, which is a continuous challenge in how to encode / decode video data more efficiently while maintaining the image quality of the decoded video data.
[0007] Certain video content, e.g., screen content video, is encoded in 4:4:4 chrominance format, where all three components (luminance and two chrominance components) have the same resolution. Although the 4:4:4 chrominance format contains more redundancy compared to the 4:2:0 and 4:2:2 chrominance formats (which are detrimental to achieving good compression efficiency), the 4:4:4 chrominance format is still the preferred encoding format for many applications where high fidelity is required to preserve color information such as sharp edges in the decoded video. Given the redundancy present in the 4:4:4 chrominance video format, there is evidence that significant coding improvements can be achieved by exploiting the correlations between the three color components of 4:4:4 video (e.g., Y, Cb, and Cr in the YCbCr domain, or G, B, and R in the RGB domain). Due to these correlations, during the development of the HEVC screen content coding (SCC) extension, an adaptive color space transformation (ACT) tool is employed to exploit the correlations between the three color components. Summary of the Invention [Problem to be solved by the invention]
[0008] This application describes implementations related to encoding and decoding video data, and more particularly, describes methods and systems that perform a clipping operation before adaptive color space conversion (ACT). [Means for solving the problem]
[0009] For video signals originally captured in 4:4:4 color format, if high fidelity is desired for the decoded video signal, it is preferable to encode the video in the original space, where information redundancy is abundant, for example RGB video. Some of the inferences in the current VVC standard, such as cross-component linear model prediction (CCLM), Although the three-component coding tool can improve the efficiency of 4:4:4 video coding, it does not fully eliminate the redundancy among the three components, because only the Y / G component is used to predict the Cb / B component and the Cr / R component, and the correlation between the Cb / B component and the Cr / R component is not taken into account. Meanwhile, further de-correlation of the three color components may improve the coding performance of 4:4:4 video coding.
[0010] In the current VVC standard, the existing inter and intra tool designs are mainly targeted at videos captured in 4:2:0 chrominance format. Therefore, to achieve a better complexity / performance tradeoff, it is necessary to improve the coding tools (e.g., position-dependent intra prediction combining (PDPC)). Many of the proposed methods, such as multi-reference line (MRL), sub-split prediction (ISP), is applicable only to the luma component and is invalid for the chroma component, or uses different operations (e.g., an interpolation filter applied to motion compensation prediction) for the luma and chroma components. However, a video signal in 4:4:4 chroma format exhibits very different characteristics compared to 4:2:0 video. For example, the Cb / B and Cr / R components of 4:4:4 YCbCr and RGB videos exhibit richer color information and have higher frequency information (e.g., edges and textures) than the chroma components of 4:2:0 video. Due to such considerations, it may always be optimal to use the same design of some existing encoding tools for both 4:2:0 and 4:4:4 videos in VVC.
[0011] According to a first aspect of the present application, a method for decoding video data includes receiving, from a bitstream, video data corresponding to a coding unit that is coded using an intra-prediction mode or an inter-prediction mode; receiving from the video data a first syntax element indicating whether the first coding unit is coded using adaptive color space transformation (ACT); processing the video data to generate a residual of the coding unit; performing a clipping operation on the residual of the coding unit in response to a determination based on the first syntax element that the coding unit is coded using ACT; and applying an inverse ACT to the residual of the coding unit after the clipping operation.
[0012] In some embodiments, the clipping operation limits the dynamic range of the coding unit residual to within a predetermined range for processing by the inverse ACT.
[0013] According to a second aspect of the present application, an electronic device includes one or more processors, a memory, and a number of programs stored in the memory, which, when executed by the one or more processors, cause the electronic device to perform a method for decoding video data as described above.
[0014] According to a third aspect of the present application, a non-transitory computer-readable recording medium stores a number of programs for execution by an electronic device having one or more processors, the programs, when executed by the one or more processors, causing the electronic device to perform the method for decoding video data as described above.
[0015] The accompanying drawings, which are included to provide a further understanding of the embodiments, are incorporated in and constitute a part of this specification, illustrate the described embodiments, and together with the description, serve to explain basic principles, with like reference numerals referring to corresponding parts. [Brief description of the drawings]
[0016] [Figure 1] FIG. 1 is a block diagram illustrating an example video encoding and decoding system according to some embodiments of the present disclosure. [Diagram 2] 1 is a block diagram illustrating an example video encoder according to some embodiments of the present disclosure. [Diagram 3] 1 is a block diagram illustrating an example video decoder according to some embodiments of the present disclosure. [Figure 4A] 2 is a block diagram illustrating how a frame is recursively divided into multiple video blocks of different sizes and shapes in accordance with some embodiments of this disclosure. [Figure 4B] 2 is a block diagram illustrating how a frame is recursively divided into multiple video blocks of different sizes and shapes in accordance with some embodiments of this disclosure. [Figure 4C] 2 is a block diagram illustrating how a frame is recursively divided into multiple video blocks of different sizes and shapes in accordance with some embodiments of this disclosure. [Figure 4D] 2 is a block diagram illustrating how a frame is recursively divided into multiple video blocks of different sizes and shapes in accordance with some embodiments of this disclosure. [Figure 4E] 2 is a block diagram illustrating how a frame is recursively divided into multiple video blocks of different sizes and shapes in accordance with some embodiments of this disclosure. [Figure 5A] FIG. 1 is a block diagram illustrating an example of applying an adaptive color space conversion (ACT) technique to convert residuals between RGB color space and YCgCo color space in accordance with some implementations of the present disclosure. [Figure 5B] FIG. 1 is a block diagram illustrating an example of applying an adaptive color space conversion (ACT) technique to convert residuals between RGB color space and YCgCo color space in accordance with some implementations of the present disclosure. [Figure 6]FIG. 1 is a block diagram of an exemplary video data decoding process that applies a technique of luma mapping with chroma scaling (LMCS) in accordance with certain implementations of the present disclosure. [Figure 7] 1 is a block diagram illustrating an example video decoding process in which a video decoder implements a technique of inverse adaptive color space conversion (ACT) in accordance with certain implementations of the present disclosure. [Figure 8A] A block diagram showing an example video decoding process in which a video decoder implements inverse adaptive color space conversion (ACT) and luma mapping with chroma scaling (LMCS) techniques in accordance with some implementations of the present disclosure. [Figure 8B] A block diagram showing an example video decoding process in which a video decoder implements inverse adaptive color space conversion (ACT) and luma mapping with chroma scaling (LMCS) techniques in accordance with some implementations of the present disclosure. [Figure 9] 1 is a block diagram illustrating example decoding logic between performing adaptive color space conversion (ACT) and performing block differential pulse coded modulation (BDPCM) in accordance with some implementations of the present disclosure. [Figure 10] 1 is a decoding flow diagram for applying different quantization parameter (QP) offsets to different components when the internal bit depth of luma and the internal bit depth of chroma are different in accordance with some implementations of the present disclosure. [Figure 11A] A block diagram showing an example video decoding process in which a video decoder implements a clipping technique to limit the dynamic range of a coding unit residual within a predetermined range for processing by inverse ACT, in accordance with certain implementations of the present disclosure. [Figure 11B]A block diagram showing an example video decoding process in which a video decoder implements a clipping technique to limit the dynamic range of a coding unit residual within a predetermined range for processing by inverse ACT, in accordance with certain implementations of the present disclosure. [Figure 12] 1 is a flow diagram illustrating an example process in which a video decoder decodes video data by performing a clipping operation to limit the dynamic range of a residual of an encoding unit within a predetermined range for processing by inverse ACT, in accordance with certain implementations of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0017] Reference will now be made in detail to specific embodiments, examples of which are illustrated in the accompanying drawings. In the following detailed description, numerous non-limiting specific details are set forth to aid in the understanding of the subject matter presented herein. However, it will be apparent to one skilled in the art that various alternatives may be used without departing from the scope of the claims, and that the subject matter may be practiced without these specific details. For example, it will be apparent to one skilled in the art that the subject matter presented herein may be practiced in many types of electronic devices with digital video capabilities.
[0018] In some embodiments, a method is provided for improving the coding efficiency of the VVC standard for 4:4:4 video. In general, the main features of the techniques in this disclosure are summarized as follows:
[0019] In some embodiments, these methods are implemented to improve existing ACT designs that enable adaptive color space conversion in the residual domain. In particular, special care is taken to handle the interaction of ACT with some existing coding tools in VVC.
[0020] In some embodiments, the methods are implemented to improve the efficiency of some existing inter- and intra-coding tools in the VVC standard for 4:4:4 video, and include the steps of 1) enabling an 8-tap interpolation filter for the chrominance components, 2) enabling PDPC for intra-prediction of the chrominance components, 3) enabling MRL for intra-prediction of the chrominance components, and 4) enabling ISP splitting for the chrominance components.
[0021] FIGURE 1 is a block diagram illustrating an example system 10 for concurrently encoding and decoding video blocks in accordance with some embodiments of the present disclosure. As shown in FIGURE 1, system 10 includes a source device 12 that generates and encodes video data that is subsequently decoded by a destination device 14. Source device 12 and destination device 14 may comprise any of a variety of electronic devices, including desktop or laptop computers, tablet computers, smartphones, set-top boxes, digital televisions, cameras, displays, digital media players, video game consoles, video streaming devices, and the like. In some embodiments, source device 12 and destination device 14 are equipped with wireless communication capabilities.
[0022] In some embodiments, destination device 14 may receive the encoded video data to be decoded over link 16. Link 16 may comprise any type of communication medium or device capable of transferring encoded video data from source device 12 to destination device 14. In one example, link 16 may comprise a communication medium to enable source device 12 to directly transmit encoded video data to destination device 14 in real time. The encoded video data may be modulated according to a communication standard, such as a wireless communication protocol, and transmitted to destination device 14. The communication medium may comprise any wireless or wired communication medium, such as the radio frequency (RF) spectrum or one or more physical transmission paths. The communication medium may form part of a packet-based network, such as a local area network, a wide area network, or a global network, such as the Internet. The communication medium may include routers, switches, base stations, or other devices that help facilitate communication from source device 12 to destination device 14. The device may include:
[0023] In some other embodiments, the encoded video data may be transmitted from output interface 22 to recording device 32. The encoded video data in recording device 32 may then be accessed by destination device 14 via input interface 28. Recording device 32 may be a hard drive, a Blu-ray disc, a DVD, a CD-ROM, The encoded video data may include any of a variety of distributed or locally accessed data storage media, such as a USB flash memory, a volatile or non-volatile memory, or other digital storage media suitable for storing the encoded video data. In a further example, recording device 32 may represent a file server or another intermediate storage device that may hold the encoded video data generated by source device 12. Destination device 14 may access the stored video data by streaming or downloading from recording device 32. The file server may be any type of computer capable of storing the encoded video data or transmitting the encoded video data to destination device 14. Exemplary file servers include a web server (e.g., for a website), an FTP server, a network attached storage (NAS) device, or a local disk drive. Destination device 14 may access the encoded video data through any standard data connection, including a wireless channel (e.g., a Wi-Fi connection), a wired connection (e.g., DSL, cable modem, etc.), or a combination of both, suitable for accessing the encoded video data stored in the file server. The transmission of the encoded video data from recording device 32 may be a streaming transmission, a download transmission, or a combination of both.
[0024] 1, source device 12 includes a video source 18, a video encoder 20, and an output interface 22. Video source 18 may include sources such as a video capture device, e.g., a video camera, a video archive containing previously captured video, a video feed interface for receiving video from a video content provider, and / or a computer graphics system for generating computer graphics data as the source video, or a combination of such sources. As an example, if video source 18 is a video camera in a security surveillance system, source device 12 and destination device 14 may form a camera phone or video phone. However, the embodiments described herein may be applicable to video coding in general and may be applied to wireless and / or wired applications.
[0025] Captured, pre-captured, or computer-generated video may be encoded by video encoder 20. The encoded video data may be transmitted directly to destination device 14 through output interface 22 of source device 12. The encoded video data may also (or alternatively) be stored in recording device 32 for later access by destination device 14 or other devices for decoding and / or playback. Output interface 22 may further include a modem and / or a transmitter.
[0026] Destination device 14 includes an input interface 28, a video decoder 30, and a display device 34. Input interface 28 may include a receiver and / or a modem and receives encoded video data over link 16. The encoded video data communicated over link 16 or provided by recording device 32 may include various syntax elements generated by video encoder 20 for use in decoding the video data by video decoder 30. The encoded video data, which may include such syntax elements, is transmitted over a communications medium and stored on a recording medium or a file server.
[0027] In some embodiments, the destination device 14 may include a display device 34 that is an integrated display device. , and an external display device configured to communicate with destination device 14. Display device 34 displays the decoded video data to a user and may be a liquid crystal display (LCD). crystal display, plasma display, organic light-emitting diode (OL The display may comprise any of a variety of display devices, such as a digital still camera (ED) display, or another type of display device.
