CN108989819A - Each mode uses the data compression method and device of respectively corresponding color space - Google Patents
Each mode uses the data compression method and device of respectively corresponding color space Download PDFInfo
- Publication number
- CN108989819A CN108989819A CN201710410669.1A CN201710410669A CN108989819A CN 108989819 A CN108989819 A CN 108989819A CN 201710410669 A CN201710410669 A CN 201710410669A CN 108989819 A CN108989819 A CN 108989819A
- Authority
- CN
- China
- Prior art keywords
- bitdepth
- color space
- coding mode
- coding
- data
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/61—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/597—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding specially adapted for multi-view video sequence encoding
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Color Television Systems (AREA)
Abstract
The present invention provides a kind of each coding modes using the data compression method and device of the data color space progress encoding and decoding for meeting each mode self character.In this method and device, by usual 2≤L≤10 scheduled L() kind coding mode be classified as scheduled K(2≤K≤L) group, the corresponding mutually different data color space of K kind is respectively adopted in K group coding mode, and the coding mode with group all uses same data color space.In this way, making data color spatial adaptation in the characteristic of each coding mode, binary encoding efficiency is improved.When the color space of original input data plants color space different from the kth (1≤k≤K) among the K kind color space, need to convert the data that original input data is converted to the kth kind color space using color space.
Description
Technical field
It is damaged the present invention relates to a kind of pair of data or the encoding and decoding system of lossless compression, especially image and view
The method and apparatus of the encoding and decoding of frequency evidence.
Background technique
As human society enters big data, cloud computing, mobile computing, cloud-mobile computing, ultra high-definition (4K) and special superelevation
The epoch of (8K) video image resolution ratio, 4G/5G communication, virtual reality, artificial intelligence clearly, to various data, including big data,
Image data, video data, the data compression for carrying out superelevation compression ratio and very high-quality become essential technology.
Data set be by data element (such as: byte, bit, pixel, pixel component, spatial sampling point, transform domain system
Number) composition set.To be arranged in certain space (one-dimensional, two-dimentional or multidimensional) shape data set (such as: a dimension
According to queue, a two-dimensional data files, a frame image, a video sequence, a transform domain, a transform block, multiple transformation
The sequence of block, a three-dimensional scenic, three-dimensional scenic persistently changed), especially two dimension or above data collection carries out data
When coding (and corresponding decoding) of compression, usually this data set is divided into several with predetermined shape and/or size
Subset, referred to as encoding block (from decoded angle i.e. decoding block, are referred to as encoding and decoding block), as unit of encoding and decoding block, with
Scheduled sequence, one piece one piece is encoded or is decoded.At any one time, the encoding block encoded is known as present encoding
Block.At any one time, the decoding block decoded is known as current decoding block.Present encoding block or current decoding block are referred to as working as
Preceding encoding and decoding block or referred to as current block.Data element (being also referred to as element sometimes) just in encoding or decoding is known as working as
Preceding encoding data elements or current decoded data element, are referred to as current data element, referred to as currentElement.Work as the list of elements
When showing color, element is made of N number of component (usual 1≤N≤5), for example, being made of 3 components: G(green) component, B
(blue) component, R(are red) component or Y(brightness) component, U(coloration 1) component, V(coloration 2) component or Y(brightness) component, Cb
(colour difference blue) component, Cr(colour difference are red) component or Y(brightness) component, Cg(coloration is green) and component, Co(coloration orange) component or H
(form and aspect) component, S(saturation degree) component, V(lightness) component or H(form and aspect) component, S(saturation degree) component, L(brightness) component;
Or it is made of 4 components: (C, M, Y, K) i.e. (cyan, magenta, yellow, black) or (R, G, B, A) i.e. (red, green, blue, A Er
Method) or (Y, U, V, A) i.e. (brightness, coloration 1, coloration 2, Alpha) or (Y, Cb, Cr, A) i.e. (brightness, colour difference is blue, colour difference
It is red, Alpha) or (brightness, coloration is green, coloration orange, Alpha) (Y, Cg, Co, A) i.e..These different representation in components of color
Method is known as color space.It usually can mutually lossless (not losing any color information) or damage (can between different color space
A part of color information can be lost) convert.The nothing between the different representations of same color is set up in color space conversion
Damage is i.e. mathematically accurate or damages the relationship for mathematically having error to a certain degree.
For one with certain shapes (being not necessarily limited to square or rectangle, can be any reasonable other shapes)
A encoding and decoding block needs to be divided into many occasions finer primitive (basic unit), according to scheduled time sequencing,
One primitive, one primitive is encoded or is decoded.To all sample values in a primitive, same type of coding is usually implemented
Or decoding operate.At any one time, the primitive in encoding or decoding is known as current primitive.One primitive is encoded
The result is that one or more coding parameters, finally generate the compressed data code stream containing these coding parameters.One primitive is carried out
Decoding is exactly to parse the compressed data code stream to obtain one or more coding parameters, multiple from one or more of coding parameters
Original goes out the sample value of the data of reconstruct.
