CN106303535B

CN106303535B - Image compression method and device with reference pixels taken from different-degree reconstruction pixels

Info

Publication number: CN106303535B
Application number: CN201510308137.8A
Authority: CN
Inventors: 林涛
Original assignee: Shanghai Tianhe Electronic Information Co ltd
Current assignee: Shanghai Tianhe Electronic Information Co ltd
Priority date: 2015-06-08
Filing date: 2015-06-08
Publication date: 2022-12-13
Anticipated expiration: 2035-06-08
Also published as: CN106303535A; CN115499663A

Abstract

The invention provides a method or a device for coding and decoding an image, wherein reference pixels are taken from reconstructed pixels with at least two different perfections. Preferably, in the present invention, the first part of the reference pixels are the reconstructed pixels of the perfection I, the second part of the reference pixels are the reconstructed pixels of the perfection II, the third part of the reference pixels are the reconstructed pixels of the perfection III, and so on. Preferably, the reconstructed pixels of different perfection are taken from different regions of the image. Preferably, the reference pixels of a reference pixel sample segment are comprised of reconstructed pixels of at least two different perfections. According to the invention, the reference pixel range of the copy mode is divided into at least two different regions having different levels of sophistication of the reconstructed pixels, and the reference pixels of a reference pixel sample segment corresponding to a current pixel sample segment are taken from at least two different regions having different levels of sophistication of the reconstructed pixels.

Description

Image compression method and device with reference pixels taken from different-degree reconstruction pixels

Technical Field

The invention relates to a digital video compression coding and decoding system, in particular to a method and a device for coding and decoding a composite image and a video containing a computer screen image.

Background

As televisions and displays enter ultra high definition (4K) and ultra high definition (8K) resolutions and new generation cloud computing and information processing models and platforms, typically represented by remote desktops, are developed and popularized, the demand for video image data compression is also moving towards higher resolutions and composite images including camera-captured images and computer screen images. Ultra-high compression ratio and extremely high quality data compression for video images are indispensable techniques.

The method fully utilizes the characteristics of 4K/8K images and computer screen images to carry out ultrahigh-Efficiency compression on Video images, and is also a main target of the latest international Video compression standard HEVC (High Efficiency Video Coding) and other international standards, domestic standards and industrial standards in the process of preparation.

The natural form of a digital video signal of an image is a sequence of images. A frame of image is usually a rectangular area composed of several pixels, and a digital video signal is a video image sequence composed of tens of frames to thousands of frames of images, sometimes referred to as a video sequence or sequence for short. Encoding a digital video signal is to encode one frame by one frame of image. At any one time, the frame of picture being encoded is referred to as the current encoded picture. Similarly, decoding a compressed video stream (referred to as a bitstream for short) of a digital video signal is to decode a bitstream of a frame-by-frame compressed image. At any one time, the frame of picture being decoded is referred to as the current decoded picture. The currently encoded picture or the currently decoded picture are collectively referred to as the current picture.

In encoding (and corresponding decoding) of a frame of picture in almost all international standards for video picture Coding, such as MPEG-1/2/4, h.264/AVC and HEVC, a frame of picture is divided into sub-pictures of blocks MxM pixels, called Coding blocks (i.e., decoding blocks from the decoding point of view, collectively called Coding blocks) or "Coding units (Coding units abbreviated as CUs)", and the sub-pictures are encoded block by block, with the CU as the basic Coding Unit. Commonly used sizes of M are 4,8, 16, 32, 64. Therefore, encoding a sequence of video images is to encode each coding unit, i.e., CU, of each frame image, one CU after another. At any one time, the CU being encoded is referred to as the current encoding CU. Similarly, decoding a code stream of a video image sequence also decodes each CU of each frame image one by one, and finally reconstructs the entire video image sequence. At any one time, the CU being decoded is referred to as the currently decoded CU. The current encoded CU or the current decoded CU is collectively referred to as the current CU.

In order to adapt to the difference of the content and the property of each part of the image in a frame of image, the most effective coding is carried out in a targeted manner, the size of each CU in a frame of image can be different, some are 8x8, some are 64x64, and the like. In order to enable CUs of different sizes to be seamlessly spliced, a frame of image is usually divided into "Largest Coding Units (LCUs)" having NxN pixels and having the same size, and then each LCU is further divided into a plurality of CUs having a tree structure and not necessarily the same size. Accordingly, an LCU is also referred to as a "Coding Tree Unit (CTU)". For example, one frame image is first divided into LCUs of 64 × 64 pixels (N = 64) of identical size. One LCU is composed of 3 CUs of 32 × 32 pixels and 4 CUs of 16 × 16 pixels, and thus 7 CUs in a tree structure constitute one CTU. And another LCU consists of 2 CU of 32x32 pixels, 3 CU of 16x16 pixels, and 20 CU of 8x8 pixels. Such 25 treelized CUs constitute another CTU. A frame of picture is coded, i.e. a CU in a CTU is coded in sequence. In the HEVC international standard, LCU is synonymous with CTU. A CU equal in size to a CTU is referred to as a CU with depth 0. A CU obtained by dividing a CU having a depth of 0 into upper, lower, left, and right halves is referred to as a CU having a depth of 1. A CU obtained by dividing a CU having a depth of 1 into upper, lower, left, and right halves is referred to as a CU having a depth of 2. A CU obtained by dividing a CU having a depth of 2 in upper, lower, left, and right directions is referred to as a CU having a depth of 3. At any one time, the CTU being encoded is referred to as the current encoded CTU. At any one time, the CTU being decoded is referred to as the current decoding CTU. The currently encoded CTU or the currently decoded CTU is collectively referred to as the current CTU.

A CU may also be further divided into sub-regions. Sub-regions include, but are not limited to, prediction Units (PUs), transform Units (TUs), asymmetrically partitioned (AMP) regions.

A color pixel typically consists of 3 components. The two most commonly used pixel color formats are the GBR color format, which consists of a green component, a blue component, a red component, and the YUV color format, which consists of one luminance (luma) component and two chrominance (chroma) components. The color format commonly referred to as YUV actually includes a variety of color formats, such as the YCbCr color format. Therefore, when encoding one CU, one CU may be divided into 3 component planes (G plane, B plane, R plane, or Y plane, U plane, V plane), and the 3 component planes may be encoded separately; it is also possible to combine 3 component bundles of a pixel into one 3-tuple and encode the whole CU consisting of these 3-tuples. The former arrangement of pixels and their components is called planar format (planar format) of the image (and its CU), while the latter arrangement of pixels and their components is called packed format (packed format) of the image (and its CU). Both the GBR color format and the YUV color format of a pixel are 3-component representation formats of the pixel.

In addition to the 3-component representation format of pixels, another common prior art representation format of pixels is the palette index representation format. In the palette index representation format, the value of one pixel may also be represented by the index of the palette. The palette space stores the numerical or approximate numerical values of the 3 components of the color of the pixel that needs to be represented, and the address of the palette is referred to as the index of the color of the pixel stored in this address. One index may represent one component of the color of a pixel, and one index may also represent 3 components of the color of a pixel. The palette may be one or more. In the case of multiple palettes, a complete index is actually composed of both a palette index (indicating which one of the multiple palettes) and an index of the palette of that index. The index representation format of a pixel is to represent the pixel by an index. If pixels in an image region (e.g., a coded block or a decoded block) cannot all be represented by palette colors (i.e., for at least one pixel in the image region, there is no palette color and its index where the 3 components have the same or approximately the same value as the pixel), then there is usually a special index in the palette called escape color to represent pixels that cannot be represented by normal palette colors. Therefore, if the index of a pixel is the index of an escape color, the pixel needs to represent its color with an additional dedicated 3 components. The normal color and the escape color in the palette are called palette colors, but the escape color is a virtual color, there is no physical space in the palette for storing the color, and there is only a special virtual index. The index of the escape color is typically the last index of the palette. The index representation format of a pixel is also known in the art as the index color (extended color) or pseudo color (pseudo color) representation format of the pixel, or often directly as an index pixel (extended pixel) or pseudo pixel (pseudo pixel) or pixel index or index. The index is sometimes also referred to as an index. Rendering pixels in their indexed rendering format is also referred to as indexing or indexing.

Other common prior art pixel representation formats include the CMYK representation format and the grayscale representation format.

The YUV color format can be subdivided into a number of seed formats depending on whether the chrominance components are down-sampled: 1 pixel is YUV4:4 pixel color format composed of 1Y component, 1U component and 1V component; 2 pixels adjacent to the left and right are in YUV4:2 pixel color format consisting of 2Y components, 1U component and 1V component; the left, right, upper and lower adjacent 4 pixels arranged according to 2x2 spatial positions are in YUV4:2 pixel color format consisting of 4Y components, 1U component and 1V component. One component is typically represented by 1 number of 8 to 16 bits. The YUV4: 2. A pixel component is also referred to as a pixel sample (sample) or simply a sample (sample).

The most basic element in encoding or decoding may be one pixel, one pixel component, or one pixel index (i.e., index pixel). A pixel or a pixel component or an index pixel, which is the most basic element of encoding or decoding, is collectively referred to as a pixel sample, sometimes also commonly referred to as a pixel value, or simply a sample.

In the present invention and the present patent application, the terms "pixel sample", "pixel value", "sample", "index pixel" and "pixel index" are synonymous, and depending on the context, it may be clear whether "pixel" or "one pixel component" or "index pixel" or any of the three at the same time. If it is not clear from the context, any of the three is represented at the same time.

In the present invention and the present patent application, a coding block or a decoding block (collectively called a coding block) is a region composed of several pixel values. The shape of the codec block may be rectangular, square, parallelogram, trapezoid, polygon, circle, ellipse, and other various shapes. The rectangle also includes a rectangle whose width or height is one pixel value and degenerates into a line (i.e., a line segment or a line shape). Each codec block may have a different shape and size from one another in a frame image. In a frame of image, some or all of the coding and decoding blocks may have mutually overlapped parts, or all of the coding and decoding blocks may not overlap each other. A coding/decoding block may be composed of "pixels", or "components of pixels", or "index pixels", or a mixture of these 3, or any 2 of these 3. From the perspective of video image encoding or decoding, a coding/decoding block refers to a region of a frame of image for which encoding or decoding is performed, and includes but is not limited to at least one of the following: a maximum coding unit LCU, a coding tree unit CTU, a coding unit CU, a sub-region of a CU, a prediction unit PU, a transform unit TU.

A significant feature of computer screen images is that there are usually many similar or even identical pixel patterns (pixel patterns) within the same frame image. For example, chinese characters or foreign language characters frequently appearing in computer screen images are composed of a few basic strokes, and many similar or identical strokes can be found in the same frame of image. Menus, icons, etc. that are commonly found in computer screen images also have many similar or identical patterns. Therefore, various copying methods are generally adopted in the existing image and video compression technology, including at least the following copying methods:

1) Intra block copy is intra block matching or intra motion compensation or block matching or block copy. The basic operation of block copy encoding or decoding is to copy, for a current encoded block or a current decoded block (referred to simply as a current block), a reference block of the same size (the same number of pixel samples) as the current block from the reconstructed set of reference pixel samples, and assign the value of the reference block to the current block. The copy parameter of the block copy manner includes a displacement vector of the current block, indicating a relative position between the reference block and the current block. A current block has a displacement vector.

2) The intraframe micro block copying is intraframe micro block matching or micro block copying. In the microblock copy, a current block (e.g., 8x8 pixel sample) is divided into several microblocks (e.g., a microblock of 4x2 pixel sample or a microblock of 8x2 pixel sample or a microblock of 2x4 pixel sample or a microblock of 2x8 pixel sample), and the basic operation of microblock copy encoding or decoding is to copy a reference microblock from within a reconstructed set of reference pixel samples and assign the value of the reference microblock to the current microblock for each encoded or decoded microblock in the current block (referred to simply as the current microblock). The copy parameter of the micro-block copy mode comprises a displacement vector of the current micro-block, and represents the relative position between the reference micro-block and the current micro-block. A current tile has a displacement vector. There are as many displacement vectors as there are as many tiles into which a current block is divided.

3) The intra-frame line (or stripe) copy is the intra-frame stripe matching or stripe copying. A bar is a tile of height 1 or width 1, such as a tile of 4x1 or 8x1 or 1x4 or 1x8 pixel samples. The basic operation of slice copy coding or decoding is to copy a reference slice from within the reconstructed set of reference pixel samples for each coded or decoded slice (referred to as the current slice) in the current block and assign the value of the reference slice to the current slice. Obviously, stripe replication is a special case of micro-tile replication. The replication parameters of the stripe replication approach include a displacement vector of the current stripe, representing the relative position between the reference stripe and the current stripe. A current bar has a displacement vector. There are as many displacement vectors as there are how many slices into which a current block is divided.

4) Intra-string replication, i.e., intra-string matching or pixel string replication. In pixel string replication, a currently encoded block or a currently decoded block (referred to simply as the current block) is divided into several variable-length pixel sample strings. Here, a string is a string of pixels in a two-dimensional region of an arbitrary shape arranged to have a length much greater than its width (e.g., a string of 1 pixel sample in width and 37 pixel samples in length or a string of 2 pixel samples in width and 111 pixel samples in length, typically but not limited to, a length being an independent encoding or decoding parameter and a width being a predetermined or derived parameter from other encoding or decoding parameters). The basic operation of string copy coding or decoding is to copy, for each coded or decoded string in the current block (called simply the current string), a reference string from within the reconstructed set of reference pixel samples and assign the value of the reference string to the current string. The copy parameters of the string copy mode include the displacement vector and copy length, i.e., copy size, of the current string, which respectively represent the relative position between the reference string and the current string and the length, i.e., the number of pixel samples, of the current string. The length of the current string is also the length of the reference string. A current string has a displacement vector and a copy length. There are as many displacement vectors and as many copy lengths as there are strings into which a current block is divided.

5) Palette index string replication, or palette or index string replication. In the palette coding and corresponding decoding method, a palette is first constructed or obtained, then part or all of the pixels of a current coding block or a current decoding block (referred to as a current block) are represented by indexes of the palette, and then the indexes are coded and decoded, including but not limited to: the index of a current block is divided into several index strings with variable length, namely, the index string copy coding and decoding are carried out. The basic operation of index string copy coding or decoding is to copy a reference index string from the set of indexed reconstructed reference pixel samples for each index coded string or index decoded string in the current block (referred to as the current index string for short), and assign the index value of the reference index string to the current index string. The copy parameters of the index string copy mode include a displacement vector and a copy length, i.e., a copy size, of the current index string, which respectively represent a relative position between the reference index string and the current index string and a length of the current index string, i.e., the number of corresponding pixel samples. The length of the current index string is also the length of the reference index string. A current index string has a displacement vector and a copy length. There are as many displacement vectors and as many copy lengths as there are index strings into which a current block is divided.

6) The index string replication and pixel string replication mixed fusion replication mode is called an index-pixel string fusion replication mode for short. When a current coding block or a current decoding block (called as a current block for short) is coded or decoded, a pixel string copy mode is adopted for part or all pixels, and an index string copy mode is adopted for part or all pixels.

Other copy methods include a rectangular copy method, a copy method in which a plurality of copy methods are mixed, and the like.

The block in the block copy mode, the micro block in the micro block copy mode, the bar in the bar copy mode, the string in the string copy mode, the rectangle in the rectangle copy mode, and the pixel index string in the palette index mode are collectively referred to as a pixel sample segment, which is simply referred to as a sample segment. The basic constituent element of a sample segment is a pixel or pixel component or pixel index. A sample segment has a copy parameter indicative of a relationship between the current pixel sample segment and the reference pixel sample segment. Therefore, one sample segment is the minimum unit of one copy operation having the same copy relationship. A replication parameter comprises a number of replication parameter components, the replication parameter components including at least: displacement vector horizontal component, displacement vector vertical component, 1-dimensional displacement vector, linear address, relative linear address, index, palette linear address, relative index, palette relative linear address, copy length, copy width, copy height, rectangle width, rectangle length, unmatched pixel (also known as no reference pixel, i.e., non-copied pixel that is not copied from elsewhere).

In various copy schemes, the pixel samples or indices need to be arranged in a certain order. The arrangement is also referred to as a scanning mode. The scanning methods can be classified into the following types according to the path shape:

a) The horizontal zigzag scanning mode is also called a horizontal raster scanning mode. The pixel samples or indices of a coding block or decoding block (collectively referred to as a coding block) are arranged in rows and columns, all in the same direction (all left to right or all right to left) within all rows. The rows can be arranged from top to bottom or from bottom to top.

B) The vertical zigzag scanning is also called vertical raster scanning. The pixel samples or indices of a coding block or decoding block (collectively referred to as a coding block or a decoding block) are arranged in columns, and are arranged in the same direction (all from top to bottom or all from bottom to top) in all columns. The columns may be arranged from left to right or from right to left.

C) A horizontal arcuate scan pattern. The pixel samples or indices of a coding block or decoding block (collectively referred to as a coding block) are arranged in one row, in one direction (e.g., from left to right) in odd rows and in the other (opposite) direction (e.g., from right to left) in even rows. The rows can be arranged from top to bottom or from bottom to top.

D) A vertical arcuate scan pattern. The pixel samples or indices of a coding block or decoding block (collectively referred to as a coding block) are arranged in columns in one direction (e.g., top-down) in odd columns and in the other (opposite) direction (e.g., bottom-up) in even columns. The columns may be arranged from left to right or from right to left.

It should be noted that "copy" is an operation of reconstruction and decoding, and the corresponding encoding operation is "match". Therefore, various copying methods such as a block matching method, a micro-block copying method, a line copying method, a pixel string copying method, an index string copying method, and the like are also referred to as a block matching method, a micro-block matching method, a line matching method, a pixel string matching method, an index string matching method, and the like.

In various existing copy modes, a reference pixel is an imperfect reconstructed pixel which is neither processed by a Deblocking Filtering (DF) step nor a Sample Adaptive Offset (SAO) step. There may be a large error between the imperfect reconstructed pixel and the final perfect reconstructed pixel, resulting in a large error between the reference pixel and the original pixel, which reduces the compression efficiency of the image.

Disclosure of Invention

To solve this problem in the prior art of image video encoding and decoding, the present invention provides a method or apparatus for image encoding and decoding in which reference pixels are taken from at least two reconstructed pixels of different perfection. That is, a first portion of the reference pixels are refinement I reconstructed pixels, a second portion of the reference pixels are refinement II reconstructed pixels, a third portion of the reference pixels are refinement III reconstructed pixels, and so on. Preferably, reconstructed pixels of different perfection are taken from different locations, i.e. different regions, of the image. Preferably, the reference pixels of a reference pixel sample segment are comprised of reconstructed pixels of at least two different perfections. Preferably, the reference pixels of a reference pixel sample segment are taken from at least two different regions of the image, the different regions having different degrees of perfection of the reconstructed pixels. According to the invention, the reference pixel range of the copy mode is divided into at least two different regions having reconstructed pixels of different perfection. According to the invention, in the copy mode, reference pixels of a reference pixel sample segment corresponding to a current pixel sample segment of a current coding or decoding block (collectively referred to as a coding or decoding block) are taken from at least two different regions having different levels of perfection of reconstructed pixels.

The invention is characterized in that the reference pixel is taken from K (K is more than or equal to 2, and usually K is less than or equal to 4) reconstructed pixels with different perfection degrees. For example, as shown in fig. 1, the reconstructed pixels are taken from the following 3 different perfections:

1) A reconstructed pixel that has not been subjected to either DF or SAO processing;

2) Reconstructed pixels processed by vertical edge DF;

3) All DF and SAO processed reconstructed pixels.

In the present invention, preferably, as shown in FIG. 1, a reference pixel sample segment (one reference pixel sample string or another reference pixel sample block in FIG. 1) corresponding to a current pixel sample segment (one current pixel sample string or another current pixel sample block in FIG. 1) in the current CU is composed of K (K ≧ 2, usually K ≦ 4) different perfection reconstruction pixels. That is, a portion of the reference pixels of the reference pixel sample segment are reconstructed pixels of perfection I, and the remaining portion of the reference pixels are reconstructed pixels of other K-1 perfections different from the perfection I.

In the present invention, it is preferred that the legal desirable reference pixel range for a current segment of pixel samples in the current CU is comprised of K (K ≧ 2, usually K ≦ 4) regions of different perfection reconstructed pixels. For example, as shown in FIG. 1, it is composed of 3 regions.

In the present invention, preferably, the reference pixels of a reference pixel sample segment corresponding to a current pixel sample segment in the current CU are taken from K (K ≧ 2, usually K ≦ 4) regions of reconstructed pixels of different degrees of sophistication in the legal and desirable reference pixel range. For example, as shown in FIG. 1, the reference pixels of one reference pixel sample string are taken from regions of 3 different perfection reconstructed pixels, while the reference pixels of another reference pixel sample block are taken from regions of 2 different perfection reconstructed pixels.

The most basic unique technical characteristic of the coding method or the device is that when a current coding block is coded in a copying mode, reference pixels of copying operation are taken from K (K is more than or equal to 2, and usually K is less than or equal to 4) reconstructed pixels with different perfection degrees. Fig. 2 is a schematic diagram of an encoding method or apparatus of the present invention. Preferably, the replication means is one of the following means or a fusion thereof: prediction mode (including intra-frame prediction or inter-frame prediction), intra-frame block copy mode, micro-block copy mode, stripe copy mode, string copy mode, index string copy mode; the copy operation is one of the following operations or a fusion thereof: prediction operation, intra block copy operation, micro block copy operation, stripe copy operation, string copy operation, index string copy operation; accordingly, the reference pixels are reference pixels in a predictor (block), a reference block, a reference micro-block, a reference strip, a reference string, a palette, respectively.

The most basic special technical characteristic of the decoding method or the device is that when a video code stream of a current decoding block is decoded in a copying mode, reference pixels of copying operation are taken from K (K is more than or equal to 2, and usually K is less than or equal to 4) reconstructed pixels with different perfection degrees. Fig. 3 is a schematic diagram of a decoding method or apparatus of the present invention. Preferably, the replication mode is one of the following modes or a fusion thereof: prediction mode (including intra-frame prediction or inter-frame prediction), intra-frame block copy mode, micro-block copy mode, stripe copy mode, string copy mode, index string copy mode; the copy operation is one of the following operations or a fusion thereof: prediction operation, intra block copy operation, micro block copy operation, stripe copy operation, string copy operation, index string copy operation; accordingly, the reference pixels are reference pixels in a predictor (block), a reference block, a reference micro-block, a reference strip, a reference string, a palette, respectively.

According to an aspect of the present invention, there is provided an image encoding method or apparatus, including at least steps or modules for performing the following functions and operations:

at least 2 reconstructed pixels with different perfection degrees are used as reference pixels to carry out copy coding on the current coding block, and a video code stream containing information of copy parameters is generated.

According to another aspect of the present invention, there is also provided an image decoding method or apparatus, including at least the steps or modules for performing the following functions and operations:

analyzing the video code stream to obtain the information of the copy parameters, and copying and decoding the current decoding block by adopting at least 2 reconstructed pixels with different perfection degrees as reference pixels.

The invention is suitable for encoding and decoding the images in the pack format. The present invention is also equally applicable to encoding and decoding of component plane format images.

The technical features of the present invention are explained above by specific embodiments. Other advantages and effects of the present invention will be readily apparent to those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention.

Drawings

FIG. 1 is a schematic representation of the present invention with reference pixels taken from a plurality of different perfection reconstructed pixels.

Fig. 2 is a schematic diagram of an encoding method or apparatus of the present invention.

Fig. 3 is a schematic diagram of a decoding method or apparatus of the present invention.

Detailed Description

The following are further implementation details or variations of the present invention.

Examples or modifications 1

In the encoding method or apparatus or the decoding method or apparatus, the duplication encoding or decoding is one of the following encoding or decoding modes or a combination thereof: predictive encoding or decoding, intra block copy encoding or decoding, micro block copy encoding or decoding, slice copy encoding or decoding, string copy encoding or decoding, index string copy encoding or decoding; the copy operation is one of the following operations or a fusion thereof: prediction operation, intra block copy operation, micro block copy operation, stripe copy operation, string copy operation, index string copy operation; accordingly, the reference pixels are reference pixels in a predictor (block), a reference block, a reference micro-block, a reference strip, a reference string, a palette, respectively.

Examples of embodiment or modification 2

In the coding method or device or the decoding method or device, the legal and desirable reference pixel range is composed of K (K is more than or equal to 2, and usually K is less than or equal to 4) regions with different degrees of perfection for reconstructing pixels.

Examples of embodiment or modification 3

In the coding method or device or the decoding method or device, a legal and desirable reference pixel range of one current pixel sample value segment in the current coding block or the current decoding block consists of K (K is more than or equal to 2, and usually K is less than or equal to 4) regions with reconstructed pixels with different perfection degrees.

Examples of embodiment or modification 4

In the encoding method or device or the decoding method or device, the reference pixel of a reference pixel sample segment corresponding to a current pixel sample segment in the current encoding block or decoding block is taken from K (K is more than or equal to 2, and usually K is less than or equal to 4) regions of reconstructed pixels with different perfection degrees in a legal and desirable reference pixel range.

Examples of embodiment or modification 5

In the encoding method or apparatus or the decoding method or apparatus, the at least 2 reconstructed pixels with different perfection degrees at least include at least 2 reconstructed pixels of the following 3 reconstructed pixels:

1) Perfecting a reconstructed pixel of the degree I;

2) Perfecting reconstructed pixels of degree II;

3) And perfecting the reconstructed pixels of the degree III.

Examples of embodiment or modification 6

In the encoding method or apparatus or the decoding method or apparatus according to embodiment or variation 5, the reconstructed pixels of the perfection degrees I, II, and III are:

1) Reconstructed pixels that have not been DF nor SAO processed;

2) Reconstructing the pixel after vertical edge DF treatment;

3) All DF and SAO processed reconstructed pixels.

Examples of embodiment or modification 7

In the encoding method or apparatus or the decoding method or apparatus according to embodiment or variation 5 or 6, the reconstruction pixels of the perfection levels I, II, III are respectively from the following regions in the current image:

1) The current CTU and the rightmost four columns of CTUs to the left and more left of the current CTU;

2) Four rows above the area not yet reconstructed or the area of 1);

3) The legal desirable reference pixel range does not belong to the area of either 1) or 2).

Claims

1. An image encoding method, characterized by comprising at least the steps of:

1) Constructing at least 2 reconstructed pixels with different perfection degrees;

2) Adopting at least the reconstructed pixel as a reference pixel to carry out coding at least in a copying mode on the current coding block;

3) Generating a video code stream at least containing information of the copy parameter;

the at least 2 different levels of sophistication include at least one level of sophistication that has not been subjected to at least deblocking filtering and sample adaptive compensation.

2. An image encoding apparatus, comprising at least the following modules:

1) The reconstruction pixel construction module is used for constructing at least 2 reconstruction pixels with different perfection degrees;

2) The duplication encoding module is used for encoding the current encoding block at least in a duplication mode by using at least the reconstruction pixel as a reference pixel;

3) The video code stream generating module is used for generating a video code stream at least containing information of the copy parameters;

the at least 2 different perfection levels include at least one perfection level that has not been subjected to at least deblocking filtering and sample adaptive compensation processing.

3. An image decoding method, comprising at least the steps of:

1) Analyzing the video code stream to obtain information at least containing copy parameters;

2) Constructing at least 2 reconstructed pixels with different perfection degrees;

3) Decoding the current decoding block at least in a copy mode by using at least the copy parameter and at least the reconstructed pixel as a reference pixel;

4. The decoding method according to claim 3, wherein:

the decoded block is a decoded region of the picture, including at least one of: a largest coding unit LCU, a coding tree unit CTU, a coding unit CU, a sub-region of a CU, a prediction unit PU.

5. The decoding method according to claim 3, wherein:

the decoding adopting the copy mode is one of the following decoding modes or the combination thereof: intra block copy decoding, micro block copy decoding, stripe copy decoding, string copy decoding, index string copy decoding;

accordingly, the operation involved in the replication is one of the following or a fusion thereof: intra block copy operation, micro block copy operation, stripe copy operation, string copy operation, index string copy operation;

accordingly, the reference pixel is one of the following or a fusion thereof: reference pixels of a reference block, reference pixels of a reference micro-block, reference pixels of a reference strip, reference pixels of a reference string, reference pixels in a palette.

6. The decoding method according to claim 3, wherein:

a legal desirable reference pixel range consists of regions of at least two different perfection reconstructed pixels.

7. The decoding method according to claim 3, wherein:

a legal and desirable reference pixel range for a current segment of pixel samples in said current decoded block is comprised of regions of at least two different perfection reconstructed pixels.

8. The decoding method according to claim 3, wherein:

the reference pixels of a reference pixel sample segment corresponding to a current pixel sample segment in said current decoding block are taken from regions of at least two different perfection reconstructed pixels in a legal desirable reference pixel range.

9. The decoding method according to claim 3, wherein:

the at least 2 different perfection reconstructed pixels comprise at least 2 of the following 3 reconstructed pixels:

1) A first perfection of reconstructed pixels;

2) A second perfection of reconstructed pixels;

3) A third perfection of reconstructed pixels.

10. The decoding method according to claim 9, wherein:

the reconstructed pixels with the first perfection degree, the reconstructed pixels with the second perfection degree and the reconstructed pixels with the third perfection degree are respectively as follows:

2) Reconstructing the pixel after vertical edge DF treatment;

3) All DF and SAO processed reconstructed pixels.

11. The decoding method according to claim 9 or claim 10, characterized in that:

the first perfection degree reconstructed pixel, the second perfection degree reconstructed pixel and the third perfection degree reconstructed pixel are respectively from the following areas in the current image:

1) The current CTU and the CTU positioned on the left side of the current CTU are called as a left CTU, and the rightmost four columns in the CTUs positioned on the left side of the left CTU are called as left CTUs;

2) Four rows above the region that has not been reconstructed or the region of the 1);

12. The decoding method according to claim 3, wherein:

the copying mode is the fusion of an inter-frame prediction mode and an intra-frame block copying mode;

the at least 2 reconstructed pixels with different perfections at least comprise the following 2 reconstructed pixels:

1) The first perfection of reconstructed pixels is those that have not been both DF nor SAO processed;

2) The second perfection of reconstructed pixels is the reconstructed pixels processed by all DF and SAO;

the reconstructed pixels of different perfection degrees are respectively taken from different areas of the image.

13. An image decoding apparatus, comprising at least the following modules:

1) The video code stream analyzing module analyzes the video code stream to obtain information at least containing the copy parameters;

2) The reconstruction pixel construction module is used for constructing at least 2 reconstruction pixels with different perfection degrees;

3) A copy decoding module, which uses at least the copy parameter and adopts at least the reconstructed pixel as a reference pixel to decode the current decoding block at least in a copy mode;

14. The decoding device according to claim 13, wherein:

the decoded block is a decoded region of the picture, including at least one of: maximum coding unit LCU, coding tree unit CTU, coding unit CU, sub-regions of CU, prediction unit PU.

15. The decoding device according to claim 13, wherein:

the decoding adopting the copy mode is one of the following decoding modes or the combination thereof: intra block copy decoding, micro block copy decoding, strip copy decoding, string copy decoding, index string copy decoding;

16. The decoding apparatus according to claim 13, wherein:

17. The decoding device according to claim 13, wherein:

a legal desirable reference pixel range for a current segment of pixel samples in said current decoding block is comprised of regions of at least two different perfection reconstructed pixels.

18. The decoding apparatus according to claim 13, wherein:

the reference pixels of a reference pixel sample segment corresponding to a current pixel sample segment in said current decoded block are taken from regions of at least two different perfection reconstructed pixels in a legal, desirable range of reference pixels.

19. The decoding apparatus according to claim 13, wherein:

1) A first perfection of reconstructed pixels;

2) A second perfection of reconstructed pixels;

3) A third perfection of reconstructed pixels.

20. The decoding apparatus according to claim 19, wherein:

1) Reconstructed pixels that have not been DF nor SAO processed;

2) Reconstructed pixels processed by vertical edge DF;

3) All DF and SAO processed reconstructed pixels.

21. The decoding apparatus according to claim 19 or claim 20, wherein:

the first perfection reconstructed pixel, the second perfection reconstructed pixel and the third perfection reconstructed pixel are respectively from the following areas in the current image:

1) The current CTU and the CTU positioned on the left side of the current CTU are called as a left CTU, and the rightmost four columns in the CTUs positioned on the left side of the left CTU are called as the right CTU;

22. The decoding apparatus according to claim 13, wherein:

1) The reconstructed pixels of the first perfection are reconstructed pixels which are not subjected to DF nor SAO processing;

2) The reconstructed pixels of the second perfection are the reconstructed pixels processed by all DF and SAO;