[0028] The video encoder 20 and the video decoder 30 are compatible with VVC, HEVC, MPEG-4 Pa rt 10 AVC (Advanced Video Coding), or extensions of these standards. It should be understood that the present application is not limited to a particular video encoding / decoding standard and may be applicable to other video encoding / decoding standards. It is generally contemplated that video encoder 20 of source device 12 may be configured to encode video data in accordance with any of these current or future standards. Similarly, it is generally contemplated that video decoder 30 of destination device 14 may be configured to decode video data in accordance with any of these current or future standards.
[0029] Video encoder 20 and video decoder 30 may each be implemented using one or more microprocessors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other similar devices. GA: Field programmable gate array, discrete Logic (discrete logic), software, hardware The video encoder 20 and the video decoder 30 may be implemented as any of a variety of suitable encoding circuitry, such as hardware, firmware, or any combination thereof. When implemented partially in software, the electronic device may store instructions for the software on a suitable non-transitory computer-readable medium and execute the instructions in hardware using one or more processors to perform the video encoding / decoding operations disclosed in this disclosure. Each of the video encoder 20 and the video decoder 30 may be included in one or more encoders or decoders, any of which may be integrated as part of a combined encoder / decoder (CODEC) in the respective device.
[0030] 2 is a block diagram illustrating an example video encoder 20 according to some embodiments described herein. Video encoder 20 may perform intra-predictive and inter-predictive coding of video blocks within video frames. Intra-predictive coding relies on spatial prediction to reduce or remove spatial redundancy in video data within a given video frame or picture. Inter-predictive coding relies on temporal prediction to reduce or remove temporal redundancy in video data within adjacent video frames or pictures of a video sequence.
[0031] As shown in FIG. 2, the video encoder 20 includes a video data memory 40, a prediction processor 41, a decoded picture buffer (DPB) 64, an adder 50, a transform processor 52, a quantizer 54, and an entropy coder 56. The prediction processor 41 further includes a motion estimator 42, a motion compensation processor 44, a partitioner 45, an intra prediction processor 46, and an intra block copy (BC) unit 48. In some embodiments, the video encoder 20 also includes an inverse quantizer 58 for reconstructing video blocks, an inverse transform processor 60, and an adder 62. A deblocking filter (not shown) may be disposed between the adder 62 and the DPB 64 to filter block boundaries from the reconstructed video to remove block artifacts. In addition to the deblocking filter, an in-loop filter (not shown) may also be used to filter the output of the adder 62. The video encoder 20 may be implemented as a hardware unit, such as a fixed or programmable hardware unit. It may take the form of, or may be divided into, one or more fixed or programmable hardware units.
[0032] Video data memory 40 may store video data to be encoded by components of video encoder 20. The video data in video data memory 40 may be obtained, for example, from video source 18. DPB 64 is a buffer that records reference video data used to encode the video data by video encoder 20 (e.g., in an intra-predictive or inter-predictive encoding mode). Video data memory 40 and DPB 64 may be formed by any of a variety of storage devices. In various examples, video data memory 40 may be on-chip with other components of video encoder 20 or off-chip relative to those components.
[0033] As shown in FIG. 2, a partitioner 45 within the prediction processor 41 partitions the received video data into video blocks. This partitioning may include partitioning the video frame into slices, tiles, or other larger coding units (CUs) according to a predetermined partitioning structure, such as a quadtree structure, associated with the video data. The video frame may be partitioned into a number of video blocks (or sets of video blocks referred to as tiles). The prediction processor 41 may select one of a number of possible predictive coding modes, such as one of a number of intra-predictive coding modes or one of a number of inter-predictive coding modes, for the current video block based on the error result (e.g., coding rate or distortion level). The prediction processor 41 may provide the resulting intra-predictive or inter-predictive coded block to an adder 50 to generate a residual block, and may provide the coded block to an adder 62 for later reconstructing for use as part of a reference frame. The prediction unit 41 also provides syntax elements, such as motion vectors, intra mode indicators, partitioning information, and other such syntax information, to the entropy coding unit 56 .
[0034] To select an appropriate intra-prediction coding mode for a current video block, an intra-prediction unit 46 within prediction unit 41 may perform intra-prediction coding of the current video block relative to one or more neighboring blocks in the same frame as the current block being coded, resulting in spatial prediction. A motion estimation unit 42 and a motion compensation unit 44 within prediction unit 41 perform inter-prediction coding of the current video block relative to one or more predictive blocks in one or more reference frames, resulting in temporal prediction. Video encoder 20 may perform multiple coding passes, for example to select an appropriate coding mode for each block of video data.
[0035] In some embodiments, motion estimator 42 determines an inter prediction mode for a current video frame by generating a motion vector indicating a displacement of a prediction unit (PU) of a video block within a current video frame relative to a predictive block within a reference video frame according to a predetermined pattern within a sequence of video frames. Motion prediction performed by motion estimator 42 is a process of generating a motion vector that estimates the motion of a video block. The motion vector may indicate, for example, a displacement of a PU of a video block within a current video frame or picture relative to a predictive block within a reference frame (or other coding unit) relative to a current block to be coded within the current frame (or other coding unit). The predetermined pattern may designate the video frame as a P frame or a B frame in the sequence. Intra BC unit 48 may determine vectors, e.g., block vectors, for intra BC coding in a manner similar to the determination of motion vectors by motion estimator 42 for inter prediction, or may utilize motion estimator 42 to determine block vectors.
[0036] A prediction block is a block of a reference frame that is considered to closely correspond to a PU of a video block to be coded in terms of pixel difference, which may be determined by sum of absolute difference (SAD), sum of square difference (SSD), or other difference metric. In some embodiments, video encoder 20 may calculate values for sub-integer pixel locations of the reference frame stored in DPB 64. For example, video encoder 20 may interpolate values for quarter-pixel, eighth-pixel, or other fractional pixel locations of the reference frame. Thus, motion estimation unit 42 may perform motion search for whole pixel locations and fractional pixel locations to output motion vectors with fractional pixel accuracy.
[0037] The motion estimation unit 42 calculates a motion vector for a PU of a video block of an inter-predictive coded frame by comparing the position of the PU with the position of a predictive block of a reference frame selected from a first reference frame list (List 0) or a second reference frame list (List 1), where the first reference frame list or the second reference frame list respectively identifies one or more reference frames stored in the DPB 64. The motion estimation unit 42 sends the calculated motion vector to the motion compensation unit 44 and then to the entropy coding unit 56.
[0038] The motion compensation performed by motion compensation unit 44 may include fetching or generating a predictive block based on the motion vector determined by motion estimation unit 42. When motion compensation unit 44 receives a motion vector for a PU of a current video block, it locates the predictive block to which the motion vector points in one of the reference frame lists, retrieves the predictive block from DPB 64, and forwards the predictive block to summer 50. Summer 50 then forms a residual video block of pixel difference values by subtracting pixel values of the predictive block provided by motion compensation unit 44 from pixel values of the current video block to be coded. The pixel difference values forming the residual video block may include luma or chroma difference components, or both. Motion compensation unit 44 may also generate syntax elements related to the video blocks of the video frames for use by video decoder 30 in decoding the video blocks of the video frames. The syntax elements may include, for example, syntax elements defining the motion vectors used to describe the predictive blocks, any flags indicating prediction modes, or other syntax information described herein. Note that motion estimation unit 42 and motion compensation unit 44 may be largely integrated together, but are shown separately for conceptual purposes.
[0039] In some embodiments, the intra BC unit 48 may generate a vector to capture a predictive block in a manner similar to that described above with respect to the motion estimation unit 42 and the motion compensation unit 44, but the predictive block is in the same frame as the current block to be coded, and the vector is referred to as a block vector, as opposed to a motion vector. In particular, the intra BC unit 48 may determine to use an intra prediction mode to code the current block. In some examples, the intra BC unit 48 may code the current block using various intra prediction modes, e.g., during separate coding passes, and analyze the performance of the intra prediction modes by rate-distortion analysis. The intra BC unit 48 may then select an appropriate intra prediction mode among the various intra prediction modes tested to use in generating the intra mode indicator. For example, the intra BC unit 48 may use the rate-distortion analysis to calculate rate-distortion values for the various intra prediction modes tested, and select the intra prediction mode with the best rate-distortion characteristics among the tested modes as the appropriate intra prediction mode to use. Rate-distortion analysis generally determines the bit rate (i.e., number of bits) used to generate a coded block, along with the amount of distortion (or error) between the coded block and the original block that was coded to generate the coded block. Unit 48 may calculate distortion-to-rate ratios for various coding blocks to determine the intra-prediction mode that yields the best rate-distortion value for that block.
[0040] In other examples, the intra BC unit 48 may use, in whole or in part, the motion estimation unit 42 and the motion compensation unit 44 to perform such functions for intra BC prediction in accordance with the embodiments described herein. In either case, for intra block copying, the predictive block may be a block that is considered to be a close equivalent to the block to be coded in terms of pixel differences, and may be determined by sum of absolute differences (SAD), sum of squared differences (SSD), or other difference measures. Identifying the predictive block may include computing values for sub-integer pixel positions.
[0041] Whether the predictive block is from the same frame via intra prediction or a different frame via inter prediction, video encoder 20 may form a residual video block by subtracting pixel values of the predictive block from pixel values of the current video block being encoded, forming pixel difference values. The pixel difference values that form the residual video block may include both luma and chroma difference components.
[0042] The intra prediction unit 46 may intra predict the current video block as an alternative to inter prediction performed by the motion estimation unit 42 and the motion compensation unit 44 or intra block copy prediction performed by the intra BC unit 48, as described above. In particular, the intra prediction unit 46 may determine to use an intra prediction mode to encode the current block. To do so, the intra prediction unit 46 may encode the current block using various intra prediction modes, for example during separate encoding passes, and the intra prediction unit 46 (or a mode selection unit in some examples) may select an appropriate intra prediction mode to use from the tested intra prediction modes. The intra prediction unit 46 may provide information indicative of the selected intra prediction mode for the block to the entropy coder 56. The entropy coder 56 may encode the information indicative of the selected intra prediction mode in the bitstream.
[0043] After prediction processor 41 determines a predictive block for the current video block, either by inter prediction or intra prediction, adder 50 generates a residual video block by subtracting the predictive block from the current video block. The residual video data in the residual block may be included in one or more transform units (TUs) and provided to transform processor 52. Transform processor 52 performs a discrete cosine transform (DCT) to generate a residual video block. The residual video data is transformed into residual transform coefficients using a transform such as the residual video transform (RfD) or a conceptually similar transform.
[0044] The transform processor 52 may send the resulting transform coefficients to a quantizer 54, which quantizes the transform coefficients to further reduce the bit rate. The quantization process may also reduce the bit depth associated with some or all of the coefficients. The degree of quantization may be varied by adjusting a quantization parameter. In some examples, the quantizer 54 may then perform a scan of a matrix containing the quantized transform coefficients. Alternatively, the entropy coder 56 may perform the scan.
[0045] Following quantization, the entropy coding unit 56 may, for example, use context adaptive variable length coding (CAVLC). h coding), context adaptive binary arithmetic coding (CABAC), syntax Syntax-based Context Adaptive Binary Arithmetic Coding (SBAC) ntext-adaptive binary arithmetic coding), Probability interval entropy coding (PIPE) l partitioning entropy coding), or another entropy An encoding technique or technology is used to entropy encode the quantized transform coefficients into a video bitstream. The encoded bitstream may then be transmitted to video decoder 30 or recorded on storage device 32 for later transmission to or retrieval by video decoder 30. Entropy encoder 56 may also entropy encode motion vectors and other syntax elements for the current video frame being encoded.
[0046] Inverse quantization unit 58 applies inverse quantization, and inverse transform unit 60 applies inverse transform to reconstruct the residual video block in the pixel domain to generate a reference block for predicting other video blocks. As mentioned above, motion compensation unit 44 may generate a motion compensated prediction block from one or more reference blocks of a frame stored in DPB 64. Motion compensation unit 44 may also apply one or more interpolation filters to the prediction block to calculate sub-integer pixel values for use in motion estimation.
[0047] Summer 62 adds the reconstructed residual block to the motion compensated prediction block generated by motion compensation unit 44 to generate a reference block for storage in DPB 64. The reference block may then be used by intra BC unit 48, motion estimation unit 42, and motion compensation unit 44 as a prediction block to inter predict another video block in a subsequent video frame.
[0048] 3 is a block diagram illustrating an exemplary video decoder 30 according to some embodiments of the present application. The video decoder 30 includes a video data memory 79, an entropy decoding unit 80, a prediction processing unit 81, an inverse quantization unit 86, an inverse transform processing unit 88, an adder 90, and a DPB 92. The prediction processing unit 81 further includes a motion compensation unit 82, an intra prediction processing unit 84, and an intra BC unit 85. The video decoder 30 may perform a decoding process that is generally inverse to the encoding process described for the video encoder 20 in connection with FIG. 2. For example, the motion compensation unit 82 may generate prediction data based on a motion vector received from the entropy decoding unit 80, while the intra prediction processing unit 84 may generate prediction data based on an intra prediction mode indicator received from the entropy decoding unit 80.