The example of primitive include encoding and decoding block (entire block is as a primitive), sub-block, microlith, string, byte serial, alpha(Ah
That method) string, pixel string, sample value string, index string, lines.
One distinguishing feature of many common data sets is the formula with many matchings (i.e. similar or even identical)
Sample.For example, usually there is many matched pixel patterns in image and video sequence.Therefore, in available data compress technique usually
By the way of matching (also referred to as predict or compensate), i.e., with " predicted value " (also referred to as " offset " or " referring to sample value ", such as
" reference pixel ") come match (also referred to as predict, indicate, represent, compensation, it is approximate, approach) sample value in present encoding or decoding
The mode of (referred to as " current sample ") achievees the effect that the lossless or lossy compressions of data.In many occasions, matching way
Basic operation be exactly duplication with reference to sample value of the sample value i.e. on reference position to the position of current sample.Therefore, matching way
Or prediction mode is also referred to as copy mode.In matching way, had been subjected at least partly encoding operation and at least partly solved
Reconstruct (also referred to as rebuild or restore) sample value composition reference set (also referred to as reference set space or with reference to space or with reference to slow of code operation
It deposits).The original samples and its position that reconstruct sample value and its position among reference set are concentrated with initial data are one-to-one.
When carrying out the encoding and decoding of a current block, current block is divided into several matchings and (also referred to as predicts) base by matching way
Member, a Matching unit have one or more matching (coding) parameter (also referred to as matching relationship or duplication parameter or replication relations
Or referring-to relation) indicate its feature.In match parameter a most important parameter be displacement vector (also referred to as motion vector,
Positional shift, offset, relative position, position, relative address, address, relative coordinate, coordinate, relative indexing, index etc.).Position
Move the current primitive of vector representation sample value and its with reference to the relative shift between sample value, be equivalent to data sample be arranged in it is one-dimensional
One-dimensional offset after data.Obviously, it can be exported with reference to where the reference position of sample value from displacement vector.The displacement of current primitive
Vector is known as current displacement vector.Other examples of match parameter: primitive mode, scan pattern, match-type, matching length, no
(prediction) sample value can be matched, etc..
The example of Matching unit includes encoding and decoding block, sub-block, microlith, string, pixel string, sample value string, index string, lines.
In available data compress technique, usually data are compressed using a variety of coding modes, such as prediction+transformation
Traditional Hybrid coding mode or its variant, for another example the various match patterns based on matching way or its.
In available data compress technique, all coding modes all compile data element using same color space
Decoding, that is, data element color space be it is fixed, may not change with the difference of coding mode.Another party
Face, a kind of coding mode may be different the code efficiency of the data in different color space, to a kind of number of color space
According to code efficiency may be much higher than the code efficiency to the data of another color space.In available data compress technique,
The characteristic that may not be changed with data color space using the code efficiency of coding mode, affects mentioning for code efficiency
It is high.
Summary of the invention
In order to solve the problems, such as this in data compression, is used the present invention provides a kind of each coding mode and meet each mode
The data color space of self character carries out the data compression method and device of encoding and decoding.In this method and device, it will make a reservation for
Usual 2≤L≤10 L() kind coding mode is classified as scheduled K(2≤K≤L) group, K group coding mode is respectively adopted
The corresponding mutually different data color space of K kind, the coding mode with group all use same data color space.In this way,
Make data color spatial adaptation in the characteristic of each coding mode, improves binary encoding efficiency.When the color of original input data is empty
Between be different from the K kind color space among kth (1≤k≤K) plant color space when, need using color space turn
Original input data of changing commanders is converted to the data of the kth kind color space.
For example, using scheduled L1(1≤L1≤ 10) Hybrid coding mode and its variant and scheduled L are planted2(1
≤ L2≤ 10) kind matching coding mode is when encode the original input data of (R, G, B) color space, by L=
L1 + L2Kind coding mode is classified as 2 groups (K=2), L1Kind Hybrid coding mode and its variant belong to first group, L2Kind matching is compiled
Pattern belongs to second group.To first group of coding mode, second group of coding mode is adopted using (Y, U, V) data color space
With (R, G, B) data color space.
The most basic peculiar technical characteristic of coding method or device of the invention is to use scheduled L(L > 1) kind coding mould
Coding mode is returned into K(K > 1 according to the characteristic of each coding mode itself and the requirement to data color space by the occasion of formula)
Group, it is corresponding with K kind color space respectively.When being encoded using a kind of coding mode to a present encoding block, according to volume
Group belonging to pattern selects corresponding color space, encodes to the present encoding block, generates at least true containing can be used for
The identification code of the coding mode of fixed corresponding color space and/or the compressed data stream of the information of equal value with it.Preferably, if
The color space of original input data is not the corresponding color space, then is converted first using color space by present encoding
The data of block are converted to the data of the corresponding color space, then encode again to the data after conversion.Preferably, make
The data of present encoding block generated in an encoding process are converted into other one or more colors skies with color space conversion
Between data.Fig. 1 is a schematic diagram of coding method or device of the invention.
