Image compression method and apparatus using single coding mode referring to different levels of reconstructed pixels

Abstract

This invention provides a method or apparatus for image encoding and decoding that integrates inter-frame prediction and intra-frame block copying into a single coding mode and uses at least two reconstructed pixels with different levels of perfection as reference pixels. The reconstructed pixels with different levels of perfection are taken from different regions of the reference pixel range. The reference pixel range of the coding mode is divided into at least two regions of reconstructed pixels with different levels of perfection. The first level of perfection reconstructed pixels are those that have not undergone either DF or SAO processing, while the other levels of perfection reconstructed pixels are those that have undergone DF and / or SAO processing.

Description

[0001] This application is a divisional application of the following original application:

[0002] The original application was filed on June 8, 2015.

[0003] Original application number: 2015103081378

[0004] The original invention application was titled: "Image Compression Method and Apparatus with Reference Pixels Taken from Pixels Reconstructed to Different Degrees." Technical Field

[0005] This invention relates to a digital video compression encoding and decoding system, and in particular to a method and apparatus for encoding and decoding composite images and videos containing computer screen images. Background Technology

[0006] With televisions and monitors reaching ultra-high definition (4K) and ultra-high definition (8K) resolutions, and the development and widespread adoption of next-generation cloud computing and information processing models and platforms, typified by remote desktops, the demand for video image data compression is shifting towards higher resolutions and composite images that include both camera-captured images and computer screen images. Ultra-high compression ratios and extremely high-quality data compression for video images have become essential technologies.

[0007] Making full use of the characteristics of 4K / 8K images and computer screen images to compress video images with ultra-high efficiency is also a major goal of the latest international video compression standard HEVC (High Efficiency Video Coding) currently under development, as well as several other international, domestic, and industry standards.

[0008] The natural form of a digital video signal is a sequence of images. A single frame is typically a rectangular area composed of several pixels, while a digital video signal is a sequence of video images ranging from tens to thousands of frames, sometimes simply referred to as a video sequence or sequence. Encoding a digital video signal means encoding each frame individually. At any given moment, the frame being encoded is called the currently encoded image. Similarly, decoding the compressed video stream (also called a bitstream) of a digital video signal means decoding each frame of the compressed bitstream. At any given moment, the frame being decoded is called the currently decoded image. Both the currently encoded image and the currently decoded image are collectively referred to as the current image.

[0009] In almost all international video image coding standards such as MPEG-1 / 2 / 4, H.264 / AVC, and HEVC, when encoding (and correspondingly decoding) a frame of image, the frame is divided into several MxM pixel sub-images, called coding blocks (or decoding blocks from a decoding perspective, collectively referred to as code-decode blocks) or "coding units (CUs)". The CU is used as the basic coding unit, and each sub-image is encoded block by block. Commonly used sizes for M are 4, 8, 16, 32, and 64. Therefore, encoding a video image sequence involves encoding each coding unit (CU) of each frame sequentially, one CU at a time. At any given moment, the CU currently being encoded is called the current coding CU. Similarly, decoding a video image sequence's bitstream also involves decoding each CU of each frame sequentially, one CU at a time, ultimately reconstructing the entire video image sequence. At any given moment, the CU currently being decoded is called the current decoding CU. The current coding CU or the current decoding CU is collectively referred to as the current CU.

[0010] To accommodate the varying content and properties of different parts of an image frame and to perform targeted and efficient encoding, the size of each CU within an image frame can differ, such as 8x8, 64x64, etc. To ensure seamless stitching of CUs of different sizes, an image frame is typically first divided into "Largest Coding Units" (LCUs) of identical size (NxN pixels). Each LCU is then further divided into multiple CUs of varying sizes in a tree structure. Therefore, LCUs are also called "Coding Tree Units" (CTUs). For example, an image frame might be first divided into identical 64x64 pixel LCUs (N=64). One LCU might consist of three 32x32 pixel CUs and four 16x16 pixel CUs, forming a tree structure of seven CUs that constitute one CTU. Another LCU might consist of two 32x32 pixel CUs, three 16x16 pixel CUs, and twenty 8x8 pixel CUs. These 25 tree-structured CUs constitute another CTU. Encoding a frame of image involves sequentially encoding each CU within a CTU. In the HEVC international standard, LCU and CTU are synonyms. A CU of size equal to the CTU is called a CU with depth 0. A CU with depth 0 divided into four equal parts (top, bottom, left, and right) is called a CU with depth 1. A CU with depth 1 divided into four equal parts (top, bottom, left, and right) is called a CU with depth 2. A CU with depth 2 divided into four equal parts (top, bottom, left, and right) is called a CU with depth 3. At any given moment, the CTU currently encoding is called the current encoding CTU. At any given moment, the CTU currently decoding is called the current decoding CTU. The current encoding CTU or the current decoding CTU is collectively referred to as the current CTU.