[0049] In some examples, units of video decoder 30 may be tasked with performing embodiments of the present application. Also, in some examples, embodiments of the present disclosure may be divided among one or more units of video decoder 30. For example, intra BC unit 85 may perform embodiments of the present application alone or in combination with other units of video decoder 30, such as motion compensation unit 82, intra prediction processing unit 84, and entropy decoding unit 80. In some examples, video decoder 30 may not include intra BC unit 85, and the functionality of intra BC unit 85 may be performed by other components of prediction processing unit 81, such as motion compensation unit 82.
[0050] Video data memory 79 may store video data, such as an encoded video bitstream, that is decoded by other components of video decoder 30. The video data stored in video data memory 79 may be obtained from recording device 32, from a local video source, such as a camera, by wired or wireless network communication of the video data, or by accessing a physical data storage medium, such as a flash drive or hard disk. Video data memory 79 may include a coded picture buffer (CPB), which stores coded video data from the coded video bitstream. A decoded picture buffer of the video decoder 30 may be included. The digital pre-frame block (DPB) 92 is generated by the video decoder 30 (e.g., in an intra-prediction coding mode or The video data memory 79 and the DPB 92 store reference video data used for encoding the video data (in an inter-prediction encoding mode). The video data memory 79 and the DPB 92 may be implemented by a synchronous dynamic random access memory (SDRAM), a magnetoresistive RAM (MRAM), Various memory devices, such as dynamic random access memory (DRAM), including resistive RAM (RRAM), or other types of memory devices For purposes of illustration, video data memory 79 and DPB 92 are depicted as two separate components of video decoder 30 in FIG. 3. However, it will be apparent to one skilled in the art that video data memory 79 and DPB 92 may be provided by the same memory device or separate memory devices. In some examples, video data memory 79 may be on-chip with other components of video decoder 30 or may be off-chip relative to those components.
[0051] During the decoding process, video decoder 30 receives an encoded video bitstream representing video blocks of encoded video frames and associated syntax elements. Video decoder 30 may receive video frame-level and / or video block-level syntax elements. An entropy decoding unit 80 of video decoder 30 entropy decodes the bitstream to generate quantized coefficients, motion vectors or intra-prediction mode indicators, and other syntax elements. Entropy decoding unit 80 then forwards the motion vectors and other syntax elements to a prediction processor 81.
[0052] When a video frame is coded as an intra-predictive coded (I) frame, or for intra-coded predictive blocks in other types of frames, the intra-prediction processing unit 84 of the prediction processing unit 81 may generate predictive data for video blocks of the current video frame based on a signaled intra-prediction mode and reference data from previously decoded blocks of the current frame.
[0053] When a video frame is coded as an inter-predictive (i.e., B or P) frame, motion compensation unit 82 of prediction unit 81 generates one or more predictive blocks for video blocks of the current video frame based on the motion vectors and other syntax elements received from entropy decoding unit 80. Each of the predictive blocks may be generated from a reference frame within one of the reference frame lists. Video decoder 30 may construct the reference frame lists, List 0 and List 1, using a default construction technique based on the reference frames stored in DPB 92.
[0054] In some examples, when a video block is encoded according to the intra BC modes described herein, intra BC unit 85 of prediction processor 81 generates a predictive block for the current video block based on the block vectors and other syntax elements received from entropy decoding unit 80. The predictive block may be within the same reconstructed region of the picture as the current video block defined by video encoder 20.
[0055] Motion compensation unit 82 and / or intra BC unit 85 determine prediction information for video blocks of a current video frame by analyzing the motion vectors and other syntax elements, and then use the prediction information to generate a predictive block for the current video block to be decoded. For example, motion compensation unit 82 may use some of the received syntax elements to determine a prediction mode (e.g., intra prediction or inter prediction) used to encode the video blocks of the video frame, an inter prediction frame type (e.g., B or P), configuration information for one or more of a reference frame list for the frame, a motion vector for each inter prediction coded video block in ... The inter-prediction unit 100 determines the inter-prediction state of the inter-prediction coded video block and other information for decoding the video block in the current video frame.
[0056] Similarly, intra BC unit 85 may use some of the received syntax elements, such as flags, to determine that the current video block was predicted using an intra BC mode, configuration information of the video blocks of the frame that are within the reconstructed region and that should be stored in DPB 92, block vectors of each intra BC predicted video block of the frame, intra BC prediction states of each intra BC predicted video block of the frame, and other information for decoding video blocks in the current video frame.
[0057] Motion compensation unit 82 may also perform the interpolation using an interpolation filter such as that used to calculate sub-integer pixel interpolated values of the reference block during encoding of the video block by video encoder 20. In this case, motion compensation unit 82 may determine from the received syntax elements the interpolation filter used by video encoder 20 and generate the predictive block using the interpolation filter.
[0058] An inverse quantizer 86 inverse quantizes the quantized transform coefficients provided in the bitstream and entropy decoded by entropy decoder 80 using the same quantization parameters calculated by video encoder 20 to determine the degree of quantization for each video block in a video frame. An inverse transform processor 88 applies an inverse transform, such as an inverse DCT, an inverse integer transform, or a conceptually similar inverse transform process, to the transform coefficients in order to reconstruct residual blocks in the pixel domain.
[0059] After the motion compensation unit 82 or the intra BC unit 85 generates a predictive block for the current video block based on the vectors and other syntax elements, the summer 90 reconstructs a decoded video block for the current video block by summing the residual block from the inverse transform processor 88 and the corresponding predictive block generated by the motion compensation unit 82 and the intra BC unit 85. An in-loop filter (not shown) may be disposed between the summer 90 and the DPB 92 to further process the decoded video block. The decoded video block in a given frame is then stored in the DPB 92, which stores reference frames used for subsequent motion compensation of the next video block. The DPB 92 or a memory device separate from the DPB 92 may also store the decoded video for later presentation to a display device, such as display device 34 of FIG. 1.
[0060] In a typical video coding process, a video sequence typically includes an ordered set of frames or pictures. Each frame may include three sample arrays, denoted as SL, SCb, and SCr. SL is a two-dimensional array of luma samples. SCb is a two-dimensional array of Cb chrominance samples. SCr is a two-dimensional array of Cr chrominance samples. In other cases, a frame may be black and white and therefore include only one two-dimensional array of luma samples.
[0061] As shown in FIG. 4A, video encoder 20 (and more specifically, splitter 45) generates a coded representation of a frame by first splitting the frame into a set of coding tree units (CTUs). A video frame may contain an integer number of CTUs sequentially ordered in a raster scan order from left to right and top to bottom. Each CTU is the largest logical coding unit, and the width and height of the CTUs are signaled by video encoder 20 in the sequence parameter set such that all CTUs in a video sequence have the same size, which may be one of 128×128, 64×64, 32×32, and 16×16. However, the present application is not necessarily limited to any particular size. It should be noted that, as shown in FIG. 4B, each CTU may include one coding tree block (CTB) of luma samples, a corresponding coding tree block of two chrominance samples, and syntax elements used to code the coding tree block samples. The syntax elements describe the properties of various types of units of coding blocks of pixels and how a video sequence may be reconstructed at video decoder 30, including inter- or intra-prediction, intra-prediction mode, motion vectors, and other parameters. For a monochrome picture or a picture with three separate color planes, a CTU may include a single coding tree block and syntax elements used to code the coding tree block samples. A coding tree block may be an N×N block of samples.
[0062] To achieve better performance, the video encoder 20 may recursively perform tree partitioning, such as binary-tree partitioning, ternary-tree partitioning, quad-tree partitioning, or a combination thereof, on the coding tree blocks of the CTU to partition the CTU into smaller coding units (CUs). As shown in FIG. 4C, a 64×64 CTU 400 is first partitioned into four smaller CUs, each with a block size of 32×32. Among the four smaller CUs, CU 410 and CU 420 are partitioned into four CUs, each with a block size of 16×16. Two 16×16 CUs 430 and 440 are further partitioned into four CUs, each with a block size of 8×8. FIG. 4D illustrates a quadtree data structure showing the final result of the partitioning process of the CTU 400 as illustrated in FIG. 4C, where each leaf node of the quadtree corresponds to one CU of a respective size ranging from 32×32 to 8×8. Each CU may include a coding block (CB) of luma samples, two corresponding coding blocks of chroma samples of the same size frame, and syntax elements used to code the samples of the coding block, similar to the CTU illustrated in FIG. 4B. In a monochrome picture or a picture with three separate color planes, the CU may include a single coding block and syntax structures used to code the samples of the coding block. It should be noted that the quadtree partitioning illustrated in FIG. 4C and FIG. 4D is merely illustrative, and one CTU may be partitioned into CUs based on quadtree / ternary / binary tree partitioning to fit various local characteristics. In the composite tree structure, one CTU is divided by a quadtree structure, and the leaf CU of each quadtree can be further divided by a binary tree structure and a ternary tree structure. As shown in Fig. 4E, there are five division types, including 4-way, horizontal 2-way, vertical 2-way, horizontal 3-way, and vertical 3-way.
[0063] In some embodiments, video encoder 20 may further divide the coding blocks of a CU into one or more M×N prediction blocks (PBs). A prediction block is a rectangular (square or non-square) block of samples to which a uniform (inter or intra) prediction is applied. A prediction unit (PU) of a CU may include a prediction block of luma samples, two corresponding prediction blocks of chroma samples, and syntax elements used to predict the prediction block. In a monochrome picture or a picture with three separate color planes, a PU may include a single prediction block and syntax structures used to predict the prediction block. Video encoder 20 may generate predicted luma, Cb and Cr blocks for luma, and Cb and Cr prediction blocks in each PU of a CU.
[0064] Video encoder 20 may use intra prediction or inter prediction to generate the predictive blocks for a PU. If video encoder 20 uses intra prediction to generate the predictive blocks for a PU, video encoder 20 may generate the predictive blocks for the PU based on decoded samples of a frame associated with the PU. If video encoder 20 uses inter prediction to generate the predictive blocks for a PU, video encoder 20 may generate the predictive blocks for the PU based on decoded samples of one or more frames other than the frame associated with the PU. possible.
[0065] After generating the predictive luma block, the predictive Cb block, and the predictive Cr block for one or more PUs in the CU, the video encoder 20 may generate a luma residual block for the CU by subtracting the predictive luma block of the CU from the original luma coding block of the CU, such that each sample in the luma residual block of the CU indicates a difference between a luma sample in one of the predictive luma blocks of the CU and a corresponding sample in the original luma coding block of the CU. Similarly, the video encoder 20 may generate a Cb residual block and a Cr residual block of the CU, respectively, such that each sample in the Cb residual block of the CU indicates a difference between a Cb sample in one of the predictive Cb blocks of the CU and a corresponding sample in the original Cb coding block of the CU, and each sample in the Cr residual block of the CU may indicate a difference between a Cr sample in one of the predictive Cr blocks of the CU and a corresponding sample in the original Cr coding block of the CU.
[0066] Moreover, as shown in FIG. 4C, video encoder 20 uses quadtree partitioning to decompose the luma, Cb, and Cr residual blocks of a CU into one or more luma, Cb, and Cr transform blocks. A transform block is a rectangular (square or non-square) block of samples to which the same transform is applied. A transform unit (TU) of a CU may include a transform block of luma samples, two corresponding transform blocks of chroma samples, and syntax elements used to predict the transform block samples. Thus, each TU of a CU may be associated with a luma transform block, a Cb transform block, and a Cr transform block. In some examples, the luma transform block associated with a TU may be a sub-block of the luma residual block of the CU. The Cb transform block may be a sub-block of the Cb residual block of the CU. The Cr transform block may be a sub-block of the Cr residual block of the CU. In a monochrome picture or a picture with three separate color planes, a TU may include a single transform block and syntax structures used to transform the samples of the transform block.
[0067] Video encoder 20 may apply one or more transforms to a luma transform block of the TU to generate a luma coefficient block for the TU. The coefficient block may be a two-dimensional array of transform coefficients. The transform coefficients may be scalar quantities. Video encoder 20 may apply one or more transforms to a Cb transform block of the TU to generate a Cb coefficient block for the TU. Video encoder 20 may apply one or more transforms to a Cr transform block of the TU to generate a Cr coefficient block for the TU.
[0068] After generating a coefficient block (e.g., a luma coefficient block, a Cb coefficient block, or a Cr coefficient block), the video encoder 20 may quantize the coefficient block. Quantization generally refers to a process in which transform coefficients are quantized to somehow reduce the amount of data used to represent the transform coefficient, resulting in further compression. After quantizing the coefficient block, the video encoder 20 may entropy code syntax elements indicating the quantized transform coefficients. For example, the video encoder 20 may perform context-adaptive binary arithmetic coding (CABAC) on the syntax elements indicating the quantized transform coefficients. Finally, the video encoder 20 may output a bitstream including a series of bits forming a representation of the encoded frame and associated data, which may be stored on a recording device 32 or transmitted to a destination device 14.
[0069] After receiving the bitstream generated by video encoder 20, video decoder 30 may parse the bitstream to obtain syntax elements from the bitstream. Video decoder 30 may reconstruct frames of video data based at least in part on the syntax elements obtained from the bitstream. The process of reconstructing video data is generally the reverse of the encoding process performed by video encoder 20. For example, video decoder 30 may perform an inverse transform on coefficient blocks associated with the TUs of the current CU to reconstruct residual blocks associated with the TUs of the current CU. Video decoder 30 also reconstructs coding blocks of the current CU by adding samples of predictive blocks for the PUs of the current CU to samples of transform blocks of corresponding TUs of the current CU. Video decoder 30 may reconstruct the frame after reconstructing the coding blocks for each CU of the frame.
[0070] As mentioned above, video coding mainly uses two modes, intra-frame prediction (i.e., intra-prediction) and inter-frame prediction (i.e., inter-prediction), to achieve video compression. Palette-based coding is another coding scheme adopted by many video coding standards. Palette-based coding is particularly suitable for coding screen generated content, in which a video coder (e.g., video encoder 20 or video decoder 30) forms a palette table of colors representing a given block of video data. The palette table includes the most dominant (e.g., frequently used) pixel values in a given block. Pixel values that are not frequently represented in the video data of a given block are not included in the palette table or are included in the palette table as avoided colors.
[0071] Each entry in the palette table includes an index for a corresponding pixel value in the palette table. A palette index for a sample in a block may be coded to indicate the entry in the palette table used to predict or reconstruct the sample. This palette mode begins with the process of generating a palette predictor for the first block of a picture, slice, tile, or such grouping of video blocks. As described below, palette predictors for subsequent video blocks are generally generated by updating a previously used palette predictor. For illustrative purposes, it is assumed that the palette predictor is defined at the picture level. In other words, a picture may contain multiple coding blocks, each with its own palette table, but there is one palette predictor for the entire picture.
[0072] To reduce the number of bits required to signal palette entries in a video bitstream, a video decoder may utilize a palette predictor to determine new palette entries for a palette table used to reconstruct a video block. For example, the palette predictor may include palette entries from a previously used palette table, or may be initialized with a most recently used palette table by including all entries of the most recently used palette table. In some embodiments, the palette predictor may include fewer than all entries from the most recently used palette table and then incorporate some entries from other previously used palette tables. The size of the palette predictor may be the same, larger, or smaller compared to the size of the palette tables used to encode different blocks. In one example, the palette predictor is implemented as a first-in-first-out (FIFO) table containing 64 palette entries.
[0073] To generate a palette table for a block of video data from the palette predictor, the video decoder may receive a one-bit flag for each input of the palette predictor from the encoded video bitstream. The one-bit flag may have a first value (e.g., a binary 1) indicating that the associated input of the palette predictor is included in the palette table or a second value (e.g., a binary 0) indicating that the associated input of the palette predictor is not included in the palette table. If the size of the palette predictor is larger than the palette table used for the block of video data, the video decoder may stop receiving further flags once the maximum size of the palette table is reached.
[0074] In some embodiments, some entries of the palette table may be directly signaled in the coded video bitstream rather than being determined using a palette predictor. For such entries, the video decoder may receive three separate m-bit values from the coded video bitstream indicating pixel values for the luma component and two chroma components associated with the entry, where m represents the bit depth of the video data. Palette entries derived from the palette predictor require only a one-bit flag, whereas multiple m-bit values are required for directly signaled palette entries. Thus, signaling some or all of the palette entries using the palette predictor can significantly reduce the number of bits required to signal new palette table entries, thereby improving the overall coding efficiency of palette mode coding.
[0075] In many cases, a palette predictor for a block is determined based on a palette table used to code one or more previously coded blocks. However, when coding the first coding tree unit in a picture, slice, or tile, the palette table of the previously coded block may not be available. Thus, the palette predictor cannot be generated using entries in the previously used palette table. In such cases, a set of palette predictor initializers, values used to generate a palette predictor when the previously used palette table is not available, may be signaled in a sequence parameter set (SPS) and / or a picture parameter set (PPS). An SPS generally refers to a syntax structure of syntax elements that fits into a set of consecutive coded video pictures, called a coded video sequence (CVS), as determined by the content of syntax elements found in a PPS referenced by syntax elements found in each slice segment header. A PPS generally refers to a syntax structure of syntax elements that fits into one or more individual pictures within a CVS, as determined by the content of syntax elements found in each slice segment header. Thus, an SPS is generally considered a higher level syntax structure than a PPS, meaning that the syntax elements contained in an SPS generally change less frequently and fit a larger portion of the video data compared to the syntax elements contained in a PPS.
[0076] 5A-5B are block diagrams illustrating examples of applying an adaptive color space conversion (ACT) technique to convert residuals between RGB color space and YCgCo color space in accordance with some implementations of the present disclosure.
[0077] In the HEVC screen content coding extension, ACT is applied to adaptively convert the residual from one color space (e.g., RGB) to another color space (e.g., YCgCo), such that the correlation (e.g., redundancy) between the three color components (e.g., R, G, and B) is significantly reduced in the YCgCo color space. Furthermore, in the existing ACT design, the adaptation of different color spaces is performed at the transform unit (TU) level by signaling one flag, tu_act_enabled_flag, for each TU. The flag tu_act_enabled_flag is 1 to indicate that the residual of the current TU is coded in the YCgCo space, otherwise (i.e., the flag is 0), the residual of the current TU is coded in the original color space (i.e., without color space conversion). In addition, different color space conversion formulas are applied depending on whether the current TU is coded in a lossless mode or a lossy mode. Specifically, the forward and reverse color space conversion formulas for the lossy mode between the RGB color space and the YCgCo color space are defined in FIG. 5A.
[0078] For the lossless mode, a lossless version of the RGB-YCgCo transform (also known as YCgCo-LS) is used, which is implemented based on the lifting operation depicted in FIG. 5B and the associated description.
[0079] As shown in FIG. 5A, the forward and inverse color transform matrices used in the lossy mode are not normalized. Therefore, the YCgCo signal after the color transform is applied is smaller than the original signal. To compensate for the magnitude reduction caused by the forward color transform, an adjusted quantization parameter is applied to the residual in the YCgCo domain. Specifically, when the color space transform is applied, the QP used to quantize the residual in the YCgCo domain is Y , Q.P. Cg and QP Co are set to QP-5, QP-5 and QP-3, respectively, where QP is the quantization parameter used in the original color space.
[0080] FIG. 6 is a block diagram of an application of a technique of Luminance Mapping with Chrominance Scaling (LMCS) in an example video data decoding process in accordance with some implementations of the present disclosure.
[0081] In VVC, LMCS is used as a new coding tool applied before in-loop filters (e.g., deblocking filter, SAO and ALF). In general, LMCS has two main modules: 1) in-loop mapping of luma components based on an adaptive piecewise linear model, and 2) luma-dependent chroma residual scaling. Figure 6 shows a modified decoding process in which LMCS is applied. In Figure 6, the decoding modules processed in the mapped domain include an entropy decoding module, an inverse quantization module, an inverse transform module, a luma intra prediction module, and a luma sample reconstruction module (i.e., addition of luma prediction samples and luma residual samples). The decoding modules processed in the original (i.e., unmapped) domain include a motion compensation prediction module, a chroma intra prediction module, a chroma sample reconstruction module (i.e., addition of chroma prediction samples and chroma residual samples), and all in-loop filter modules, such as a deblocking module, an SAO module, and an ALF module. New available modules introduced by LMCS include a forward mapping of luma samples module 610 , a backward mapping of luma samples module 620 , and a chroma residual scaling module 630 .
[0082] The in-loop mapping of LMCS can adjust the dynamic range of the input signal to improve the coding efficiency. The in-loop mapping of luma samples in the existing LMCS design is built on two mapping functions: one forward mapping function FwdMap and one corresponding backward mapping function InvMap. The forward mapping function is signaled from the encoder to the decoder using one piecewise linear model with 16 equal-sized parts. The backward mapping function can be directly derived from the forward mapping function and therefore does not need to be signaled.
[0083] The parameters of the luminance mapping model are signaled at the slice level. A presence flag is first signaled to indicate whether a luminance mapping model should be signaled for the current slice. If a luminance mapping model exists for the current slice, the corresponding piecewise linear model parameters are further signaled. In addition, another LMCS control flag is signaled at the slice level to enable / disable LMCS for the slice.
[0084] The chroma residual scaling module 630 is designed to compensate for the interaction of quantization precision between the luma signal and its corresponding chroma signal when in-loop mapping is applied to the luma signal. Whether chroma residual scaling is enabled or disabled for the current slice is determined by the chroma residual scaling module 630. It is also signaled in the slice header whether luma mapping is enabled. If luma mapping is enabled, an additional flag is signaled indicating whether luma-dependent chroma residual scaling is applied or not. When luma mapping is not used, luma-dependent chroma residual scaling is always disabled and no additional flag is required. In addition, chroma residual scaling is always disabled for CUs that contain four or fewer chroma samples.
[0085] FIG. 7 is a block diagram illustrating an example video decoding process in which a video decoder implements an inverse adaptive color space conversion (ACT) technique in accordance with certain implementations of this disclosure.
[0086] ACT in VVC, like the ACT design in HEVC SCC, is based on the original color space The inverse ACT module 710 converts the intra / inter prediction residual of one CU in 4:4:4 chrominance format from (e.g., RGB color space) to YCgCo color space. As a result, the redundancy among the three color components can be reduced for better coding efficiency. FIG. 7 illustrates a decoding flow diagram showing how inverse ACT is applied by adding an inverse ACT module 710 to a VVC configuration. When processing a CU coded with ACT enabled, entropy decoding, inverse quantization, and inverse DCT / DST-based transform are first applied to the CU. Then, as illustrated in FIG. 7, inverse ACT is called to convert the decoded residual from YCgCo color space to the original color space (e.g., RGB or YCbCr). In addition, since ACT is not normalized in lossy mode, a QP adjustment of (-5, -5, -3) is applied to Y, Cg, and Co components to compensate for the magnitude change of the transformed residual.
[0087] In some embodiments, the ACT method reuses the same ACT core transform of HEVC to perform color transformation between different color spaces. Specifically, two different versions of color transformation are applied depending on whether the current CU is lossy or lossy in its encoding. Forward and inverse color transformation in the lossy case uses the lossy YCgCo transformation matrix as shown in FIG. 5A. In the lossless case, the lossless color transformation YCgCo-LS is applied as shown in FIG. 5B. Moreover, different from the existing ACT design, the following modifications are introduced to the ACT scheme to handle the interaction of the ACT scheme with other encoding tools in the VVC standard:
[0088] For example, since the residual of one CU in HEVC may be split into multiple TUs, an ACT control flag is signaled separately for each TU to indicate whether or not a color space transformation needs to be applied. However, as described above in relation to FIG. 4E, in VVC, a quadtree with nested bipartition and tripartition structures is applied to replace the concept of multiple partition types, thereby eliminating the separate CU, PU and TU partitions in HEVC. This means that in most cases, one CU leaf node is also used as a unit of prediction and transformation processing without further partitioning, unless the maximum supported transform size is smaller than the width or height of one component of the CU. Based on such partition structure, ACT can be adaptively enabled / disabled at the CU level. Specifically, one flag cu_act_enabled_flag is signaled for each CU to select between the original color space and the YCgCo color space to code the residual of the CU. If this flag is 1, it indicates that all of the residuals of the TUs inside the CU are coded in the YCgCo color space. Otherwise, if the flag cu_act_enabled_flag is 0, all residuals of the CU are coded in the original color space.
[0089] In some embodiments, there are various scenarios for disabling ACT. When ACT is enabled for a CU, it needs to access the residuals of all three components to perform color space conversion. However, the VVC design cannot guarantee that each CU always contains information of all three components. According to the embodiments of the present disclosure, if a CU has access to all three components, the ACT can be used to convert the residuals of all three components. If the ACT does not contain information for all three components, the ACT is forced to be invalid.
[0090] In some embodiments, first, when a separate tree partitioning structure is applied, luma samples and chroma samples in one CTU are partitioned into CUs based on the separate partitioning structure. As a result, the CU in the luma partitioning tree only contains the coding information of the luma component, and the CU in the chroma partitioning tree only contains the coding information of the two chroma components. According to the current VVC, the switching between the separate tree partitioning structure and the separate tree partitioning structure is performed at the slice level. Therefore, according to an embodiment of the present disclosure, when a slice is found to be applied to a separate tree, ACT is always disabled for all CUs (both luma CUs and chroma CUs) in this slice without signaling the ACT flag, and the ACT flag is inferred to be zero instead.
[0091] Second, in some embodiments, when ISP mode (described further below) is enabled, TU splitting is applied only to luma samples, and chroma samples are coded without further splitting into multiple TUs. Assuming the number of ISP subpartitions (i.e., TUs) for one intra CU is N, according to the current ISP design, only the last TU contains both luma and chroma components, and the first N-1 ISP TUs consist of luma components only. According to an embodiment of the present disclosure, ACT is disabled in ISP mode. There are two ways to disable ACT for ISP mode. In the first method, an ACT enable / disable flag (i.e., cu_act_enabled_flag) is signaled before signaling the ISP mode syntax. In such a case, when the flag cu_act_enabled_flag is 1, the ISP mode is not signaled in the bitstream and is always inferred to be zero (i.e., off). In the second method, ISP mode signaling is used to avoid signaling the ACT flag. Specifically, in this method, the ISP mode is signaled prior to the flag cu_act_enabled_flag. When the ISP mode is selected, the flag cu_act_enabled_flag is not signaled and is inferred to be zero. Otherwise (the ISP mode is not selected), the flag cu_act_enabled_flag is still signaled to adaptively select a color space for residual coding of the CU.