The most basic peculiar technical characteristic of coding/decoding method or device of the invention is parsing compressed data stream, obtains coding mode
Information, according to the information of the coding mode using the corresponding color space in one of scheduled multicolour space to one
Current decoding block is decoded.Preferably, if the color space for rebuilding data (is equal to the original input data of encoder
Color space) it is not the corresponding color space, then it is converted current decoding block in decoding process using color space
The data of generation are converted to the data of the corresponding color space.Preferably, it is converted using color space by current decoding block
The data generated in decoding process be converted to the data of other one or more color spaces.Fig. 2 is decoding of the invention
One schematic diagram of method or apparatus.
According to an aspect of the invention, there is provided coding method or the device of a kind of data compression, it is following to include at least completion
The step of function and operation or module:
According to scheduled L(L > 1) characteristic of coding mode itself and the requirement to data color space are planted, coding mode is returned into
K(K > 1) group, it is corresponding with scheduled K kind color space respectively.One present encoding block is being compiled using a kind of coding mode
When code, according to group belonging to coding mode, corresponding color space is selected, the present encoding block is encoded, generated extremely
The compressed data of few identification code containing the coding mode that can be used for determining corresponding color space and/or the information of equal value with it
Stream.
According to another aspect of the present invention, the coding/decoding method or device of a kind of data compression are additionally provided, includes at least and completes
The step of following function and operation or module:
Parsing uses scheduled L(L > 1) compressed data stream that coding mode generates is planted, the information of coding mode is obtained, according to institute
The information of coding mode is stated using scheduled K(K > 1) the corresponding color space of one of color space is planted to a current decoding
Block is decoded.
The present invention is suitable for carrying out data the coding and decoding of lossy compression, and present invention is equally applicable to carry out nothing to data
Damage the coding and decoding of compression.The present invention is suitable for one-dimensional data such as string data or byte string data or one-dimensional pattern or divides
The coding and decoding for tieing up figure, present invention is equally applicable to the reconciliation of the coding of two dimension or above data such as image or video data
Code.
In the present invention, data involved in data compression include one or a combination set of data of following classes
1) one-dimensional data;
2) 2-D data;
3) multidimensional data;
4) figure;
5) divide dimension figure;
6) image;
7) sequence of image;
8) video;
9) three-dimensional scenic;
10) sequence of the three-dimensional scenic persistently changed;
11) scene of virtual reality;
12) sequence of the scene of the virtual reality persistently changed
13) image of pixel form;
14) the transformation numeric field data of image;
15) set of two dimension or the above byte of two dimension;
16) two dimension or the set of the above bit of two dimension;
17) set of pixel;
18) set of four component pixels (C, M, Y, K);
19) set of four component pixels (R, G, B, A);
20) set of four component pixels (Y, U, V, A);
21) set of four component pixels (Y, Cb, Cr, A);
22) set of four component pixels (Y, Cg, Co, A);
23) set of three-component pixel (R, G, B);
24) set of three-component pixel (Y, U, V);
25) set of three-component pixel (Y, Cb, Cr);
26) set of three-component pixel (Y, Cg, Co);
27) set of three-component pixel (H, S, V);
28) set of three-component pixel (H, S, L).
In the present invention, the coding mode includes one or a combination set of following coding mode or its variant:
1) include intra prediction Hybrid coding mode;
2) include inter-prediction Hybrid coding mode;
3) include wavelet transformation coding mode;
4) include residual coding coding mode;
5) include matching way coding mode;
6) include Block- matching coding mode;
It 7) include the matched coding mode of sub-block;
It 8) include the matched coding mode of microlith;
It 9) include the matched coding mode of lines;
10) include String matching coding mode;
11) include pixel String matching coding mode;
12) include sample value String matching coding mode;
It 13) include the coding mode for indexing String matching;
14) include main reference buffer area String matching coding mode;
It 15) include time coding mode of reference buffer area String matching;
16) include string prediction coding mode;
It 17) include the general coding mode gone here and there and predicted;
It 18) include the coding mode for deviating string prediction;
19) include coordinate string prediction coding mode;
20) include non-matched pixel coding mode;
21) include non-matched pixel string coding mode;
22) include unpredictable pixel coding mode;
23) include unpredictable pixel string coding mode.
It is the situation of image, the sequence of image, video etc. in data in the present invention, encoding block or decoding block are one of image
Coding region or a decoding region, comprise at least one of the following: entire image, the subgraph of image, macro block, maximum coding are single
First LCU, coding tree unit CTU, the subregion of coding unit CU, CU, predicting unit PU, converter unit TU.
In the present invention, the primitive includes one or a combination set of following state: encoding and decoding block, sub-block, microlith, string, byte serial,
Alpha(Alpha) string, pixel string, sample value string, index string, lines, match block, matched sub-block, matching microlith, matching go here and there,
With pixel string, matching sample value string, match index string, matching item, matching lines, offset string, coordinate string, unpredictable pixel, no
Predictable pixel string, coordinate or unpredictable pixel string.
Illustrate technical characteristic of the invention above by several specific specific examples.Those skilled in the art can be by this specification
Other advantages and efficacy of the present invention can be easily understood for revealed content.The present invention can also pass through in addition different specific realities
The mode of applying is embodied or practiced, and the various details in this specification can also be based on different viewpoints and application, without departing substantially from this
Various modifications or alterations are carried out under the spirit of invention.
It is more implementation details or variant of the invention below.
Implementation or variant example 1
The coding method or device perhaps the color space of original input data or rebuild data in coding/decoding method or device
Color space be not the corresponding color space, use color space conversion carry out these color spaces between mutually turn
It changes, including one or a combination set of following conversion:
The color space of the original input data is converted into the corresponding color space
Or
The color space for rebuilding data is converted into the corresponding color space
Or
The corresponding color space is converted to the color space of the original input data
Or
The corresponding color space is converted into the color space for rebuilding data.
Implementation or variant example 2
In the coding method or device or coding/decoding method or device, using color space conversion by current block in coding or
It is one or more that the data of before person's decoding process and/or centre and/or a kind of color space being related to later are converted to other
The data of color space.
Implementation or variant example 3
In the coding method or device or coding/decoding method or device, the scheduled L kind coding mode includes at least predetermined
L1Kind Hybrid coding mode or its variant and scheduled L2Kind matching coding mode;The scheduled K kind color space is at least
Including one or a combination set of following color space:
(R, G, B) color space or its variant;
(Y, U, V) color space or its variant;
(Y, Cb, Cr) color space or its variant;
(Y, Cg, Co) color space or its variant;
(H, S, V) color space or its variant;
(H, S, L) color space or its variant;
(C, M, Y, K) color space or its variant;
(R, G, B, A) color space or its variant;
(Y, U, V, A) color space or its variant;
(Y, Cb, Cr, A) color space or its variant;
(Y, Cg, Co, A) color space or its variant.
Implementation or variant example 4
In the coding method or device or coding/decoding method or device, the scheduled L kind coding mode includes at least predetermined
L1Hybrid coding mode or its variant and scheduled L of the kind based on prediction and transformation2Coding mould of the kind based on general string prediction
Formula or its variant;The L1Kind coding mode is classified as K1Group, respectively with scheduled K1Kind color space is corresponding;The L2Kind coding
Mode is classified as K2Group, respectively with scheduled K2Kind color space is corresponding;The scheduled K1Kind of color space include at least (Y, U,
V) color space or its variant;The scheduled K2Kind color space includes at least (R, G, B) color space or its variant.
Implementation or variant example 5
In the coding method or device or coding/decoding method or device, using color space conversion by current block in coding or
In person's decoding after the prediction that generates after data (i.e. forecast sample data) and/or matching data (reference data replicated) and/
Or it data and/or rebuilds data after compensation and from a kind of color space is converted to other one or more color spaces.
Implementation or variant example 6
In the coding method or device or coding/decoding method or device, the scheduled K kind color space includes (R, G, B) color
Color space or its variant and (Y, U, V) color space or its variant;(R, G, the B) color space or its variant and (Y, U, V)
Color space conversion between color space or its variant is following five one or a combination set of the conversions or its change for applying a formula expression
Body:
The first direct transform formula that applies a formula is
t = 0.2126*R + 0.7152*G + 0.0722*B
Y = floor ((16 + 219 * t / (2BitDepth–1)) * 2BitDepth–8 + 0.5)
U = floor ((128 + 112 * (B – t) / ((2BitDepth–1) * (1 – 0.0722))) * 2BitDepth–8
+ 0.5)
V = floor ((128 + 112 * (R – t) / ((2BitDepth–1) * (1 – 0.2126))) * 2BitDepth–8
+ 0.5)
The first inverse transformation formula that applies a formula is
t = floor ((Y - 16 * 2BitDepth–8) * (2BitDepth–1) / (219 * 2BitDepth–8) + 0.5)
R = clamp [t + (Cr - 2BitDepth–1) * (2BitDepth–1) * (1-0.2126)/(112 * 2BitDepth–8)
+ 0.5]
B = clamp [t + (Cb - 2BitDepth–1) * (2BitDepth–1) * (1-0.0722)/(112 * 2BitDepth–8)
+ 0.5]
G = clamp [( t - 0.2126*R - 0.0722*B)/0.7152 + 0.5]
Wherein, t is floating number, and R, G, B, Y, U, V are integers, and number of bits is BitDepth.Floor (x) is less than
Maximum integer equal to x, clamp [x] is the maximum for first taking 0 and floor (x), then takes itself and 2BitDepth- 1 reckling;
The second direct transform formula that applies a formula is
t = 0.2126*R + 0.7152*G + 0.0722*B
Y = floor (t)
U = floor (2BitDepth–1 + 0.5 * (B – t) / (1 – 0.0722))
V = floor (2BitDepth–1 + 0.5 * (R – t) / (1 – 0.2126))
The second inverse transformation formula that applies a formula is
R = clamp [Y + (Cr - 2BitDepth–1) * 2 * (1-0.2126) + 0.5]
B = clamp [Y + (Cb - 2BitDepth–1) * 2 * (1-0.0722) + 0.5]
G = clamp [(Y - 0.2126*R - 0.0722*B)/0.7152 + 0.5]
The direct transform formula that third applies a formula is
Y = (G + ((R + B) >> 1) + 1) >> 1
U = ((R - B - 1) >> 1) + (1<<(BitDepth-1))
V = (G – Y) + (1<<(BitDepth-1))
The inverse transformation formula that third applies a formula is
R = clip3(0, (1<<BitDepth) - 1, Y – V + U)
G = clip3(0, (1<<BitDepth) - 1, Y + V - (1<<(BitDepth-1)))
B = clip3(0, (1<<BitDepth) - 1, Y – V – U + (1<<BitDepth))
Wherein,<<is binary shift left operation, and>>is binary shift right operation, clip3 (0, (1<<BitDepth) -1, x)
It is first to take the maximum of 0 and x, then take itself and 2BitDepth- 1 reckling;
The 4th direct transform formula that applies a formula is
Y = (G + ((R + B) >> 1) + 1) >> 1
U = ((G - B - 1) >> 1) + (1<<(BitDepth-1))
V = ((G - R - 1) >> 1) + (1<<(BitDepth-1))
The 4th inverse transformation formula that applies a formula is
G = clip3(0, (1<<BitDepth) - 1, Y + ((U + V - (1<<BitDepth)) >> 1))
B = clip3(0, (1<<BitDepth) - 1, G - 2U + (1<<BitDepth) - 1)
R = clip3(0, (1<<BitDepth) - 1, G - 2V + (1<<BitDepth) - 1)
The 5th direct transform formula that applies a formula is
Y = (G + ((R + B) >> 1) + 1) >> 1
U = ((Y - B) >> 1) + (1<<(BitDepth-1))
V = ((Y - R) >> 1) + (1<<(BitDepth-1))
The 5th inverse transformation formula that applies a formula is
G = clip3(0, (1<<BitDepth) - 1, Y + U + V - (1<<BitDepth))
B = clip3(0, (1<<BitDepth) - 1, Y - 2U + (1<<BitDepth))
R = clip3(0, (1<<BitDepth) - 1, Y - 2V + (1<<BitDepth))。
Implementation or variant example 7
In the coding method or device or coding/decoding method or device, the encoding and decoding block has one in the compressed data stream
A coding mode identification code directly or indirectly or directly mixed indirectly, it is empty using corresponding color according to the value of the identification code
Between to the encoding and decoding block carry out encoding and decoding;The direct coding mode identification code is by one or more of compressed data stream
Bit string (binary symbols string) is formed;The indirect coding mode identification code is from other codec parameters and/or compression number
Coding mode identification code derived from other syntactic elements according to stream or scheduled identification code default value;It is described directly indirectly to mix
Coding mode identification code be part directly (being made of one or more bit strings in compressed data stream) part it is indirect (i.e. from
Other syntactic elements of other codec parameters and/or compressed data stream export or make a reservation for default) mixing coding mode mark
Code.
Detailed description of the invention
Fig. 1 is a schematic diagram of coding method or device of the invention.
Fig. 2 is a schematic diagram of coding/decoding method or device of the invention.
Claims (10)
1. coding method or the device of a kind of data compression, which is characterized in that including at least the step for completing following function and operation
Rapid or module:
According to scheduled L(L > 1) characteristic of coding mode itself and the requirement to data color space are planted, coding mode is returned into
K(K > 1) group, it is corresponding with scheduled K kind color space respectively;One present encoding block is being compiled using a kind of coding mode
When code, according to group belonging to coding mode, corresponding color space is selected, the present encoding block is encoded, generated extremely
The compressed data of few identification code containing the coding mode that can be used for determining corresponding color space and/or the information of equal value with it
Stream.
2. the coding/decoding method or device of a kind of data compression, which is characterized in that including at least the step for completing following function and operation
Rapid or module:
Parsing uses scheduled L(L > 1) compressed data stream that coding mode generates is planted, the information of coding mode is obtained, according to institute
The information of coding mode is stated using scheduled K(K > 1) the corresponding color space of one of color space is planted to a current decoding
Block is decoded.
3. coding method according to claim 1 or device or coding/decoding method according to claim 2 or device,
It is characterized in that the coding mode includes one or a combination set of following coding mode or its variant:
Hybrid coding mode including intra prediction;
Hybrid coding mode including inter-prediction;
Coding mode including wavelet transformation;
Coding mode including residual coding;
Coding mode including matching way;
Coding mode including Block- matching;
Including the matched coding mode of sub-block;
Including the matched coding mode of microlith;
Including the matched coding mode of lines;
Coding mode including String matching;
Coding mode including pixel String matching;
Coding mode including sample value String matching;
Coding mode including indexing String matching;
Coding mode including main reference buffer area String matching;
Coding mode including secondary reference buffer area String matching;
Coding mode including string prediction;
Coding mode including general string prediction;
Coding mode including offset string prediction;
Coding mode including the prediction of coordinate string;
Coding mode including non-matched pixel;
Coding mode including non-matched pixel string;
Coding mode including unpredictable pixel;
Coding mode including unpredictable pixel string.
4. coding method according to claim 1 or device or coding/decoding method according to claim 2 or device,
It is characterized by:
Data involved in the data compression are image data or image sequence data or video data;
The encoding block or the decoding block be image a coding region or a decoding region, including following at least one
Kind: entire image, the subgraph of image, macro block, maximum coding unit LCU, coding tree unit CTU, coding unit CU, CU son
Region, predicting unit PU, converter unit TU.
5. coding method according to claim 1 or device or coding/decoding method according to claim 2 or device,
It is characterized in that being converted current block before encoding or decoding process using color space and/or intermediate and/or later
A kind of data for the color space being related to are converted to the data of other one or more color spaces.
6. coding method according to claim 1 or device or coding/decoding method according to claim 2 or device,
It is characterized in that the scheduled L kind coding mode includes at least scheduled L1It plants Hybrid coding mode or its variant and makes a reservation for
L2Kind matching coding mode;The scheduled K kind color space includes at least one or a combination set of following color space:
(R, G, B) color space or its variant;
(Y, U, V) color space or its variant;
(Y, Cb, Cr) color space or its variant;
(Y, Cg, Co) color space or its variant;
(H, S, V) color space or its variant;
(H, S, L) color space or its variant;
(C, M, Y, K) color space or its variant;
(R, G, B, A) color space or its variant;
(Y, U, V, A) color space or its variant;
(Y, Cb, Cr, A) color space or its variant;
(Y, Cg, Co, A) color space or its variant.
7. coding method according to claim 1 or device or coding/decoding method according to claim 2 or device,
It is characterized in that, the scheduled L kind coding mode includes at least scheduled L1Hybrid coding of the kind based on prediction and transformation
Mode or its variant and scheduled L2Coding mode or its variant of the kind based on general string prediction;The L1Kind coding mode is classified as
K1Group, respectively with scheduled K1Kind color space is corresponding;The L2Kind coding mode is classified as K2Group, respectively with scheduled K2Kind color
Space is corresponding;The scheduled K1Kind color space includes at least (Y, U, V) color space or its variant;The scheduled K2Kind
Color space includes at least (R, G, B) color space or its variant.
8. coding method according to claim 1 or device or coding/decoding method according to claim 2 or device,
It is characterized in that being converted using color space by data (i.e. pre- test sample after the prediction of current block generated in encoding or decoding
Notebook data) and/or matching after data and/or rebuild data from a kind of color after data (reference data replicated) and/or compensation
Color space is converted to other one or more color spaces.
9. coding method according to claim 1 or device or coding/decoding method according to claim 2 or device,
It is characterized in that, the scheduled K kind color space includes (R, G, B) color space or its variant and (Y, U, V) color space
Or its variant;Color space between (R, G, the B) color space or its variant and (Y, U, V) color space or its variant turns
Change be following five apply a formula expression one or a combination set of conversion or its variant:
The first direct transform formula that applies a formula is
t = 0.2126*R + 0.7152*G + 0.0722*B
Y= floor ((16 + 219 * t / (2BitDepth–1)) * 2BitDepth–8 + 0.5)
U = floor ((128 + 112 * (B – t) / ((2BitDepth–1) * (1 – 0.0722))) * 2BitDepth–8
+ 0.5)
V = floor ((128 + 112 * (R – t) / ((2BitDepth–1) * (1 – 0.2126))) * 2BitDepth–8
+ 0.5)
The first inverse transformation formula that applies a formula is
t = floor ((Y - 16 * 2BitDepth–8) * (2BitDepth–1) / (219 * 2BitDepth–8) + 0.5)
R = clamp [t + (Cr - 2BitDepth–1) * (2BitDepth–1) * (1-0.2126)/(112 * 2BitDepth–8)
+ 0.5]
B = clamp [t + (Cb - 2BitDepth–1) * (2BitDepth–1) * (1-0.0722)/(112 * 2BitDepth–8)
+ 0.5]
G = clamp [( t - 0.2126*R - 0.0722*B)/0.7152 + 0.5]
Wherein, t is floating number, and R, G, B, Y, U, V are integers, and number of bits is BitDepth;Floor (x) is less than
Maximum integer equal to x, clamp [x] is the maximum for first taking 0 and floor (x), then takes itself and 2BitDepth- 1 reckling;
The second direct transform formula that applies a formula is
t = 0.2126*R + 0.7152*G + 0.0722*B
Y = floor (t)
U = floor (2BitDepth–1 + 0.5 * (B – t) / (1 – 0.0722))
V = floor (2BitDepth–1 + 0.5 * (R – t) / (1 – 0.2126))
The second inverse transformation formula that applies a formula is
R = clamp [Y + (Cr - 2BitDepth–1) * 2 * (1-0.2126) + 0.5]
B = clamp [Y + (Cb - 2BitDepth–1) * 2 * (1-0.0722) + 0.5]
G = clamp [(Y - 0.2126*R - 0.0722*B)/0.7152 + 0.5]
The direct transform formula that third applies a formula is
Y = (G + ((R + B) >> 1) + 1) >> 1
U = ((R - B - 1) >> 1) + (1<<(BitDepth-1))
V = (G – Y) + (1<<(BitDepth-1))
The inverse transformation formula that third applies a formula is
R = clip3(0, (1<<BitDepth) - 1, Y – V + U)
G = clip3(0, (1<<BitDepth) - 1, Y + V - (1<<(BitDepth-1)))
B = clip3(0, (1<<BitDepth) - 1, Y – V – U + (1<<BitDepth))
Wherein,<<is binary shift left operation, and>>is binary shift right operation, clip3 (0, (1<<BitDepth) -1, x)
It is first to take the maximum of 0 and x, then take itself and 2BitDepth- 1 reckling;
The 4th direct transform formula that applies a formula is
Y = (G + ((R + B) >> 1) + 1) >> 1
U = ((G - B - 1) >> 1) + (1<<(BitDepth-1))
V = ((G - R - 1) >> 1) + (1<<(BitDepth-1))
The 4th inverse transformation formula that applies a formula is
G = clip3(0, (1<<BitDepth) - 1, Y + ((U + V - (1<<BitDepth)) >> 1))
B = clip3(0, (1<<BitDepth) - 1, G - 2U + (1<<BitDepth) - 1)
R = clip3(0, (1<<BitDepth) - 1, G - 2V + (1<<BitDepth) - 1)
The 5th direct transform formula that applies a formula is
Y = (G + ((R + B) >> 1) + 1) >> 1
U = ((Y - B) >> 1) + (1<<(BitDepth-1))
V = ((Y - R) >> 1) + (1<<(BitDepth-1))
The 5th inverse transformation formula that applies a formula is
G = clip3(0, (1<<BitDepth) - 1, Y + U + V - (1<<BitDepth))
B = clip3(0, (1<<BitDepth) - 1, Y - 2U + (1<<BitDepth))
R = clip3(0, (1<<BitDepth) - 1, Y - 2V + (1<<BitDepth))。
10. coding method according to claim 1 or device or coding/decoding method according to claim 2 or dress
It sets, which is characterized in that the encoding and decoding block has one directly or indirectly or directly to mix indirectly in the compressed data stream
Coding mode identification code carries out encoding and decoding to the encoding and decoding block using corresponding color space according to the value of the identification code;
The direct coding mode identification code is made of one or more bit strings (binary symbols string) in compressed data stream;It is described
Indirect coding mode identification code is from coding derived from other syntactic elements of other codec parameters and/or compressed data stream
Pattern identification code or scheduled identification code default value;The coding mode identification code directly mixed indirectly is that part is direct (i.e.
It is made of one or more bit strings in compressed data stream) partially indirectly (i.e. from other codec parameters and/or compressed data
Stream other syntactic elements export or make a reservation for default) mixing coding mode identification code.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710410669.1A CN108989819B (en) | 2017-06-03 | 2017-06-03 | Data compression method and device adopting respective corresponding color spaces for modes |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710410669.1A CN108989819B (en) | 2017-06-03 | 2017-06-03 | Data compression method and device adopting respective corresponding color spaces for modes |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108989819A true CN108989819A (en) | 2018-12-11 |
CN108989819B CN108989819B (en) | 2023-04-07 |
Family
ID=64502706
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710410669.1A Active CN108989819B (en) | 2017-06-03 | 2017-06-03 | Data compression method and device adopting respective corresponding color spaces for modes |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108989819B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110312136A (en) * | 2019-05-10 | 2019-10-08 | 同济大学 | A kind of coding and decoding methods of pair of multi-component data |
CN112622276A (en) * | 2020-11-30 | 2021-04-09 | 北京恒创增材制造技术研究院有限公司 | Color model maximum color difference representation method for 3DP multi-color printing |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008271411A (en) * | 2007-04-24 | 2008-11-06 | Canon Inc | Image coding unit, control method of image coding unit, program and recording medium |
US20100097394A1 (en) * | 2008-10-16 | 2010-04-22 | Ming-Hsun Lu | Method for clearing blur images of a monitor |
CN103077396A (en) * | 2013-01-11 | 2013-05-01 | 上海电机学院 | Method and device for extracting vector space feature points of color image |
CN103096092A (en) * | 2013-02-07 | 2013-05-08 | 上海国茂数字技术有限公司 | Method and system of conducting coding and decoding error correction based on color space transformation |
US20150181223A1 (en) * | 2013-12-20 | 2015-06-25 | Canon Kabushiki Kaisha | Method and apparatus for transition encoding in video coding and decoding |
US20150271509A1 (en) * | 2014-03-19 | 2015-09-24 | Arris Enterprises, Inc. | Scalable coding of video sequences using tone mapping and different color gamuts |
WO2016202470A1 (en) * | 2015-06-18 | 2016-12-22 | Gurulogic Microsystems Oy | Encoder, decoder and method employing palette utilization and compression |
US20170105012A1 (en) * | 2015-10-08 | 2017-04-13 | Mediatek Inc. | Method and Apparatus for Cross Color Space Mode Decision |
US20170105011A1 (en) * | 2015-10-08 | 2017-04-13 | Mediatek Inc. | Method and Apparatus for Syntax Signaling in Image and Video Compression |
CN106657945A (en) * | 2016-12-30 | 2017-05-10 | 上海集成电路研发中心有限公司 | Nonlinear segmented gamma correction implementation method |
-
2017
- 2017-06-03 CN CN201710410669.1A patent/CN108989819B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008271411A (en) * | 2007-04-24 | 2008-11-06 | Canon Inc | Image coding unit, control method of image coding unit, program and recording medium |
US20100097394A1 (en) * | 2008-10-16 | 2010-04-22 | Ming-Hsun Lu | Method for clearing blur images of a monitor |
CN103077396A (en) * | 2013-01-11 | 2013-05-01 | 上海电机学院 | Method and device for extracting vector space feature points of color image |
CN103096092A (en) * | 2013-02-07 | 2013-05-08 | 上海国茂数字技术有限公司 | Method and system of conducting coding and decoding error correction based on color space transformation |
US20150181223A1 (en) * | 2013-12-20 | 2015-06-25 | Canon Kabushiki Kaisha | Method and apparatus for transition encoding in video coding and decoding |
US20150271509A1 (en) * | 2014-03-19 | 2015-09-24 | Arris Enterprises, Inc. | Scalable coding of video sequences using tone mapping and different color gamuts |
WO2016202470A1 (en) * | 2015-06-18 | 2016-12-22 | Gurulogic Microsystems Oy | Encoder, decoder and method employing palette utilization and compression |
US20170105012A1 (en) * | 2015-10-08 | 2017-04-13 | Mediatek Inc. | Method and Apparatus for Cross Color Space Mode Decision |
US20170105011A1 (en) * | 2015-10-08 | 2017-04-13 | Mediatek Inc. | Method and Apparatus for Syntax Signaling in Image and Video Compression |
CN106973296A (en) * | 2015-10-08 | 2017-07-21 | 联发科技股份有限公司 | Video or method for encoding images and relevant apparatus |
CN106657945A (en) * | 2016-12-30 | 2017-05-10 | 上海集成电路研发中心有限公司 | Nonlinear segmented gamma correction implementation method |
Non-Patent Citations (2)
Title |
---|
A.M. TOURAPIS: "A new tool for Color Gamut Analysis of MPEG video content", 《JCTVC》 * |
向方明: "YUV到RGB颜色空间转换算法研究", 《现代电子技术》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110312136A (en) * | 2019-05-10 | 2019-10-08 | 同济大学 | A kind of coding and decoding methods of pair of multi-component data |
CN112622276A (en) * | 2020-11-30 | 2021-04-09 | 北京恒创增材制造技术研究院有限公司 | Color model maximum color difference representation method for 3DP multi-color printing |
Also Published As
Publication number | Publication date |
---|---|
CN108989819B (en) | 2023-04-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108632607B (en) | A kind of point cloud genera compression method based on multi-angle self-adaption intra-frame prediction | |
CN105284111B (en) | Dynamic image encoding device, moving image decoding apparatus, dynamic image encoding method, dynamic image decoding method and program | |
CN105704491B (en) | Image encoding method, decoding method, encoding device, and decoding device | |
CN104702962B (en) | Decoding method, encoder and decoder in frame | |
CN104144343B (en) | A kind of digital image compression encrypts joint coding method | |
CN109076221A (en) | Method for video coding and its device | |
KR101868247B1 (en) | Image encoding and decoding method and device | |
CN108702501A (en) | The method and device that the compartmentalization luma prediction modes of colorimetric prediction for Video coding are inherited | |
CN107071450A (en) | Coding, coding/decoding method and the device of data compression | |
CN107835431A (en) | Method and apparatus to Video coding and the method and apparatus to video decoding | |
CN106030651A (en) | Advanced screen content coding with improved color table and index map coding methods | |
CN110830803B (en) | Image compression method combining block matching and string matching | |
CN105100814B (en) | Image coding and decoding method and device | |
CN106797468A (en) | Improved pallet mode in efficient video coding (HEVC) screen content coding (SCC) | |
CN113784124B (en) | Block matching encoding and decoding method for fine division using multi-shape sub-blocks | |
CN103957412B (en) | Base color and index map algorithm for correcting inter-frame residual errors of screen video | |
CN108495135A (en) | A kind of fast encoding method of screen content Video coding | |
CN110087083A (en) | The selection method of prediction mode for chroma, image processing equipment and storage equipment in frame | |
CN110087090A (en) | Using data encoding, the coding/decoding method of mixing String matching and intra prediction | |
CN110392256A (en) | Method, coding side, decoding end and the system coded and decoded | |
CN105847842A (en) | Image encoding method and apparatus and image decoding method and apparatus | |
CN104539970A (en) | 3D-HEVC interframe coding merge mode fast decision making method | |
CN105392009A (en) | Low bit rate image coding method based on block self-adaptive sampling and super-resolution reconstruction | |
CN107770540B (en) | Data compression method and device for fusing multiple primitives with different reference relations | |
CN108989819A (en) | Each mode uses the data compression method and device of respectively corresponding color space |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | 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 |