[0011] The CU can be further divided into several sub-regions. Sub-regions include, but are not limited to, the regions of prediction units (PU), transform units (TU), and asymmetric partitions (AMP).

[0012] A color pixel is typically composed of three components. The two most commonly used pixel color formats are the GBR color format, which consists of green, blue, and red components, and the YUV color format, which consists of one luma component and two chroma components. The commonly referred to YUV color format actually includes several other color formats, such as YCbCr. Therefore, when encoding a CU (Color Cube), it can be divided into three component planes (G plane, B plane, R plane or Y plane, U plane, V plane), and each plane can be encoded separately; alternatively, the three components of a pixel can be bundled into a triplet, and the CU composed of these triplets can be encoded as a whole. The former arrangement of pixels and their components is called the planar format of the image (and its CU), while the latter is called the packed format. Both the GBR and YUV color formats are three-component representations of pixels.

[0013] Besides the three-component pixel representation format, another commonly used existing pixel representation format is the palette indexed representation format. In the palette indexed representation format, the value of a pixel can also be represented using a palette index. The palette space stores the values ​​or approximate values ​​of the three components of the color of the pixel to be represented, and the address of the palette is called the index of the pixel's color stored at that address. An index can represent one component of a pixel's color, or it can represent all three components. There can be one palette or multiple palettes. In the case of multiple palettes, a complete index actually consists of two parts: the palette number (indicating which of the multiple palettes) and the index of that palette. The pixel indexed representation format uses the index to represent the pixel. If not all pixels in an image region (such as an encoded or decoded block) can be represented using palette colors (i.e., for at least one pixel in the image region, there is no palette color and its index whose three component values ​​are equal to or approximately equal to that pixel), then the palette typically has a special index called the escape color to represent pixels that cannot be represented by normal palette colors. Therefore, if a pixel's index is an escape color index, that pixel needs to be represented using additional dedicated three components. Both normal colors and escape colors in the palette are called palette colors, but the escape color is a virtual color; there is no physical space to store this color in the palette, only a special virtual index. The escape color index is usually the last index of the palette. The pixel index representation format is also known in existing technology as indexed color or pseudo color representation format, or often simply as indexed pixel, pseudo pixel, pixel index, or index. The index is sometimes also called the exponent. Representing pixels using their indexed representation format is also known as indexing or exponentiation.

[0014] Other commonly used existing pixel representation formats include CMYK representation format and grayscale representation format.

[0015] The YUV color format can be further subdivided into several sub-formats based on whether the chroma components are downsampled: YUV4:4:4 pixel color format, where one pixel consists of one Y component, one U component, and one V component; YUV4:2:2 pixel color format, where two adjacent pixels consist of two Y components, one U component, and one V component; and YUV4:2:0 pixel color format, where four adjacent pixels arranged in a 2x2 spatial position consist of four Y components, one U component, and one V component. A component is typically represented by an 8-16 bit number. Both YUV4:2:2 and YUV4:2:0 pixel color formats are obtained by downsampling the chroma components of the YUV4:4:4 pixel color format. A pixel component is also called a pixel sample or simply a sample.

[0016] The most basic element in encoding or decoding can be a pixel, a pixel component, or a pixel index (i.e., an indexed pixel). A pixel, a pixel component, or an indexed pixel that serves as the most basic element in encoding or decoding is collectively referred to as a pixel sample, sometimes also simply as a pixel value, or simply as a sample.

[0017] In this invention and its patent application, "pixel sample value," "pixel value," "sample value," "indexed pixel," and "pixel index" are synonyms. Depending on the context, it can be clearly determined whether they refer to "pixel," "a pixel component," "indexed pixel," or any of the three. If the context does not specify, then they refer to any of the three.