[0092] In some embodiments, when the luma and chroma split structure is misaligned, in addition to forcibly disabling ACT for the CU, LMCS for the CU to which ACT is applied is disabled. In one embodiment, when a CU selects a YCgCo color space to code its residual, both luma mapping and chroma residual scaling are disabled (i.e., ACT is one). In another embodiment, when ACT is enabled for a CU, only chroma residual scaling is disabled, and luma mapping may still be applied to adjust the dynamic range of the output luma samples. In a final embodiment, both luma mapping and chroma residual scaling are enabled for a CU that applies ACT for coding its residual. There may be multiple ways to enable chroma residual scaling for a CU that applies ACT. In one way, chroma residual scaling is applied before inverse ACT in the decoding. This means that when ACT is applied, chroma residual scaling is applied to the chroma residual (i.e., Cg and Co residual) in the YCgCo domain. In another way, chroma residual scaling is applied after inverse ACT. Specifically, in the second method, chroma scaling is applied to the residual in the original color space. Assuming the input video is captured in RGB format, this means that chroma residual scaling is applied to the residuals of the B and R components.
[0093] In some embodiments, a sequence parameter set (SPS), for example sps_act_enabled_flag, is added to the syntax elements to indicate whether ACT is enabled at the sequence level. In addition, when color space conversion is applied to video content whose luma and chroma components have the same resolution (e.g., 4:4:4 chroma format), one bitstream conformance requirement needs to be added so that ACT can only be enabled for 4:4:4 chroma format. Table 1 shows the modified SPS syntax table with the above syntax added.
[0094] [Table 1]
[0095] Specifically, sps_act_enabled_flag equal to 1 indicates that ACT is enabled, and sps_act_enabled_flag equal to 0 indicates that ACT is disabled such that the flag cu_act_enabled_flag is not signaled to the CU referencing the SPS and is inferred to be 0. When ChromaArrayType is not 3, it is a requirement for bitstream conformance that the value of sps_act_enabled_flag is 0.
[0096] In another embodiment, instead of always signaling sps_act_enabled_flag, the signaling of the flag is conditional on the input signal of chrominance type. Specifically, if ACT can be applied only when the luma and chrominance components are of the same resolution, then the flag sps_act_enabled_flag is signaled only when the input video is captured in 4:4:4 chrominance format. The modified SPS syntax table with such a change is as follows:
[0097] [Table 2]
[0098] In some embodiments, a syntax for decoding video data using ACT is provided. The design specifications are shown in the table below.
[0099] [Table 3-1] [Table 3-2] [Table 3-3] [Table 3-4] [Table 3-5] [Table 3-6] [Table 3-7] [Table 3-8] [Table 3-9]
[0100] The flag cu_act_enabled_flag indicates that the residual of the coding unit is coded in the YCgCo color space when it is 1, and indicates that the residual of the coding unit is coded in the original color space (e.g., RGB or YCbCr) when it is 0. The flag cu_act_enabled_flag is inferred to be 0 when it is not present.
[0101] In the current VVC working draft, when the input video is captured in 4:4:4 chrominance format, the transform skip mode can be applied to both the luma and chrominance components. Based on such a design, in some embodiments, three methods are used below to handle the interaction between ACT and transform skip.
[0102] In one method, when transform skip mode is enabled for one ACT·CU, the transform skip mode is applied only to the luma component and not to the chroma components. In some embodiments, syntax design specifications for such a method are shown in the following table.
[0103] [Table 4-1] [Table 4-2]
[0104] In another method, the transform skip mode is applied to both the luma and chroma components. In some embodiments, the syntax design specification for such a method is shown in the following table.
[0105] [Table 5-1] [Table 5-2]
[0106] In another method, when ACT is enabled for one CU, transform skip mode is always disabled. In some embodiments, syntax design specifications for such a method are shown in the following table.
[0107] [Table 6-1] [Table 6-2]
[0108] 8A and 8B are block diagrams illustrating an example video decoding process in which a video decoder implements techniques of inverse adaptive color space transformation (ACT) and luma mapping with chroma scaling, according to some implementations of this disclosure. In some embodiments, the video bitstream is encoded using both ACT (e.g., inverse ACT 710 in FIG. 7) and chroma residual scaling (e.g., chroma residual scaling 630 in FIG. 6). In some other embodiments, the video bitstream is encoded using only chroma residual scaling without ACT, thereby making inverse ACT 710 unnecessary.
[0109] 8A illustrates an embodiment in which the video encoder performs chroma residual scaling 630 before the inverse ACT 710. As a result, the video encoder performs luma mapping in the color space transformed domain using chroma residual scaling 630. For example, assuming the input video is captured in RGB format and transformed to YCgCo color space, the video encoder performs chroma residual scaling 630 on the chroma residuals Cg and Co according to the luma residual Y in the YCgCo color space.
[0110] 8B illustrates an alternative embodiment in which the video encoder performs chroma residual scaling 630 after the inverse ACT 710. As a result, the video encoder performs luma mapping in the original color space domain using chroma residual scaling 630. For example, assuming the input video is captured in RGB format, the video encoder applies chroma residual scaling to the B and R components.
[0111] FIG. 9 is a block diagram illustrating example decoding logic between performing adaptive color space conversion (ACT) and performing block differential pulse coded modulation (BDPCM) in accordance with some implementations of the present disclosure.
[0112] BDPCM is an encoding tool for screen content encoding. In some embodiments, the BDPCM enable flag is signaled in the SPS at the sequence level. The BDPCM enable flag is signaled only if the transform skip mode is enabled in the SPS.
[0113] When BDPCM is enabled, if the CU size is less than or equal to MaxTsSize×MaxTsSize for luma samples and the CU is intra-coded, a flag is sent at the CU level, where MaxTsSize is the maximum block size for which transform skip mode is enabled. This flag indicates whether normal intra-coding or BDPCM is used. If BDPCM is used, another BDPCM prediction direction flag is further transmitted to indicate whether the prediction is horizontal or vertical. The block is then predicted using a normal horizontal or vertical intra-prediction process with unfiltered reference samples. The residuals are quantized, and the difference between each quantized residual and its predictor, i.e., the previously coded residual of a horizontally or vertically neighboring position (depending on the BDPCM prediction direction), is coded.
[0114] For a block of size M (height) × N (width), 0 ≤ i ≤ M-1, 0 ≤ j ≤ N-1, r i,j Let be the prediction residual. Q(r i,j ) is the residual r i,j Let s denote the quantized version of . The BDPCM is applied to the quantized residual values to obtain the modified M × N array
number
number
number
number
number
[0115] For horizontal BDPCM prediction modes, 0 <i<(M-1)について、
number
number
[0116] At the decoder side, Q(r i,j ), 0≦i≦M-1, 0≦j≦N-1, the above process is reversed as follows:
number
[0117] To generate the reconstructed sample values, we dequantize the residual Q -1 (Q(r i,j )) is added.
[0118] The predicted quantized residual value
number
[0119] In some embodiments, the video encoder implements various logic to handle interactions between ACT and BDPCM. For example, when ACT is applied to an intra CU, BDPCM is enabled for the luma component but disabled for the chroma component (910). In some embodiments, when ACT is applied to an intra CU, BDPCM is enabled for the luma component but disabled for signaling of BDPCM for the chroma component. When signaling of chroma BDPCM is avoided, in one embodiment, the values of intra_bdpcm_chroma_flag and intra_bdpcm_chroma_dir are equal to those of the luma component, i.e., intra_bdpcm_flag and intra_bdpcm_dir_flag (i.e., use the same luma BDPCM direction for chroma BDPCM). In another embodiment, the values of intra_bdpcm_chroma_flag and intra_bdpcm_chroma_dir_flag are set to zero when signaled for ACT mode (i.e., disable chroma BDPCM mode for the chroma components). The tax table is shown as follows:
[0120] [Table 7-1] [Table 7-2] [Table 7-3]
[0121] In some embodiments, when ACT is applied to one intra CU, BDPCM is enabled for both luma and chroma components (920). The corresponding modified syntax table of the coding unit is shown as follows:
[0122] [Table 8-1] [Table 8-2] [Table 8-3]
[0123] In some embodiments, when ACT is applied to one intra CU, BDPCM is disabled for both luma and chroma components (930). In such a case, there is no need to signal syntax elements related to BDPCM. The corresponding modified syntax table of the coding unit is shown as follows:
[0124] [Table 9-1] [Table 9-2] [Table 9-3]
[0125] In some embodiments, for ACT mode, one suppressed chroma BDPCM signaling method is used. Specifically, when ACT is applied, the signaling of the chroma BDPCM enable / disable flag, i.e., intra_bdpcm_chroma_flag, is adjusted conditional on the presence of the luma BDPCM, i.e., intra_bdpcm_flag. The flag intra_bdpcm_chroma_flag is signaled only if the flag intra_bdpcm_flag is 1 (i.e., the luma BDPCM mode is enabled). Otherwise, the flag intra_bdpcm_chroma_flag is inferred as zero (i.e., chroma BDPCM is disabled). When the flag intra_bdpcm_chroma_flag is 1 (i.e., chroma BDPCM is enabled), the direction of the BDPCM applied to the chroma components is always set equal to the direction of the luma BDPCM, i.e., the value of the flag intra_bdpcm_chroma_dir_flag is always set equal to the value of intra_bdpcm_dir_flag. The corresponding modified syntax table for the coding unit is shown as follows:
[0126] [Table 10-1] [Table 10-2] [Table 10-3]
[0127] In some embodiments, when ACT is applied, the signaling of the chroma BDPCM mode is conditionally enabled when the intra prediction mode of the luma component is horizontal or vertical, rather than adjusting the presence of the flag intra_bdpcm_chroma_flag based on the value of intra_bdpcm_flag. Specifically, according to this method, the flag intra_bdpcm_chroma_flag is signaled only when the luma intra prediction direction is purely horizontal or vertical. Otherwise, the flag intra_bdpcm_chroma_flag is inferred as zero (meaning that chroma BDPCM is disabled). When the flag intra_bdpcm_chroma_flag is 1 (i.e., chroma BDPCM is enabled), the direction of the BDPCM applied to the chroma component is always set equal to the direction of luma intra prediction. The following table shows the signaling of the BDPCM mode for the chroma component. Here, the numbers 18 and 50 represent the current intra prediction indexes for horizontal and vertical intra prediction in the current VVC draft. The corresponding modified syntax table of the coding unit is shown as follows:
[0128] [Table 11-1] [Table 11-2] [Table 11-3]
[0129] In some embodiments, context modeling for luma / chroma BDPCM mode is implemented. In the current BDPCM design in VVC, the signaling of BDPCM for luma and chroma components reuses the same context modeling. In particular, one single context is shared by the luma BDPCM enable / disable flag (i.e., intra_bdpcm_flag) and the chroma BDPCM enable / disable flag (i.e., intra_bdpcm_chroma_flag), and another single context is shared by the luma BDPCM direction flag (i.e., intra_bdpcm_dir_flag) and the chroma BDPCM direction flag (i.e., intra_bdpcm_chroma_dir_flag).
[0130] In some embodiments, to improve coding efficiency, in one method separate contexts are used to signal BDPCM enable / disable for luma and chroma components. In another embodiment, separate contexts are used to signal BDPCM direction flags for luma and chroma components. In another embodiment, two additional contexts are used to code the chroma BDPCM enable / disable flag: a first context used to signal intra_bdpcm_chroma_flag when luma BDPCM mode is enabled, and a second context used to signal intra_bdpcm_chroma_flag when luma BDPCM mode is disabled.
[0131] In some embodiments, ACT is handled using lossless coding. In the HEVC standard, the lossless mode of one CU is indicated by signaling one CU level flag cu_transquant_bypass_flag set to 1. However, in the ongoing VVC standardization process, a different lossless enabling method is applied. Specifically, when one CU is coded in lossless mode, it is only necessary to skip the transform and use a quantization step size to 1. This can be achieved by signaling the QP value of the CU level to 1 and signaling the transform_skip_flag of the TU level to 1. Thus, in one embodiment of the present disclosure, for one CU / TU, the lossy ACT and the lossless ACT are switched according to the value of the transform_skip_flag and the QP value. As shown below, if the flag transform_skip_flag is 1 and the QP value is 4, the lossless ACT is applied, otherwise the lossy version of ACT is applied.
[0132] If transform_skip_flag is 1 and QP is 4, the residual sample r Y , r Cb and r Cr The (nTbW) × (nTbH) array (x=0..nTbW-1, y=0..nTbH-1) is tmp=r Y [x][y]-(r Cb [x][y]>>1) r Y [x][y]=tmp+r Cb [x][y] r Cb [x][y]=tmp-(r Cr [x][y]>>1) r Cr [x][y]=r Cb [x][y]+r Cr [x][y] will be changed to:
[0133] Otherwise, the residual sample r Y , r Cb and rCr The (nTbW) × (nTbH) array (x=0..nTbW-1, y=0..nTbH-1) is tmp=r Y [x][y]-r Cb [x][y] r Y [x][y]=r Y [x][y]+r Cb [x][y] r Cb [x][y]=tmp-r Cr [x][y] r Cr [x][y]=tmp+r Cr [x][y] will be changed to:
[0134] In the above description, separate ACT matrices are used for lossy coding and lossless coding. To realize another unified design, a lossless ACT matrix is used for both lossy coding and lossless coding. In addition, if the lossless ACT increases the dynamic range of the Cg and Co components by 1 bit, an additional 1-bit right shift is applied to the Cg and Co components after the forward ACT, and a 1-bit left shift is applied to the Cg and Co components before the reverse ACT. As described below,
[0135] If transform_skip_flag is 0 or QP is not 4, the residual sample r Cb and r Cr (nTbW) × (nTbH) array (x=0..nT bW-1, y=0..nTbH-1) is r Cb [x][y]=r Cb [x][y]<<1 r Cr [x][y]=r Cr [x][y]<<1 will be changed to:
[0136] Residual sample r Y , r Cb and r CrThe (nTbW) × (nTbH) array (x=0..nTbW-1, y=0..nTbH-1) is tmp=r Y [x][y]-(r Cb [x][y]>>1) r Y [x][y]=tmp+r Cb [x][y] r Cb [x][y]=tmp-(r Cr [x][y]>>1) r Cr [x][y]=r Cb [x][y]+r Cr [x][y] will be changed to:
[0137] In addition, as can be seen from above, when ACT is applied, a QP offset (-5, -5, -3) is applied to the Y, Cg and Co components. Therefore, for small input QP values (e.g., <5), an undefined negative QP will be used for quantization / dequantization of ACT coefficients. To solve such a problem, one clipping operation, i.e., QP'=max(QP), is applied after the QP adjustment of ACT so that the applied QP value is always equal to or greater than zero. org -QP offset ,0) is added, where QP is the original QP, and QP offset is the ACT·QP offset and QP' is the adjusted QP value.
[0138] Although the method as described above uses the same ACT matrix for lossy and lossless coding (ie, the lossless ACT matrix), the following two problems can still be identified.
[0139] Depending on whether the current CU is a lossy or lossless CU, different inverse ACT operations are subsequently applied. Specifically, inverse ACT is applied to lossless CUs, and an additional right shift must be applied to lossy CUs before inverse ACT. In addition, the decoder needs to know whether the current CU is coded in a lossy or lossless mode. This is inconsistent with the current VVC lossless design. In particular, unlike the HEVC lossless design in which the lossless mode of a CU is represented by signaling a cu_transquant_bypass_flag, lossless coding in VVC is done in a purely non-normative way, i.e., skipping the transform of prediction residuals (by enabling transform skip mode for luma and chroma components), selecting an appropriate QP value (i.e., 4), and explicitly disabling coding tools such as in-loop filters that prevent lossless coding.
[0140] Here, the QP offset used to normalize the ACT is fixed. However, the selection of the optimal QP offset in terms of coding efficiency may depend on the content itself. Therefore, when the ACT tool is enabled, it may be more beneficial to enable flexible QP offset signaling to maximize its coding gain.
[0141] Based on the above considerations, one unified ACT design is implemented as follows: First, for CUs coded in both lossy and lossless modes, lossless ACT and lossless inverse ACT are applied. Second, instead of using a fixed QP offset, the QP offsets applied to the ACT·CUs in the bitstream (i.e., three QP offsets applied to the Y, Cg and Co components) are explicitly signaled. Third, to prevent possible overflow issues of the QPs applied to the ACT·CUs, a clipping operation to a valid QP range is applied to the resulting QPs of each ACT·CU. As can be seen, based on the above methods, lossy and lossless coding are implemented in a unified ACT design. The choice between lossy and lossless coding may be realized purely by changing the encoder (i.e., using different encoder settings). The decoding process is identical for lossy and lossless coding of ACT·CU. Specifically, to enable lossless coding, the encoder only needs to signal the values of three QP offsets as zero in addition to the existing lossless configuration on the encoder side. On the other hand, to enable lossy coding, the encoder may signal non-zero QP offsets. For example, in one embodiment, QP offsets (-5, 1, 3) may be signaled for Y, Cg, and Co components when ACT is applied to compensate for the dynamic range change brought about by lossless ACT in lossy coding. On the other hand, the ACT·QP offsets may be signaled at different coding levels, e.g., at the Sequence Parameter Set (SPS), Picture Parameter Set (PPS), picture header, coded block group level, etc., resulting in different QP adaptations at different granularities. The following table shows an example where QP offset signaling is performed at SPS.
[0142] [Table 12]
[0143] In another embodiment, one high-level control flag is added in the SPS or PPS (e.g., picture_header_act_qp_offset_present_flag). If this flag is zero, it means that the QP offset signaled in the SPS or PPS applies to all CUs coded in ACT mode. Otherwise, when the flag is 1, additional QP offset syntax (e.g., picture_header_y_qp_offset_plus5, picture_header_cg_qp_offset_minus1, and picture_header_co_qp_offset_minus3) can be further signaled in the picture header to separately control the QP value applied to the ACT·CU in one particular picture.
[0144] On the other hand, the signaled QP offset should also be applied to clip the final ACT·QP value to the valid dynamic range. In addition, different clipping ranges may be applied for CUs coded with transform and CUs coded without transform. For example, when no transform is applied, the final QP should not be smaller than 4. Assuming that the ACT·QP offset is signaled at the SPS level, the derivation process of the corresponding QP value for the ACT·CU can be described as follows:
[0145]
number
[0146] In another embodiment, the ACT enable / disable flag is signaled at the SPS level while signaling the ACT QP offset at the PPS level, so that the encoder can more flexibly adjust the QP offset applied to the ACT CU to improve the coding efficiency. Specifically, the syntax tables of SPS and PPS with the described changes are shown below.
[0147] [Table 13]
[0148] pps_act_qp_offset_present_flag equal to 1 indicates the presence of pps_act_y_qp_offset_plus5, pps_act_cg_qp_offset_minus1 and pps_act_co_qp_offset_minus3 in the bitstream. When pps_act_qp_offset_present_flag is equal to 0, the syntax elements pps_act_y_qp_offset_plus5, pps_act_cg_qp_offset_minus1 and pps_act_co_qp_offset_minus3 are not present in the bitstream. A value of 0 for pps_act_qp_offset_present_flag when sps_act_enabled_flag is equal to 0 indicates bitstream compliance.
[0149] pps_act_y_qp_offset_plus5, pps_act_cg_qp_offset_minus1 and pps_act_co_qp_offset_minus3 are used to determine offsets applied to the values of the quantization parameters used for the luma and chroma components of coding blocks where cu_act_enabled_flag is 1. When not present, pps_act_y_qp_offset_plus5, pps_act_cg_qp_offset_minus1 and pps_act_cr_qp_offset_minus3 are inferred to have a value of 0.
[0150] In the above PPS signaling, the same QP offset value is applied to each ACT·CU when joint coding of chrominance residual mode (JCCR) is applied or not applied. Such a design may not be optimal when only the residual of the chrominance component of one signal is coded in JCCR mode. Therefore, to achieve a better coding gain, another QP offset may be applied to code the residual of the chrominance component when JCCR mode is applied to one ACT·CU. Based on such consideration, one separate QP offset signaling is added in the PPS for JCCR mode, as specified below.
[0151] [Table 14]
[0152] The quantizer used for the chrominance residual of a coding block for which joint chrominance residual coding is applied The pps_joint_cbcr_qp_offset is used to determine the offset to be applied to the value of the quantization parameter. When pps_joint_cbcr_qp_offset is not present, a value of zero is inferred.
[0153] 10 illustrates a method for handling ACT when the internal bit depths of luma and chroma are different in some embodiments. Specifically, FIG. 10 illustrates a decoding flow diagram for applying different QP offsets to different components when the internal bit depths of luma and chroma are different in some implementations of the present disclosure.
[0154] According to the existing VVC specification, it is allowed to use separate internal bit depths (denoted as BitDepthY and BitDepthC) for the luma and chroma components for encoding. However, the existing ACT design assumes that the luma internal bit depth and the chroma internal bit depth are always the same. In the following paragraphs, a method is realized to improve the ACT design when BitDepthY is different from BitDepthC.
[0155] In the first method, the ACT tool is disabled whenever the internal bit depth of the luma component differs from that of the chroma components.
[0156] In the second method, in the second solution, the smaller bit depth of the luma and chroma components is left-shifted to match its bit depth with that of the other component, and then the scaled component is readjusted to its original bit depth by bit-shifting it to the right after color conversion.
[0157] Similar to HEVC, the quantization step size is approximately 2 for each increment of QP. 1 / 6 10 , the QP value is increased by 100 times and exactly doubled every 6 increments. Based on such design, in the second method, to compensate for the internal bit depth between luma and chroma, the QP value used for the component with the smaller internal bit depth is increased by 6Δ, where Δ is the difference between the internal bit depth of luma and the internal bit depth of chroma. Then, the residual of the component is shifted to the original dynamic range by applying a right shift of Δ bits. Figure 10 shows the corresponding decoding process when the above method is applied. For example, assuming that the input QP value is qp, the default QP values applied for the Y, Cg and Co components are qp-5, qp-5 and qp-3. It is further assumed that the internal bit depth of luma is larger than that of chroma, and Δ=BitDepthY-BitDepthC. Then, the final QP values applied for the luma and chroma components are qp-5, qp-5+6Δ and qp-3+6Δ.
[0158] In some embodiments, an encoder acceleration logic is implemented. The simplest approach to select a color space for residual coding of one CU is for the encoder to check each coding mode (e.g., intra coding mode, inter coding mode, and IBC mode) twice, with ACT enabled and disabled. This nearly doubles the coding complexity. To further reduce the coding complexity of ACT, the following encoder acceleration logic is implemented in this disclosure.
[0159] First, because YCgCo space is more compact than RGB space, when the input video is in RGB format, we first check the rate distortion (RD) cost of enabling the ACT tool, and then check the RD cost of disabling the ACT tool. In addition, when ACT is enabled, the RD cost of disabling color space conversion is calculated only if there is at least one non-zero coefficient. Alternatively, when the input video is in YCbCr format, the RD cost of disabling ACT is checked, followed by the RD check of enabling ACT. The second (ACT-enabling) RD check is performed. only when ACT is disabled and there is at least one non-zero coefficient.
[0160] Secondly, to reduce the number of coding mode tests, the same coding mode is used for two color spaces. More specifically, for intra mode, the intra prediction mode selected for full RD cost comparison is shared between two color spaces, for inter mode, the selected motion vector, reference picture, motion vector predictor and merge index (used for inter merge mode) are shared between two color spaces, and for IBC mode, the selected block vector, block vector predictor and merge index (used for IBC merge mode) are shared between two color spaces.
[0161] Third, due to the quaternary / binary / trifurcated tree partitioning structure used in VVC, one same block partitioning can be obtained by a combination of various partitions. In order to speed up color space selection, when one same block is realized by various partitioning paths, the ACT enable / disable decision is used. Specifically, when a CU is first encoded, the color space selected to encode the residual of one particular CU is stored. Then, when the same CU is obtained by another partitioning path, the stored color space decision is directly reused instead of selecting one of the two spaces.
[0162] Fourth, when there is a strong correlation between one CU and its spatial neighbors, the color space selection information of its spatial neighboring blocks is used to determine how many color spaces need to be considered for the residual coding of the current CU. For example, if a sufficient number of spatial neighboring blocks select the YCgCo space to code their residuals, it is reasonable to infer that the current CU is likely to select the same color space. Correspondingly, the RD check of the residual coding of the current CU in the original color space can be skipped. If a sufficient number of spatial neighbors select the original color space, the RD check of the residual coding in the YCgCo domain can be avoided. Otherwise, both of the two color spaces need to be tested.
[0163] Fifth, if there is a strong correlation between CUs in the same region, one CU can select the same color space as the parent CU to encode its residual. Or, the child CU can derive the color space from the information of the parent CU, such as the selected color space and the RD cost of each color space. Therefore, to simplify the encoding complexity, for one CU, if the residual of the parent CU is encoded in the YCgCo domain, the check of the RD cost of the residual encoding in the RGB domain is skipped. In addition, if the residual of the parent CU is encoded in the RGB domain, the check of the RD cost of the residual encoding in the YCgCo domain is skipped. Another traditional way is to use the RD cost of the parent CU in the two color spaces when two color spaces are tested in the encoding of the parent CU. If the parent CU selects the YCgCo color space, and the RD cost of YCgCo is much smaller than that of RGB, the RGB color space is skipped, and vice versa.
