CN106303535A - Reference pixel takes from method for compressing image and the device of reconstructed pixel in various degree - Google Patents
Reference pixel takes from method for compressing image and the device of reconstructed pixel in various degree Download PDFInfo
- Publication number
- CN106303535A CN106303535A CN201510308137.8A CN201510308137A CN106303535A CN 106303535 A CN106303535 A CN 106303535A CN 201510308137 A CN201510308137 A CN 201510308137A CN 106303535 A CN106303535 A CN 106303535A
- Authority
- CN
- China
- Prior art keywords
- pixel
- decoding
- block
- coding
- reconstructed
- 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.)
- Granted
Links
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention provides the method or apparatus of picture coding and decoding that a kind of reference pixel takes from the reconstructed pixel of at least two difference degree of perfection.Preferably, in the present invention, Part I reference pixel is the reconstructed pixel of degree of perfection I, and Part II reference pixel is the reconstructed pixel of degree of perfection II, and Part III reference pixel is the reconstructed pixel of degree of perfection III, etc..Preferably, the reconstructed pixel of different degree of perfections takes from the zones of different of image.Preferably, the reference pixel of a reference pixel sample value section is made up of the reconstructed pixel of at least two difference degree of perfection.According to the present invention, the reference pixel scope of copy mode is divided at least two parts and has the zones of different of different degree of perfection reconstructed pixel, and the reference pixel of the reference pixel sample value section corresponding to a current pixel sample value section is taken from least two parts and had the zones of different of different degree of perfection reconstructed pixel.
Description
Technical field
The present invention relates to a kind of compression of digital video encoding and decoding system, particularly contain the combination picture of computer screen image and the method and apparatus of the encoding and decoding of video.
Background technology
Along with television set and display enter ultra high-definition (4K) and the resolution of special ultra high-definition (8K), and the cloud computing of new generation with remote desktop as classic manifestations and messaging model and the development of platform and universal, the demand of video image data compression is also moved towards higher resolution and comprised the combination picture of camera pickuping image and computer screen image.The data compression that video image carries out superelevation compression ratio and very high-quality becomes requisite technology.
Make full use of the feature of 4K/8K image and computer screen image, video image is carried out the compression of Ultra-High Efficiency, is also the up-to-date international video compression standards HEVC(High Efficiency Video Coding formulated) and some other international standards, domestic standard, a main target of industry standard.
The natural form of the digital video signal of image is the sequence of image.One two field picture is typically the rectangular area being made up of some pixels, and digital video signal is exactly the sequence of video images being made up of tens frames to thousands of two field pictures, the most also referred to as video sequence or sequence.Image one by one is encoded by numeral encoding video signal exactly.At any one time, that two field picture encoded is referred to as current encoded image.Equally, it is decoded exactly the code stream compressing image one by one being decoded to the video code flow (be called for short code stream and be also referred to as bit stream) after the compression of digital video signal.At any one time, that two field picture decoded is referred to as currently decoding image.Current encoded image or current decoding image are all referred to as present image.
International standard such as MPEG-1/2/4 at nearly all encoding video pictures, H.264/AVC and in HEVC, when one two field picture is encoded (and decoding accordingly), one two field picture is divided into the subimage of some pieces of MxM pixels, it is referred to as encoding block and (namely solves code block from the angle of decoding, it is referred to as encoding and decoding block) or " coding unit (Coding Unit is called for short CU) ", with CU for basic coding unit, subimage one piece a piece is encoded.The size of conventional M is 4,8,16,32,64.Therefore, a sequence of video images carries out coding exactly each coding unit i.e. CU CU the most one by one of each two field picture to be encoded.At any one time, the CU encoded is referred to as present encoding CU.Equally, it is decoded also being that each CU CU the most one by one to each two field picture is decoded to the code stream of a sequence of video images, finally reconstructs whole sequence of video images.At any one time, the CU decoded is referred to as currently decoding CU.Present encoding CU or current decoding CU are referred to as current CU.
