CN102801973B - Video image deblocking filter method and device - Google Patents

Video image deblocking filter method and device Download PDF

Info

Publication number
CN102801973B
CN102801973B CN201210236491.0A CN201210236491A CN102801973B CN 102801973 B CN102801973 B CN 102801973B CN 201210236491 A CN201210236491 A CN 201210236491A CN 102801973 B CN102801973 B CN 102801973B
Authority
CN
China
Prior art keywords
block
filtering
sub
current
memory
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201210236491.0A
Other languages
Chinese (zh)
Other versions
CN102801973A (en
Inventor
熊富贵
谢成兴
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Allwinner Technology Co Ltd
Original Assignee
Allwinner Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Allwinner Technology Co Ltd filed Critical Allwinner Technology Co Ltd
Priority to CN201210236491.0A priority Critical patent/CN102801973B/en
Publication of CN102801973A publication Critical patent/CN102801973A/en
Application granted granted Critical
Publication of CN102801973B publication Critical patent/CN102801973B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

The invention provides a video image deblocking filter method and a video image deblocking filter device. According to the method, a video or image frame is divided into a plurality of 16*16 macro blocks, and each macro block is divided into 16 4*4 subblocks. The method comprises the following steps of: reading the pixel data of the current macro blocks from a first storage device and a second storage device to a specified buffer group according to the instruction information described by a control vector when the pixel data of the current macro blocks is read; and executing operations of determining the number of corresponding subblocks, calculating the filter strength, determining the filter type, performing filtering operation processing and performing data write-back processing according to the instruction information described by the control vector in the process of executing filtering processing of each current subblock. The essence is that the universal filtering processing operation on different coding and decoding standards is executed according to the instruction information described by the universal control vector in a control table. According to the video image deblocking filter method and the video image deblocking filter device, standard deblocking filter processing can be performed on different videos, the universality is high, and the filtering speed is increased.