[0164] In some embodiments, the efficiency of 4:4:4 video encoding is improved by enabling luma-only coding tools for the chroma components. Because VVC design is primarily focused on video captured in 4:2:0 chroma format, most existing inter / intra coding tools are only effective for the luma component and are ineffective for the chroma components. However, as previously discussed, video signals in 4:4:4 chroma format exhibit significantly different characteristics compared to 4:2:0 video signals. For example, the Cb / B and Cr / R components of 4:4:4 YCbCr / RGB video, similar to the luma component, often contain useful high frequency texture and edge information. This is different from the chroma components in 4:2:0 video, which are usually very smooth and contain much less information than the luma component. Based on such analysis, it is considered that when the input video is in 4:4:4 chroma format, the Cb / B and Cr / R components of 4:4:4 YCbCr / RGB video ... When this is done, the following method is implemented to extend some of the current VVC coding tools that are only for luma to the chroma components.
[0165] First, a luma interpolation filter is enabled for the chroma components. The VVC standard, like HEVC, utilizes motion compensation prediction techniques to exploit redundancy between temporally neighboring pictures, and supports motion vectors with 1 / 16 pixel accuracy for the Y component and 1 / 32 pixel accuracy for the Cb and Cr components. Fractional samples are interpolated using a set of separable 8-tap filters. The fractional interpolation of the Cb and Cr components is essentially the same as that of the Y component, except that a separable 4-tap filter is used for the 4:2:0 video format. This is because, for 4:2:0 video, the Cb and Cr components contain much less information than that of the Y component, and the 4-tap interpolation filter can reduce the filtering complexity of the fractional interpolation without truncating the efficiency of the motion compensation prediction for the Cb and Cr components when compared with the 8-tap interpolation filter.
[0166] As previously pointed out, existing 4-tap chroma interpolation filters may not be efficient for interpolating fractional samples for motion-compensated prediction of chroma components in 4:4:4 video. Therefore, in one embodiment of the present disclosure, the same set of 8-tap interpolation filters (used for the luma component of 4:2:0 video) is used for fractional sample interpolation of both luma and chroma components in 4:4:4 video. In another embodiment, adaptive interpolation filter selection for chroma samples of 4:4:4 video is enabled for a better tradeoff between coding efficiency and complexity. For example, one interpolation filter selection flag may be signaled at SPS level, PPS level, and / or slice level to indicate whether an 8-tap interpolation filter (or other interpolation filter) or a default 4-tap interpolation filter is used for the chroma component at various coding levels.
[0167] Second, PDPC and MRL are enabled for the chrominance components.
[0168] The Position-Dependent Intra Prediction Combining (PDPC) tool in VVC extends the above concept by employing a weighted combination of intra prediction samples with unfiltered reference samples. In the current VVC working draft, PDPC is enabled for unsignaled intra modes such as planar, DC, horizontal (i.e., mode 18), vertical (i.e., mode 50), near-bottom-left diagonal (i.e., modes 2, 3, 4, ..., 10), and near-top-right diagonal (i.e., modes 58, 59, 60, ..., 66). Considering a predicted sample whose coordinates (x,y) are located as pred(x,y), the corresponding value after PDPC is pred(x,y)=(wL×R-1,y+wT×Rx,-1-wTL×R-1,-1+(64-wL-wT+wTL)×pred(x,y)+32)>>6 where Rx,-1, R-1,y represent the top and left reference samples of the current sample (x,y), respectively, and R-1,-1 represent the reference sample in the upper left corner of the current block. The weights wL, wT, and wTL in the above equations are adaptively selected depending on the prediction mode and sample position, and can be written as follows assuming that the size of the current coding block is W×H:
[0169] For DC mode: wT=32>>((y<<1)>>shift), wL=32>>((x<<1)>>shift), wTL=(wL>>4)+(wT>>4)
[0170] For the planar mode: wT=32>>((y<<1)<<shift)、wL=32> >((x<<1)>>shift), wTL=0
[0171] For the horizontal mode: wT=32>>((y<<1)>>shift), wL=32>>((x<<1)>>shift), wTL=wT
[0172] For the vertical mode: wT=32>>((y<<1)>>shift), wL=32>>((x<<1)>>shift), wTL=wL
[0173] For the bottom left diagonal direction: wT=16>>((y<<1)>>shift), wL=16>>((x<<1)>>shift), wTL=0
[0174] For the upper right diagonal direction: wT=16>>((y<<1)>>shift), wL=16>>((x<<1)>>shift), wTL=0 Here, shift = (log2(W)-2+log2(H)-2+2)>>2.
[0175] Unlike HEVC, where only the nearest row / column of reconstructed samples is used as reference, VVC introduces a multi-reference line (MRL) where two additional rows / columns are used for intra prediction. The index of the selected reference row / column is signaled from the encoder to the decoder. When a non-nearest row / column is selected, the planar and DC modes are excluded from the set of intra modes that can be used to predict the current block.
[0176] In the current VVC design, only the PDPC tool is employed by the luma component to reduce / eliminate discontinuity between intra-prediction samples and their reference samples derived from reconstructed neighboring samples. However, as mentioned before, a video signal in 4:4:4 chroma format may have rich texture information in chroma blocks. Therefore, a tool such as PDPC, which uses a weighted average of unfiltered reference samples and intra-prediction samples to improve prediction quality, should also be beneficial to improve chroma coding efficiency of 4:4:4 video. Based on such consideration, in one embodiment of the present disclosure, the PDPC process is enabled for intra-prediction of chroma components in 4:4:4 video.
[0177] The same considerations can be extended to MRL tools. In current VVC, MRLs may not be applied to chrominance components. Based on one embodiment of the present disclosure, MRLs for chrominance components of 4:4:4 video are enabled by signaling one MRL index for chrominance components of one intra CU. Based on this embodiment, various methods can be used. In one method, one additional MRL index is signaled and can be shared by both Cb / B and Cr / R components. In another method, two MRL indexes are signaled, one for each chrominance component. In a third method, luma MRL indexes for intra prediction of chrominance components are reused so that no additional MRL signaling is required to enable MRLs for chrominance components.
[0178] Third, an ISP is enabled for the chrominance components.
[0179] In some embodiments, to further improve intra-coding efficiency, a coding tool called sub-split prediction (ISP) is introduced into VVC. Conventional intra-modes only utilize the reconstructed samples adjacent to one CU to generate intra-predicted samples of a block. Based on such design, the spatial correlation between the predicted samples and the reference samples is approximately proportional to the distance between them. Therefore, the prediction quality of samples in the inner part (especially the samples at the lower right corner of the block) is usually poorer than that of samples close to the block boundary. ISP divides the current CU into two or four sub-blocks in either horizontal or vertical direction depending on the block size, and each sub-block contains at least 16 samples. The reconstructed samples in one sub-block can be used as reference to predict samples in the next sub-block. The above process is repeated until all the sub-blocks in the current CU are coded. In addition, to reduce the signaling overhead, all the sub-blocks in one ISPCU share the same intra-mode. In addition, according to the existing ISP design, sub-block division is only applicable to luma component. Specifically, only the luma samples of one ISPCU can be further divided into multiple sub-blocks (or TUs), and each luma sub-block is coded separately. However, the chroma samples of the ISPCU are not divided. In other words, for the chroma components, the CU is used as a processing unit for intra prediction, transformation, quantization and entropy coding without further division.
[0180] In current VVC, when ISP mode is enabled, TU splitting is applied only to luma samples, and chroma samples are coded without further splitting into multiple TUs. According to one embodiment of the present disclosure, ISP mode is also enabled for chroma coding in 4:4:4 video, since there is rich texture information in the chroma plane. Based on this embodiment, various methods can be used. In one method, one additional ISP index is signaled and shared by the two chroma components. In another method, two additional ISP indexes are signaled separately, one for Cb / B and one for Cr / R. In a third method, the ISP index used for the luma component is reused for ISP prediction of the two chroma components.
[0181] Fourth, a novel intra prediction technique, matrix-based intra prediction (MIP), is enabled for the chrominance components.
[0182] To predict samples of a rectangular block of width W and height H, MIP takes as input one reconstructed H line of border samples from the block's left neighbor and one reconstructed W line of border samples from the block's top neighbor. If reconstructed samples are not available, they are generated as is done in conventional intra prediction.
[0183] In some embodiments, only MIP mode is enabled for the luma component. In one embodiment, MIP is enabled for the chroma components of 4:4:4 video for the same reasons as enabling ISP mode for the chroma components. Two signaling methods may be applied. In the first method, two MIP modes are signaled separately: the MIP mode used for the luma component and the MIP mode used for the two chroma components. In the second method, only one single MIP mode shared by the luma and chroma components is signaled.
[0184] Fifth, multiple transform selection (MTS) is enabled for the chrominance components.
[0185] In addition to the DCT-II adopted in HEVC, the MTS scheme is used for residual coding of both inter-coded and intra-coded blocks. It uses several transforms selected from T8 / DST7. The newly introduced transform matrices are DST-VII and DCT-VIII.
[0186] In current VVC, MTS tool is only enabled for luma component. In one embodiment of the present disclosure, MIP is enabled for chroma component of 4:4:4 video. Two signaling methods can be applied. In the first method, when MTS is enabled for one CU, two transform indexes are signaled separately: the MIP mode used for luma component and the MIP mode used for two chroma components. In the second method, when MTS is enabled, one transform index shared by luma and chroma components is signaled.
[0187] In some embodiments, unlike the HEVC standard, in which a fixed lookup table is used to derive the quantization parameter (QP) used from the chrominance component based on the luma QP, in the VVC standard, a luma to chrominance mapping table is sent from the encoder to the decoder, which is defined by several pivot points of a piecewise linear function. Specifically, the syntax elements and reconstruction process of the luma to chrominance mapping table are described as follows:
[0188] [Table 15]
[0189] same_qp_table_for_chroma is 1, which specifies that only one chroma QP mapping table is signaled, and this table applies to both Cb and Cr residuals. sps_joint_cbcr_enabled_flag is 1. also applies to the joint Cb-Cr residual when sps_joint_cbcr_enabled_flag is set to 1. When same_qp_table_for_chroma is 0, it specifies that two chroma QP mapping tables are signaled in the SPS: one for Cb and one for Cr, plus an additional one for joint Cb-Cr when sps_joint_cbcr_enabled_flag is set to 1. When no identical qp_table_for_chroma exists in the bitstream, the value of same_qp_table_for_chroma is inferred to be 1.
[0190] qp_table_start_minus26[i] plus 26 specifies the starting luma and chroma QPs used to describe the i-th chroma QP mapping table. The value of qp_table_start_minus26[i] shall be in the range of -26-QpBdOffset to 36, inclusive. When qp_table_start_minus26[i] is not present in the bitstream, the value of qp_table_start_minus26[i] is inferred to be 0.
[0191] num_points_in_qp_table_minus1[i] plus 1 specifies the number of points used to describe the i-th chroma QP mapping table. The value of num_points_in_qp_table_minus1[i] shall be in the range 0 to 63 + QpBdOffset, inclusive. When num_points_in_qp_table_minus1[0] is not present in the bitstream, the value of num_points_in_qp_table_minus1[0] is inferred to be 0.
[0192] delta_qp_in_val_minus1[i][j] specifies the increment value used to derive the input coordinate of the jth pivot point of the ith chroma QP mapping table. When delta_qp_in_val_minus1[0][j] is not present in the bitstream, the value of delta_qp_in_val_minus1[0][j] is inferred to be 0.
[0193] delta_qp_diff_val[i][j] specifies the increment value used to derive the output coordinate of the jth pivot point of the ith chroma QP mapping table.
[0194] The i-th chroma QP mapping table, ChromaQpTable[i] for i=0..numQpTables-1, is derived as follows:
number
[0195] In some embodiments, an improved luma-to-chroma mapping function for RGB video is disclosed herein.
[0196] In some embodiments, when the input video is in RGB format, a two-part linear function is sent from the encoder to the decoder to map the luma QP to the chroma QP. This is done by setting the syntax elements: same_qp_table_for_chroma=1, qp_table_start_minus26[0]=0, num_points_in_qp_table_minus1[0]=0, delta_qp_in_val_minus1[0][0]=0, delta_qp_diff_val[0][0]=0. Specifically, the corresponding luma to chroma QP mapping function is defined as follows:
number
[0197] Table 16 shows the luma to chroma QP mapping function applied to RGB encoding, assuming the intra encoding bit depth is 10 bits. [Table 16]
[0198] As shown in equation (5), unequal QP values are used to code the luma and chroma components when the luma QP is greater than 26. In some embodiments, the impact of unequal QP values extends not only to the quantization / dequantization process, but also to the decisions made during rate-distortion (RD) optimization if weighted chroma distortion is used when computing the RD cost for the mode decision as defined in the following equation:
[0199] J mode =(SSE luma +W chroma · SSE chroma )+λ mode ·R mode (6)
[0200] In equation (6), SSE luma and SSE chroma are the distortions of the luminance and chrominance components, respectively, and R mode is the number of bits, and λ mode is the Lagrange multiplier, and W chroma is a weighting parameter for chrominance distortion and is calculated as follows:
number
[0201] However, compared to YCbCr / YUV video, the correlation between the three channels of RGB video is stronger. Therefore, when video content is captured in RGB format, there is usually strong texture and high frequency information in all three components, i.e., R, G and B information are equally important. Therefore, in one embodiment of the present disclosure, for RGB encoding, equal QP values are applied for all three channels. This can be done by setting the corresponding luma to chroma QP mapping syntax elements as same_qp_table_for_chroma=1, qp_table_start_minus26[0]=0, num_points_in_qp_table_minus1[0]=0, delta_qp_in_val_minus1[0][0]=0, delta_qp_diff_val[0][0]=1. Correspondingly, the RGB luma to chroma mapping function can be calculated using the method disclosed herein as follows: 8) and Table 17.
[0202] QP C =QP L (8)
[0203] [Table 17]
[0204] 11A and 11B are block diagrams illustrating an example video decoding process in which a video decoder implements a clipping technique to limit the dynamic range of a coding unit residual within a predetermined range for processing by inverse ACT, in accordance with some implementations of the present disclosure.
[0205] More specifically, Figure 11A shows an example where a clipping operation 1102 is applied to the input of the inverse ACT 1104. Figure 11B shows an example where a clipping operation 1102 is applied to the input of the inverse ACT 1104 and a clipping operation 1106 is applied to the output. In some embodiments, the clipping operations 1102 and 1106 are identical. In some embodiments, the clipping operations 1102 and 1106 are different.
[0206] In some embodiments, the bit depth control method is implemented using ACT. According to existing ACT designs, the input of the inverse ACT process in the decoder is the output residual from other residual decoding processes (e.g., inverse transform, inverse BDPCM, inverse JCCR). In current implementations, those residual samples may reach the maximum value of a 16-bit signed integer. In such designs, the inverse ACT cannot be implemented by a 16-bit implementation, which is very expensive for hardware implementation. To solve such problems, a clipping operation is applied once to the input residual of the inverse ACT process. In one embodiment, the following clipping operation is applied to the input residual of the inverse ACT process: Clip input =Clip(-(2 Bitdepth -1),2 Bitdepth -1,M) Bitdepth is the internal coding bit depth.
[0207] In another embodiment, a different clipping operation is applied to the input residual of the inverse ACT process: Clip input =Clip(-(2 15 -1),2 15 -1,M)
[0208] Additionally, in another embodiment, a clipping operation is applied to the output residual of the inverse ACT. In one embodiment, a clipping operation is applied to the output of the inverse ACT: Clip output =Clip(-(2 Bitdepth -1),2Bitdepth -1,M)
[0209] In another embodiment, a clipping operation is applied to the output of the inverse ACT: Clip output =Clip(-(2 15 -1),2 15 -1,M)
[0210] As shown in the lifting operation illustrated in FIG. 5B, when the reversible YCgCo transform is applied, the lifting operation will increase the dynamic range of the Cg and Co components by 1 bit. Therefore, in one embodiment, in order to preserve the accuracy of the residual samples output from the reversible YCgCo transform, the input residual of the inverse ACT is clipped according to the following formula: Clip input =Clip(-2 Bitdepth+1 ,2 Bitdepth+1 -1,M)
[0211] In another embodiment, a different clipping operation is applied to the input Y, Cg, and Co residuals to the inverse ACT. Specifically, since the bit depth of the Y component remains unchanged before and after the reversible ACT, the input luma residual to the inverse ACT is clipped by the following operation: Clip input =Clip(-2 Bitdepth ,2 Bitdepth -1,M)
[0212] For the Cg and Co components, as the bit depth increases, the corresponding input residuals to the inverse ACT are clipped by the operation: Clip input =Clip(-2 Bitdepth+1 ,2 Bitdepth+1 -1,M)
[0213] In another embodiment, the input residual to the inverse ACT is clipped by the operation: Clip input =Clip(-2 C ,2 C -1,M) C is a constant.
[0214] FIG. 12 is a flow diagram 1200 illustrating an example process in which a video decoder (e.g., video decoder 30) decodes video data by performing a clipping operation to limit the dynamic range of a residual of a coding unit within a predetermined range for processing by inverse ACT, in accordance with certain implementations of the present disclosure.
[0215] Video decoder 30 receives video data from the bitstream that corresponds to a coding unit that has been coded using an intra-prediction mode or an inter-prediction mode (1210).
[0216] Video decoder 30 then receives a first syntax element from the video data, the first syntax element indicating whether the coding unit is coded using adaptive color space transformation (ACT) (1220).
[0217] Video decoder 30 then processes the video data to generate a residual for the coding unit (1230). In response to determining, based on the first syntax element, that the coding unit is coded using ACT, video decoder 30 performs a clipping operation on the residual for the coding unit (1240). Video decoder 30 applies inverse ACT to the residual for the coding unit after the clipping operation (1250).
[0218] In some embodiments, the clipping operation limits the dynamic range of the coding unit residual to within a predetermined range for processing by the inverse ACT.
[0219] In some embodiments, the clipping operation comprises: Clip input =Clip(-2 Bitdepth+1 ,2 Bitdepth+1 -1,M) where M is the input to the clipping operation, Bitdepth is the internal coding bit depth, and Clip input is -2 Bitdepth+1 ~(2 Bitde pth+1 In some embodiments, the input M is the output of the clipping operation in the range of -2 Bitdepth+1 If it is less than this, the output of the clipping operation will be -2 Bitdepth+1 In some embodiments, the input M is set to (2 Bitdepth+1 -1), the output of the clipping operation is (2 Bitdepth+1 -1).
[0220] In some embodiments, the clipping operation comprises: Clip input =Clip(-2 Bitdepth ,2 Bitdepth -1,M) where M is the input to the clipping operation, Bitdepth is the internal coding bit depth, and Clip input is -2 Bitdepth ~(2 Bitdepth In some embodiments, the input M is the output of the clipping operation in the range of -2 Bitdepth If it is less than this, the output of the clipping operation will be -2 Bitdepth In some embodiments, the input M is set to (2 Bitdepth -1), the output of the clipping operation is Bitdepth -1).
[0221] In some embodiments, video decoder 30 applies an inverse transform to the residual of the coding unit before performing the clipping operation.
[0222] In some embodiments, video decoder 30 applies a second clipping operation to the residual of the coding unit after applying the inverse ACT to the residual of the coding unit.
[0223] In some embodiments, the clipping operation adjusts the dynamic range of the coding unit residual to within a fixed intra-coding bit depth implemented by the inverse ACT.
[0224] In some embodiments, the fixed intra-coding bit depth is 15.
[0225] In some embodiments, video decoder 30 receives a second syntax element from the video data after receiving the first syntax element from the video data, the second syntax element indicating a variable intra-coding bit depth used in the clipping operation.
[0226] In some embodiments, after receiving the first syntax element from the video data, video decoder 30 receives a second syntax element from the video data, the second syntax element indicating a first intra-coding bit depth, and the second intra-coding bit depth used in the clipping operation is the first intra-coding bit depth plus one.
[0227] In some embodiments, the clipping operation further comprises a first clipping operation applied to a luma component of the residual of the coding unit and a second clipping operation applied to a chroma component of the residual of the coding unit.
[0228] In some embodiments, the first clipping operation limits the dynamic range of a luma component of the residual of the coding unit to within a first intra-coding bit depth plus one, and the second clipping operation limits the dynamic range of a chroma component of the residual of the coding unit to within a second intra-coding bit depth plus one, the second intra-coding bit depth being the first intra-coding bit depth plus one.
[0229] In one or more examples, the functions described may be implemented in hardware, software, firmware, or other The functions may be implemented in software, software, or any combination thereof. If implemented in software, the functions may be stored on or transmitted as one or more instructions or codes on a computer-readable medium and executed by a hardware-based processing unit. A computer-readable medium may include a computer-readable medium corresponding to a tangible medium, such as a data storage medium, or a communication medium including any medium that facilitates transfer of a computer program from one place to another, for example via a communication protocol. Thus, a computer-readable medium may generally correspond to (1) a tangible, non-transitory computer-readable medium, or (2) a communication medium, such as a signal or carrier wave. A data storage medium may be any available medium that can be accessed by one or more computers or one or more processors to retrieve instructions, codes and / or data structures for implementing the embodiments described herein. A computer program product may include a computer-readable medium.
[0230] The terminology used in the description of the embodiments herein is for the purpose of describing only certain embodiments and is not intended to limit the scope of the claims. As used in the description of the embodiments and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise. The term "and / or" as used herein is also understood to refer to and include any and all possible combinations of one or more of the associated listed items. It is further understood that the terms "comprise" and / or "comprising" as used herein specify the presence of stated features, elements, and / or components, but do not exclude the presence or addition of one or more other features, elements, components, and / or groups thereof.
[0231] Terms such as first, second, etc. may be used herein to describe various elements, but it will be understood that these elements should not be limited by these terms. These terms are merely used to distinguish one element from another. For example, a first electrode may be referred to as a second electrode, and similarly, a second electrode may be referred to as a first electrode, without departing from the scope of the embodiment. The first electrode and the second electrode are both electrodes, but are not the same electrode.
[0232] The description in this application has been presented for purposes of illustration and description and is not intended to be exhaustive or limited to the invention in the form disclosed. Many modifications, variations, and alternative embodiments will be apparent to those skilled in the art having the benefit of the teachings presented in the foregoing description and the associated drawings. The embodiments have been selected and described in order to best explain the principles and practical applications of the invention, to enable others skilled in the art to understand the invention in terms of various embodiments, and to best utilize the basic principles and various embodiments with various modifications as appropriate for the particular use contemplated. It is therefore to be understood that the claims are not limited to the specific examples of the disclosed embodiments and modifications thereof, but that other embodiments are contemplated to be included within the scope of the appended claims.
Claims
1. 1. A method for decoding video data, comprising the steps of: receiving, from a bitstream, video data corresponding to a coding unit that is coded in an intra-prediction mode or an inter-prediction mode; receiving from the video data a first syntax element indicating whether the coding unit is coded using adaptive color space transformation (ACT); processing the video data to generate a residual for the coding unit; In response to determining, based on the first syntax element, that the coding unit is coded using the ACT, performing a clipping operation on the residual of the coding unit; applying an inverse ACT to the residual of the coding unit after the clipping operation; A method for providing the above.
2. The method of claim 1 , wherein the clipping operation limits the dynamic range of the residual of the coding unit to within a predetermined range for processing by the inverse ACT.
3. 2. The method of claim 1, wherein the clipping operation comprises: Clip input =Clip(-2 Bitdepth+1 , 2 Bitdepth+1 -1, M) where M is the input to the clipping operation, Bitdepth is the intra-coding bit depth, and Clip input is -2 Bitdepth+1 ~ (2 Bitdepth+1 -1) (output of the clipping operation).
4. The method of claim 1 , further comprising the step of applying an inverse transform to the residual of the coding unit before performing the clipping operation.
5. The method of claim 1 , further comprising the step of applying a second clipping operation to the residual of the coding unit after applying the inverse ACT to the residual of the coding unit.
6. 2. The method of claim 1, wherein the clipping operation adjusts the dynamic range of the residual of the coding unit to within a fixed intra-coding bit depth implemented by the inverse ACT.
7. The method of claim 6 , wherein the fixed intra-coding bit depth is 15.
8. 2. The method of claim 1, further comprising the step of receiving a second syntax element from the video data after receiving a first syntax element from the video data, the second syntax element indicating a variable intra-coding bit depth used in the clipping operation.
9. 2. The method of claim 1, further comprising the step of receiving a second syntax element from the video data after receiving a first syntax element from the video data, the second syntax element indicating a first intra-coding bit depth, and a second intra-coding bit depth used in the clipping operation is the first intra-coding bit depth plus one.
10. 2. The method of claim 1 , wherein the clipping operation further comprises a first clipping operation applied to a luminance component of the residual of the coding unit and a second clipping operation applied to a chrominance component of the residual of the coding unit.
11. 11. The method of claim 10, the first clipping operation limits a dynamic range of the luminance component of the residual of the coding unit to within a range of a first intra-coding bit depth plus one; the second clipping operation limits a dynamic range of the chrominance components of the residual of the coding unit to within a range of a second intra-coding bit depth plus one; The method of claim 1, wherein the second intra coding bit depth is the first intra coding bit depth plus one.
12. one or more processing units; a memory coupled to the one or more processing units; a plurality of programs stored in the memory; The plurality of programs, when executed by the one or more processors, cause the electronic device to perform the method of any one of claims 1 to 11.
13. 12. A non-transitory computer readable storage medium storing a plurality of programs for execution by an electronic device having one or more processing units, the plurality of programs, when executed by the one or more processing units, causing the electronic device to perform a method according to any one of claims 1 to 11.
Citation Information
Patent Citations
Method and device for encoding a high dynamic range image and / or decoding a bitstream
JP2017508332A
Cross-component prediction for video coding and clipping for adaptive color conversion
JP2018507622A
Method and apparatus for video coding using adaptive clipping
JP2019521571A