[0018] In this invention and its patent application, an encoding block or decoding block (collectively referred to as an encoding / decoding block) is a region composed of several pixel values. The shape of an encoding / decoding block can be rectangular, square, parallelogram, trapezoidal, polygonal, circular, elliptical, and various other shapes. A rectangle also includes a rectangle whose width or height is a single pixel value, degenerated into a line (i.e., a line segment or line shape). In a frame of an image, each encoding / decoding block can have different shapes and sizes. In a frame of an image, some or all encoding / decoding blocks can overlap, or all encoding / decoding blocks can be non-overlapping. An encoding / decoding block can be composed of "pixels," "pixel components," "index pixels," a mixture of these three, or any two of these three. From the perspective of video image encoding or decoding, an encoding / decoding block refers to a region in a frame of an image that is encoded or decoded, including but not limited to at least one of the following: Maximum Coding Unit (LCU), Coding Tree Unit (CTU), Coding Unit (CU), sub-regions of a CU, Prediction Unit (PU), and Transform Unit (TU).

[0019] A prominent characteristic of computer screen images is the presence of numerous similar or even identical pixel patterns within a single frame. For example, Chinese or foreign language characters frequently appearing on computer screens are composed of only a few basic strokes, and many similar or identical strokes can be found within the same frame. Menus and icons commonly found on computer screens also exhibit many similar or identical patterns. Therefore, existing image and video compression technologies typically employ various copying methods, including at least the following:

[0020] 1) Intra-frame block copying, also known as intra-frame block matching or intra-frame motion compensation, is a basic operation of block copying encoding or decoding. It involves copying a reference block of the same size (the same number of pixel samples) from the reconstructed reference pixel sample set to a current coded block or current decoded block (referred to as the current block), and then assigning the values ​​of the reference block to the current block. The copying parameters include the displacement vector of the current block, representing the relative position between the reference block and the current block. Each current block has one displacement vector.

[0021] 2) Intra-frame micro-block copying, also known as intra-frame micro-block matching or micro-block copying, involves dividing a current block (e.g., an 8x8 pixel sample) into several micro-blocks (e.g., 4x2, 8x2, 2x4, or 2x8 pixel samples). The basic operation for micro-block copying encoding or decoding is to copy a reference micro-block from the reconstructed reference pixel sample set for each encoded or decoded micro-block (referred to as the current micro-block) in the current block, and then assign the value of the reference micro-block to the current micro-block. The copying parameters for micro-block copying include the displacement vector of the current micro-block, representing the relative position between the reference micro-block and the current micro-block. Each current micro-block has one displacement vector. The number of displacement vectors corresponds to the number of micro-blocks into which the current block is divided.

[0022] 3) Intra-frame line (strip) copying, also known as intra-frame strip matching or strip copying, involves using a strip as a micro-block with a height or width of 1, such as a 4x1, 8x1, 1x4, or 1x8 pixel sample micro-block. The basic operation of strip copying encoding or decoding is to copy a reference strip from the reconstructed reference pixel sample set for each encoded or decoded strip (referred to as the current strip) in the current block, and then assign the value of the reference strip to the current strip. Clearly, strip copying is a special case of micro-block copying. The copying parameters for strip copying include the displacement vector of the current strip, representing the relative position between the reference strip and the current strip. Each current strip has one displacement vector. The number of displacement vectors corresponds to the number of strips in a current block.

[0023] 4) Intra-frame string copying, also known as intra-frame string matching, string matching, string copying, or pixel string copying, involves dividing a current coded block or a current decoded block (referred to as the current block) into several variable-length pixel sample strings. Here, a string refers to arranging the pixel samples within a two-dimensional region of arbitrary shape into a string whose length is much greater than its width (e.g., a string with a width of 1 pixel sample and a length of 37 pixel samples, or a string with a width of 2 pixel samples and a length of 111 pixel samples; typically, but not limited to, the length being an independent encoding or decoding parameter while the width is a predetermined parameter or a parameter derived from other encoding or decoding parameters). The basic operation of string copying encoding or decoding is to copy a reference string from the reconstructed reference pixel sample set for each coded string or decoded string (referred to as the current string) in the current block, and assign the value of the reference string to the current string. The copying parameters for string copying include the displacement vector of the current string and the copying length (copying size), representing the relative position between the reference string and the current string, and the length of the current string (the number of pixel samples), respectively. The length of the current string is also the length of the reference string. A current string has one displacement vector and one copy length. The number of displacement vectors and copy lengths corresponds to the number of strings into which a current block is divided.

