KR20090073112A - Method and apparatus for multiple pass video coding and decoding - Google Patents

Abstract

There are provided a video encoder, a video decoder and corresponding method for encoding and decoding video signal data using a multiple-pass video encoding scheme. The video encoder includes a motion estimator (116) and a decomposition module (174). The motion estimator performs motion estimation on the video signal data to obtain a motion residual corresponding to the video signal data in a first encoding pass. The decomposition module, in signal communication with the motion estimator, decomposes the motion residual in a subsequent encoding pass.

Description

METHOD AND APPARATUS FOR MULTIPLE PASS VIDEO CODING AND DECODING}

This application is directed to PCT International Application No. PCT / US2006 / 037139, entitled “Methods and Apparatus for Multi-Path Video Coding and Decoding,” filed September 22, 2006, which is hereby incorporated by reference in its entirety. Insist on priority.

The present invention relates generally to video encoding and decoding, and in particular, to a method and apparatus for multipath video encoding and decoding.

International Organization for Standardization / International Electrotechnical Commission (ISO / IEC) Moving Picture Experts Group-4 (MPEG-4) Part 10 Advanced Video Coding (AVC) Standard / International Telecommunications Union, Telecommunication Sector (ITU-T) H.264 standard (hereafter “MPEG4 / H.264 standard” or simply “H.264 standard”) is the most powerful and up-to-date video coding standard. . Like all other video coding standards, the H.264 standard uses block-based motion compensation and Discrete Cosine Transform (DCT) type transform coding. It is well known that DCT is efficient in video coding and suitable for high performance applications such as broadcast high definition television (HDTV). However, the DCT algorithm is not suitable for applications that require very low bit rates, such as dedicated video cell phones. At very low bit rates, the DCT transform will result in blocking artifacts, even when using a deblocking filter, with very few coefficients that can be coded at very low bit rates, with each coefficient being very spacing. This is because there is a tendency to have this wide quantization step.

Matching Pursuit (MP) is a greedy algorithm that decomposes an arbitrary signal into a linear extension of selected waveforms from a redundant dictionary of functions. These waveforms are chosen to optimally match the signal structure.

Suppose we have a 1-D signal f (t) and want to decompose this signal using a basis vector from an over-complete preset G.

Here, γ is an indexing parameter associated with a particular dictionary element. Decomposition begins by choosing γ to maximize the absolute value of the dot product as follows:

The residual signal is then calculated as follows:

This residual signal is then extended in the same way as the original signal. The procedure is repeated repeatedly until a set number of expansion factors is generated or a predetermined energy threshold for the remainder is reached. Each step n produces a dictionary function γ _n . After the entire M step, the signal can be approximated by the linear function of the dictionary elements as follows:

The complexity of MP decomposition of the signal of n samples is proved to be of order k.N.d.nlog ₂ n. Where d depends on the size of the dictionary without considering the transformation, N is the number of expansion coefficients selected, and constant k follows the strategy for selecting the dictionary function. Given the high-over Complete Dictionary, MP is, its complexity requires more computation than the 8 × 8 and 4 × 4 DCT integer transform used in, H.264 standard, which is defined as o (nlog ₂ n).

In general, the MP algorithm is suitable for any set of residual basis forms. It is proposed to extend the signal using the overcomplete basis of the Gabor function. The 2-D Gabor dictionary is very redundant, and each form may be at any integer-pixel location of the coded residual picture. Since the MP has a much larger dictionary set, and each coded basis function is well matched to the structure of the residual signal, the frame-based Gabor dictionary does not have an artificial block structure.

The Gabor Redundant Dictionary Set uses a MP algorithm (hereinafter referred to as "a Gabor-based MP video coding approach of the prior art") for a very low bit rate video coding based on MP for a proposed video coding system. Has been adapted. The proposed system is based on a simulation model for very low bit rate picture coding, or a framework of a low bit rate hybrid DCT system, referred to simply as "SIM3", where the DCT residual coder is replaced with an MP coder. do. This coder uses MP to decompose the residual motion picture into pre-separable 2-D Gabor functions. This proposed system is shown to work well in low bit rate low motion sequences.

A smooth 16x16 sine-squared window was applied to the predictive picture for 8x8 partitions of the Gabor-based MP video coding approach of the prior art. The MP video codec of the Gabor-based MP video coding approach of the prior art is based on the ITU-T H.263 codec. However, the H.264 standard enables small block size variable block-size motion compensation, which can be as small as 4 × 4 for luma motion compensation. In addition, the H.264 standard is based primarily on 4 × 4 DCT type transforms, rather than 8 × 8 as in most other significant conventional video coding standards, for baseline and main profiles. Directional spatial prediction for intra coding improves the quality of the prediction signal. All of these important design features make the H.264 standard more effective, but require more complex situations when applying MP to the H.264 standard. A smooth 16 × 16 sine squared window would look like this:

A hybrid coding technique (hereafter “hybrid coding technique of the prior art”) has been proposed that benefits from some features introduced by the H.264 standard for motion compensation and replaces the transform in the spatial domain. The prediction error is coded using the MP algorithm, which decomposes the signal into a properly designed two-dimensional anisotropic redundant dictionary. In addition, fast atom search techniques have been introduced. However, the proposed prior art hybrid coding technique does not describe whether to use the 1-pass or 2-pass technique. In addition, the proposed prior art hybrid coding scheme described that the motion compensation unit is in accordance with the H.264 standard, but any deblocking filter was used in the coding scheme, or caused by the predictive picture at a very low bit rate. It does not describe whether any other methods have been used to smooth the blocking artifacts.

The above and other drawbacks and drawbacks of the prior art are solved by the present invention with respect to methods and apparatus for multipath video encoding and decoding.

According to one aspect of the present invention, a video encoder is provided that encodes video signal data using a multipath video encoding technique. The video encoder includes a motion estimator and a decomposition module. The motion estimator performs motion estimation on the video signal data to obtain a motion residual corresponding to the video signal data in the first encoding path. A decomposition module in signal communication with the motion estimator decomposes the motion residuals in subsequent encoding paths.

According to another aspect of the present invention, a method of encoding video signal data using a multipath video encoding technique is provided. The method includes performing motion estimation on the video signal data to obtain a motion residual corresponding to the video signal data in the first encoding path, and decomposing the motion residual in the subsequent encoding path.

According to another aspect of the invention, a video decoder is provided for decoding a video bitstream. Video decoders include entropy decoders, atomic decoders, inverse transformers, motion compensators, deblocking filters, and synthesizers. The entropy decoder decodes the video bitstream to obtain a decomposed video bitstream. An atomic decoder in signal communication with the entropy decoder decodes the resolved atoms corresponding to the resolved bitstream to obtain decoded atoms. The inverse transformer in signal communication with the atomic decoder applies an inverse transform to the decoded atoms to form a reproduced residual image. A motion compensator in signal communication with the entropy decoder performs motion compensation using a motion vector corresponding to the decomposed bitstream to form a reproduced predicted picture. The deblocking filter in signal communication with the motion compensator performs deblocking filtering on the reproduced predictive image to smooth the reproduced predictive image. A synthesizer in signal communication with the inverse transformer and the overlapping block motion compensator synthesizes the reproduced prediction picture and the residual picture to obtain a reproduced picture.

According to another aspect of the present invention, a method of decoding a video bitstream is provided. The method includes decoding a video bitstream to obtain a decomposed video bitstream, decoding a decomposed atom corresponding to the decomposed bitstream to obtain a decoded atom, and applying an inverse transform to the decoded atom to reproduce the residual. Forming an image, performing motion compensation using a motion vector corresponding to the decomposed bitstream, forming a reproduced prediction image, and performing deblocking filtering on the reproduced prediction image to smoothly reproduce the reproduced prediction image And synthesizing the reproduced predictive picture and the residual picture to obtain a reproduced picture.

These and other aspects, features, and advantages of the present invention will become apparent from the following detailed description of exemplary embodiments to be described in conjunction with the accompanying drawings.

The invention will be better understood according to the exemplary drawings described below.

1A and 1B illustrate exemplary first and second path portions of an encoder of an MP encoder / decoder (CODEC) based on a two-path H.264 standard, in which the principles of the present invention may be applied in accordance with one embodiment of the present invention. It is a drawing about.

FIG. 2 is a diagram of an exemplary decoder of a dual path H.264 standard based MP encoder / decoder (CODEC) in which the principles of the present invention may be applied in accordance with an embodiment of the present invention.

3 is a diagram of an exemplary method for encoding an input video sequence in accordance with an embodiment of the present invention.

4 is a diagram of an exemplary method for decoding an input video sequence in accordance with an embodiment of the present invention.

The present invention relates to a method and apparatus for multipath video encoding and decoding. Advantageously, the present invention corrects the blocking artifacts introduced by the DCT transform used in, for example, the H.264 standard in very low bit rate applications. It will also be appreciated that the present invention is not limited to low bit rate applications, but may be used for other (higher) bit rates while maintaining the scope of the present invention.

This description describes the principles of the invention. Therefore, those skilled in the art will be able to devise various configurations that embody the principles of the present invention and fall within the spirit and scope, even if not explicitly described or illustrated herein.

All examples and conditional languages mentioned herein are for pedagogical purposes that aid the reader in understanding the principles and concepts of the present invention as provided by the inventors and to enhance their skills, and such specifically mentioned examples and conditions. It should not be construed as limited to these.

Moreover, not only the specific examples of the present invention, but all descriptions citing the principles, aspects, and embodiments of the present invention are intended to include both structural and functional equivalents thereof. In addition, these equivalents are intended to include not only equivalents now known, but also equivalents to be developed in the future, that is, any components developed to perform the same function regardless of structure.

Thus, for example, those skilled in the art will understand that the block diagrams presented herein represent conceptual diagrams of exemplary circuits implementing the present invention. Likewise, any flowchart, flowchart, state transition diagram, pseudocode, or the like may represent various processes that can be represented by a computer-readable medium and executed by a computer or processor, regardless of whether the computer or processor is clearly shown. Will be understood.

The functionality of the various components shown in the figures may be provided through the use of dedicated hardware as well as hardware capable of executing software in combination with appropriate software. When provided by a processor, the functions may be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which may be shared. In addition, the explicit use of the term "processor" or "controller" should not be interpreted exclusively to refer to hardware capable of executing software, and is not limited to digital signal processor (DSP) hardware, ROM (read to store software). only memory (RAM), random access memory (RAM), and non-volatile storage devices may be implicitly included.

Other hardware, tolerances and / or orders may also be included. Likewise, any switches shown in the figures are merely conceptual. Their functions may be performed through the operation of program logic, through dedicated logic, through program control and interaction of dedicated logic, or manually, and certain techniques are implemented as will be more clearly understood from the context. It is selectable by ruler.

In the claims of this specification, any component indicated as a means for performing a particular function is, for example, a) a combination of circuit elements that perform the function or b) a suitable circuit for causing the software to perform that function. It is intended to cover any means for carrying out its functions, including any form of software, including firmware, microcode, etc., in conjunction with. The invention as defined by these claims is based on the fact that the functionality provided by the various recited means is obtained by combining together in a manner claimed by the claims. Thus, any means capable of providing these functionalities is considered to be the same as shown herein.

In accordance with the principles of the present invention, a multipath video encoding and decoding technique is provided. Matching tracking (MP) may be used for the multipath video encoding and decoding techniques. In an exemplary embodiment, a two path H.264 based coding technique is disclosed for MP video coding.

The H.264 standard applies block-based motion compensation and DCT type transformation similarly to other video compression standards. At very low bit rates, the DCT transform will result in blocking artifacts, even with a deblocking filter, which has few coefficients that can be coded at very low bit rates, and each coefficient tends to have a very spaced quantization step. Because of this. In accordance with the principles of the present invention, an MP using an overcomplete basis is applied to code the residual picture. The motion compensation and mode determiner follows the H.264 standard. Overlapped block motion compensation (OBMC) is applied to smooth the predicted picture. In addition, a new approach is provided to select a base other than MP.

In accordance with the principles of the present invention, the video encoder and / or decoder applies OBMC to the predictive picture, thereby reducing the blocking artifacts caused by the predictive model. The MP algorithm is used to code the residual picture. The advantage of MP is that it is not block based but frame based, so there are no blocking artifacts caused by the coding residual difference.

1A and 1B, exemplary first and second path portions of an encoder of a dual path H.264 standard based MP encoder / decoder (CODEC) are indicated generally by reference numerals 110 and 160. The encoder is indicated generally by the reference numeral 190 and the decoder portion is indicated generally by the reference numeral 191.

1A, the input of the first path portion 110 is connected in signal communication with a non-inverting input of the synthesizer 112, an input of the encoder control module 114, and a first input of the motion estimator 116. . The first output of synthesizer 112 is connected in signal communication with a first input of buffer 118. The second output of the synthesizer 112 is connected in signal communication with an input of an integer conversion / scaling / quantization module 120. An output of the integer conversion / scaling / quantization module 120 is connected in signal communication with a first input of the scale / inverse conversion module 122.

A first output of the encoder control module 114 is connected in signal communication with a first input of the intra frame predictor 126. A second output of the encoder control module 114 is connected in signal communication with a first input of the motion compensator 124. A third output of the encoder control module 114 is connected in signal communication with a second input of the motion estimator 116. A fourth output of the encoder control module 114 is connected in signal communication with a second input of the scale / inverse transform module 122. A fifth output of the encoder control module 114 is connected in signal communication with a first input of a buffer 118.

An output of the motion estimator 116 is connected in signal communication with a second input of the motion compensator 124 and a second input of the buffer 128. The inverting input of the synthesizer 112 is selectively connected in signal communication with the output of the motion compensator 124 or the output of the intra frame predictor 126. An output of a selected one of the motion compensator 124 or the intra frame predictor 126 is connected in signal communication with a first input of the synthesizer 128. An output of the scale / inverse transform module 122 is connected in signal communication with a second input of the synthesizer 128. An output of the synthesizer 128 is connected in signal communication with a second input of the intra frame predictor 126, a third input of the motion estimator 116, and an input / output of the motion compensator 124. The output of the buffer 118 is available as the output of the first path portion 110.

In the first path unit 110, the encoder control module 114, the integer transform / size adjustment / quantization module 120, the buffer 118, and the motion estimator 116 are included in the encoder 190. In addition, in the first path unit, the scale / inverse transform module 122, the intra frame predictor 126, and the motion compensator 124 are included in the decoder unit 191.

An input of the first path unit 110 receives the input video 111 and buffers control data (eg, a motion vector, a mode selection, a predictive picture, etc.) for use in the second path unit 160. Stored at 118.

Referring to FIG. 1B, the first input of the second path unit 160 is connected in signal communication with an input of the entropy coder 166. This first input receives control data 162 (eg, mode selection, etc.) and motion vector 164 from first path portion 110. The second input of the second path portion 160 is connected in signal communication with the non-inverting input of the synthesizer 168. The third input of the second path unit 160 is connected in signal communication with an input of the overlapping block motion compensation (OBMC) / deblocking module 170. The second input of the second path unit 160 receives the input video 111, and the third input of the second path unit 160 receives the predicted image 187 from the first path unit 110.

An output of synthesizer 168 providing residual 172 is connected in signal communication with an input of atom finder 174. The output of the atomic finder 174 providing the coded residual 178 is connected in signal communication with the input of the atomic coder 176 and the first non-inverting input of the synthesizer 180. An output of the OBMC / deblocking module 170 is connected in signal communication with an inverting input of the synthesizer 168 and a second non-inverting input of the synthesizer 180. The output of synthesizer 180, which provides output video, is connected in signal communication with an input of reference buffer 182. An output of the atomic coder 176 is connected in signal communication with an input of an entropy coder 166. The output of entropy coder 166 is available as the output of second path portion 160 and provides an output bitstream.

In the second path portion 160, an entropy coder is included in the encoder 190, the synthesizer 168, the OBMC module 170, the atomic finder 174, the atomic coder 176, and the reference buffer 182. Is included in the decoder unit 191.

With reference to FIG. 2, an exemplary decoder of a dual path H.264 standard based MP encoder / decoder (CODEC) is indicated generally by the reference numeral 200.

An input of the decoder 200 is connected in signal communication with an input of the entropy decoder 210. The output of the entropy decoder is connected in signal communication with the input of the atomic decoder 220 and the input of the motion compensator 250. An output of inverse transform module 230 providing residual is connected in signal communication with a first non-inverting input of synthesizer 270. The output of the motion compensator 250 is connected in signal communication with an input of the OBMC / deblocking module 260. An output of the OBMC / deblocking module 260 is connected in signal communication with a second non-inverting input of the synthesizer 270. The output of the synthesizer is available as the output of the decoder 200.

Unlike the MP video codec of the Gabor based MP video coding approach of the prior art based on the H.263 codec, the present invention is applicable to the ITU-T H.264 / AVC coding system. Due to the frame-based residual coding, OBMC is applied to the predictive picture, which is not implemented in the H.264 / AVC codec.

In one embodiment according to the invention, the first path of the video encoding technique is in accordance with the H.264 standard. There is no actual coding in the first path. For example, all control data such as mode selection, predictive pictures and motion vectors are stored in the buffer for the second path. DCT conversion is still applied in the first path for mode selection and motion compensation using Rate Distortion Optimization (RDO). Instead of coding the residual picture using the DCT coefficients, all residual pictures are stored for the second path. In one embodiment of the present invention, it has been proposed to apply 16x16 restricted intra coding or H.264 standard compliant limited intra coding, and in particular to process the boundary between intra coded macroblock and inter coded macroblock.

In the second path, the motion vector and control data can be coded by entropy coding. The residual picture may be coded by the MP. Atomic retrieval and parametric coding can be performed, for example, according to the Gabor based MP video coding approach of the prior art. The reproduced picture is stored for the reference frame.

One of the advantages of MP video coding is that MP is not block based and therefore has no blocking artifacts. However, if motion prediction is performed block-based and inaccurate, MP video coding still generates certain blocking artifacts at very low bit rates. The simulation shows that the contour in which the atoms move and the motion vector (MV) appear in areas that are not very accurate. Improvements in motion estimation allow atoms to better represent residuals.

To remove artifacts from motion prediction, one method involves smoothing the blunt boundaries of the predictive picture using an H.264 type or an improved deblocking filter. In another approach, a smoother motion model using overlapping blocks (OBMC) is employed. In the Gabor-based MP video coding approach of the prior art, a 16x16 sine squared window has been adapted. The N × N sine squared window can be defined according to, for example, a hybrid coding technique of the prior art. A 16x16 sine squared window is designed for 8x8 blocks, and the 16x16 blocks are treated as four 8x8 blocks.

However, in the H.264 standard, partitions with luma block size 16x16, 16x8, 8x16, and 8x8 samples are supported. If partitions with 8x8 samples are selected, the 8x8 partition is further divided into 8x4, 4x8 or 4x4 luma samples and corresponding chroma samples. Here, four approaches are proposed for dealing with further partition forms. The first approach is to use 8x8 sine squared windows for 4x4 partitions. For all other partitions above 4x4, divide these partitions into multiple 4x4 partitions. The second approach uses a 16x16 sine squared window for 8x8 and partitions above it, but does not touch partitions smaller than 8x8. The third approach is to use adaptive OBMC for all partitions. All three approaches implement only OBMC, not deblocking filter, and the fourth approach is to combine OBMC and deblocking filter (s).

In addition to redundant Gabor dictionary sets of prior art Gabor based MP video coding approaches, implemented for residual coding, we propose the use of a more overcomplete base. At low bit rates, the transform motion model does not accurately represent the natural movement of related visual characteristics such as moving edges. Thus, most of the residual error energy is located in these areas. Therefore, it is important to represent an error image using the edge detection redundant dictionary. Low redundancy discrete wavelet transform (eg, 2-D Dual-Tree Discrete Wavelet Transform (DDWT)), or any other edge detection dictionary, in which a 2-D Gaver dictionary may be utilized may be used. 2-D DDWT has more subbands / directions than 2-D DWT. Each subband represents one direction, which is edge detective. After noise shaping, the 2-D DDWT achieves higher PSNR compared to standard 2-D DWT with coefficients that remain the same. Thus, 2-D DDWT is more suitable for coding edge information. After applying the OBMC to the predictive picture, the error picture will have smoother edges. Parametric overcomplete 2-D dictionaries can be used to provide smoother edges.

Referring to FIG. 3, an exemplary method for encoding an input video sequence is indicated generally at 300. The method 300 includes initiation block 305 that passes control to a decision block 310. Decision block 310 determines whether the current frame is an I frame. If the current frame is an I frame, control is passed to a function block 355. Otherwise, control is passed to a function block 315.

Function block 355 performs an H.264 standard compliant frame coding to provide an output bitstream, and passes control to end block 370.

The function block 315 performs H.264 standard compliant motion compensation, and passes control to a function block 320. The function block 320 stores the motion vector (MV), the control data, and the prediction block, and passes control to a decision block 325. The decision block 325 determines whether the end of the frame has been reached. If the end of the frame has been reached, control is passed to a function block 330. Otherwise, control returns to function block 315.

The function block 330 performs OBMC and / or deblocking filtering on the predictive picture, and passes control to a function block 335. The function block 335 obtains the residual picture from the original picture and the predictive picture, and passes control to a function block 340. The function block 340 codes the residual using the MP, and passes control to a function block 345. The function block 345 performs entropy coding to provide an output bitstream, and passes control to an end block 370.

Referring to FIG. 4, an exemplary method for decoding an input video sequence is indicated generally at 400. The method 400 includes initiation block 405, which passes control to a decision block 410. Decision block 410 determines whether the current frame is an I frame. If the current frame is an I frame, control is passed to a function block 435. Otherwise, control is passed to a function block 415.

The function block 435 performs H.264 standard compliant decoding to provide a reproduced picture, and passes control to an end block 470.

Function block 415 decodes the motion vectors, control data, and MP atoms, and passes control to function block 420 and function block 425. The function block 420 reproduces the residual picture using the decoded atoms, and passes control to a function block 430. The function block 425 reproduces the predictive picture by decoding motion vectors and other control data and applying OBMC and / or deblocking filtering, and passes control to a function block 430. The function block 430 synthesizes the reproduced residual picture and the reproduced predictive picture to provide a reproduced picture, and passes control to an end block 470.

Now, some of the many minor advantages / features of the present invention will be described, some of which have been described above. For example, one advantage / feature is a video encoder that encodes video signal data using a multipath video encoding technique, which includes a motion estimator and a decomposition module. The motion estimator performs motion estimation on the video signal data to obtain a motion residual corresponding to the video signal data in the first encoding path. A decomposition module in signal communication with the motion estimator decomposes the motion residuals in subsequent encoding paths.

Another advantage / feature is the video encoder as described above, wherein the multipath video coding technique is a double path video encoding technique. The video encoder further includes a buffer in signal communication with the motion estimator and decomposition module to store the motion residue obtained in the first encoding path for subsequent use in the second encoding path. The decomposition module decomposes the motion residual using a redundant Gabor dictionary set in the second encoding path.

Another advantage / feature is a video encoder employing the two-path video encoding technique as described above, wherein the motion estimator is in the first encoding path the International Telecommunication Union, Department of Telecommunications (ITU-T) H.264 standard. According to the motion estimation and coding mode selection.

Yet another advantage / feature is a video encoder using a dual path video encoding technique as described above, wherein the video encoder further comprises a prediction module and a nested block motion compensator. The prediction module in signal communication with the buffer forms a predictive picture corresponding to the video signal data in the first encoding path. An overlapping block motion compensator in signal communication with the buffer performs overlapping block motion compensation (OBMC) on the predictive picture using a 16 × 16 sine squared window to smooth the predictive picture in the second encoding path. The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path.

Further, another advantage / feature is a video encoder using a dual path video encoding technique as described above, wherein the video encoder further comprises a prediction module and an overlapping block motion compensator. The prediction module in signal communication with the buffer forms a predictive picture corresponding to the video signal data in the first encoding path. The overlapping block motion compensator in signal communication with the buffer performs overlapping block motion compensation (OBMC) only for partitions of 8x8 and larger of the predicted picture in the second encoding path. The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path.

Further, another advantage / feature is a video encoder using a dual path video encoding technique as described above, wherein the video encoder further comprises a prediction module and an overlapping block motion compensator. The prediction module in signal communication with the buffer forms a predictive picture corresponding to the video signal data in the first encoding path. An overlapping block motion compensator in signal communication with the buffer performs overlapping block motion compensation (OBMC) using an 8x8 sine square window for the 4x4 partition of the predictive picture in the second encoding path. All partitions of the predictive picture are divided into 4x4 partitions when OBMC is performed in the second encoding path. The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path.

Further, another advantage / feature is a video encoder using a dual path video encoding technique as described above, wherein the video encoder further comprises a prediction module and an overlapping block motion compensator. The prediction module in signal communication with the buffer forms a predictive picture corresponding to the video signal data in the first encoding path. An overlapping block motion compensator in signal communication with the buffer performs adaptive overlapping block motion compensation (OBMC) on all partitions of the predictive picture in the second encoding path. The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path.

Further, another advantage / feature is a video encoder using a dual path video encoding technique as described above, wherein the video encoder further comprises a prediction module and a deblocking filter. The prediction module in signal communication with the buffer forms a predictive picture corresponding to the video signal data in the first encoding path. The deblocking filter in signal communication with the buffer performs a deblocking operation on the predictive picture in the second encoding path. The buffer stores the prediction picture therein in the first encoding path for subsequent use in the second encoding path.

Also, another advantage / feature is a video encoder using the dual path video encoding technique as described above, wherein the decomposition module performs a dual-tree wavelet transform to decompose the motion residuals.

Another advantage / feature is a video encoder using a dual path video encoding technique and dual tree wavelet transform as described above, wherein the decomposition module selects the coefficients of the dual tree wavelet transform using noise shaping.

Another advantage / feature is a video encoder using a dual path video encoding technique as described above, wherein the decomposition module applies a parametric overcomplete 2-D dictionary to decompose the motion residuals in the second encoding path.

Another advantage / feature is a video decoder that decodes a video bitstream, wherein the video decoder includes an entropy decoder, an atomic decoder, an inverse transformer, a motion compensator, a deblocking filter, and a synthesizer. The entropy decoder decodes the video bitstream to obtain a decomposed video bitstream. An atomic decoder in signal communication with the entropy decoder decodes the resolved atoms corresponding to the resolved bitstream to obtain decoded atoms. The inverse transformer in signal communication with the atomic decoder applies an inverse transform to the decoded atoms to form a reproduced residual image. A motion compensator in signal communication with the entropy decoder performs motion compensation using a motion vector corresponding to the decomposed bitstream to form a reproduced predicted picture. The deblocking filter in signal communication with the motion compensator performs deblocking filtering on the reproduced predictive image to smooth the reproduced predictive image. A synthesizer in signal communication with the inverse transformer and the overlapping block motion compensator synthesizes the reproduced prediction picture and the residual picture to obtain a reproduced picture.

These and other features and advantages of the present invention will be readily apparent to those skilled in the art based on the teachings herein. It will be appreciated that the teachings of the present invention may be implemented in various forms of hardware, software, firmware, dedicated processors, or a combination thereof.

Most preferably, the teachings of the present invention are implemented as a combination of hardware and software. In addition, the software may be implemented as an application program that is actually implemented in the program storage unit. The application program can be uploaded to and executed by a machine having any suitable structure. Preferably, the machine is implemented with a computer platform having hardware such as one or more central processing units (CPUs), RAM, and input / output (I / O) interfaces. The computer platform may also include an operating system and microinstruction code. The various processes and functions described herein may be part of microinstruction code or part of an application program, or any combination thereof, that may be executed by a CPU. In addition, various other peripherals, such as additional data storage and printing, may be connected to the computer platform.

In addition, since some of the component system elements and methods shown in the accompanying drawings are preferably implemented in software, the actual connection between system components or process functional blocks may vary depending on how the invention is programmed. Through the teachings herein, those skilled in the art will be able to devise the above and similar implementations or configurations of the present invention.

Although exemplary embodiments have been described herein with reference to the accompanying drawings, the invention is not limited to these specific embodiments, and various changes and modifications may be made by those skilled in the art without departing from the scope or spirit of the invention. It must be understood. All such changes and modifications are intended to be included within the scope of the present invention as defined in the appended claims.

Claims (24)

A video encoder for encoding video signal data using a multipath video encoding technique, A motion estimator 116 for performing motion estimation on the video signal data to obtain a motion residual corresponding to the video signal data in a first encoding path; Decomposition module 174 in signal communication with the motion estimator and for decomposing the motion residual in subsequent encoding paths Video encoder comprising a. The method of claim 1, The multipath video coding technique is a double path video encoding technique, The video encoder further comprises a buffer 118 in signal communication with the motion estimator and the decomposition module and for storing the motion residual obtained in the first encoding path for subsequent use in a second encoding path, The decomposition module (174) decomposes the motion residual using a set of redundant Gabor dictionaries in the second encoding path. The method of claim 2, The motion estimator 116 performs the motion estimation and coding mode selection according to the International Telecommunication Union (ITU-T: Telecommunication Sector) H.264 standard in the first encoding path. Video encoder. The method of claim 2, A prediction module (124, 126) in signal communication with the buffer and forming a predictive picture corresponding to the video signal data in the first encoding path; In signal communication with the buffer, and performing overlapping block motion compensation (OBMC) on the predicted picture using a 16 × 16 sine square window to smooth the predicted picture in the second encoding path. Further comprises an overlapping block motion compensator (170), The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path. The method of claim 2, A prediction module (124, 126) in signal communication with the buffer and forming a predictive picture corresponding to the video signal data in the first encoding path; And an overlapping block motion compensator 170 in signal communication with the buffer and performing overlapping block motion compensation (OBMC) only for 8 × 8 and larger partitions of the predicted picture in the second encoding path. , The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path. The method of claim 2, A prediction module (124, 126) in signal communication with the buffer and forming a predictive picture corresponding to the video signal data in the first encoding path; An overlapping block motion compensator 170 in signal communication with the buffer and performing overlapping block motion compensation (OBMC) using an 8 × 8 sine squared window for the 4 × 4 partition of the predicted picture in the second encoding path. Including more, All partitions of the predictive picture are divided into 4x4 partitions when OBMC is performed in the second encoding path, The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path. The method of claim 2, A prediction module (124, 126) in signal communication with the buffer and forming a predictive picture corresponding to the video signal data in the first encoding path; An overlapping block motion compensator 170 in signal communication with the buffer and performing adaptive overlapping block motion compensation (OBMC) on all partitions of the predictive picture in the second encoding path, The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path. The method of claim 2, A prediction module (124, 126) in signal communication with the buffer and forming a predictive picture corresponding to the video signal data in the first encoding path; And a deblocking filter 170 in signal communication with the buffer, the deblocking filter 170 performing a deblocking operation on the predicted picture in the second encoding path, The buffer stores therein the predictive picture in the first encoding path for subsequent use in the second encoding path. The method of claim 2, The decomposition module (174) performs a dual-tree wavelet transform to decompose the residual motion. The method of claim 9, The decomposition module (174) selects coefficients of the dual tree wavelet transform using noise shaping. The method of claim 2, The decomposition module (174) applies a parametric over-complete 2-D dictionary to decompose the motion residual in the second encoding path. A method for encoding video signal data using a multipath video encoding technique, Performing motion estimation on the video signal data to obtain a motion residual corresponding to the video signal data in a first encoding path (315); Decomposing (340) the motion residue in a subsequent encoding path How to include. The method of claim 12, The multipath video coding technique is a double path video encoding technique, The method further includes storing 320 the motion residual obtained in the first encoding path for subsequent use in a second encoding path, Said decomposition step (340) decomposing said motion residual using a redundant Gabor dictionary set in said second encoding path. The method of claim 13, The motion estimation and coding mode selection is performed according to the International Telecommunication Union, Telecommunications Sector (ITU-T) H.264 standard in the first encoding path. The method of claim 13, Forming a predictive picture corresponding to the video signal data in the first encoding path (315); Storing (320) the prediction picture in the first encoding path; Performing overlapping block motion compensation (OBMC) on the prediction picture using a 16 × 16 sine squared window to smooth the prediction picture in the second encoding path (330) How to include more. The method of claim 13, Forming a predictive picture corresponding to the video signal data in the first encoding path (315); Storing (320) the prediction picture in the first encoding path; Performing overlap block motion compensation (OBMC) only on 8x8 and larger partitions of the predictive picture in the second encoding path (330) How to include more. The method of claim 13, Forming a predictive picture corresponding to the video signal data in the first encoding path (315); Storing (320) the prediction picture in the first encoding path; Performing an overlapping block motion compensation (OBMC) on a 4x4 partition of the predictive picture using an 8x8 sine square window in the second encoding path (330), All partitions of the predictive picture are divided into 4x4 partitions when OBMC is performed in the second encoding path. The method of claim 13, Forming a predictive picture corresponding to the video signal data in the first encoding path (315); Storing (320) the prediction picture in the first encoding path; Performing adaptive overlap block motion compensation (OBMC) on all partitions of the prediction picture in the second encoding path (330) How to include more. The method of claim 13, Forming a predictive picture corresponding to the video signal data in the first encoding path (315); Storing (320) the prediction picture in the first encoding path; Performing a deblocking operation on the prediction picture in the second encoding path (330) How to include more. The method of claim 13, The decomposing step (340) decomposes the residual motion by performing a dual tree wavelet transform. The method of claim 20, The decomposing step (340) selects coefficients of the dual tree wavelet transform using noise shaping. The method of claim 13, The decomposing step (340) applies a parametric overcomplete 2-D dictionary to decompose the motion residual in the second encoding path. A video decoder for decoding a video bitstream, An entropy decoder 210 for decoding the video bitstream to obtain a decomposed video bitstream; An atomic decoder (220) in signal communication with the entropy decoder and obtaining a decoded atom by decoding a decomposed atom corresponding to the decomposed bitstream; An inverse transformer 230 in signal communication with the atomic decoder and applying an inverse transform to the decoded atoms to form a reproduced residual image; A motion compensator 250 in signal communication with the entropy decoder and performing motion compensation using a motion vector corresponding to the decomposed bitstream to form a reproduced prediction image; A deblocking filter 260 in signal communication with the motion compensator and performing deblocking filtering on the reproduced predictive picture to smooth the reproduced predictive picture; A synthesizer 270 that is in signal communication with the inverse transformer and the overlapping block motion compensator, and synthesizes the reproduced prediction picture and the residual picture to obtain a reproduced picture. Video decoder comprising a. A method for decoding a video bitstream, Decoding the video bitstream to obtain a decomposed video bitstream (405); Decoding a decomposed atom corresponding to the decomposed bitstream to obtain a decoded atom (415); Applying an inverse transform to the decoded atoms to form a reproduced residual image (420); Performing motion compensation using a motion vector corresponding to the decomposed bitstream to form a reproduced prediction image (425); Performing deblocking filtering on the reproduced prediction picture to smooth the reproduced prediction picture (425); Synthesizing the reproduced prediction picture and the residual picture to obtain a reproduced picture (430) How to include.

Images