Different from character for each several part picture material in adapting to a two field picture, carry out maximally effective coding targetedly, in a two field picture, the size of each CU can be different, has plenty of 8x8, has plenty of 64x64, etc..In order to enable different size of CU seamless spliced, one two field picture is the most first divided into identical " the maximum coding unit (Largest Coding Unit is called for short LCU) " with NxN pixel of size, and the most each LCU is further divided into the multiple not of uniform size fixed identical CU of tree.Therefore, LCU is also referred to as " code tree unit (Coding Tree Unit is called for short CTU) ".Such as, a two field picture is first divided into the LCU(N=64 of size identical 64x64 pixel).Wherein certain LCU is made up of the CU of 3 32x32 pixels and the CU of 4 16x16 pixels, and such 7 CU becoming tree constitute a CTU.And another LCU is made up of the CU of the CU of 2 32x32 pixels, the CU of 3 16x16 pixels and 20 8x8 pixels.Such 25 CU becoming tree constitute another CTU.One two field picture is encoded, it is simply that successively the CU one by one in CTU one by one is encoded.In HEVC international standard, LCU Yu CTU is synonym.The size CU referred to as degree of depth equal to CTU is the CU of 0.The degree of depth be the CU referred to as degree of depth that obtains of the CU quartering up and down of 0 be the CU of 1.The degree of depth be the CU referred to as degree of depth that obtains of the CU quartering up and down of 1 be the CU of 2.The degree of depth be the CU referred to as degree of depth that obtains of the CU quartering up and down of 2 be the CU of 3.At any one time, the CTU encoded is referred to as present encoding CTU.At any one time, the CTU decoded is referred to as currently decoding CTU.Present encoding CTU or current decoding CTU are referred to as current CTU.
CU can also further be divided into some subregions.Subregion includes but not limited to predicting unit (PU), converter unit (TU), the region of asymmetric division (AMP).
One colour element generally has 3 component (component) compositions.GBR color format that the most frequently used two kinds of pixel color forms (pixel color format) are made up of green component, blue component, red component and the YUV color format being made up of brightness (luma) component and two colourity (chroma) components.It is commonly referred to as that the color format of YUV is actual includes multicolour form, such as YCbCr color format.Therefore, when a CU is encoded, a CU can be divided into 3 component planes (G plane, B plane, R plane or Y plane, U plane, V plane), 3 component planes are encoded respectively;The 3 of one pixel component binding combination can also be become 3 tuples, the CU entirety being made up of these 3 tuples is encoded.The arrangement mode of former pixel and component thereof is referred to as the planar format (planar of image (and CU)
Format), and the arrangement mode of latter pixel and component thereof is referred to as the folded bag form (packed format) of image (and CU).The GBR color format of pixel and YUV color format are all 3 component representation formats of pixel.
Except 3 component representation formats of pixel, the representation formats of the prior art that the another kind of pixel is conventional is that palette indexes representation formats.In palette index representation formats, the numerical value of a pixel can also show with the index of palette.Storing numerical value or the numerical approximation of 3 components of the color of the pixel that needs are showed in palette space, the address of palette is referred to as in this address the index of the color of the pixel of storage.One index can show the one-component of the color of pixel, and index can also show 3 components of the color of pixel.Palette can be one, it is also possible to is multiple.In the situation of multiple palettes, complete index is actually made up of index two parts of palette sequence number (represent in multiple palette which) and the palette of this sequence number.The index representation formats of pixel is exactly that index of reference is to show this pixel.If the pixel in an image-region (such as encoding block or solution code block) can not all show (i.e. at least one pixel in this image-region by palette of colors, there is no palette of colors and index thereof that the numerical value of 3 components is equal or approximately equal with this pixel), palette then generally has a special index being referred to as escape color, is used for showing the pixel that can not show by normal palette of colors.Therefore, if the index of a pixel is the index of escape color, then this pixel needs to show its color with other special 3 component.Normal color and escape color in palette are referred to as palette of colors, but escape color is a false color, and in palette, a physical space does not deposits this color, the special virtual index that only one of which is special.The index of escape color is typically last index of palette.The index representation formats of pixel is also referred to as the indexed color (indexed of pixel in the prior art
Or imitative color (pseudo color) representation formats, or be usually directly referred to as indexing pixel (indexed pixel) or imitative pixel (pseudo pixel) or pixel index or index color).Index is also sometimes referred to as index.Pixel is showed also referred to as indexation or indexation by its index representation formats.
The pixel performance form of other conventional prior art includes CMYK representation formats and expressing gradation form.
YUV color format again can be according to whether carry out down-sampling to be subdivided into several subformat to chromatic component: the YUV4:4:4 pixel color form that 1 pixel is made up of 1 Y-component, 1 U component, 1 V component;The YUV4:2:2 pixel color form that adjacent 2 pixels in left and right are made up of 2 Y-component, 1 U component, 1 V component;The neighbouring YUV4:2:0 pixel color form being made up of 4 Y-component, 1 U component, 1 V component by 4 pixels of 2x2 spatial arrangements in left and right.One-component typically represents by the numeral of 18 ~ 16 bit.YUV4:2:2 pixel color form and YUV4:2:0 pixel color form are all that the down-sampling that YUV4:4:4 pixel color form is implemented chromatic component obtains.One pixel component is also referred to as a pixel sample value (pixel
Or be simply referred as a sample value (sample) sample).