[0024] 5) Palette index string copying, also known as palette or index string copying, involves first constructing or acquiring a palette in the palette encoding and corresponding decoding method. Then, some or all pixels of the current encoded or decoded block (referred to as the current block) are represented using the palette's indices. The indices are then encoded and decoded, including but not limited to: dividing the indices of a current block into several variable-length index strings, i.e., performing index string copying encoding and decoding. The basic operation of index string copying encoding or decoding involves copying a reference index string from the indexed reconstructed reference pixel sample set for each index encoded or decoded string (referred to as the current index string) in the current block, and assigning the index value of the reference index string to the current index string. The copying parameters for the index string copying method include the displacement vector of the current index string and the copying length (i.e., the copy size), representing the relative position between the reference index string and the current index string, and the length of the current index string (i.e., the number of corresponding pixel samples), respectively. The length of the current index string is also the length of the reference index string. A current index string has one displacement vector and one copying length. The number of index strings a current block is divided into determines the number of displacement vectors and the number of copy lengths.

[0025] 6) A hybrid copying method combining index string copying and pixel string copying, referred to as the index-pixel string hybrid copying method. When encoding or decoding a current encoding block or current decoding block (referred to as the current block), some or all pixels are copied using the pixel string copying method, and some or all pixels are copied using the index string copying method.

[0026] Other copying methods include rectangular copying and a combination of several copying methods.

[0027] In block copying, the block is the same as the micro-block in micro-block copying, the bar in bar copying, the string in string copying, the rectangle in rectangle copying, and the pixel index string in palette indexing. These are collectively called pixel sample segments, or simply sample segments. The basic components of a sample segment are pixels, pixel components, or pixel indices. A sample segment has a copy parameter that represents the relationship between the current pixel sample segment and a reference pixel sample segment. Therefore, a sample segment is the smallest unit of a copy operation with the same copy relationship. A copy parameter includes several copy parameter components, which at least include: horizontal component of the displacement vector, vertical component of the displacement vector, 1D 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, and unmatched pixels (also called non-reference pixels, i.e., non-copy pixels not copied from elsewhere).

[0028] In various copying methods, pixel samples or indices need to be arranged in a certain order. This arrangement is also called the scanning method. Based on their path shape, scanning methods can be categorized as follows:

[0029] A) Horizontal Z-scanning, also known as horizontal raster scanning, involves arranging pixel samples or indices of an encoding or decoding block (collectively referred to as an encoding / decoding block) row by row, with all rows arranged in the same direction (either entirely from left to right or entirely from right to left). Rows can be arranged from top to bottom or bottom to top.

[0030] B) Vertical Z-scanning, also known as vertical raster scanning, involves arranging pixel samples or indices of an encoding or decoding block (collectively called an encoding / decoding block) column by column, all in the same direction (either all from top to bottom or all from bottom to top). Columns can be arranged from left to right or right to left.

[0031] C) Horizontal bow-shaped scanning method. The pixel samples or indices of an encoding block or decoding block (collectively referred to as an encoding / decoding block) are arranged row by row. In odd-numbered rows, they are arranged in one direction (e.g., from left to right), while in even-numbered rows, they are arranged in the opposite direction (e.g., from right to left). The rows can be arranged from top to bottom or from bottom to top.

[0032] D) Vertical bow-shaped scanning method. The pixel samples or indices of an encoding block or decoding block (collectively referred to as an encoding / decoding block) are arranged column by column. In odd-numbered columns, they are arranged in one direction (e.g., from top to bottom), while in even-numbered columns, they are arranged in the opposite direction (e.g., from bottom to top). Columns can be arranged from left to right or from right to left.

[0033] It should be noted that "copying" is the operation of reconstruction and decoding, while the corresponding encoding operation is "matching". Therefore, various copying methods such as block matching, micro-block copying, line copying, pixel string copying, and index string copying are also called block matching, micro-block matching, line matching, pixel string matching, and index string matching, respectively.

[0034] In existing copying methods, the reference pixels are imperfect reconstructed pixels that have neither undergone the Deblocking Filter (DF) step nor the Sample Adaptive Offset (SAO) step. The imperfect reconstructed pixels may have significant errors compared to the final perfect reconstructed pixels, resulting in a large error between the reference pixels and the original pixels, thus reducing the compression efficiency of the image. Summary of the Invention

[0035] To address this problem in existing image and video encoding and decoding technologies, this invention provides a method or apparatus for image encoding and decoding, where reference pixels are taken from reconstructed pixels with at least two different levels of refinement. Specifically, the first portion of reference pixels is reconstructed pixels with refinement level I, the second portion is reconstructed pixels with refinement level II, the third portion is reconstructed pixels with refinement level III, and so on. Preferably, the reconstructed pixels with different levels of refinement are taken from different locations, i.e., different regions, of the image. Preferably, the reference pixels of a reference pixel sample segment consist of reconstructed pixels with at least two different levels of refinement. Preferably, the reference pixels of a reference pixel sample segment are taken from at least two different regions in the image, where the different regions have reconstructed pixels with different levels of refinement. According to this invention, the reference pixel range in the copying method is divided into at least two different regions with reconstructed pixels of different levels of refinement. According to this invention, in the copying method, the reference pixels of the reference pixel sample segment corresponding to a current pixel sample segment of a current encoding block or decoding block (collectively referred to as an encoding / decoding block) are taken from at least two different regions with reconstructed pixels of different levels of refinement.

[0036] The primary technical feature of this invention is that the reference pixel is taken from K (K≥2, typically K≤4) reconstructed pixels with different levels of perfection. For example, it is taken from the following three reconstructed pixels with different levels of perfection:

[0037] 1) Reconstructed pixels that have neither undergone DF nor SAO processing;

[0038] 2) Reconstructed pixels after vertical edge DF processing;

[0039] 3) Reconstructed pixels after full DF and SAO processing.

[0040] In this invention, preferably, a reference pixel sample segment (reference pixel sample string or reference pixel sample block) corresponding to a current pixel sample segment (current pixel sample string or current pixel sample block) in the current CU is composed of K (K≥2, usually K≤4) reconstructed pixels with different levels of perfection. That is, a portion of the reference pixels in the reference pixel sample segment are reconstructed pixels with perfection level I, and the remaining reference pixels are reconstructed pixels with other K-1 levels of perfection that are different from perfection level I.

[0041] In this invention, preferably, the legally acceptable reference pixel range of a current pixel sample segment in the current CU consists of K (K≥2, usually K≤4) regions of reconstructed pixels with different levels of perfection. For example, it consists of 3 types of regions.

[0042] In this invention, preferably, the reference pixel of a reference pixel sample segment corresponding to a current pixel sample segment in the current CU is taken from the regions of K (K≥2, usually K≤4) different levels of reconstructed pixels within a legally acceptable range of reference pixels. For example, the reference pixel of a reference pixel sample string is taken from the regions of 3 different levels of reconstructed pixels, while the reference pixel of another reference pixel sample block is taken from the regions of 2 different levels of reconstructed pixels.

[0043] The most fundamental and unique technical feature of the encoding method or apparatus of the present invention is that when encoding the current encoding block using a copying method, the reference pixel for the copying operation is taken from K (K≥2, usually K≤4) reconstructed pixels with different degrees of perfection. Figure 1 This is a schematic diagram of the encoding method or apparatus of the present invention. Preferably, the copying method is one of the following or a combination thereof: prediction method (including intra-frame prediction or inter-frame prediction), intra-frame block copying method, micro-block copying method, strip copying method, string copying method, index string copying method; the copying operation is one of the following operations or a combination thereof: prediction operation, intra-frame block copying operation, micro-block copying operation, strip copying operation, string copying operation, index string copying operation; correspondingly, the reference pixels are respectively the prediction value (block), reference block, reference micro-block, reference strip, reference string, and reference pixel in the palette.

[0044] The most fundamental and unique technical feature of the decoding method or apparatus of the present invention is that when decoding the video bitstream of the current decoding block using a copying method, the reference pixel for the copying operation is taken from K (K≥2, usually K≤4) reconstructed pixels with different degrees of perfection. Figure 2 This is a schematic diagram of the decoding method or apparatus of the present invention. Preferably, the copying method is one of the following or a combination thereof: prediction method (including intra-frame prediction or inter-frame prediction), intra-frame block copying method, micro-block copying method, strip copying method, string copying method, index string copying method; the copying operation is one of the following operations or a combination thereof: prediction operation, intra-frame block copying operation, micro-block copying operation, strip copying operation, string copying operation, index string copying operation; correspondingly, the reference pixels are respectively the prediction value (block), reference block, reference micro-block, reference strip, reference string, and reference pixel in the palette.

[0045] According to one aspect of the present invention, an image encoding method or apparatus is provided, comprising at least the steps or modules that perform the following functions and operations:

[0046] At least two reconstructed pixels with different levels of perfection are used as reference pixels to copy the current coding block, generating a video stream containing copying parameters.

[0047] According to another aspect of the present invention, an image decoding method or apparatus is also provided, comprising at least the steps or modules that perform the following functions and operations:

[0048] Parse the video bitstream to obtain copy parameter information, and use at least two reconstructed pixels with different levels of perfection as reference pixels to copy and decode the current decoding block.

[0049] This invention is applicable to the encoding and decoding of overlay format images. It is also applicable to the encoding and decoding of component plane format images.

[0050] The technical features of the present invention have been illustrated above through several specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the encoding method or apparatus of the present invention.

[0052] Figure 2 This is a schematic diagram of the decoding method or apparatus of the present invention. Detailed Implementation

[0053] The following are further implementation details or variations of the present invention.

[0054] Implementation or variant example 1

[0055] In the encoding method or apparatus or decoding method or apparatus, the copy encoding or decoding is one of the following encoding or decoding methods or a combination thereof: prediction encoding or decoding, intra-block copy encoding or decoding, micro-block copy encoding or decoding, strip 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 combination thereof: prediction operation, intra-block copy operation, micro-block copy operation, strip copy operation, string copy operation, index string copy operation; correspondingly, the reference pixels are respectively the prediction value (block), reference block, reference micro-block, reference strip, reference string, and reference pixel in the palette.

[0056] Implementation or variant example 2

[0057] In the encoding method or apparatus or decoding method or apparatus, the legally acceptable range of reference pixels consists of regions of K (K≥2, usually K≤4) reconstructed pixels with different degrees of perfection.

[0058] Implementation or variant example 3

[0059] In the encoding method or apparatus or decoding method or apparatus, the legal and desirable reference pixel range of a current pixel sample segment in the current encoding block or decoding block is composed of K (K≥2, usually K≤4) regions of reconstructed pixels with different degrees of perfection.

[0060] Implementation or variant example 4

[0061] In the encoding method or apparatus or decoding method or apparatus, 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 the region of K (K≥2, usually K≤4) different degrees of perfection of reconstructed pixels in the range of legally available reference pixels.

[0062] Implementation or variant example 5

[0063] In the encoding method or apparatus or the decoding method or apparatus, the at least two reconstructed pixels with different levels of perfection include at least two of the following three types of reconstructed pixels:

[0064] 1) Reconstructed pixels at perfection level I;

[0065] 2) Reconstructed pixels at perfection level II;

[0066] 3) Reconstructed pixels at perfection level III.

[0067] Implementation or variant example 6

[0068] In the encoding method or apparatus or decoding method or apparatus described in Implementation or Variation Example 5, the reconstructed pixels of perfection levels I, II, and III are respectively:

[0069] 1) Reconstructed pixels that have neither undergone DF nor SAO processing;

[0070] 2) Reconstructed pixels after vertical edge DF processing;

[0071] 3) Reconstructed pixels after full DF and SAO processing.

[0072] Implementation or variant example 7

[0073] In the encoding method or apparatus or decoding method or apparatus described in implementation or variation 5 or 6, the reconstructed pixels of perfection levels I, II, and III respectively come from the following regions in the current image:

[0074] 1) The rightmost four columns of the current CTU, the CTU to the left of the current CTU, and the CTU further to the left;

[0075] 2) The four rows above the region that has not yet been reconstructed or the region described in 1);

[0076] 3) The region within the legal and desirable reference pixel range that does not belong to either 1) or 2).

Claims

1. An image coding method that integrates inter-frame prediction and intra-frame block copying into a single coding mode, characterized in that, It should include at least the following steps: 1) Construct at least two reconstructed pixels with different levels of refinement; 2) Encode the coding block using the coding mode by using at least the reconstructed pixels as reference pixels; 3) Generate a video stream containing at least the information of the encoding mode; The at least two types of reconstructed pixels with different levels of perfection include at least one type of reconstructed pixel with level II perfection and at least one of the following two types of reconstructed pixels: Reconstructed pixels at perfection level I; Reconstructed pixels at perfection level III; The reconstructed pixels for perfection levels I, II, and III are respectively: Reconstructed pixels that have neither undergone DF nor SAO processing; Reconstructed pixels after vertical edge DF processing; Reconstructed pixels after full DF and SAO processing.

2. An image coding apparatus that integrates inter-frame prediction and intra-frame block copying into a single coding mode, characterized in that, It should include at least the following modules: 1) Reconstruct pixel building blocks to build at least two reconstructed pixels with different levels of refinement; 2) The encoding module uses at least the reconstructed pixels as reference pixels to encode the encoding block using the encoding mode; 3) A video stream generation module that generates a video stream containing at least the information of the encoding mode; The at least two types of reconstructed pixels with different levels of perfection include at least one type of reconstructed pixel with level II perfection and at least one of the following two types of reconstructed pixels: Reconstructed pixels at perfection level I; Reconstructed pixels at perfection level III; The reconstructed pixels for perfection levels I, II, and III are respectively: Reconstructed pixels that have neither undergone DF nor SAO processing; Reconstructed pixels after vertical edge DF processing; Reconstructed pixels after full DF and SAO processing.

3. An image decoding method that integrates inter-frame prediction and intra-frame block copying into a single coding mode, characterized in that, It should include at least the following steps: 1) Parse the video stream to obtain information containing at least the encoding mode; 2) Construct at least two reconstructed pixels with different levels of refinement; 3) Decode the decoding block using at least the encoding mode and employing at least the reconstructed pixels as reference pixels; The at least two types of reconstructed pixels with different levels of perfection include at least one type of reconstructed pixel with level II perfection and at least one of the following two types of reconstructed pixels: Reconstructed pixels at perfection level I; Reconstructed pixels at perfection level III; The reconstructed pixels for perfection levels I, II, and III are respectively: Reconstructed pixels that have neither undergone DF nor SAO processing; Reconstructed pixels after vertical edge DF processing; Reconstructed pixels after full DF and SAO processing.

4. The decoding method according to claim 3, characterized in that: The decoding block is a decoding region of an image, including at least one of the following: maximum coding unit (LCU), coding tree unit (CTU), coding unit (CU), sub-regions of CU, and prediction unit (PU).

5. The decoding method according to claim 3, characterized in that: The reference pixel range of the encoding mode is divided into at least two regions of reconstructed pixels with different degrees of perfection.

6. The decoding method according to claim 3, characterized in that: The reconstructed pixels with different levels of perfection are taken from different regions of the reference pixel range of the image.

7. An image decoding apparatus that integrates inter-frame prediction and intra-frame block copying into a single coding mode, characterized in that, It should include at least the following modules: 1) Video stream parsing module, which parses the video stream to obtain information containing at least the encoding mode; 2) Reconstruct pixel building blocks to build at least two reconstructed pixels with different levels of refinement; 3) Decoding module, which decodes the decoding block using at least the encoding mode and employs at least the reconstructed pixels as reference pixels; The at least two types of reconstructed pixels with different levels of perfection include at least one type of reconstructed pixel with level II perfection and at least one of the following two types of reconstructed pixels: Reconstructed pixels at perfection level I; Reconstructed pixels at perfection level III; The reconstructed pixels for perfection levels I, II, and III are respectively: Reconstructed pixels that have neither undergone DF nor SAO processing; Reconstructed pixels after vertical edge DF processing; Reconstructed pixels after full DF and SAO processing.

8. The decoding device according to claim 7, characterized in that: The decoding block is a decoding region of an image, including at least one of the following: maximum coding unit (LCU), coding tree unit (CTU), coding unit (CU), sub-regions of CU, and prediction unit (PU).

9. The decoding device according to claim 7, characterized in that: The reference pixel range of the encoding mode is divided into at least two regions of reconstructed pixels with different degrees of perfection.

10. The decoding device according to claim 7, characterized in that: The reconstructed pixels with different levels of perfection are taken from different regions of the reference pixel range of the image.

Images