Description

Video image block-removal filtering method and device
Technical field
The present invention relates to digital image processing techniques field, particularly relate to and be applied in video image block-removal filtering method in video coding and decoding technology and device.
Background technology
In prior art, HD video encoding and decoding process proposes unprecedented requirement to block elimination filtering process.
Block elimination filtering process is all Video codings and the considerable link of decoding, and the Video processing of all pursuit compression efficiencies and quality all can carry out relevant treatment.But because different video standard has different operation methods in the process of block elimination filtering, also therefore become the typical characteristics that each video standard distinguishes, this also challenges to the versatility of deblocking filtering apparatus simultaneously; Due to the complexity of block elimination filtering operation method, process a macro block or a frame needs the more time, this is obviously unfavorable for the encoding and decoding process of high definition or ultra high-definition video.
The practices well of current reply the problems referred to above is: carry out independent process to different video form, only stores link in data and shares; Or in order to pick up speed, adopting parallel mode, meeting the demands by increasing computing unit.
Such way poor universality, processing speed is slow, also can add general-purpose chip cost simultaneously, reduce efficiency.
Generally speaking, existing deblocking filtering apparatus cannot realize carrying out filtering to different video formats and standard, and in prior art, filtering method filtration efficiency is low.Therefore for prior art, how to solve the block elimination filtering for various criterion, and to improve filtering processing speed be problem demanding prompt solution.
Summary of the invention
The invention provides a kind of video image block-removal filtering method and device, in order to solve the block elimination filtering process problem for various criterion, and improve the processing speed of block elimination filtering.
Based on the problems referred to above, a kind of video image block-removal filtering method provided by the invention, a video or picture frame are divided into several macro blocks of 16 × 16 by described block-removal filtering method, and each macro block is divided into the sub-block of 16 4 × 4, comprises the following steps:
Step S1, the pixel data of current macro to be preserved in a first memory, and the pixel data of each sub-block on the top adjacent with current macro, the left side and limit, upper left is kept in second memory;
Step S2, transfer the command information described by dominant vector of control table;
Steps A, when reading current macro block pixels data, the pixel data of described current macro is read in the group of buffers of specifying by the command information described by described dominant vector from first memory and second memory;
Step B, when performing each sub-block filtering in current macro, the command information described by filtering dominant vector is determined to describe the data interaction information of current filter window and group of buffers and is determined the block number of current filter sub-block;
Step C, calculate the intensity of current filter sub-block according to described data interaction information and judge whether to need to carry out filtering to current sub-block;
If step D judges to need to carry out filtering to current sub-block, then according to the filtering type of described data interaction information determination current sub-block, and pass through the method for the corresponding filtering operation of filtering type determination current sub-block; Do not need to carry out filtering to current sub-block if judge, then directly perform step G;
Step e, according to the method for the filtering operation of the current sub-block determined limit filtering is carried out to current sub-block and the limit filtering operation result of current sub-block is kept in group of buffers;
Step F, command information described by dominant vector, to be written back to the described operation result of current sub-block limit filtering described in group of buffers in second memory and to terminate the filtering operation of current filter sub-block;
Step G, judge that whether current sub-block is last sub-block of filtering operation in current macro, if the determination result is YES, then perform following step H; If judged result is no, returns step B execution and the filtering process of next sub-block is operated;
The filtering process operation of step H, end current macro;
Wherein, described control table is applicable to different video encoding and decoding standards, and the command information described by the dominant vector of described control table is the decomposition control procedure of general-purpose operation method in different video encoding and decoding standard.
Further, as a kind of embodiment, in described step C, specific implementation step is as follows:
C1, calculate the intensity of current filter sub-block according to described data interaction information;
If the boundary intensity of the current sub-block that C2 calculates is 0, then judge not need to carry out limit filtering to current sub-block; If the boundary intensity of the current sub-block calculated is not 0, then judge to need to carry out limit filtering to current sub-block.
Preferably, the data interaction information of described current filter window and group of buffers comprises current macro information, the adjacent pixel data information of the adjacent sub-blocks on current sub-block information and limit to be filtered.
Correspondingly, the embodiment of the present invention also provides a kind of video image deblocking filtering apparatus, comprise first memory, second memory, bank of filters, group of buffers, the 3rd memory, reading unit, determining unit, filtering strength judging unit, filtering operation unit, processing unit, judging unit, end unit, integrated treatment unit and dominant vector table, wherein:
Described first memory, for preserving the pixel data of current filter macro block;
Described second memory, for preserving the pixel data of each sub-block on the top adjacent with current macro, the left side and limit, upper left; And preserve the operation result read in group of buffers;
Described group of buffers, for preserving the pixel data read in from first memory and second memory;
Described reading unit, for when reading current macro block pixels data, the pixel data of described current macro is read into the group of buffers of specifying by the command information described by described dominant vector from first memory and second memory;
Described determining unit, for when performing each sub-block filtering in current macro, the command information described by filtering dominant vector is determined to describe the data interaction information of current filter window and group of buffers and is determined the block number of current filter sub-block;
Described filtering strength judging unit, for calculating the intensity of current filter sub-block according to described data interaction information and judging whether to need to carry out filtering to current sub-block; Need if judge, then the filtering type of the described data interaction information determination current sub-block in the command information described by described dominant vector, and the method passing through the corresponding filtering operation of filtering type determination current sub-block; Do not need if judge, then directly return judging unit and perform corresponding operation;