Most basic element when coding or decoding can be a pixel, it is also possible to be a pixel component, it is also possible to be a pixel index (i.e. indexing pixel).It is referred to as a pixel sample value as coding or the pixel of most basic element of decoding or a pixel component or an index pixel, is the most also commonly referred to as a pixel value, or is simply referred as a sample value.
In the present invention and present patent application, " pixel sample value ", " pixel value ", " sample value ", " index pixel ", " pixel index " are synonyms, based on context, can clearly represent " pixel " or represent " pixel component " or represent " index pixel " or represent that three's is arbitrary simultaneously.If can not be clear and definite from context, then be exactly to represent that three's is arbitrary simultaneously.
In the present invention and present patent application, the region that encoding block or solution code block (being referred to as encoding and decoding block) are made up of some pixel values.The shape of encoding and decoding block can be rectangle, square, parallelogram, trapezoidal, polygon, circle, ellipse and other are variously-shaped.Rectangle also includes width or the rectangle deteriorating to line (i.e. line segment or linear) that height is a pixel value.In one two field picture, each encoding and decoding block can have different shapes and sizes.In one two field picture, some or all encoding and decoding block can overlap each other part, it is also possible to all encoding and decoding block all non-overlapping copies.One encoding and decoding block, can be made up of " pixel ", it is also possible to be made up of " component of pixel ", it is also possible to be made up of " index pixel ", it is also possible to be made up of this mixing, it is also possible to be made up of any 2 kinds of mixing in this.From encoding video pictures or the angle of decoding, encoding and decoding block refers to, in a two field picture, it is implemented coding or a region of decoding, include but not limited to following at least one: maximum coding unit LCU, code tree unit CTU, the subregion of coding unit CU, CU, predicting unit PU, converter unit TU.
One distinguishing feature of computer screen image is generally to have the most identical the most similar pixel pattern (pixel in same two field picture
Pattern).Such as, the Chinese often occurred in computer screen image or foreign language, it is all to be made up of a few basic strokes, the most similar or identical stroke in same two field picture, can be found.Menu common in computer screen image, icon etc., it may have the most similar or identical pattern.Therefore, existing image and video compression technology generally use various copy modes, at least include following copy mode:
1)
The i.e. intra block of intra block duplication mates or claims intra frame motion compensation or claims Block-matching or claim block to replicate.The elementary operation of block replica code or decoding is to a present encoding block or currently to solve code block (referred to as current block), from the reference block replicating onesize with current block the number of pixel sample value (as) in reconstructed reference pixel sample value collection, and the numerical value assignment of described reference block is given current block.The duplication parameter of block copy mode includes the displacement vector of current block, represents the relative position between reference block and current block.One current block has a displacement vector.
2)
In frame, in microlith duplication i.e. frame, microlith coupling or title microlith coupling or title microlith replicate.In microlith replicates, one current block (such as 8x8 pixel sample value) is divided into several microlith (such as the microlith of 4x2 pixel sample value or the microlith of 8x2 pixel sample value or the microlith of 2x4 pixel sample value or the microlith of 2x8 pixel sample value), the elementary operation of microlith replica code or decoding is to each the coding microlith in current block or decoding microlith (referred to as current microlith), in reconstructed reference pixel sample value collection, replicate a reference microlith, and the described numerical value assignment with reference to microlith is given current microlith.The duplication parameter of microlith copy mode includes the displacement vector of current microlith, represents with reference to the relative position between microlith with current microlith.One current microlith has a displacement vector.One current block is divided into how many microliths just how many displacement vectors.
3)
In frame, in lines (abbreviation bar) duplication i.e. frame, bar coupling or title bar coupling or title bar replicate.Bar is that height is 1 or width is the microlith of 1, such as 4x1 or 8x1 or the microlith of 1x4 or 1x8 pixel sample value.The elementary operation of bar replica code or decoding is to each encoding strip in current block or to solve code-bar (referred to as working as preceding article), replicates a reference bar, and the numerical value assignment of described reference bar given and work as preceding article in reconstructed reference pixel sample value collection.Obviously, bar duplication is a kind of special circumstances that microlith replicates.The duplication parameter of bar copy mode includes the displacement vector when preceding article, represents reference bar and the relative position between preceding article.One has a displacement vector when preceding article.One current block is divided into how many bars just how many displacement vectors.
4)
In frame, in string duplication i.e. frame, String matching or title String matching or title string replicate or claim the duplication of pixel string.In pixel string replicates, current to a present encoding block or one solution code block (referred to as current block) is divided into several adjustable length pixel sample value string.Here string refers to the pixel sample value in the 2 dimensional region of an arbitrary shape to be arranged in the length string (as width is 1 pixel sample value and the string of a length of 37 pixel sample values or width are 2 pixel sample values and the string of a length of 111 pixel sample values, be typically but not limited to that length is an absolute coding or decoding parametric and width is a parameter that is predetermined or that derived) much larger than width by other codings or decoding parametric.The elementary operation of string replica code or decoding is to each coded strings in current block or to solve sequence (the most currently string), replicates a reference strings, and the numerical value assignment of described reference strings is given current string in reconstructed reference pixel sample value collection.The duplication parameter of string copy mode includes that the displacement vector currently gone here and there and duplication length i.e. replicate size, represents the relative position between reference strings with current string and the current length i.e. number of pixel sample value gone here and there respectively.The length of current string is also the length of reference strings.One current string has a displacement vector and a duplication length.One current block is divided into how many strings just how many displacement vectors and how many duplication length.
5)
Palette index string duplication i.e. palette or title index string replicate.In palette coding and corresponding decoding process, first construct or obtain a palette, then the index of present encoding block or the part or all of pixel palette currently solving code block (referred to as current block) is showed, again index encoded and decode, include but not limited to: the index of a current block is divided into several adjustable length index string, is i.e. indexed string replica code and decoding.The elementary operation of index string replica code or decoding is that each the index coded strings in current block or index are solved sequence (the most currently indexing string), in the reconstructed reference pixel sample value collection of indexation, replicate a reference key string, and the index numerical value assignment of described reference key string is given currently index string.The duplications parameter of index string copy mode includes currently indexing the displacement vector of string and replicates length and i.e. replicate size, represents the number of pixel sample value corresponding to length that the relative position between reference key string and current index string and current index are gone here and there respectively.The length of current index string is also the length of reference key string.One current index string has a displacement vector and a duplication length.One current block is divided into how many index strings just how many displacement vectors and how many duplication length.
6)
Index string replicates and replicates, with pixel string, the fusion copy mode mixed, and is called for short index-pixel string and merges copy mode.To a present encoding block or currently solve code block (referred to as current block) encode or decode time, part or all of pixel use pixel string copy mode, part or all of pixel use index string copy mode.
Other copy mode also includes rectangle copy mode and the copy mode etc. of several copy mode mixing.
Block in block copy mode, the microlith in microlith copy mode, the bar in bar copy mode, the string in string copy mode, the rectangle in rectangle copy mode, the pixel index string in palette indexed mode, it is referred to as pixel sample value section, is called for short sample value section.The basic component of sample value section is pixel or pixel component or pixel index.One sample value Duan Youyi replicates parameter, is used for representing the relation between current pixel sample value section and reference pixel sample value section.Therefore, a sample value section is the least unit once replicating operation with same replication relation.One is replicated parameter and includes some duplication parametric component, replicate parametric component at least to include: displacement vector horizontal component, displacement vector vertical component, 1 dimension displacement vector, linear address, opposite linear address, index, palette linear address, relative indexing, palette opposite linear address, duplication length, duplication width, duplication height, rectangle width, rectangle length, non-matched pixel (also known as without reference pixel, be not the most from the not replicated pixel replicated elsewhere).
In various copy modes, pixel sample value or index need according to a definite sequence arrangement.Arrangement mode is also referred to as scan mode.Scan mode, according to its path shape, can be divided into following several:
A) horizontal zigzag scan mode is also known as horizontal raster scan mode.One encoding block or solve pixel sample value or the index of code block (being referred to as encoding and decoding block), a line a line arranges, all according to same direction (the most from left to right or all the most from right to left) arrangement in all of row.Can arrange from the top down between row and row, it is also possible to arrange from bottom to top.
B) vertical Z shape scan mode is also known as vertical raster scan mode.One encoding block or solve pixel sample value or the index of code block (being referred to as encoding and decoding block), string string arranges, all according to same direction (the most from the top down or all the most from bottom to top) arrangement in all of row.Can arrange from left to right between the column and the column, it is also possible to arrange from right to left.
C) horizontal arcuate scanning mode.One encoding block or solve code block (being referred to as encoding and decoding block) pixel sample value or index, a line a line arrange, in odd-numbered line according to a direction (such as: from left to right) arrangement and in even number line according to another (on the contrary) direction (such as: from right to left) arrange.Can arrange from the top down between row and row, it is also possible to arrange from bottom to top.
D) vertical arcuate scan mode.One encoding block or solve code block (being referred to as encoding and decoding block) pixel sample value or index, string string arrange, in odd column according to a direction (such as: from the top down) arrangement and in even column according to another (on the contrary) direction (such as: from bottom to top) arrange.Can arrange from left to right between the column and the column, it is also possible to arrange from right to left.
It should be noted that " duplication " is reconstruct and the operation of decoding, corresponding encoding operation is " coupling ".Therefore, also referred to as Block-matching mode, microlith matching way, lines matching way, pixel String matching mode, the index String matching mode etc. such as various copy modes such as Block-matching mode, microlith copy mode, lines copy mode, pixel string copy mode, index string copy mode.
In existing various copy mode, reference pixel be all both without block-eliminating effect filtering (Deblocking Filter is called for short DF) step also without sample value adaptive equalization (Sample Adaptive
Offset is called for short SAO) the imperfection reconstructed pixel of step process.Imperfection reconstructed pixel and final improving may have bigger error between reconstructed pixel, cause there is bigger error between reference pixel and original pixels, reduce the compression efficiency to image.
Summary of the invention
In order to solve this problem of the prior art of image/video coding and decoding, the invention provides the method or apparatus of picture coding and decoding that a kind of reference pixel takes from the reconstructed pixel of at least two difference degree of perfection.It is to say, Part I reference pixel is the reconstructed pixel of degree of perfection I, Part II reference pixel is the reconstructed pixel of degree of perfection II, and Part III reference pixel is the reconstructed pixel of degree of perfection III, etc..Preferably, the reconstructed pixel of different degree of perfections takes from the diverse location i.e. zones of different of image.Preferably, the reference pixel of a reference pixel sample value section is made up of the reconstructed pixel of at least two difference degree of perfection.Preferably, the reference pixel of a reference pixel sample value section takes from the region that in image, at least two places are different, and described different region has the reconstructed pixel of different degree of perfection.According to the present invention, the reference pixel scope of copy mode is divided at least two parts and has the zones of different of different degree of perfection reconstructed pixel.According to the present invention, in copy mode, the reference pixel of the reference pixel sample value section corresponding to a current pixel sample value section of a present encoding block or solution code block (being referred to as encoding and decoding block) is taken from least two parts and is had the zones of different of different degree of perfection reconstructed pixel.
The primary technical characteristic of the present invention is that reference pixel takes from K(K >=2, usual K≤4) plant the reconstructed pixel of different degree of perfection.Such as, shown in Fig. 1, take from the reconstructed pixel of following 3 kinds of different degree of perfections:
1)
The reconstructed pixel not only without DF but also processed without SAO;
2)
Through the reconstructed pixel that vertical edge DF processes;
3)
Through the reconstructed pixel that whole DF and SAO process.
In the present invention, preferably, as shown in Figure 1, currently the reference pixel sample value section (a reference pixel sample value string in Fig. 1 or another reference pixel sample value block) corresponding to a current pixel sample value section (a current pixel sample value string in Fig. 1 or another current pixel sample value block) in CU is by K(K >=2, usual K≤4) plant different degree of perfection reconstructed pixel composition.It is to say, a part of reference pixel of described reference pixel sample value section is the reconstructed pixel of degree of perfection I, remainder reference pixel is the reconstructed pixel of other the K-1 kind degree of perfections different from described degree of perfection I.
In the present invention, it is preferable that the legal desirable reference pixel scope of a current pixel sample value section in current CU is by K(K >=2, usual K≤4) plant the region composition of different degree of perfection reconstructed pixel.Such as, shown in Fig. 1, it is made up of 3 kinds of regions.
In the present invention, it is preferable that the reference pixel of the reference pixel sample value section corresponding to a current pixel sample value section in current CU takes from K(K >=2 in legal desirable reference pixel scope, usual K≤4) plant the region of different degree of perfection reconstructed pixel.Such as, shown in Fig. 1, the reference pixel of a reference pixel sample value string takes from the region of 3 kinds of different degree of perfection reconstructed pixel, and the reference pixel of another reference pixel sample value block takes from the region of 2 kinds of different degree of perfection reconstructed pixel.
The coded method of the present invention or the most basic peculiar technical characteristic of device are when using a kind of copy mode to encode present encoding block, and the reference pixel replicating operation takes from K(K >=2, usual K≤4) plant the reconstructed pixel of different degree of perfection.Fig. 2 is coded method or a schematic diagram of device of the present invention.Preferably, described copy mode is one of following manner or its fusion: prediction mode (including infra-frame prediction or inter prediction), intra block copy mode, microlith copy mode, bar copy mode, string copy mode, index string copy mode;Described duplication operation be one of following operation or its merge: predicted operation, intra block replicate operation, microlith replicate operation, bar replicate operation, string replicate operation, index string replicate operation;Correspondingly, the reference pixel during described reference pixel is predictive value (block), reference block, reference microlith, reference bar, reference strings, palette respectively.
The coding/decoding method of the present invention or the most basic peculiar technical characteristic of device are when using a kind of copy mode to be decoded the current video code flow solving code block, the reference pixel replicating operation takes from K(K >=2, usual K≤4) plant the reconstructed pixel of different degree of perfection.Fig. 3 is coding/decoding method or a schematic diagram of device of the present invention.Preferably, described copy mode is one of following manner or its fusion: prediction mode (including infra-frame prediction or inter prediction), intra block copy mode, microlith copy mode, bar copy mode, string copy mode, index string copy mode;Described duplication operation be one of following operation or its merge: predicted operation, intra block replicate operation, microlith replicate operation, bar replicate operation, string replicate operation, index string replicate operation;Correspondingly, the reference pixel during described reference pixel is predictive value (block), reference block, reference microlith, reference bar, reference strings, palette respectively.
According to an aspect of the invention, it is provided a kind of method for encoding images or device, at least include following function and the step of operation or module:
The reconstructed pixel using at least 2 kinds of different degree of perfections carries out replica code as reference pixel to present encoding block, produces the video code flow containing the information replicating parameter.
According to another aspect of the present invention, additionally provide a kind of picture decoding method or device, at least include following function and the step of operation or module:
Resolve video code flow and obtain the information replicating parameter, use the reconstructed pixel of at least 2 kinds of different degree of perfections to carry out replicating decoding to the current code block that solves as reference pixel.
The present invention is applicable to coding and the decoding of folded bag format-pattern.Present invention is equally applicable to coding and the decoding of component planar format image.
Technical characteristic above by some specific instantiations explanation present invention.Those skilled in the art can be understood other advantages and effect of the present invention easily by content disclosed in the present specification.The present invention can also be carried out by the most different detailed description of the invention or apply, and the every details in this specification can also carry out various modification or change based on different viewpoints and application under the spirit without departing substantially from the present invention.
The following is more implementation detail or the variant of the present invention.
Implement or variant example 1
In described coded method or device or coding/decoding method or device, described replica code or decoding be one of following coding or decoding process or its merge: predictive coding or decoding, intra block replica code or decoding, microlith replica code or decoding, bar replica code or decoding, string replica code or decoding, index string replica code or decoding;Described duplication operation be one of following operation or its merge: predicted operation, intra block replicate operation, microlith replicate operation, bar replicate operation, string replicate operation, index string replicate operation;Correspondingly, the reference pixel during described reference pixel is predictive value (block), reference block, reference microlith, reference bar, reference strings, palette respectively.
Implement or variant example 2
In described coded method or device or coding/decoding method or device, legal desirable reference pixel scope is by K(K >=2, usual K≤4) plant the region composition of different degree of perfection reconstructed pixel.
Implement or variant example 3
In described coded method or device or coding/decoding method or device, described present encoding block or solve the legal desirable reference pixel scope of a current pixel sample value section in code block by K(K >=2, usual K≤4) plant the region composition of different degree of perfection reconstructed pixel.
Implement or variant example 4
In described coded method or device or coding/decoding method or device, described present encoding block or solve the reference pixel of the reference pixel sample value section corresponding to a current pixel sample value section in code block and take from K(K >=2 in legal desirable reference pixel scope, usual K≤4) plant the region of different degree of perfection reconstructed pixel.
Implement or variant example 5
In described coded method or device or coding/decoding method or device, the reconstructed pixel of described at least 2 kinds of different degree of perfections at least includes at least 2 kinds of following 3 kinds of reconstructed pixel:
1)
The reconstructed pixel of degree of perfection I;
2)
The reconstructed pixel of degree of perfection II;
3)
The reconstructed pixel of degree of perfection III.
Implement or variant example 6
Implement or in coded method or device described in variant example 5 or coding/decoding method or device, the reconstructed pixel of described degree of perfection I, II, III respectively:
1)
The reconstructed pixel not only without DF but also processed without SAO;
2)
Through the reconstructed pixel that vertical edge DF processes;
3)
Through the reconstructed pixel that whole DF and SAO process.
Implement or variant example 7
Implement or in coded method or device described in variant example 5 or 6 or coding/decoding method or device, the reconstructed pixel of described degree of perfection I, II, III be respectively from the following areas in present image:
1)
The rightest four row of current CTU and the CTU on the current CTU left side and more left side CTU;
2)
The region not yet reconstructed or described 1) four row of top in region;
3)
Both be not belonging to described 1 in the range of legal desirable reference pixel) be also not belonging to described 2) region.
Accompanying drawing explanation
Fig. 1 is the schematic diagram that in the present invention, reference pixel takes from multiple different degree of perfection reconstructed pixel.
Fig. 2 is coded method or a schematic diagram of device of the present invention.
Fig. 3 is coding/decoding method or a schematic diagram of device of the present invention.
Claims (10)
1. a method for encoding images or device, it is characterised in that at least included following function and the step of operation or module:
The reconstructed pixel using at least 2 kinds of different degree of perfections carries out replica code as reference pixel to present encoding block, produces the video code flow containing the information replicating parameter.
2. a picture decoding method or device, it is characterised in that at least included following function and the step of operation or module:
Resolve video code flow and obtain the information replicating parameter, use the reconstructed pixel of at least 2 kinds of different degree of perfections to carry out replicating decoding to the current code block that solves as reference pixel.
Coded method the most according to claim 1 or device or coding/decoding method according to claim 2 or device, it is characterised in that:
Described encoding block or described solution code block are a coding region of image or a decoding region, including following at least one: maximum coding unit LCU, code tree unit CTU, the subregion of coding unit CU, CU, predicting unit PU, converter unit TU.
Coded method the most according to claim 1 or device or coding/decoding method according to claim 2 or device, it is characterised in that:
Described replica code or decoding are one of following coding or decoding process or its fusion: predictive coding or decoding, intra block replica code or decoding, microlith replica code or decoding, bar replica code or decoding, string replica code or decoding, index string replica code or decoding;Described duplication operation be one of following operation or its merge: predicted operation, intra block replicate operation, microlith replicate operation, bar replicate operation, string replicate operation, index string replicate operation;Correspondingly, the reference pixel during described reference pixel is predictive value (block), reference block, reference microlith, reference bar, reference strings, palette respectively.
Coded method the most according to claim 1 or device or coding/decoding method according to claim 2 or device, it is characterised in that:
Legal desirable reference pixel scope is made up of the region of at least two difference degree of perfection reconstructed pixel.
Coded method the most according to claim 1 or device or coding/decoding method according to claim 2 or device, it is characterised in that:
The legal desirable reference pixel scope of a current pixel sample value section in described present encoding block or solution code block is made up of the region of at least two difference degree of perfection reconstructed pixel.
Coded method the most according to claim 1 or device or coding/decoding method according to claim 2 or device, it is characterised in that:
The reference pixel of the reference pixel sample value section corresponding to a current pixel sample value section in described present encoding block or solution code block takes from the region of at least two difference degree of perfection reconstructed pixel in legal desirable reference pixel scope.
Coded method the most according to claim 1 or device or coding/decoding method according to claim 2 or device, it is characterised in that:
The reconstructed pixel of described at least 2 kinds of different degree of perfections at least includes at least 2 kinds of following 3 kinds of reconstructed pixel:
The reconstructed pixel of degree of perfection I;
The reconstructed pixel of degree of perfection II;
The reconstructed pixel of degree of perfection III.
Coded method the most according to claim 8 or device or coding/decoding method or device, it is characterised in that:
The reconstructed pixel of described degree of perfection I, II, III is respectively:
The reconstructed pixel not only without DF but also processed without SAO;
Through the reconstructed pixel that vertical edge DF processes;
Through the reconstructed pixel that whole DF and SAO process.
Coded method the most according to Claim 8 or described in claim 9 or device or coding/decoding method or device, it is characterised in that:
The reconstructed pixel of described degree of perfection I, II, III is respectively from the following areas in present image:
The rightest four row of current CTU and the CTU on the current CTU left side and more left side CTU;
The region not yet reconstructed or described 1) four row of top in region;
Both be not belonging to described 1 in the range of legal desirable reference pixel) be also not belonging to described 2) region.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210467859.8A CN115499663A (en) | 2015-06-08 | 2015-06-08 | Image compression method and device for different-degree reconstructed pixels with reference to single coding mode |
CN201510308137.8A CN106303535B (en) | 2015-06-08 | 2015-06-08 | Image compression method and device with reference pixels taken from different-degree reconstruction pixels |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510308137.8A CN106303535B (en) | 2015-06-08 | 2015-06-08 | Image compression method and device with reference pixels taken from different-degree reconstruction pixels |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210467859.8A Division CN115499663A (en) | 2015-06-08 | 2015-06-08 | Image compression method and device for different-degree reconstructed pixels with reference to single coding mode |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106303535A true CN106303535A (en) | 2017-01-04 |
CN106303535B CN106303535B (en) | 2022-12-13 |
Family
ID=57659082
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210467859.8A Pending CN115499663A (en) | 2015-06-08 | 2015-06-08 | Image compression method and device for different-degree reconstructed pixels with reference to single coding mode |
CN201510308137.8A Active CN106303535B (en) | 2015-06-08 | 2015-06-08 | Image compression method and device with reference pixels taken from different-degree reconstruction pixels |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210467859.8A Pending CN115499663A (en) | 2015-06-08 | 2015-06-08 | Image compression method and device for different-degree reconstructed pixels with reference to single coding mode |
Country Status (1)
Country | Link |
---|---|
CN (2) | CN115499663A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019127484A1 (en) * | 2017-12-29 | 2019-07-04 | 深圳市大疆创新科技有限公司 | Video coding method, video decoding method, and related device |
CN112333446A (en) * | 2020-11-03 | 2021-02-05 | 中山大学 | Intra-frame block copy reference block compression method |
WO2022067805A1 (en) * | 2020-09-30 | 2022-04-07 | Oppo广东移动通信有限公司 | Image prediction method, encoder, decoder, and computer storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101682784A (en) * | 2007-04-19 | 2010-03-24 | 汤姆逊许可证公司 | Adaptive reference picture data generation for intra prediction |
CN103124349A (en) * | 2011-11-19 | 2013-05-29 | 捷讯研究有限公司 | Multi-level significance map scanning |
EP2699001A1 (en) * | 2012-08-13 | 2014-02-19 | Politechnika Poznanska | A method and a system for video signal encoding and decoding with motion estimation |
CN104284199A (en) * | 2013-07-11 | 2015-01-14 | Nxp股份有限公司 | Video decoding method and device through deblocking effect operation for reducing complexity |
-
2015
- 2015-06-08 CN CN202210467859.8A patent/CN115499663A/en active Pending
- 2015-06-08 CN CN201510308137.8A patent/CN106303535B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101682784A (en) * | 2007-04-19 | 2010-03-24 | 汤姆逊许可证公司 | Adaptive reference picture data generation for intra prediction |
CN103124349A (en) * | 2011-11-19 | 2013-05-29 | 捷讯研究有限公司 | Multi-level significance map scanning |
EP2699001A1 (en) * | 2012-08-13 | 2014-02-19 | Politechnika Poznanska | A method and a system for video signal encoding and decoding with motion estimation |
CN104284199A (en) * | 2013-07-11 | 2015-01-14 | Nxp股份有限公司 | Video decoding method and device through deblocking effect operation for reducing complexity |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019127484A1 (en) * | 2017-12-29 | 2019-07-04 | 深圳市大疆创新科技有限公司 | Video coding method, video decoding method, and related device |
WO2022067805A1 (en) * | 2020-09-30 | 2022-04-07 | Oppo广东移动通信有限公司 | Image prediction method, encoder, decoder, and computer storage medium |
CN112333446A (en) * | 2020-11-03 | 2021-02-05 | 中山大学 | Intra-frame block copy reference block compression method |
Also Published As
Publication number | Publication date |
---|---|
CN106303535B (en) | 2022-12-13 |
CN115499663A (en) | 2022-12-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105704491B (en) | Image encoding method, decoding method, encoding device, and decoding device | |
CN106797468B (en) | Efficient video coding (HEVC) screen content encodes improved pallet mode in (SCC) | |
KR102588425B1 (en) | Adaptive partition coding | |
CN105491376B (en) | Image coding and decoding method and device | |
KR20220165694A (en) | A method of constructing the tiles and an apparatus having the same | |
KR101946598B1 (en) | Image coding and decoding method and device | |
WO2020216303A1 (en) | Selective use of secondary transform in coded video | |
CN105847842B (en) | Image encoding method and device, and image decoding method and device | |
CN104853211A (en) | Image compression method and apparatus employing various forms of reference pixel storage spaces | |
KR20220006055A (en) | Interactions between multiple intra-coding methods | |
WO2021083376A1 (en) | Derivation of linear parameter in cross-component video coding | |
CN113784124B (en) | Block matching encoding and decoding method for fine division using multi-shape sub-blocks | |
WO2021110018A1 (en) | Separable secondary transform processing of coded video | |
EP3306931A1 (en) | Image encoding and decoding method, image processing device, and computer storage medium | |
CN106254888A (en) | A kind of picture coding and coding/decoding method, image processing equipment | |
KR102532391B1 (en) | Video encoding method and apparatus and video decoding method and apparatus | |
CN104811731A (en) | Multilayer sub-block matching image compression method | |
CN106303535A (en) | Reference pixel takes from method for compressing image and the device of reconstructed pixel in various degree | |
CN106254878B (en) | Image encoding and decoding method and image processing equipment | |
CN105992003A (en) | Method and device for image compression of palette color number based on order or frequency | |
CN106303534A (en) | Multiple index string and pixel string merge method for compressing image and the device of copy mode | |
WO2016119746A1 (en) | Image coding method and apparatus, and image decoding method and apparatus | |
CN115066897A (en) | Coding and decoding of low bit depth visual media data | |
EP3571842B1 (en) | Devices and methods for video coding | |
CN105828080A (en) | Method and device for coding and decoding image |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |