Patent Application
20160241876
Engineers use compression (also called source coding or source encoding) to reduce the bit rate of digital video. Compression decreases the cost of storing and transmitting video information by converting the information into a lower bit rate form. Decompression (also called decoding) reconstructs a version of the original information from the compressed form. A “codec” is an encoder/decoder system.
Over the last two decades, various video codec standards have been adopted, including the ITU-T H.261, H.262 (MPEG-2 or ISO/IEC 13818-2), H.263 and H.264 (MPEG-4 AVC or ISO/IEC 14496-10) standards, the MPEG-1 (ISO/IEC 11172-2) and MPEG-4 Visual (ISO/IEC 14496-2) standards, and the SMPTE 421M standard. More recently, the HEVC standard (ITU-T H.265 or ISO/IEC 23008-2) has been approved. Extensions to the HEVC standard (e.g., for scalable video coding/decoding, for coding/decoding of video with higher fidelity in terms of sample bit depth or chroma sampling rate, or for multi-view coding/decoding) are currently under development. A video codec standard typically defines options for the syntax of an encoded video bitstream, detailing parameters in the bitstream when particular features are used in encoding and decoding. In many cases, a video codec standard also provides details about the decoding operations a decoder should perform to achieve conforming results in decoding. Aside from codec standards, various proprietary codec formats define other options for the syntax of an encoded video bitstream and corresponding decoding operations.
Intra block copy (“BC”) is a prediction mode under consideration for HEVC extensions. For intra BC prediction mode, the sample values of a current block of a picture are predicted using previously reconstructed sample values in the same picture. A block vector (“BV”) indicates a displacement from the current block to a region of the picture that includes the previously reconstructed sample values used for prediction. The BV is signaled in the bitstream. Intra BC prediction is a form of intra-picture prediction—intra BC prediction for a block of a picture does not use any sample values other than sample values in the same picture.
As currently specified in the HEVC standard and implemented in some reference software for the HEVC standard, intra BC prediction mode has several problems. For example, encoder-side decisions about how to use intra BC prediction are not made effectively.
In summary, the detailed description presents innovations in hash-based block matching. For example, some of the innovations relate to encoding of blocks with hash values, in which an encoder uses hash values for blocks when searching for a block that matches a current block being encoded. A corresponding decoder uses signaled hash values to determine which blocks to use during reconstruction.
The innovations for hash-based block matching can be implemented as part of a method, as part of a computing device adapted to perform the method or as part of a tangible computer-readable media storing computer-executable instructions for causing a computing device to perform the method. The various innovations can be used in combination or separately.
The foregoing and other objects, features, and advantages of the invention will become more apparent from the following detailed description, which proceeds with reference to the accompanying figures.
The detailed description presents innovations in the use of hash-based block matching. In particular, the detailed description presents innovations for encoding blocks with hash values and reconstructing blocks using hash values.
Although operations described herein are in places described as being performed by a video encoder, in many cases the operations can be performed by another type of media processing tool (e.g., image encoder).
Some of the innovations described herein are illustrated with reference to syntax elements and operations specific to the HEVC standard. For example, reference is made to the draft version JCTVC-N1005 of the HEVC standard—“High Efficiency Video Coding (HEVC) Range Extensions Text Specification: Draft 4,” JCTVC-N1005, July 2013. The innovations described herein can also be implemented for other standards or formats.
More generally, various alternatives to the examples described herein are possible. For example, some of the methods described herein can be altered by changing the ordering of the method acts described, by splitting, repeating, or omitting certain method acts, etc. The various aspects of the disclosed technology can be used in combination or separately. Different embodiments use one or more of the described innovations. Some of the innovations described herein address one or more of the problems noted in the background. Typically, a given technique/tool does not solve all such problems.
With reference to
A computing system may have additional features. For example, the computing system (100) includes storage (140), one or more input devices (150), one or more output devices (160), and one or more communication connections (170). An interconnection mechanism (not shown) such as a bus, controller, or network interconnects the components of the computing system (100). Typically, operating system software (not shown) provides an operating environment for other software executing in the computing system (100), and coordinates activities of the components of the computing system (100).
The tangible storage (140) may be removable or non-removable, and includes magnetic disks, magnetic tapes or cassettes, CD-ROMs, DVDs, or any other medium which can be used to store information and which can be accessed within the computing system (100). The storage (140) stores instructions for the software (180) implementing one or more innovations for hash-based block matching and/or hash-based block reconstruction.
The input device(s) (150) may be a touch input device such as a keyboard, mouse, pen, or trackball, a voice input device, a scanning device, or another device that provides input to the computing system (100). For video, the input device(s) (150) may be a camera, video card, TV tuner card, or similar device that accepts video input in analog or digital form, or a CD-ROM or CD-RW that reads video samples into the computing system (100). The output device(s) (160) may be a display, printer, speaker, CD-writer, or another device that provides output from the computing system (100).
The communication connection(s) (170) enable communication over a communication medium to another computing entity. The communication medium conveys information such as computer-executable instructions, audio or video input or output, or other data in a modulated data signal. A modulated data signal is a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media can use an electrical, optical, RF, or other carrier.
The innovations can be described in the general context of computer-readable media. Computer-readable media are any available tangible media that can be accessed within a computing environment. By way of example, and not limitation, with the computing system (100), computer-readable media include memory (120, 125), storage (140), and combinations of any of the above.
The innovations can be described in the general context of computer-executable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The functionality of the program modules may be combined or split between program modules as desired in various embodiments. Computer-executable instructions for program modules may be executed within a local or distributed computing system.
The terms “system” and “device” are used interchangeably herein. Unless the context clearly indicates otherwise, neither term implies any limitation on a type of computing system or computing device. In general, a computing system or computing device can be local or distributed, and can include any combination of special-purpose hardware and/or general-purpose hardware with software implementing the functionality described herein.
The disclosed methods can also be implemented using specialized computing hardware configured to perform any of the disclosed methods. For example, the disclosed methods can be implemented by an integrated circuit (e.g., an ASIC (such as an ASIC digital signal process unit (“DSP”), a graphics processing unit (“GPU”), or a programmable logic device (“PLD”), such as a field programmable gate array (“FPGA”)) specially designed or configured to implement any of the disclosed methods.
For the sake of presentation, the detailed description uses terms like “determine” and “use” to describe computer operations in a computing system. These terms are high-level abstractions for operations performed by a computer, and should not be confused with acts performed by a human being. The actual computer operations corresponding to these terms vary depending on implementation.
In the network environment (201) shown in
A real-time communication tool (210) manages encoding by an encoder (220).
In the network environment (202) shown in
The video source (310) can be a camera, tuner card, storage media, or other digital video source. The video source (310) produces a sequence of video frames at a frame rate of, for example, 30 frames per second. As used herein, the term “frame” generally refers to source, coded or reconstructed image data. For progressive video, a frame is a progressive video frame. For interlaced video, in example embodiments, an interlaced video frame is de-interlaced prior to encoding. Alternatively, two complementary interlaced video fields are encoded as an interlaced video frame or separate fields. Aside from indicating a progressive video frame, the term “frame” or “picture” can indicate a single non-paired video field, a complementary pair of video fields, a video object plane that represents a video object at a given time, or a region of interest in a larger image. The video object plane or region can be part of a larger image that includes multiple objects or regions of a scene.
An arriving source frame (311) is stored in a source frame temporary memory storage area (320) that includes multiple frame buffer storage areas (321, 322, . . . , 32n). A frame buffer (321, 322, etc.) holds one source frame in the source frame storage area (320). After one or more of the source frames (311) have been stored in frame buffers (321, 322, etc.), a frame selector (330) periodically selects an individual source frame from the source frame storage area (320). The order in which frames are selected by the frame selector (330) for input to the encoder (340) may differ from the order in which the frames are produced by the video source (310), e.g., a frame may be ahead in order, to facilitate temporally backward prediction. Before the encoder (340), the encoder system (300) can include a pre-processor (not shown) that performs pre-processing (e.g., filtering) of the selected frame (331) before encoding. The pre-processing can also include color space conversion into primary and secondary components for encoding. Typically, before encoding, video has been converted to a color space such as YUV, in which sample values of a luma (Y) component represent brightness or intensity values, and sample values of chroma (U, V) components represent color-difference values. The chroma sample values may be sub-sampled to a lower chroma sampling rate (e.g., for YUV 4:2:0 format), or the chroma sample values may have the same resolution as the luma sample values (e.g., for YUV 4:4:4 format). Or, the video can be encoded in another format (e.g., RGB 4:4:4 format).
The encoder (340) encodes the selected frame (331) to produce a coded frame (341) and also produces memory management control operation (“MMCO”) signals (342) or reference picture set (“RPS”) information. If the current frame is not the first frame that has been encoded, when performing its encoding process, the encoder (340) may use one or more previously encoded/decoded frames (369) that have been stored in a decoded frame temporary memory storage area (360). Such stored decoded frames (369) are used as reference frames for inter-frame prediction of the content of the current source frame (331). Generally, the encoder (340) includes multiple encoding modules that perform encoding tasks such as partitioning into tiles, intra prediction estimation and prediction, motion estimation and compensation, frequency transforms, quantization and entropy coding. The exact operations performed by the encoder (340) can vary depending on compression format. The format of the output encoded data can be a variation or extension of HEVC format, Windows Media Video format, VC-1 format, MPEG-x format (e.g., MPEG-1, MPEG-2, or MPEG-4), H.26x format (e.g., H.261, H.262, H.263, H.264), or another format.
The encoder (340) can partition a frame into multiple tiles of the same size or different sizes. For example, the encoder (340) splits the frame along tile rows and tile columns that, with frame boundaries, define horizontal and vertical boundaries of tiles within the frame, where each tile is a rectangular region. Tiles are often used to improve options for parallel processing. A frame can also be organized as one or more slices, where a slice can be an entire frame or region of the frame. A slice can be decoded independently of other slices in a frame, which improves error resilience. The content of a slice or tile is further partitioned into blocks or other sets of samples for purposes of encoding and decoding.
For syntax according to the HEVC standard, the encoder splits the content of a frame (or slice or tile) into coding tree units. A coding tree unit (“CTU”) includes luma sample values organized as a luma coding tree block (“CTB”) and corresponding chroma sample values organized as two chroma CTBs. The size of a CTU (and its CTBs) is selected by the encoder, and can be, for example, 64×64, 32×32 or 16×16 sample values. A CTU includes one or more coding units. A coding unit (“CU”) has a luma coding block (“CB”) and two corresponding chroma CBs. For example, a CTU with a 64×64 luma CTB and two 64×64 chroma CTBs (YUV 4:4:4 format) can be split into four CUs, with each CU including a 32×32 luma CB and two 32×32 chroma CBs, and with each CU possibly being split further into smaller CUs. Or, as another example, a CTU with a 64×64 luma CTB and two 32×32 chroma CTBs (YUV 4:2:0 format) can be split into four CUs, with each CU including a 32×32 luma CB and two 16×16 chroma CBs, and with each CU possibly being split further into smaller CUs. The smallest allowable size of CU (e.g., 8×8, 16×16) can be signaled in the bitstream.
Generally, a CU has a prediction mode such as inter or intra. A CU includes one or more prediction units for purposes of signaling of prediction information (such as prediction mode details, displacement values, etc.) and/or prediction processing. A prediction unit (“PU”) has a luma prediction block (“PB”) and two chroma PBs. For an intra-predicted CU, the PU has the same size as the CU, unless the CU has the smallest size (e.g., 8×8). In that case, the CU can be split into four smaller PUs (e.g., each 4×4 if the smallest CU size is 8×8) or the PU can have the smallest CU size, as indicated by a syntax element for the CU. A CU also has one or more transform units for purposes of residual coding/decoding, where a transform unit (“TU”) has a transform block (“TB”) and two chroma TBs. A PU in an intra-predicted CU may contain a single TU (equal in size to the PU) or multiple TUs. As used herein, the term “block” can indicate a CB, PB, TB or other set of sample values, depending on context. The encoder decides how to partition video into CTUs, CUs, PUs, TUs, etc.
Returning to
The encoder (340) represents an inter-coded, predicted block of a source frame (331) in terms of prediction from reference frames. A motion estimator estimates motion of the block with respect to one or more reference frames (369). When multiple reference frames are used, the multiple reference frames can be from different temporal directions or the same temporal direction. A motion-compensated prediction reference region is a region of samples in the reference frame(s) that are used to generate motion-compensated prediction values for a block of samples of a current frame. The motion estimator outputs motion information such as motion vector information, which is entropy coded. A motion compensator applies motion vectors to reference frames (369) to determine motion-compensated prediction values.
The encoder determines the differences (if any) between a block's prediction values (intra or inter) and corresponding original values. These prediction residual values are further encoded using a frequency transform, quantization and entropy encoding. For example, the encoder (340) sets values for quantization parameter (“QP”) for a picture, tile, slice and/or other portion of video, and quantizes transform coefficients accordingly. The entropy coder of the encoder (340) compresses quantized transform coefficient values as well as certain side information (e.g., motion vector information, BV values, hash values, QP values, mode decisions, parameter choices). Typical entropy coding techniques include Exp-Golomb coding, arithmetic coding, differential coding, Huffman coding, run length coding, variable-length-to-variable-length (“V2V”) coding, variable-length-to-fixed-length (“V2F”) coding, LZ coding, dictionary coding, probability interval partitioning entropy coding (“PIPE”), and combinations of the above. The entropy coder can use different coding techniques for different kinds of information, and can choose from among multiple code tables within a particular coding technique.
The coded frames (341) and MMCO/RPS information (342) are processed by a decoding process emulator (350). The decoding process emulator (350) implements some of the functionality of a decoder, for example, decoding tasks to reconstruct reference frames. The decoding process emulator (350) uses the MMCO/RPS information (342) to determine whether a given coded frame (341) needs to be reconstructed and stored for use as a reference frame in inter-frame prediction of subsequent frames to be encoded. If the MMCO/RPS information (342) indicates that a coded frame (341) needs to be stored, the decoding process emulator (350) models the decoding process that would be conducted by a decoder that receives the coded frame (341) and produces a corresponding decoded frame (351). In doing so, when the encoder (340) has used decoded frame(s) (369) that have been stored in the decoded frame storage area (360), the decoding process emulator (350) also uses the decoded frame(s) (369) from the storage area (360) as part of the decoding process.
The decoded frame temporary memory storage area (360) includes multiple frame buffer storage areas (361, 362, . . . , 36n). The decoding process emulator (350) uses the MMCO/RPS information (342) to manage the contents of the storage area (360) in order to identify any frame buffers (361, 362, etc.) with frames that are no longer needed by the encoder (340) for use as reference frames. After modeling the decoding process, the decoding process emulator (350) stores a newly decoded frame (351) in a frame buffer (361, 362, etc.) that has been identified in this manner.
The coded frames (341) and MMCO/RPS information (342) are buffered in a temporary coded data area (370). The coded data that is aggregated in the coded data area (370) contains, as part of the syntax of an elementary coded video bitstream, encoded data for one or more pictures. The coded data that is aggregated in the coded data area (370) can also include media metadata relating to the coded video data (e.g., as one or more parameters in one or more supplemental enhancement information (“SEI”) messages or video usability information (“VUI”) messages).
The aggregated data (371) from the temporary coded data area (370) are processed by a channel encoder (380). The channel encoder (380) can packetize the aggregated data for transmission as a media stream (e.g., according to a media stream multiplexing format such as ISO/IEC 13818-1), in which case the channel encoder (380) can add syntax elements as part of the syntax of the media transmission stream. Or, the channel encoder (380) can organize the aggregated data for storage as a file (e.g., according to a media container format such as ISO/IEC 14496-12), in which case the channel encoder (380) can add syntax elements as part of the syntax of the media storage file. Or, more generally, the channel encoder (380) can implement one or more media system multiplexing protocols or transport protocols, in which case the channel encoder (380) can add syntax elements as part of the syntax of the protocol(s). The channel encoder (380) provides output to a channel (390), which represents storage, a communications connection, or another channel for the output.
The decoder system (400) includes a channel (410), which can represent storage, a communications connection, or another channel for coded data as input. The channel (410) produces coded data that has been channel coded. A channel decoder (420) can process the coded data. For example, the channel decoder (420) de-packetizes data that has been aggregated for transmission as a media stream (e.g., according to a media stream multiplexing format such as ISO/IEC 13818-1), in which case the channel decoder (420) can parse syntax elements added as part of the syntax of the media transmission stream. Or, the channel decoder (420) separates coded video data that has been aggregated for storage as a file (e.g., according to a media container format such as ISO/IEC 14496-12), in which case the channel decoder (420) can parse syntax elements added as part of the syntax of the media storage file. Or, more generally, the channel decoder (420) can implement one or more media system demultiplexing protocols or transport protocols, in which case the channel decoder (420) can parse syntax elements added as part of the syntax of the protocol(s).
The coded data (421) that is output from the channel decoder (420) is stored in a temporary coded data area (430) until a sufficient quantity of such data has been received. The coded data (421) includes coded frames (431) and MMCO/RPS information (432). The coded data (421) in the coded data area (430) contain, as part of the syntax of an elementary coded video bitstream, coded data for one or more pictures. The coded data (421) in the coded data area (430) can also include media metadata relating to the encoded video data (e.g., as one or more parameters in one or more SEI messages or VUI messages).
In general, the coded data area (430) temporarily stores coded data (421) until such coded data (421) is used by the decoder (450). At that point, coded data for a coded frame (431) and MMCO/RPS information (432) are transferred from the coded data area (430) to the decoder (450). As decoding continues, new coded data is added to the coded data area (430) and the oldest coded data remaining in the coded data area (430) is transferred to the decoder (450).
The decoder (450) periodically decodes a coded frame (431) to produce a corresponding decoded frame (451). As appropriate, when performing its decoding process, the decoder (450) may use one or more previously decoded frames (469) as reference frames for inter-frame prediction. The decoder (450) reads such previously decoded frames (469) from a decoded frame temporary memory storage area (460). Generally, the decoder (450) includes multiple decoding modules that perform decoding tasks such as entropy decoding, inverse quantization, inverse frequency transforms, intra prediction, motion compensation and merging of tiles. The exact operations performed by the decoder (450) can vary depending on compression format.
For example, the decoder (450) receives encoded data for a compressed frame or sequence of frames and produces output including decoded frame (451). In the decoder (450), a buffer receives encoded data for a compressed frame and, at an appropriate time, makes the received encoded data available to an entropy decoder. The entropy decoder entropy decodes entropy-coded quantized data as well as entropy-coded side information, typically applying the inverse of entropy encoding performed in the encoder. A motion compensator applies motion information to one or more reference frames to form motion-compensated prediction values for any inter-coded blocks of the frame being reconstructed. An intra prediction module can spatially predict sample values of a current block from neighboring, previously reconstructed sample values or, for intra BC prediction, predict sample values of a current block using previously reconstructed sample values of an intra-prediction region in the frame. The intra-prediction region can be indicated with a BV value or, alternatively, a hash value that indicates a reconstruction block in a block hash dictionary. The decoder (450) also reconstructs prediction residuals. An inverse quantizer inverse quantizes entropy-decoded data. For example, the decoder (450) sets values for QP for a picture, tile, slice and/or other portion of video based on syntax elements in the bitstream, and inverse quantizes transform coefficients accordingly. An inverse frequency transformer converts the quantized, frequency domain data into spatial domain information. For an inter-predicted block, the decoder (450) combines reconstructed prediction residuals with motion-compensated predictions. The decoder (450) can similarly combine prediction residuals with predictions from intra prediction. A motion compensation loop in the video decoder (450) includes an adaptive de-blocking filter to smooth discontinuities across block boundary rows and/or columns in the decoded frame (451).
The decoded frame temporary memory storage area (460) includes multiple frame buffer storage areas (461, 462, . . . , 46n). The decoded frame storage area (460) is an example of a decoded picture buffer. The decoder (450) uses the MMCO/RPS information (432) to identify a frame buffer (461, 462, etc.) in which it can store a decoded frame (451). The decoder (450) stores the decoded frame (451) in that frame buffer.
An output sequencer (480) uses the MMCO/RPS information (432) to identify when the next frame to be produced in output order is available in the decoded frame storage area (460). When the next frame (481) to be produced in output order is available in the decoded frame storage area (460), it is read by the output sequencer (480) and output to the output destination (490) (e.g., display). In general, the order in which frames are output from the decoded frame storage area (460) by the output sequencer (480) may differ from the order in which the frames are decoded by the decoder (450).
The encoder (500) is block-based and uses a block format that depends on implementation. Blocks may be further sub-divided at different stages, e.g., at the prediction, frequency transform and/or entropy encoding stages. For example, a picture can be divided into 64×64 blocks, 32×32 blocks or 16×16 blocks, which can in turn be divided into smaller blocks of sample values for coding and decoding. In implementations of encoding for the HEVC standard, the encoder partitions a picture into CTUs (CTBs), CUs (CBs), PUs (PBs) and TU (TBs).
The encoder (500) compresses pictures using intra-picture coding and/or inter-picture coding. Many of the components of the encoder (500) are used for both intra-picture coding and inter-picture coding. The exact operations performed by those components can vary depending on the type of information being compressed.
A tiling module (510) optionally partitions a picture into multiple tiles of the same size or different sizes. For example, the tiling module (510) splits the picture along tile rows and tile columns that, with picture boundaries, define horizontal and vertical boundaries of tiles within the picture, where each tile is a rectangular region.
The general encoding control (520) receives pictures for the input video signal (505) as well as feedback (not shown) from various modules of the encoder (500). Overall, the general encoding control (520) provides control signals (not shown) to other modules (such as the tiling module (510), transformer/scaler/quantizer (530), scaler/inverse transformer (535), intra-picture estimator (540), motion estimator (550) and intra/inter switch) to set and change coding parameters during encoding. In particular, the general encoding control (520) can decide whether and how to use hash-based block matching during encoding. The general encoding control (520) can also evaluate intermediate results during encoding, for example, performing rate-distortion analysis. The general encoding control (520) produces general control data (522) that indicates decisions made during encoding, so that a corresponding decoder can make consistent decisions. The general control data (522) is provided to the header formatter/entropy coder (590).
If the current picture is predicted using inter-picture prediction, a motion estimator (550) estimates motion of blocks of sample values of the current picture of the input video signal (505) with respect to one or more reference pictures. The decoded picture buffer (570) buffers one or more reconstructed previously coded pictures for use as reference pictures. When multiple reference pictures are used, the multiple reference pictures can be from different temporal directions or the same temporal direction. The motion estimator (550) produces as side information motion data (552) such as motion vector data and reference picture selection data. The motion data (552) is provided to the header formatter/entropy coder (590) as well as the motion compensator (555).
The motion compensator (555) applies motion vectors to the reconstructed reference picture(s) from the decoded picture buffer (570). The motion compensator (555) produces motion-compensated predictions for the current picture.
In a separate path within the encoder (500), an intra-picture estimator (540) determines how to perform intra-picture prediction for blocks of sample values of a current picture of the input video signal (505). The current picture can be entirely or partially coded using intra-picture coding. Using values of a reconstruction (538) of the current picture, for intra spatial prediction, the intra-picture estimator (540) determines how to spatially predict sample values of a current block of the current picture from neighboring, previously reconstructed sample values of the current picture. Or, for intra BC prediction using BV values, the intra-picture estimator (540) estimates displacement of the sample values of the current block to different candidate regions within the current picture. For hash-based block matching during the BV estimation, the intra-picture estimator (540) can use a block hash dictionary (541) to find a BV value for a current block. The block hash dictionary (541) is a data structure that organizes candidate blocks for hash-based block matching, as described below. The block hash dictionary (541) can be updated during encoding to store information about new candidate blocks, as those candidate blocks become available for use in hash-based block matching. Alternatively, instead of using BV values, intra BC prediction can use hash values that represent blocks. In this case, for hash-based block matching, the intra-picture estimator (540) can use a block hash dictionary (541) to find a hash value to use to represent a current block, as described below.
The intra-picture estimator (540) produces as side information intra prediction data (542), such as information indicating whether intra prediction uses spatial prediction or BC prediction (e.g., a flag value per intra block), prediction mode direction (for intra spatial prediction) and BV values or hash values (for intra BC prediction). The intra prediction data (542) is provided to the header formatter/entropy coder (590) as well as the intra-picture predictor (545).
According to the intra prediction data (542), the intra-picture predictor (545) spatially predicts sample values of a current block of the current picture from neighboring, previously reconstructed sample values of the current picture. Or, for intra BC prediction, the intra-picture predictor (545) predicts the sample values of the current block using previously reconstructed sample values of an intra-prediction region, which is indicated by a BV value for the current block (or is indicated by the hash value in a block hash dictionary (541)). When the chroma data for a picture has the same resolution as the luma data (e.g. when the format is YUV 4:4:4 format or RGB 4:4:4 format), the BV value that is applied for the chroma block may be the same as the BV value applied for the luma block. On the other hand, when the chroma data for a picture has reduced resolution relative to the luma data (e.g. when the format is YUV 4:2:0 format), the BV value that is applied for the chroma block may be scaled down and possibly rounded to adjust for the difference in chroma resolution (e.g. by dividing the vertical and horizontal components of the BV value by two and truncating or rounding them to integer values). When a hash value represents a reconstruction block (from block hash dictionary (541)) for luma sample values of a frame, reconstruction blocks for chroma sample values at corresponding locations in the frame can be used.
The intra/inter switch selects values of a motion-compensated prediction or intra-picture prediction for use as the prediction (558) for a given block. The difference (if any) between a block of the prediction (558) and corresponding part of the original current picture of the input video signal (505) provides values of the residual (518). During reconstruction of the current picture, reconstructed residual values are combined with the prediction (558) to produce a reconstruction (538) of the original content from the video signal (505). In lossy compression, however, some information is still lost from the video signal (505).
In the transformer/scaler/quantizer (530), a frequency transformer converts spatial domain video information into frequency domain (i.e., spectral, transform) data. For block-based video coding, the frequency transformer applies a discrete cosine transform (“DCT”), an integer approximation thereof, or another type of forward block transform to blocks of prediction residual data (or sample value data if the prediction (558) is null), producing blocks of frequency transform coefficients. The encoder (500) may also be able to indicate that such transform step is skipped. The scaler/quantizer scales and quantizes the transform coefficients. For example, the quantizer applies non-uniform, scalar quantization to the frequency domain data with a step size that varies on a frame-by-frame basis, tile-by-tile basis, slice-by-slice basis, block-by-block basis or other basis. The quantized transform coefficient data (532) is provided to the header formatter/entropy coder (590).
In the scaler/inverse transformer (535), a scaler/inverse quantizer performs inverse scaling and inverse quantization on the quantized transform coefficients. An inverse frequency transformer performs an inverse frequency transform, producing blocks of reconstructed prediction residuals or sample values. The encoder (500) combines reconstructed residuals with values of the prediction (558) (e.g., motion-compensated prediction values, intra-picture prediction values) to form the reconstruction (538).
For intra-picture prediction, the values of the reconstruction (538) can be fed back to the intra-picture estimator (540) and intra-picture predictor (545). Values of the reconstruction (538) can also be used to update the block hash dictionary (541). Also, the values of the reconstruction (538) can be used for motion-compensated prediction of subsequent pictures. The values of the reconstruction (538) can be further filtered. A filtering control (560) determines how to perform deblock filtering and sample adaptive offset (“SAO”) filtering on values of the reconstruction (538), for a given picture of the video signal (505). The filtering control (560) produces filter control data (562), which is provided to the header formatter/entropy coder (590) and merger/filter(s) (565).
In the merger/filter(s) (565), the encoder (500) merges content from different tiles into a reconstructed version of the picture. The encoder (500) selectively performs deblock filtering and SAO filtering according to the filter control data (562), so as to adaptively smooth discontinuities across boundaries in the frames. Tile boundaries can be selectively filtered or not filtered at all, depending on settings of the encoder (500), and the encoder (500) may provide syntax within the coded bitstream to indicate whether or not such filtering was applied. The decoded picture buffer (570) buffers the reconstructed current picture for use in subsequent motion-compensated prediction.
The header formatter/entropy coder (590) formats and/or entropy codes the general control data (522), quantized transform coefficient data (532), intra prediction data (542), motion data (552) and filter control data (562). For example, the header formatter/entropy coder (590) uses context-adaptive binary arithmetic coding for entropy coding of various syntax elements.
The header formatter/entropy coder (590) provides the encoded data in the coded video bitstream (595). The format of the coded video bitstream (595) can be a variation or extension of HEVC format, Windows Media Video format, VC-1 format, MPEG-x format (e.g., MPEG-1, MPEG-2, or MPEG-4), H.26x format (e.g., H.261, H.262, H.263, H.264), or another format.
Depending on implementation and the type of compression desired, modules of the encoder can be added, omitted, split into multiple modules, combined with other modules, and/or replaced with like modules. In alternative embodiments, encoders with different modules and/or other configurations of modules perform one or more of the described techniques. Specific embodiments of encoders typically use a variation or supplemented version of the encoder (500). The relationships shown between modules within the encoder (500) indicate general flows of information in the encoder; other relationships are not shown for the sake of simplicity.
The decoder (600) is block-based and uses a block format that depends on implementation. Blocks may be further sub-divided at different stages. For example, a picture can be divided into 64×64 blocks, 32×32 blocks or 16×16 blocks, which can in turn be divided into smaller blocks of sample values. In implementations of decoding for the HEVC standard, a picture is partitioned into CTUs (CTBs), CUs (CBs), PUs (PBs) and TU (TBs).
The decoder (600) decompresses pictures using intra-picture decoding and/or inter-picture decoding. Many of the components of the decoder (600) are used for both intra-picture decoding and inter-picture decoding. The exact operations performed by those components can vary depending on the type of information being decompressed.
A buffer receives encoded data in the coded video bitstream (605) and makes the received encoded data available to the parser/entropy decoder (610). The parser/entropy decoder (610) entropy decodes entropy-coded data, typically applying the inverse of entropy coding performed in the encoder (500) (e.g., context-adaptive binary arithmetic decoding). As a result of parsing and entropy decoding, the parser/entropy decoder (610) produces general control data (622), quantized transform coefficient data (632), intra prediction data (642), motion data (652) and filter control data (662).
The general decoding control (620) receives the general control data (622) and provides control signals (not shown) to other modules (such as the scaler/inverse transformer (635), intra-picture predictor (645), motion compensator (655) and intra/inter switch) to set and change decoding parameters during decoding.
If the current picture is predicted using inter-picture prediction, a motion compensator (655) receives the motion data (652), such as motion vector data and reference picture selection data. The motion compensator (655) applies motion vectors to the reconstructed reference picture(s) from the decoded picture buffer (670). The motion compensator (655) produces motion-compensated predictions for inter-coded blocks of the current picture. The decoded picture buffer (670) stores one or more previously reconstructed pictures for use as reference pictures.
In a separate path within the decoder (600), the intra-prediction predictor (645) receives the intra prediction data (642), such as information indicating whether intra prediction uses spatial prediction or BC prediction (e.g., a flag value per intra block), prediction mode direction (for intra spatial prediction) and BV values or hash values (for intra BC prediction). For intra spatial prediction, using values of a reconstruction (638) of the current picture, according to prediction mode data, the intra-picture predictor (645) spatially predicts sample values of a current block of the current picture from neighboring, previously reconstructed sample values of the current picture. Or, for intra BC prediction using BV values, the intra-picture predictor (645) predicts the sample values of the current block using previously reconstructed sample values of an intra-prediction region, which is indicated by a BV value for the current block. Alternatively, for intra BC prediction using hash values, the intra-picture predictor (645) determines a reconstruction block for a current block using a signaled hash value for the current block. In this case, the intra-picture predictor (645) uses a block hash dictionary (641) to find the reconstruction block for a signaled hash value. The block hash dictionary (641) is a data structure that organizes candidate block values according to hash value.
The intra/inter switch selects values of a motion-compensated prediction or intra-picture prediction for use as the prediction (658) for a given block. For example, when HEVC syntax is followed, the intra/inter switch can be controlled based on a syntax element encoded for a CU of a picture that can contain intra-predicted CUs and inter-predicted CUs. The decoder (600) combines the prediction (658) with reconstructed residual values to produce the reconstruction (638) of the content from the video signal.
To reconstruct the residual, the scaler/inverse transformer (635) receives and processes the quantized transform coefficient data (632). In the scaler/inverse transformer (635), a scaler/inverse quantizer performs inverse scaling and inverse quantization on the quantized transform coefficients. An inverse frequency transformer performs an inverse frequency transform, producing blocks of reconstructed prediction residuals or sample values. For example, the inverse frequency transformer applies an inverse block transform to frequency transform coefficients, producing sample value data or prediction residual data. The inverse frequency transform can be an inverse DCT, an integer approximation thereof, or another type of inverse frequency transform.
For intra-picture prediction, the values of the reconstruction (638) can be fed back to the intra-picture predictor (645). Values of the reconstruction (638) can be used to update the block hash dictionary (641). For inter-picture prediction, the values of the reconstruction (638) can be further filtered. In the merger/filter(s) (665), the decoder (600) merges content from different tiles into a reconstructed version of the picture. The decoder (600) selectively performs deblock filtering and SAO filtering according to the filter control data (662) and rules for filter adaptation, so as to adaptively smooth discontinuities across boundaries in the frames. Tile boundaries can be selectively filtered or not filtered at all, depending on settings of the decoder (600) or a syntax indication within the encoded bitstream data. The decoded picture buffer (670) buffers the reconstructed current picture for use in subsequent motion-compensated prediction.
The decoder (600) can also include a post-processing deblock filter. The post-processing deblock filter optionally smoothes discontinuities in reconstructed pictures. Other filtering (such as de-ring filtering) can also be applied as part of the post-processing filtering.
Depending on implementation and the type of decompression desired, modules of the decoder can be added, omitted, split into multiple modules, combined with other modules, and/or replaced with like modules. In alternative embodiments, decoders with different modules and/or other configurations of modules perform one or more of the described techniques. Specific embodiments of decoders typically use a variation or supplemented version of the decoder (600). The relationships shown between modules within the decoder (600) indicate general flows of information in the decoder; other relationships are not shown for the sake of simplicity.
This section presents various features of hash-based block matching. Some of the features relate to encoding blocks with hash values, while other features relate to reconstructing blocks using hash values. These features can facilitate intra BC prediction that is effective in terms of rate-distortion performance and/or computational efficiency.
In particular, hash-based block matching can improve rate-distortion performance when encoding certain “artificially” created video content such as screen-capture content. Screen-capture content typically includes repeated structures (e.g., graphics, text characters), which provide opportunities for hash-based block matching to improve performance. Screen capture content is usually encoded in a format (e.g., YUV 4:4:4 or RGB 4:4:4) with high chroma sampling resolution, although it may also be encoded in a format with lower chroma sampling resolution (e.g., YUV 4:2:0). Common scenarios for encoding/decoding of screen-capture content include remote desktop conferencing and encoding/decoding of graphical overlays on natural video or other “mixed content” video.
A. Intra BC Prediction Mode and BV Values—Introduction.
For intra BC prediction, the sample values of a current block of a picture are predicted using sample values in the same picture. A BV value indicates a displacement from the current block to a region of the picture that includes the sample values used for prediction. The sample values used for prediction are previously reconstructed sample values. The BV value is signaled in the bitstream, and a decoder can use the BV value to determine the region of the picture to use for prediction, which is also reconstructed at the decoder. Intra BC prediction is a form of intra-picture prediction—intra BC prediction for a block of a picture does not use any sample values other than sample values in the same picture.
The BV (740) indicates a displacement (or offset) from the current block (730) to a region (750) of the picture that includes the sample values used for prediction. The intra-prediction region (750) indicated by BV (740) is sometimes termed the “matching block” for the current block (730). The matching block can be identical to the current block (730), or it can be an approximation of the current block (730). Suppose the top left position of a current block is at position (x0, y0) in the current frame, and suppose the top left position of the intra-prediction region is at position (x1, y1) in the current frame. The BV indicates the displacement (x1-x0, y1-y0). For example, if the top left position of the current block is at position (320, 256), and the top left position of the intra-prediction region is at position (295, 270), the BV value is (−25, 14). In this example, a negative horizontal displacement indicates a position to the left of the current block, and a negative vertical displacement indicates a position above the current block.
Intra BC prediction can improve coding efficiency by exploiting redundancy (such as repeated patterns inside a frame) using block copy operations. Finding a matching block for a current block can be computationally complex and time consuming, however, considering the number of candidate blocks that the encoder may evaluate. Comparing the sample values of a current block to the sample values of a single candidate block in sample-wise block matching can involve thousands of operations. For example, comparing the sample values of a 64×64 current block and 64×64 candidate block can involve 64×64=4096 comparison operations (e.g., subtraction operations) as well as operations to tally results of the comparison operations.
B. Hash-Based Block Matching.
When an encoder uses hash-based block matching, the encoder determines a hash value for each of multiple candidate blocks. The encoder also determines a hash value for a current block. If two blocks are identical, their hash values are the same. Using hash values, an encoder can quickly and efficiently identify candidate blocks that have the same hash value as the current block, and filter out candidate blocks that have different hash values. The encoder can then further evaluate those candidate blocks having the same hash value as the current block. Hash-based block matching speeds up the process of finding a matching block for a current block.
In this way, the possible candidate blocks are distributed into n categories. For example, if the hash function h( ) produces 12-bit hash values, the 953,474 candidate blocks in
The hash function h( ) can be a cyclic redundancy check (“CRC”) function or other hash function. Some types of hash function (e.g., CRC function) map similar blocks to different hash values, which may be efficient when seeking a matching block that exactly corresponds with a current block. Other types of hash function (e.g., locality-sensitive hash function) map similar blocks to the same hash value, which may be suitable when a matching block can approximate the current block. Example hash functions are described below.
With the hash function h( ), the encoder determines the hash value for the current block Bcurrent. In
Overall, since hash value comparisons are much simpler than sample-wise block matching, hash-based block matching can make the process of evaluating the candidate blocks in a large search range much more efficient. Also, hash values for candidate blocks can be reused in hash-based block matching for different blocks within a picture, so the cost of computing the hash values for the candidate blocks can be amortized across hash-based block matching operations for the entire picture.
C. Representing Blocks with Hash Values.
In some example implementations, instead of sending a BV value for the matching block for a current block, an encoder signals the hash value of the matching block that represents the current block. A decoder, using the signaled hash value, reconstructs the current block using its own data structure that organizes candidate blocks according to hash value. In this approach, the decoder creates and updates the data structure that organizes the candidate blocks. Using the signaled hash value for the current block, the decoder finds the appropriate block in the data structure, then uses that block for the current block. In some cases, signaling hash values consumes fewer bits than signaling BV values.
In this approach, the encoder and decoder determine the hash values for the candidate blocks using reconstructed sample values, so the hash values are the same at the encoder and decoder. The organization of the data structure depends on implementation. The encoder can signal multiple hash values per current block, to iteratively filter out non-matching blocks. Or, the encoder can signal a single hash value per current block. Or, in addition to one or more hash values, the encoder can signal an index value that indicates a matching block among multiple remaining candidate blocks.
This approach works well when the hash value(s) signaled for the current block identify, with a high probability, an identical block among the candidate blocks. Performance suffers when the hash value(s) signaled for the current block identify a non-identical matching block among the candidate blocks. In this case, however, the encoder can instead use another form of encoding for the current block (e.g., intra spatial prediction or other intra-picture prediction).
The encoder encodes (1110) data for a picture, using BC prediction with hash-based block matching. The BC prediction can be intra BC prediction (if the candidate blocks are in the same picture as a current block), or the BC prediction can reference candidate blocks in another picture (in the same video sequence, in another video sequence, or for a sprite or generated reference picture). The candidate blocks are part of a search range, which can include all previously reconstructed content in the picture with the current block (or another picture), or can include a subset of the previously reconstructed content (e.g., reconstructed content in the current CTU and one or more other CTUs in the picture with the current block, or reconstructed content in the same tile as the current block, or reconstructed content in the same slice as the current block). When it encodes (1110) the data for the picture, the encoder can perform the technique (1111) shown in
According to the technique (1111) of
The encoder identifies (1140) a matching block among multiple candidate blocks based at least in part on the hash value(s) for the current block. For example, for each of one or more of the candidate blocks, the encoder compares the hash value(s) for the current block to the hash value(s) for the candidate block. Or, when a data structure organizes the candidate blocks according to hash value, the encoder uses the hash value(s) for the current block to select a candidate block list, then determines the matching block among any candidate blocks in the selected list (e.g., performing block matching operations between sample values of the current block and a given candidate block).
If hash-based block matching succeeds, the encoder signals the hash value(s) for the matching block that represents the current block. The encoder can also signal the index value for the matching block within a candidate block list, where applicable. The encoder can encode the signaled values, e.g., using CABAC or another form of entropy coding. Otherwise, if hash-based block matching fails, the encoder can encode the current block using spatial intra prediction or another form of intra-picture prediction.
The encoder checks (1150) whether to continue with the next block of the picture. If so, when a data structure organizes candidate blocks according to hash value, the encoder can update (1160) the data structure to account for new candidate blocks that overlap the just-encoded block, then get (1120) the next block and determine (1130) the hash value(s) for it.
Returning to
The decoder receives (1210) encoded data for a picture, where the encoded data includes hash values that represent blocks of the picture. The decoder decodes (1220) data for a picture, using BC prediction with hash-based block reconstruction. The BC prediction can be intra BC prediction (if the candidate blocks are in the same picture as a current block), or the BC prediction can reference candidate blocks in another picture. When it decodes (1220) the data for the picture, the decoder can perform the technique (1221) shown in
According to the technique (1221) of
The decoder identifies (1240) a reconstruction block among multiple candidate blocks based at least in part on the hash value(s) signaled for the current block. For example, for each of one or more of the candidate blocks, the decoder compares the hash value(s) for the current block to the hash value(s) for the candidate block to find the reconstruction block. For the hash values of candidate blocks, the hash function can be a CRC function, locality-sensitive hash function, or other hash function. When a data structure organizes the candidate blocks according to hash value, the decoder can use the hash value(s) for the current block to find the reconstruction block in the data structure. Or, the decoder can use the hash value(s) for the current block to find a candidate block list in the data structure, then find the reconstruction block among any candidate blocks in the selected list (e.g., using a signaled index value).
When hash-based block reconstruction is used, the decoder uses (1250) the identified reconstruction block for the current block. Otherwise, if hash-based block reconstruction is not used, the encoder can decode the current block using spatial intra prediction or another form of intra-picture prediction.
The decoder checks (1260) whether to continue with the next block of the picture. If so, when a data structure organizes candidate blocks according to hash value, the decoder can update (1270) the data structure to account for new candidate blocks that overlap the just-decoded block, then get (1220) the hash value(s) for the next block and continue.
The decoder can repeat the technique (1200) for the next picture.
D. Example Hash Functions.
The hash function used for hash-based block matching (and, for section C, hash-based block reconstruction) depends on implementation.
A hash function can produce hash values with 8 bits, 12 bits, 16 bits, 24 bits, 32 bits, or some other number of bits. Generally, using hash values with fewer bits tends to reduce the size of the data structure that organizes candidate blocks, at least when identical blocks are eliminated. If a hash value has fewer bits, the data structure includes fewer categories, but each category may include more candidate blocks. On the other hand, using hash values with more bits tends to increase the size of the data structure that organizes candidate blocks. If a hash value has more bits, the data structure includes more categories, but each category may include fewer candidate blocks.
In some example implementations, the hash function used in hash-based block matching (and hash-based block reconstruction) is a cyclic redundancy check (“CRC”) function. CRC functions are fast and widely implemented, since they are commonly used for error checking purposes. A CRC function tends to produce very different hash values when applied to similar input values. That is, changing the sample values of a block slightly can lead to different CRC hash values. A CRC function can produce hash values with 8 bits, 12 bits, 16 bits, 24 bits, 32 bits, or some other number of bits.
When used for hash-based block matching, CRC functions work well to find a matching block that is exactly identical to a current block. A CRC function typically does not help find a matching block that only approximates a current block, however.
In other example implementations, the hash function used in hash-based block matching (and hash-based block reconstruction) uses averaging and XOR operations to determine the signature of a candidate block or current block. A hash value includes a component based on average of the block and one or more components based on directional XOR operations.
For example, for an 8×8 block B[i, j] with 8-bit sample values, the hash value is a 24-bit value with three 8-bit components. The first 8-bit component is the average μ of the sample values in the block.
For the second and third 8-bit components, a level map L[i, j] of sample values is derived using the average value μ for the block.
The second 8-bit component r is based on XOR operations along rows of the level map L[i, j]. For each row 0≦j<8 of the level map L[i, j], a bit r[j] of the second component is determined.
r[j]=L[0,j]⊕L[1,j]⊕L[2,j]⊕L[3,j]⊕L[4,j]⊕L[5,j]⊕L[6,j]⊕L[7,j],
where ⊕ represents an XOR operation.
The third 8-bit component c is based on XOR operations along columns of the level map L[i, j]. For each column 0≦i<8 of the level map L[i,j], a bit c[i] of the third component is determined.
c[i]=L[i,0]⊕L[i,1]⊕L[i,2]⊕L[i,3]⊕L[i,4]⊕L[i,5]⊕L[i,6]⊕L[i,7],
Thus, for the second component r, the jth bit is the XOR of the jth row's level map. And, for the third component c, the ith bit is the XOR of the ith column's level map. The 24-bit hash value is (μ<<16)+(r<<8)+c, where << represents a left shift operation.
For this hash function, the second and third components (based on XOR operations) are sensitive to small variations in sample values. When used for hash-based block matching, hash functions based on averaging and XOR operations work well to find a matching block that is exactly identical to a current block, however.
Instead of 8×8 blocks, a hash function that uses averaging and XOR operations can be used for a different size of block, e.g., computing XOR operations along pairs of rows or pairs of columns for a 16×16 block to determine an 8-bit component r or c, computing XOR operations along groups of four rows or four columns for a 32×32 block to determine an 8-bit component r or c, and so on. Alternatively, a hash function that uses averaging and XOR operations can produce hash values with a different number of bits.
In other example implementations, the hash function used in hash-based block matching (and hash-based block reconstruction) is a locality-sensitive hash function. A locality-sensitive hash function tends to produce the same hash value when applied to similar input values. That is, changing the sample values of a block slightly typically does not lead to different hash values. A locality-sensitive hash function can produce hash values with 8 bits, 12 bits, 16 bits, 24 bits, 32 bits, or some other number of bits.
For example, for an 8×8 block B, a 64-dimension vector {right arrow over (b)} is generated from the 64 sample values of the block B. Another 64-dimension vector {right arrow over (v)} is randomly generated. The hash function depends on the product {right arrow over (b)}·{right arrow over (v)}. For example, the hash function can be
where d is a real number chosen from the range 0 . . . drange. If two blocks are similar, then their vectors {right arrow over (b)}1 and {right arrow over (b)}2 are similar, and their hash values h({right arrow over (b)}1) and h({right arrow over (b)}2) should also be the same or similar.
In some example implementations, the hash function used in hash-based block matching (and hash-based block reconstruction) includes block width and block height as input. For example, the hash function can be a modified CRC function, modified hash function that uses averaging and XOR operations, or modified locality-sensitive hash function, for which the hash value depends in part on block height and block width as inputs. With this variation of hash function, candidate blocks of different sizes (such as 64×64, 32×32, 16×16 and 8×8) are grouped together if they have the same hash value, and candidate blocks of different sizes can be organized within a single data structure.
Alternatively, another hash function is used for hash-based block matching.
E. Intra BC Prediction Mode—Implementation Options.
A block with prediction mode of intra BC prediction can be a CB, PB or other block. When the block is a CB, the BV value for the block can be signaled at CU level (and other CBs in the CU use the same BV value or a scaled version thereof). Or, when the block is a PB, the BV value for the block can be signaled at PU level (and other PBs in the PU use the same BV value or a scaled version thereof). More generally, the BV value for an intra-BC prediction block is signaled at an appropriate syntax level for the block.
The block copying operations of prediction according to the intra BC prediction mode can be performed at the level of CB (when a BV value is signaled per CB) or PB (when a BV value is signaled per PB). For example, suppose a 16×16 CB has a single 16×16 PB. The BV value (for the PB) is applied to block copy a 16×16 region. Alternatively, the block copying operations can be performed at the level of TBs within a PB or CB, even when the BV value is signaled for the PB or CB. In this way, a BV value, as applied for a TB, can reference positions of other TBs in the same PB or CB.
In some implementations, for an intra BC predicted CU, intra BC prediction for a chroma block in a PU uses the same BV value as intra BC prediction for the luma block in the PU, possibly after scaling and rounding when the chroma data has reduced resolution relative to the luma data (e.g. when the format is YUV 4:2:0 format). Alternatively, different BV values can be signaled for the luma block and chroma blocks of a PU.
In view of the many possible embodiments to which the principles of the disclosed invention may be applied, it should be recognized that the illustrated embodiments are only preferred examples of the invention and should not be taken as limiting the scope of the invention. Rather, the scope of the invention is defined by the following claims. We therefore claim as our invention all that comes within the scope and spirit of these claims.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/CN2013/085939 | 10/25/2013 | WO | 00 |