Described filtering operation unit, carries out limit filtering for the method for the filtering operation according to the current sub-block determined to current sub-block and is kept in group of buffers by the limit filtering operation result of current sub-block;
Described bank of filters, for carrying out limit filtering to current filter sub-block;
Described processing unit, for the command information described by dominant vector, to be written back to the described operation result of current sub-block limit filtering described in group of buffers in second memory and to terminate the filtering operation of current filter sub-block;
Described judging unit, for judging that whether current sub-block is last sub-block of filtering operation in current macro, if the determination result is YES, then redirect end unit performs the operation terminating current macro filtering process; If judged result is no, then return the corresponding filtering process operation that determining unit performs next sub-block;
Described 3rd memory, for preserving described control table;
Described end unit, for terminating the filtering process operation of current macro;
Described integrated treatment unit, for the command information described by described dominant vector, coordinates each unit and performs corresponding operation;
Described deblocking filtering apparatus also comprises control unit, for transfer control table in the 3rd memory dominant vector described by command information;
Wherein, described control table is applicable to different video encoding and decoding standards, and the command information described by the dominant vector of described control table is the decomposition control procedure of general-purpose operation method in different video encoding and decoding standard.
Further, as a kind of embodiment, described filtering strength judging unit specifically comprises computation subunit and judges subelement, wherein:
Described computation subunit, for calculating the intensity of current filter sub-block according to described data interaction information;
Described judgement subelement, for when the boundary intensity of calculating current sub-block is 0, judges not need to carry out limit filtering to current sub-block; When the boundary intensity of calculating current sub-block is not 0, judge to need to carry out filtering to current sub-block.
Preferably, in described group of buffers each buffer load 4x4 block of pixels and all can interaction data.
Preferably, described dominant vector table comprises multiple basic controlling vector, described different dominant vector describes corresponding command information, the command information that described dominant vector describes comprises: the block type of the sub-block that filtering limit is adjacent, filtering direction, the numbering of the sub-block that filtering limit is adjacent, from the sub block number that first memory reads, from the sub block number that second memory reads, be written to the group number of buffer, the group number of read buffers, data interaction each source buffer group number and object buffer group number is carried out between read buffers group to the block number and group of buffers internal buffer of second memory,
Beneficial effect of the present invention comprises:
A kind of video image block-removal filtering method provided by the invention and device, wherein a video or picture frame are divided into several macro blocks of 16 × 16 by method, and each macro block is divided into the sub-block of 16 4 × 4.First, when reading current macro block pixels data, the pixel data of current macro is read in the group of buffers of specifying by the command information described by dominant vector from first memory and second memory; Then, the command information when performing the filtering process of each current sub-block described by dominant vector performs the determination of corresponding sub-block block number, the calculating of filtering strength, filtering type are determined, the operation of filtering operation process and write back data process.Its essence is that performing general filtering process by the command information execution described by dominant vector general in control table to different video encoding and decoding standard operates.
Wherein, the present invention is like this by dominant vector block elimination filtering processing method general in control table, because reasonably have adjusted the part filter order of sub-block, so reasonably decrease the quantity of group of buffers under guarantee does not affect the condition of block elimination filtering efficiency, improve filtering processing speed simultaneously, the present invention in view of the complexity of block elimination filtering and HD video encoding and decoding higher requirement is proposed to speed and encoded video form more, provide a kind of video image block-removal filtering method and device, it better tackles the complexity of block elimination filtering, versatility is good, the HD video encoding and decoding processing requirements of multiple format can be met.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of an embodiment of video image block-removal filtering method of the present invention;
Fig. 2 is macro block, sub-block and dot structure relation schematic diagram in video image block elimination filtering of the present invention;
Fig. 3 is the schematic diagram of group of buffers in the deblocking filtering apparatus of video image deblocking filtering apparatus one embodiment of the present invention;
Fig. 4 is that the block-removal filtering method of prior art adopts H.264 standard to the block elimination filtering order of the luminance pixel filtering of 16 × 16 macro blocks;
Fig. 5 is an embodiment of video image block-removal filtering method of the present invention, and the H.264 standard of employing is to the block elimination filtering order of the luminance pixel filtering of 16 × 16 macro blocks;
Fig. 6 is the structural representation of video image deblocking filtering apparatus one embodiment of the present invention.
Embodiment
Below in conjunction with Figure of description, the embodiment of video image block-removal filtering method of the present invention and device is described.
A kind of video image block-removal filtering method that the embodiment of the present invention provides, a video frame image is divided into several macro blocks of 16 × 16 by block-removal filtering method, and each macro block is divided into the sub-block of 16 4 × 4, as shown in Figure 1, specifically comprises the following steps:
Steps A, when reading current macro block pixels data, the pixel data of current macro is read in the group of buffers of specifying by the command information described by dominant vector from first memory and second memory;
Step B, when performing each sub-block filtering in current macro, the command information described by filtering dominant vector is determined to describe the data interaction information of current filter window and group of buffers and is determined the block number of current filter sub-block;
Step C, calculate the intensity of current filter sub-block according to data interaction information and judge whether to need to carry out filtering to current sub-block;
If step D judges to need to carry out filtering to current sub-block, then according to the filtering type of data interaction information determination current sub-block, and pass through the method for the corresponding filtering operation of filtering type determination current sub-block; Do not need to carry out filtering to current sub-block if judge, then directly perform step G;
Step e, according to the method for the filtering operation of the current sub-block determined limit filtering is carried out to current sub-block and the limit filtering operation result of current sub-block is kept in group of buffers;
Step F, command information described by dominant vector, to be written back to the operation result of current sub-block limit filtering in group of buffers in second memory and to terminate the filtering operation of current filter sub-block;
Step G, judge that whether current sub-block is last sub-block of filtering operation in current macro, if the determination result is YES, then perform following step H; If judged result is no, returns step B execution and the filtering process of next sub-block is operated;
The filtering process operation of step H, end current macro.
Preferably, as a kind of embodiment, also comprise before steps A and perform preparation data manipulation to be filtered, perform data manipulation to be filtered and specifically comprise the steps:
Step S1, the pixel data of current macro to be preserved in a first memory, and the pixel data of each sub-block on the top adjacent with current macro, the left side and limit, upper left is kept in second memory;
Step S2, transfer the command information described by dominant vector of control table in the 3rd memory.
Illustrate, as shown in Figure 2, the adjacent top (sub block number is 12,13,14,15) of current macro, each sub-block of the left side (sub block number is 3,7,11,15) and limit, upper left (sub block number is 15) correspondence.
Further, as a kind of embodiment, in above-mentioned steps C, specific implementation step is as follows:
C1, calculate the intensity of current filter sub-block according to data interaction information;
If the boundary intensity of the current sub-block that C2 calculates is 0, then judge not need to carry out limit filtering to current sub-block; If the boundary intensity of the current sub-block calculated is not 0, then judge to need to carry out limit filtering to current sub-block.
Preferably, the data interaction information of current filter window and group of buffers comprises current macro information, the adjacent pixel data information of the adjacent sub-blocks on current sub-block information and limit to be filtered.
Particularly, the command information described by dominant vector comprises current macro information, the adjacent pixel data information of the adjacent sub-blocks on current sub-block information and limit to be filtered; Wherein, above-mentioned information all correspond to dominant vector in control table.
Wherein, as a kind of embodiment, dominant vector table comprises multiple dominant vector, different dominant vector describes corresponding command information, the command information that dominant vector describes comprises: the block type of the sub-block that filtering limit is adjacent, filtering direction, the numbering of the sub-block that filtering limit is adjacent, from the sub block number that first memory reads, from the sub block number that second memory reads, be written to the group number of buffer, the group number of read buffers, data interaction each source buffer group number and object buffer group number etc. is carried out between read buffers group to the block number and group of buffers internal buffer of second memory, following table is dominant vector instruction card in control table, as shown in the table:
In above-mentioned table, the dominant vector of control table configures according to the coding and decoding video operation method of different-format, and is stored in the 3rd memory.Play wherein a kind of video format time, will first perform the command information described by dominant vector transferring control table in the 3rd memory.
In order to the structural relation of macro block, sub-block and each pixel is described better, as shown in Figure 2.Wherein, indicate the label that 0 ~ 15 is each sub-block working as pre-treatment interior macroblocks, in addition, when configuring pixel data to be filtered, the pixel data of each sub-block of the adjacent top of current macro (sub block number is 12,13,14,15), the left side (sub block number is 3,7,11,15) and limit, upper left (sub block number is 15) also to be configured.
Preferably, in above-mentioned control table, the command information described by the dominant vector of control table is the decomposition control procedure of general-purpose operation method in different video encoding and decoding standard.Although various video deblocking filtering method is not quite similar, all can simply be decomposed into following universal process:
The block number being read into buffer from first memory and second memory is determined according to control table.
The block number of this filtering sub-block is determined according to control table.
According to command informations such as the correlator block types that control table provides and by the filtered pixel determination filtering strength in above-mentioned information and filtering type.
Filtering is carried out according to standard agreement operation method.
The position in second memory is written back to according to control table determination filter result.
Judge that whether current sub-block is last sub-block of filtering operation in current macro, if the determination result is YES, terminate the filtering process operation of whole macro block; If judged result is no, then continue to perform to operate the filtering process of next sub-block.
After general step is determined, various video deblocking filtering method controls the dominant vector be just converted into based on framework in control table, and because different video operation method is different, corresponding dominant vector is also different.Meanwhile, a kind of video deblocking filtering method adopts different dominant vectors, and filtration efficiency is not identical yet.The key of the formation embodiment of the present invention of dominant vector in control table.Above-mentioned control table is applicable to different video encoding and decoding standards, and the command information described by the dominant vector of control table is the decomposition control procedure of general-purpose operation method in different video encoding and decoding standard.
Dominant vector table provided by the invention, follows following 3 principles in for different coding and decoding video block elimination filtering processes simultaneously; 3 principles specifically comprise following content:
Principle one, strictly follows each pixel filter sequencing in standard agreement, on this basis, and adjustable monoblock limit filter sequence;
Principle two, each pixel is is only read and write once, the quantity of reasonable arrangement group of buffers device;
Principle three, ensure that each clock cycle processes a limit, filter arithmetic device is not idle, until macro block filtering terminates.
Removing filtering strength judges and the time delay of filtering operation unit flowing water, and arithmetic element each clock cycle, namely arithmetic element was not idle in the row or column of process different edge.This preparation treating filtering data proposes higher requirement.
Wherein, second memory has data access frequently, must have buffer to preserve the result treated operational data or calculate; In embodiments of the present invention, arranging group of buffers has 5 ~ 10 groups, and often organize storage 16 pixels, i.e. a block, see Fig. 3.Wherein two groups are respectively the main window of filtering, and all group of buffers can data interaction; Determine the quantity of group of buffers.In order to make each clock cycle process a limit, group of buffers will improve abundant, instant data encasement, so its quantity is decided to be between 5 ~ 10 groups.
Ensure block elimination filtering order condition under, neatly arrange part filter order with ensures filtering data read-only once with write once, to reduce the operation to second memory; Ensure the preparation of instant filtering data simultaneously, avoid filtering operation unit idle, waste the time.Different video formats has different filter sequence regulations, and the design of part filter order is also different; And the design of filter sequence seems the very large degree of freedom;
Under its tangible 3 principle design constraints above, the degree of freedom of part filter sequence design is in fact very little.These 3 are constrained to: the number of group of buffers; Filtering data can only be read and write once; Filtering data prepares immediately, and arithmetic element is not idle.
For example, below for H.264 luminance pixel, resolve how for coding and decoding video designs block elimination filtering partial order.H.264 video compression standard of new generation has jointly been formulated for IEO/IEC/ and ITU-T Liang great International Organization for standardization.In Fig. 4 prior art, block-removal filtering method adopts the block elimination filtering order of H.264 standard, wherein heavy line represents macroblock boundaries, as shown in Figure 4, in figure, 32 luminance pixel limits are numbered, the order of numbering determines the order of filtering, and carried out numbering (number order is 0 ~ 15) to the sub-block of 16 4 × 4 by the numeral of runic with underscore, in the process of block elimination filtering, for example, to each sub-block (0 of first row, 4, 8, 12) when carrying out filtering, and respectively to 0, 1, 2, the luminance pixel of No. 3 marks carries out while filtering, and operation result is temporarily stored in group of buffers, and then be kept in second memory, to No. 4 mark luminance pixel carry out while filtering time, because No. 4 luminance pixel limits are boundary pixels of sub-block 0 and sub-block 1, so at this moment again read 1 work song block again from second memory when filtering No. 4 luminance pixel limits, can repeat like this to read second memory, simultaneously, such filter sequence needs a large amount of buffers, also add filtering data read-write number of times simultaneously, like this, have also increased dramatically design cost and the time of process needed for a macro block.
Still be filtered into example H.264 to process luminance pixel below, but, when the block-removal filtering method that the embodiment of the present invention provides carries out filtering, after the part filter order adjustment carried out, after meeting above-mentioned 3 principles mentioned, substantially reduce the time needed for process macro block, Fig. 5 for block-removal filtering method that the embodiment of the present invention provides adopt H.264 standard to the block elimination filtering order of the luminance pixel filtering of macro block, as shown in Figure 5, in figure, 32 luminance pixel limits are numbered, the order of numbering determines the order of filtering, obviously, filter sequence shown in Fig. 5 is different from the filter sequence of prior art, see Fig. 5, carry out numbering (number order is 0 ~ 15) to the sub-block of 16 4 × 4 by the numeral of runic with underscore.The block-removal filtering method that the embodiment of the present invention adopts is when iotaltered luma pixel limit, first by each luminance pixel limit filtering (0 of current sub-block, 1, 2) and preserve each luminance pixel limit (3 of the sub-block 1 that filtering is adjacent with current sub-block 0 more in a buffer, 4) and preserve in buffer, to be filtered to sub-block 4 time, because sub-block 0 is adjacent with sub-block 4, at this moment the filtering of sub-block 4 needs to obtain pixel data from sub-block 0, at this moment sub-block 0 pair of sub-block 4 has demand, therefore sub-block 0 is temporarily also kept in group of buffers and (operation result of sub-block 0 wouldn't be written back in second memory), after the complete sub-block 4 of Order Filtering, judge to draw in current macro there is no (sub-block 5-15) in each sub-block of filtering, not adjacent with sub-block 0 sub-block, at this moment sub-block 0 is non-conterminous to follow-up filtering sub-block (sub-block 5-15), just do not need from buffer, to transfer sub-block 0 more yet, at this moment the luminance pixel limit filtering operation result of sub-block in buffer 0 is saved in second memory, this avoid the pixel data repeating to transfer sub-block 0 when the sub-block that filtering is adjacent with sub-block 0, decrease the preparation quantity of group of buffers simultaneously, improve filtering processing speed, although the filter sequence of amendment specifies different from agreement, but concerning each filtered pixel data, within the scope of filter action, its filtering sequencing does not change.Reasonably have adjusted the part filter order of sub-block, so reasonably decrease the quantity of group of buffers under guarantee does not affect the condition of block elimination filtering efficiency, improve filtering processing speed.
Preferably, different video methods has different dominant vectors, and because each macro block has the limit filtering of 32+16 bar, wherein, 32 luminance pixel limits, 16 chroma pixel limits, have come so each macro block has a lot of step.These steps constitute the command information described by the dominant vector of dominant vector table, and above-mentioned control table is placed in the 3rd memory.
The core of the deblocking filtering apparatus framework of the embodiment of the present invention is in the design of control table and group of buffers.For different video form, its filter sequence, filtering operation method etc. are all different, even differ larger.But to the deblocking filtering apparatus of the embodiment of the present invention, as long as read general control table, and perform filtering operation by the command information described by dominant vector in control table, just the block elimination filtering for various criterion can be realized, it is simultaneously by the design of less buffer, accelerate filtration efficiency greatly, the realization of the full HD video decode of its very convenient full format.
The dominant vector of control table needs under the condition strictly following each pixel filter sequencing in standard agreement when controlling block elimination filtering, adjusts the method for sub-block limit filter sequence and prefetch data dexterously.
Based on same inventive concept, the embodiment of the present invention additionally provides a kind of video image deblocking filtering apparatus, the principle of dealing with problems due to this device is similar to aforementioned a kind of video image block-removal filtering method, and therefore the enforcement of this device see the enforcement of preceding method, can repeat part and repeat no more.
A kind of video image deblocking filtering apparatus that the embodiment of the present invention provides, comprise first memory 1, second memory 2, bank of filters 3, group of buffers 4, the 3rd memory 5, reading unit 601, determining unit 602, filtering strength judging unit 603, filtering operation unit 604, processing unit 605, judging unit 606, end unit 607, integrated treatment unit 608 and dominant vector table 609, as shown in Figure 6, wherein:
First memory 1, for preserving the pixel data of current filter macro block;
Second memory 2, for preserving the pixel data of each sub-block on the top adjacent with current macro, the left side and limit, upper left; And preserve the operation result read in group of buffers;
Group of buffers 4, for preserving the pixel data read in from first memory and second memory;
Reading unit 601, for when reading current macro block pixels data, the pixel data of current macro is read into the group of buffers of specifying by the command information described by dominant vector from first memory and second memory;
Determining unit 602, for when performing each sub-block filtering in current macro, the command information described by filtering dominant vector is determined to describe the data interaction information of current filter window and group of buffers and is determined the block number of current filter sub-block;
Filtering strength judging unit 603, for calculating the intensity of current filter sub-block according to data interaction information and judging whether to need to carry out filtering to current sub-block; Need if judge, then the filtering type of the data interaction information determination current sub-block in the command information described by dominant vector, and the method passing through the corresponding filtering operation of filtering type determination current sub-block; Do not need if judge, then directly return judging unit and perform corresponding operation;
Filtering operation unit 604, carries out limit filtering for the method for the filtering operation according to the current sub-block determined to current sub-block and is kept in group of buffers by the limit filtering operation result of current sub-block;
Bank of filters 3, for carrying out limit filtering to current filter sub-block;
Processing unit 605, for the command information described by dominant vector, to be written back to the operation result of current sub-block limit filtering in group of buffers in second memory and to terminate the filtering operation of current filter sub-block;
Judging unit 606, for judging that whether current sub-block is last sub-block of filtering operation in current macro, if the determination result is YES, then redirect end unit performs the operation terminating current macro filtering process; If judged result is no, then return the corresponding filtering process operation that determining unit performs next sub-block;
3rd memory 5, for preserving control table 609;
End unit 607, for terminating the filtering process operation of current macro;
Integrated treatment unit 608, for the command information described by dominant vector, coordinates each unit and performs corresponding operation.
Preferably, as a kind of embodiment, deblocking filtering apparatus also comprises control unit 610;
Control unit 610, for transfer control table in the 3rd memory dominant vector described by command information.
Further, as a kind of embodiment, filtering strength judging unit 603 specifically comprises computation subunit and judges subelement, wherein:
Computation subunit, for calculating the intensity of current filter sub-block according to data interaction information;
Judge subelement, for when the boundary intensity of calculating current sub-block is 0, judge not need to carry out limit filtering to current sub-block; When the boundary intensity of calculating current sub-block is not 0, judge to need to carry out filtering to current sub-block.
Preferably, in group of buffers each buffer load 4x4 block of pixels and all can interaction data.
Preferably, dominant vector table comprises multiple basic controlling vector, different dominant vector describes corresponding command information, the command information that dominant vector describes comprises: the block type of the sub-block that filtering limit is adjacent, filtering direction, the numbering of the sub-block that filtering limit is adjacent, from the sub block number that first memory reads, from the sub block number that second memory reads, be written to the group number of buffer, the group number of read buffers, data interaction each source buffer group number and object buffer group number is carried out between read buffers group to the block number and group of buffers internal buffer of second memory.
Preferably, control table is applicable to different video encoding and decoding standards, and the command information described by the dominant vector of control table is the decomposition control procedure of general-purpose operation method in different video encoding and decoding standard.
The video image deblocking filtering apparatus that the embodiment of the present invention provides, it can tackle the complexity of block elimination filtering preferably, possesses the beneficial effects such as speed is fast, versatility is good, can meet the HD video encoding and decoding process of various video form.
The video image deblocking filtering apparatus that the embodiment of the present invention provides, structurally main feature is the design of group of buffers and control table.Deblocking filtering apparatus adopts flowing water design, and processing speed is fast, and versatility is good.
In the video image deblocking filtering apparatus that the embodiment of the present invention provides, unit function realizes by computer program.Those skilled in the art should be understood that; above-mentioned dividing elements mode is only the one during plurality of cells divides; if be divided into other unit or without inner cells, as long as deblocking filtering apparatus has above-mentioned functions, all should within protection scope of the present invention.
A kind of video image block-removal filtering method provided by the invention and device, wherein a video frame image is divided into several macro blocks of 16 × 16 by method, and each macro block is divided into the sub-block of 16 4 × 4.First, when reading current macro block pixels data, the pixel data of current macro is read in the group of buffers of specifying by the command information described by dominant vector from first memory and second memory; Then, the command information when performing the filtering process of each current sub-block described by dominant vector performs the determination of corresponding sub-block block number, the calculating of filtering strength, filtering type are determined, the operation of filtering operation process and write back data process.Its essence is that performing general filtering process by the command information execution described by dominant vector general in control table to different video encoding and decoding standard operates.
Wherein, the present invention is like this by dominant vector block elimination filtering processing method general in control table, because reasonably have adjusted the part filter order of sub-block, so reasonably decrease the quantity of group of buffers under guarantee does not affect the condition of block elimination filtering efficiency, improve filtering processing speed simultaneously, the present invention in view of the complexity of block elimination filtering and HD video encoding and decoding higher requirement is proposed to speed and encoded video form more, provide a kind of video image deblocking filtering apparatus, it better tackles the complexity of block elimination filtering, versatility is good, the HD video encoding and decoding processing requirements of multiple format can be met.
The above embodiment only have expressed several execution mode of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection range of patent of the present invention should be as the criterion with claims.

Claims (7)

1. a video image block-removal filtering method, a video frame image is divided into several macro blocks of 16 × 16, and each macro block is divided into the sub-block of 16 4 × 4, it is characterized in that, comprises the following steps:
Step S1, the pixel data of current macro to be preserved in a first memory, and the pixel data of each sub-block on the top adjacent with current macro, the left side and limit, upper left is kept in second memory;
Step S2, transfer the command information described by dominant vector of control table;
Steps A, when reading current macro block pixels data, the pixel data of described current macro is read in the group of buffers of specifying by the command information described by described dominant vector from first memory and second memory;
Step B, when performing each sub-block filtering in current macro, the command information described by filtering dominant vector is determined to describe the data interaction information of current filter window and group of buffers and is determined the block number of current filter sub-block;
Step C, calculate the intensity of current filter sub-block according to described data interaction information and judge whether to need to carry out filtering to current sub-block;
If step D judges to need to carry out filtering to current sub-block, then according to the filtering type of described data interaction information determination current sub-block, and pass through the method for the corresponding filtering operation of filtering type determination current sub-block; Do not need to carry out filtering to current sub-block if judge, then directly perform step G;
Step e, according to the method for the filtering operation of the current sub-block determined limit filtering is carried out to current sub-block and the limit filtering operation result of current sub-block is kept in group of buffers;
Step F, command information described by dominant vector, to be written back to the described operation result of current sub-block limit filtering described in group of buffers in second memory and to terminate the filtering operation of current filter sub-block;
Step G, judge that whether current sub-block is last sub-block of filtering operation in current macro, if the determination result is YES, then perform following step H; If judged result is no, returns step B execution and the filtering process of next sub-block is operated;
The filtering process operation of step H, end current macro;
Wherein, described control table is applicable to different video encoding and decoding standards, and the command information described by the dominant vector of described control table is the decomposition control procedure of general-purpose operation method in different video encoding and decoding standard.
2. video image block-removal filtering method according to claim 1, is characterized in that, in described step C, specific implementation step is as follows:
C1, calculate the intensity of current filter sub-block according to described data interaction information;
If the boundary intensity of the current sub-block that C2 calculates is 0, then judge not need to carry out limit filtering to current sub-block; If the boundary intensity of the current sub-block calculated is not 0, then judge to need to carry out limit filtering to current sub-block.
3. video image block-removal filtering method according to claim 1, it is characterized in that, the data interaction information of described current filter window and group of buffers comprises current macro information, the adjacent pixel data information of the adjacent sub-blocks on current sub-block information and limit to be filtered.
4. a video image deblocking filtering apparatus, it is characterized in that, comprise first memory, second memory, bank of filters, group of buffers, the 3rd memory, reading unit, determining unit, filtering strength judging unit, filtering operation unit, processing unit, judging unit, end unit, integrated treatment unit and dominant vector table, wherein:
Described first memory, for preserving the pixel data of current filter macro block;
Described second memory, for preserving the pixel data of each sub-block on the top adjacent with current macro, the left side and limit, upper left; And preserve the operation result read in group of buffers;
Described group of buffers, for preserving the pixel data read in from first memory and second memory;
Described reading unit, for when reading current macro block pixels data, the pixel data of described current macro is read into the group of buffers of specifying by the command information described by dominant vector from first memory and second memory;
Described determining unit, for when performing each sub-block filtering in current macro, the command information described by filtering dominant vector is determined to describe the data interaction information of current filter window and group of buffers and is determined the block number of current filter sub-block;
Described filtering strength judging unit, for calculating the intensity of current filter sub-block according to described data interaction information and judging whether to need to carry out filtering to current sub-block; Need if judge, then the filtering type of the described data interaction information determination current sub-block in the command information described by described dominant vector, and the method passing through the corresponding filtering operation of filtering type determination current sub-block; Do not need if judge, then directly return judging unit and perform corresponding operation;
Described filtering operation unit, carries out limit filtering for the method for the filtering operation according to the current sub-block determined to current sub-block and is kept in group of buffers by the limit filtering operation result of current sub-block;
Described bank of filters, for carrying out limit filtering to current filter sub-block;
Described processing unit, for the command information described by dominant vector, to be written back to the described operation result of current sub-block limit filtering described in group of buffers in second memory and to terminate the filtering operation of current filter sub-block;
Described judging unit, for judging that whether current sub-block is last sub-block of filtering operation in current macro, if the determination result is YES, then redirect end unit performs the operation terminating current macro filtering process; If judged result is no, then return the corresponding filtering process operation that determining unit performs next sub-block;
Described 3rd memory, for preserving control table;
Described end unit, for terminating the filtering process operation of current macro;
Described integrated treatment unit, for the command information described by described dominant vector, coordinates each unit and performs corresponding operation;
Also comprise control unit, for transfer control table in described 3rd memory dominant vector described by command information;
Wherein, described control table is applicable to different video encoding and decoding standards, and the command information described by the dominant vector of described control table is the decomposition control procedure of general-purpose operation method in different video encoding and decoding standard.
5. video image deblocking filtering apparatus according to claim 4, is characterized in that, described filtering strength judging unit comprises computation subunit and judges subelement, wherein:
Described computation subunit, for calculating the intensity of current filter sub-block according to described data interaction information;
Described judgement subelement, for when the boundary intensity of calculating current sub-block is 0, judges not need to carry out limit filtering to current sub-block; When the boundary intensity of calculating current sub-block is not 0, judge to need to carry out filtering to current sub-block.
6. video image deblocking filtering apparatus according to claim 4, is characterized in that, in described group of buffers each buffer load 4x4 block of pixels and all can interaction data.
7. video image deblocking filtering apparatus according to claim 4, it is characterized in that, described dominant vector table comprises multiple basic controlling vector, described different dominant vector describes corresponding command information, the command information that described dominant vector describes comprises: the block type of the sub-block that filtering limit is adjacent, filtering direction, the numbering of the sub-block that filtering limit is adjacent, from the sub block number that first memory reads, from the sub block number that second memory reads, be written to the group number of buffer, the group number of read buffers, data interaction each source buffer group number and object buffer group number is carried out between read buffers group to the block number and group of buffers internal buffer of second memory.
CN201210236491.0A 2012-07-09 2012-07-09 Video image deblocking filter method and device Active CN102801973B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201210236491.0A CN102801973B (en) 2012-07-09 2012-07-09 Video image deblocking filter method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210236491.0A CN102801973B (en) 2012-07-09 2012-07-09 Video image deblocking filter method and device

Publications (2)

Publication Number Publication Date
CN102801973A CN102801973A (en) 2012-11-28
CN102801973B true CN102801973B (en) 2015-03-25

Family

ID=47200928

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210236491.0A Active CN102801973B (en) 2012-07-09 2012-07-09 Video image deblocking filter method and device

Country Status (1)

Country Link
CN (1) CN102801973B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105828071B (en) * 2016-03-31 2019-05-24 中国人民解放军国防科学技术大学 The deblocking filtering vectorization implementation method of vector processor-oriented
CN108574471B (en) * 2017-03-14 2021-11-23 珠海全志科技股份有限公司 Fully integrated harmonic filter for radio frequency power amplifying circuit
CN110213579A (en) * 2019-06-11 2019-09-06 上海富瀚微电子股份有限公司 The method of double-core computing unit realization loop filtering

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1921625A (en) * 2006-09-15 2007-02-28 清华大学 Filtering method and system of deblocking effect filter
CN101170701A (en) * 2007-11-16 2008-04-30 四川虹微技术有限公司 Block elimination filtering method and device for video encoding and decoding system
CN101472172A (en) * 2007-12-29 2009-07-01 深圳安凯微电子技术有限公司 Method for processing filtering previous data of video decode as well as decoder
WO2009123033A1 (en) * 2008-03-31 2009-10-08 日本電気株式会社 Deblocking filter processing device and deblocking filter processing method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5793511B2 (en) * 2010-02-05 2015-10-14 テレフオンアクチーボラゲット エル エム エリクソン(パブル) Deblocking filtering control

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1921625A (en) * 2006-09-15 2007-02-28 清华大学 Filtering method and system of deblocking effect filter
CN101170701A (en) * 2007-11-16 2008-04-30 四川虹微技术有限公司 Block elimination filtering method and device for video encoding and decoding system
CN101472172A (en) * 2007-12-29 2009-07-01 深圳安凯微电子技术有限公司 Method for processing filtering previous data of video decode as well as decoder
WO2009123033A1 (en) * 2008-03-31 2009-10-08 日本電気株式会社 Deblocking filter processing device and deblocking filter processing method

Also Published As

Publication number Publication date
CN102801973A (en) 2012-11-28

Similar Documents

Publication Publication Date Title
JP2006174486A (en) Deblocking filter for simultaneously performing horizontal and vertical filtering of video data and method of operating the same
CN103051892B (en) Embedded loop filter method and embedded loop filter
CN101330617B (en) Hardware implementing method and apparatus for anticipater within multi-standard frame based on mode mapping
CN111654710B (en) Image filtering method, device, equipment and storage medium
CN101583041A (en) Image filtering method of multi-core image encoding processing equipment and equipment
CN105791831A (en) Method and apparatus for collaborative partition coding for region based filters
CN100493196C (en) A method for quickly computing intensity of block removal filtering boundary in H.264 video standard
CN104025594A (en) Tile size in video coding
CN102801973B (en) Video image deblocking filter method and device
CN101115207B (en) Method and device for implementing interframe forecast based on relativity between future positions
CN1874516A (en) Implementation device in VLSI of filter for removing blocking effect based on AVS
CN101321290A (en) Block-removing filtering method based on digital signal processor
CN1306826C (en) Loop filter based on multistage parallel pipeline mode
CN102055981A (en) Deblocking filter for video coder and implementation method thereof
CN101635849B (en) Loop filtering method and loop filter
CN101123725A (en) Hardware implementation method for H.264 block elimination effect filter
CN104253998A (en) Hardware on-chip storage method of deblocking effect filter applying to HEVC (High Efficiency Video Coding) standard
CN102572430A (en) Method for implementing H.264 deblocking filter algorithm based on reconfigurable technique
CN101252691A (en) Method for multi-processor to parallel implement high-definition picture filtering
CN102685497A (en) Rapid interframe mode selection method and device for AVS (Advanced Audio Video Coding Standard) coder
KR100729241B1 (en) Deblocking Filter
CN101459839A (en) Deblocking effect filtering method and apparatus for implementing the method
CN100592799C (en) Rapid reading method of motion compensating data based on H.264 standard
CN101710994B (en) Method and system for video decoding
CN102075753B (en) Method for deblocking filtration in video coding and decoding

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant