An image processing method, a communication node, a storage medium and a program product

Abstract

This application provides an image processing method, a communication node, a storage medium, and a program product. The method includes: acquiring a first reconstructed image to be processed; making decisions for each first filtering region of the first reconstructed image; determining filtering control information corresponding to each first filtering region; performing adaptive loop filtering on the first filtering regions according to the filtering control information to obtain a first reconstructed value; determining a first sub-region within the first filtering region and a first adjacent filtering region corresponding to the first sub-region; when the first filtering region meets filtering adjustment conditions, performing filtering processing on the first sub-region according to the filtering control information corresponding to the first adjacent filtering region to obtain a second reconstructed value; performing filtering adjustment based on the first reconstructed value and the second reconstructed value of the first sub-region to determine the filtering result of the first filtering region; and generating an encoded image based on the filtering result of each first filtering region of the first reconstructed image, thus solving the problem of poor filtering effect.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to an image processing method, a communication node, a storage medium, and a program product. Background Technology

[0002] In H.266 / VVC, the Adaptive Loop Filter (ALF) technique adaptively selects a set of filters from a finite set of filter coefficients to filter the reconstructed video. Taking the luma ALF filtering of each Coding Tree Unit (CTU) in the video encoded bitstream as an example, the filtering methods can be as follows: not enabling ALF filtering, using a fixed subset to complete ALF filtering, and using an adaptive parameter set to complete filtering. Currently, in ECM, each CTU can select the most suitable subset from the historical 0-7 (8 APS subsets) and use the corresponding coefficients for filtering. In the actual encoding process, the encoder can select the most suitable filtering method according to application requirements, encoding configuration, and rate-distortion optimization criteria. However, when the CTU selects the historical set through rate-distortion optimization (RDO) during encoding, it means calculating the optimal APS subset for the CTU as a whole. But this set of coefficients may not be optimal locally within the CTU. For a single CTU, it is impossible to perform more refined selection of filter coefficient sets, resulting in poor filtering performance. Summary of the Invention

[0003] This application provides an image processing method, a communication node, a storage medium, and a program product to solve the problem of poor filtering effect.

[0004] To achieve the above objectives, embodiments of this application provide an image processing method applied at the encoding end, comprising:

[0005] The first reconstructed image to be processed is obtained. Decisions are made for each first filtering region of the first reconstructed image to determine the filtering control information corresponding to each first filtering region. Adaptive loop filtering is performed on the first filtering region according to the filtering control information to obtain the first reconstructed value.

[0006] A first sub-region and a first adjacent filter region corresponding to the first sub-region are determined. When the first filter region meets the filter adjustment conditions, the first sub-region is filtered according to the filter control information corresponding to the first adjacent filter region to obtain the second reconstructed value.

[0007] Based on the first reconstructed value and the second reconstructed value of the first sub-region, a filtering adjustment is performed to determine the filtering result of the first filtering region, and an encoded image is generated based on the filtering result of each first filtering region of the first reconstructed image.

[0008] To achieve the above objectives, embodiments of this application provide another image processing method applied at the decoding end, including:

[0009] The encoded image is acquired, the encoded image is decoded to obtain a second reconstructed image, and the second reconstructed image is filtered to obtain a third reconstructed value;

[0010] When it is determined that the coded image is enabled for filtering adjustment, the second sub-region of the second filtering region in the second reconstructed image and the second adjacent filtering region corresponding to the second sub-region are determined. The second sub-region is filtered according to the filtering control information corresponding to the second adjacent filtering region to obtain the fourth reconstructed value.

[0011] The filtering result of the second filtering region in the second reconstructed image is determined based on the third reconstructed value and the fourth reconstructed value of the second sub-region, and a decoded image is generated based on the filtering result of each second filtering region in the second reconstructed image.

[0012] To achieve the above objectives, embodiments of this application provide a communication node, including: a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for enabling communication between the processor and the memory. When the program is executed by the processor, it implements the steps of the image processing method as described in any one of the embodiments of this application.

[0013] To achieve the above objectives, embodiments of this application provide a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the image processing method described in any one of the embodiments of this application.

[0014] To achieve the above objectives, embodiments of this application provide a computer program product, which includes a computer program that, when executed by a processor, implements the image processing method described in any one of the embodiments of this application.

[0015] The image processing method, communication node, storage medium, and program product provided in this application embodiment acquire a first reconstructed image to be processed, make decisions on each first filtering region of the first reconstructed image, determine the filtering control information corresponding to each first filtering region, and perform adaptive loop filtering on the first filtering region according to the filtering control information to obtain a first reconstructed value; determine a first sub-region within the first filtering region and a first adjacent filtering region corresponding to the first sub-region, and when the first filtering region meets the filtering adjustment conditions, perform filtering processing on the first sub-region according to the filtering control information corresponding to the first adjacent filtering region to obtain a second reconstructed value; and perform filtering processing on the first sub-region based on the first reconstructed value and the second reconstructed value of the first sub-region. The filtering process involves adjusting the filter to determine the filtering result of the first filtering region and generating an encoded image based on the filtering result of each first filtering region in the first reconstructed image, thus solving the problem of poor filtering effect. The first filtering region in the first reconstructed image is filtered using filtering control information to obtain a first reconstructed value. If the first filtering region meets the filtering adjustment conditions, a first sub-region within the first filtering region is filtered using the filtering control information of the first adjacent filtering region. Then, filtering adjustment is performed based on the first and second reconstructed values ​​to achieve local optimization of the first filtering region, obtaining the final filtering result of the first filtering region. Finally, an encoded image is generated based on the filtering result of each first filtering region in the first reconstructed image, improving the image filtering accuracy.

[0016] Further details regarding the above embodiments and other aspects of this application, as well as their implementations, are provided in the accompanying drawings, detailed description, and claims. Attached Figure Description

[0017] Figure 1 This is a diagram of an image encoding architecture;

[0018] Figure 2 This is an architecture diagram for image decoding;

[0019] Figure 3 A flowchart illustrating an image processing method provided in one embodiment;

[0020] Figure 4 This is an example diagram illustrating the implementation of a filter adjustment method in one embodiment.

[0021] Figure 5 A flowchart of another image processing method provided in one embodiment;

[0022] Figure 6 A schematic diagram illustrating a sub-region division as provided in one embodiment;

[0023] Figure 7A schematic diagram illustrating another sub-region division provided in one embodiment;

[0024] Figure 8 This is a schematic diagram of the structure of an image processing apparatus provided in one embodiment;

[0025] Figure 9 A schematic diagram of the structure of another image processing apparatus provided in one embodiment;

[0026] Figure 10 This is a schematic diagram of the structure of a communication node provided in one embodiment. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be arbitrarily combined with each other.

[0028] For block-based hybrid coding frameworks such as HEVC (High Efficiency Video Coding), VVC (Versatile Video Coding), and AVS, in-loop filtering is typically employed to effectively reduce the distortion caused by quantization. Since these filtered reconstructed images serve as references for subsequent coded images to predict future image signals, the aforementioned filtering operation is also called in-loop filtering, i.e., filtering within the coding loop. Taking the H.266 / VVC video coding standard as an example, in-loop filtering techniques include Luma Mapping With Chroma Scaling (LMCS), Deblocking Filter (DBF), Sample Adaptive Offset (SAO), and Adaptive Loop Filter (ALF). The ALF module includes filters for the luma component, filters for the two chroma components, and a cross-component adaptive loop filter.

[0029] In H.266 / VVC, ALF technology can adaptively select a set of filters from a finite set of filter coefficients to filter the reconstructed video. The filter coefficients contain M(8) filter subsets, each filter subset contains N(25) classes of filters, and each class of filters contains K(42) coefficients. The decoder parses the signaling in the encoded bitstream, determines the ALF filter subset index used for loop filtering in each CTU, and then determines the filter class based on the content of the pixel block (where 4*4 pixel blocks are used in H.266 / VVC), thus determining the filter coefficients used for that pixel block. For the encoder, the filter class is determined based on the pixel block content, and then the optimal filter subset is determined using the rate-distortion optimization criterion. The subset index is then encoded and written into the encoded bitstream.

[0030] The luminance ALF filter coefficient set includes a fixed subset and up to eight APS (adaptive parameter set) subsets. The fixed subset refers to the filter coefficient set obtained through pre-training, specified by the standard, and does not need to be transmitted. The APS subset is the filter coefficient set generated by Wiener filtering based on the reconstructed image of the current video frame, and is transmitted through the adaptive parameter set APS. Each slice at the encoding end can generate one APS subset, each APS has an ID, and each slice can choose to use a subset of the historical ALF coefficient set. Each CTU can choose to use either a fixed subset or a historical APS subset. Therefore, in the loop filtering module during the encoding process, for the luminance component of each CTU, after RDO decision, the possible decision results include:

[0031] 1. Do not enable ALF

[0032] 2. Filter using filter coefficients of a fixed subset.

[0033] 3. Use the filtering coefficients of the historical APS subset (including the latest APS subset calculated from the current frame) for filtering.

[0034] Since most mainstream video compression is lossy, rate-distortion criterion, as a tool for measuring compression performance, is the most widely used technique in video compression encoding. Almost all modules use rate-distortion criterion to determine the final mode. Based on the principle of rate-distortion optimization, under the constraint of the encoding bit rate, the rate-distortion criterion establishes an optimization problem by minimizing distortion, as shown in the formula:

[0035] m X inD(X),stR(X)￡R c

[0036] Where D represents distortion, and since no further time-domain or frequency-domain transformation is performed in the loop filtering stage, the distortion D is uniformly represented by the minimum mean square error (MSE). R is the bit rate, and X can represent the partition structure, prediction mode, transform coefficients, etc. In the ALF module, during slice-level decision-making, X is determined within the range indicated by the above formula, including skipping ALF, using a newly calculated APS set, fixing the filter set, and using the historical APS set.

[0037] In the ALF loop filtering process at the encoder end, a series of optimizations are typically performed to improve algorithm performance and reduce computational complexity. Since ALF cannot improve the compression performance of each CTU, the encoder uses rate-distortion optimization strategies to identify CTUs that do not use ALF and excludes them, then recalculates the filter coefficients. Alternatively, after obtaining N sets of filter coefficients, different filter classes can be merged; if the resulting decrease in compression performance is limited, the merging scheme is adopted. Both optimization strategies can be iterated multiple times until each CTU determines the optimal ALF filtering scheme.

[0038] Taking image encoding and decoding based on a hybrid coding framework as an example, the image encoding and decoding process is explained. Figure 1 This paper presents an image coding architecture diagram. The H.266 / VVC coding framework, a new generation video coding standard developed by the ITU-T and ISO / IEC joint video project, includes modules such as intra-frame prediction, inter-frame prediction, transform, quantization, loop filtering, and entropy coding.

[0039] The overall framework flow of the encoding end is as follows:

[0040] (1) The input video is first divided into frames and then into blocks;

[0041] (2) The divided blocks are sent to the intra-frame / inter-frame prediction module for predictive coding. The intra-frame prediction module is mainly used to remove spatial correlations in the image; the inter-frame prediction module is mainly used to remove temporal correlations in the image.

[0042] (3) Subtract the predicted value from the original block to obtain the residual value. Then transform and quantize the residual to remove frequency domain correlation and perform lossy compression on the data.

[0043] (4) Finally, all the encoding parameters and residuals are entropy encoded to form a binary stream for storage or transmission. The output data of the entropy encoding module is the original video compressed bitstream.

[0044] (5) The predicted value and the residual after inverse quantization and inverse transformation are added together to obtain the block reconstruction value, and finally the reconstructed image is formed.

[0045] (6) The reconstructed image is filtered by a loop filter and stored in the image buffer as a reference image for the future.

[0046] Figure 2 An architecture diagram for image decoding is provided, such as... Figure 2 As shown, the overall framework flow of the decoding end is as follows:

[0047] (1) Parse the bitstream to obtain the prediction pattern and get the prediction value;

[0048] (2) Perform inverse transformation and inverse quantization on the residual obtained from the code stream parsing;

[0049] (3) The predicted value and the residual after inverse quantization and inverse transformation are added together to obtain the block reconstruction value, and finally the reconstructed image is formed.

[0050] (4) The reconstructed image is filtered by a loop filter and stored in the image buffer as a reference image for the future.

[0051] It should be noted that the technical solutions provided in this application can be applied to H.266 / VVC standards, AVS (such as AVS3), or next-generation video codec standards, and this application does not limit them.

[0052] Figure 3 A flowchart of an image processing method provided in one embodiment, such as Figure 3 As shown, the image processing method described in this application embodiment is applied to the encoding end, and the method includes S110-S130:

[0053] S110. Obtain the first reconstructed image to be processed, make a decision for each first filtering region of the first reconstructed image, determine the filtering control information corresponding to each first filtering region, and perform adaptive loop filtering on the first filtering region according to the filtering control information to obtain the first reconstructed value.

[0054] The first reconstructed image can be understood as the image obtained after encoding and other processing of the original image. The first filtering region can be understood as a portion of the first reconstructed image. For example, the first reconstructed image can be divided into regions according to certain rules, and each resulting region can serve as the first filtering region. For instance, the first reconstructed image can be divided using preset sizes, CTUs, CUs, etc., as dividing units to obtain the first filtering regions. Filtering control information can be understood as information used to control or indicate the filtering process of the first reconstructed image, such as whether filtering is initiated, and if so, the filtering parameters used. The first reconstructed value can be understood as the information obtained after filtering the image.

[0055] In the process of image encoding, the original image is acquired, encoded according to existing encoding information, and then reconstructed from the encoded data to obtain a first reconstructed image. This embodiment of the application can perform filtering processing on the first reconstructed image obtained after encoding. The first reconstructed image to be processed is acquired, a region division method is preset, and the first reconstructed image is divided into different filtering regions according to the division method. This filtering region is denoted as the first filtering region. Each first filtering region in the first reconstructed image can be processed using the method provided in this embodiment of the application. A decision is made for each first filtering region. The decision method can be preset, for example, using the RDO rate distortion optimization method to determine whether filtering needs to be enabled for each first filtering region, and if so, the parameters used. Based on the decision result, filtering control information for each first filtering region is generated. Based on the filtering control information, it is determined whether adaptive loop filtering should be performed on the first filtering region, and the parameters used for adaptive loop filtering. If adaptive loop filtering is determined based on the filtering control information, adaptive loop filtering is performed on the first filtering region according to the parameters of adaptive loop filtering to obtain a first reconstructed value. If the filter control information determines that adaptive loop filtering is not required for the first filter region, then adaptive loop filtering is not required for the first filter region.

[0056] It is important to understand that other filtering processes, such as DBF and SAO, can be performed before or after the adaptive loop filtering process on the first filtering region. In this embodiment, the filtering process can be performed before making decisions for each first filtering region of the first reconstructed image. That is, after acquiring the first reconstructed image, filtering processes such as DBF, SAO, and ALF are performed on the first reconstructed image. Here, ALF filtering refers to the calculation of the ALF filtering coefficients, which is part of the filtering process and does not complete the filtering. Decisions are made for each first filtering region of the first reconstructed image to determine the filtering control information corresponding to each first filtering region. Then, adaptive loop filtering is performed on the first filtering region according to the filtering control information to obtain the first reconstructed value.

[0057] S120. Determine the first sub-region in the first filtering region and the first adjacent filtering region corresponding to the first sub-region. If the first filtering region meets the filtering adjustment conditions, perform filtering processing on the first sub-region according to the filtering control information corresponding to the first adjacent filtering region to obtain the second reconstructed value.

[0058] The first sub-region can be understood as the area within the first filtering region that requires filtering adjustment. It can be a region at a specific location within the first filtering region, or any region at any location within the first filtering region. The first adjacent filtering region can be understood as the first filtering region adjacent to the first sub-region. The filtering adjustment conditions can be understood as the conditions used to determine whether to adjust the filtering result of the first sub-region. These conditions can be preset, for example, set based on the filtering control information of the first filtering region and the first adjacent filtering region, or determined based on information such as the size, position, and proportion of the first filtering region and / or the first sub-region in the image. The second reconstructed value can be understood as the information obtained after filtering the first sub-region in the image.

[0059] A pre-determined region division method is used to divide the first filtering region into different sub-regions. The sub-region requiring adjustment is identified and designated as the first sub-region. For example, a sub-region at a specified location within the first filtering region is selected as the first sub-region, and the other sub-regions can be designated as third sub-regions. After determining the first sub-region, other adjacent first filtering regions are identified based on their location and designated as first adjacent filtering regions. Pre-set filtering adjustment conditions are used to determine whether the first filtering region meets these conditions. For example, whether the size of the first filtering region is within a preset range, and whether the filtering control information of the first filtering region is the specified filtering control information, etc. If the first filtering region meets the filtering adjustment conditions, the filtering control information corresponding to the first adjacent filtering region is determined. Based on this information, it is determined whether filtering is enabled and, if so, the filtering parameters to be used. The first sub-region is then filtered according to the corresponding filtering parameters to obtain the second reconstructed value.

[0060] The number of first adjacent filter regions corresponding to the first sub-region can be one or more, depending on the location of the first sub-region. If there are multiple first adjacent filter regions corresponding to the first sub-region, multiple second reconstructed values ​​can be determined accordingly, or a single second reconstructed value can be determined by combining them.

[0061] S130. Filtering adjustment is performed based on the first reconstructed value and the second reconstructed value of the first sub-region to determine the filtering result of the first filtering region, and an encoded image is generated based on the filtering result of each first filtering region of the first reconstructed image.

[0062] For the first sub-region, there are two corresponding reconstructed values: a first reconstructed value and a second reconstructed value. During filtering adjustment, the adjustment can be performed directly, or the first and second reconstructed values ​​can be compared to determine if filtering adjustment is necessary. If so, the first reconstructed value is adjusted based on the second reconstructed value. The filtering adjustment method can be to directly adjust the first reconstructed value based on the second reconstructed value, for example, by using the second reconstructed value as the adjusted filtering result, or by performing a comprehensive calculation on the second and first reconstructed values, etc. After filtering adjustment, the adjusted result for the first sub-region is determined. Based on the filtered result for each first sub-region, the filtering result for the first filtered region is determined. Based on the filtering result for each first filtered region in the first reconstructed image, an encoded image is generated, completing image encoding. That is, for each first filtered region in the first reconstructed image, the original reconstructed value corresponding to the first reconstructed image is replaced with the filtering result to achieve image filtering and obtain the final encoded image.

[0063] This application provides an image processing method that solves the problem of poor filtering effect. The method filters a first filtering region in a first reconstructed image using filtering control information to obtain a first reconstructed value. When the first filtering region meets the filtering adjustment conditions, a first sub-region within the first filtering region is filtered using filtering control information from a first adjacent filtering region. Then, filtering adjustment is performed based on the first and second reconstructed values ​​to achieve local optimization of the first filtering region, resulting in the final filtering result of the first filtering region. Finally, an encoded image is generated based on the filtering result of each first filtering region in the first reconstructed image, thereby improving the image filtering accuracy.

[0064] In some embodiments, the filter control information includes at least one of the following:

[0065] Adaptive loop filter control information for the luminance component;

[0066] Adaptive loop filter control information for chrominance components;

[0067] Cross-component adaptive loop filter control information.

[0068] The adaptive loop filter control information for the luminance component is used to control the adaptive loop filter for the luminance component. For example, it controls whether to enable the adaptive loop filter for the luminance component and the parameters used when enabling it, such as the subset and subset index used for filtering. Similarly, the adaptive loop filter control information for the chrominance component is used to control the adaptive loop filter for the chrominance component. This includes controlling whether to enable the adaptive loop filter for the chrominance component and the parameters used when enabling it, such as the subset and subset index used for filtering. Finally, the adaptive loop filter control information for cross-components is used to control the adaptive loop filter for cross-components. This includes controlling whether to enable the adaptive loop filter for cross-components and the parameters used when enabling it, such as the subset and subset index used for filtering.

[0069] The filtering control information includes at least one of the above three types, which respectively realize the filtering control of the luminance component, chrominance component and cross-component.

[0070] In some embodiments, the adaptive loop filter control information includes at least one of the following:

[0071] Whether adaptive loop filtering is enabled in the filtering region or for the components of the filtering region;

[0072] The index of the fixed subset of filters corresponding to the filtering region;

[0073] Index of the adaptive parameter set corresponding to the filtering region;

[0074] The index of the optional subset in the adaptive parameter set corresponding to the filtering region.

[0075] The "Adaptive Loop Filter Enabled" setting for the filtering region indicates that adaptive loop filtering can be performed within that region. Similarly, "Adaptive Loop Filter Enabled for Components within the Filtering Region" indicates that adaptive loop filtering can be performed on the luminance, chrominance, or cross-component components of the filtering region. For example, the adaptive loop filtering control information for the luminance component might include an indication that adaptive loop filtering is enabled for the components within the filtering region, meaning that adaptive loop filtering can be performed on the luminance component of the filtering region. The adaptive parameter set index can be a historical APS subset index. The adaptive parameter set can include multiple subsets, and the optional subset index within the adaptive parameter set can indicate the subset used.

[0076] In some embodiments, the adaptive loop filter control information for the chromaticity component includes at least one of the following: adaptive loop filter control information for the first chromaticity component and adaptive loop filter control information for the second chromaticity component.

[0077] The first chromaticity component can be the Cb (blue) component, and the second chromaticity component can be the Cr (red) component.

[0078] In some embodiments, the cross-component adaptive loop filter control information includes at least one of the following: first cross-component adaptive loop filter control information and second cross-component adaptive loop filter control information.

[0079] The first cross component can be CCALF-Cb, and the second cross component can be CCALF-Cr. CCALF stands for Cross-Component Adaptive Loop Filter.

[0080] In some embodiments, adaptive loop filtering is performed on the first filtering region, including at least one of the following methods:

[0081] Adaptive loop filtering is performed on the luminance components of the first filtering region based on the adaptive loop filtering control information of the luminance components;

[0082] Based on the adaptive loop filtering control information of the chromaticity components and the adaptive loop filtering control information across components, adaptive loop filtering is performed on the chromaticity components of the first filtering region.

[0083] Based on the adaptive loop filtering control information of the luminance component, it is determined whether to perform adaptive loop filtering on the luminance component, and if adaptive loop filtering is performed, it is determined which filter subset to use. The filter subset to use can be determined based on at least one of a fixed filter subset index, an adaptive parameter set index, and an optional subset index within the adaptive parameter set. If it is determined that adaptive loop filtering is to be performed on the luminance component, adaptive loop filtering is performed on the luminance component of the first filtering region according to the determined filter subset to obtain the reconstructed value of the luminance component. If it is determined that adaptive loop filtering is not to be performed on the luminance component, adaptive loop filtering is not required for the luminance component of the first filtering region.

[0084] Based on the adaptive loop filtering control information for the chrominance components and the adaptive loop filtering control information across components, it is determined whether to perform adaptive loop filtering on the chrominance components, and if adaptive loop filtering is performed, it is determined which filter subset to use. The filter subset used can be determined based on at least one of a fixed filter subset index, an adaptive parameter set index, and an optional subset index within the adaptive parameter set. If it is determined that adaptive loop filtering is performed on the chrominance components, adaptive loop filtering is performed on the chrominance components of the first filtering region according to the determined filter subset to obtain the reconstructed values ​​of the chrominance components. If it is determined that adaptive loop filtering is not performed on the chrominance components, adaptive loop filtering is not required for the chrominance components of the first filtering region.

[0085] In some embodiments, the first sub-region includes at least one of the following:

[0086] A sub-region of a preset position and size located at the boundary of the first filtering region;

[0087] Sub-regions obtained using a preset partitioning method.

[0088] The preset position can be above, to the left, etc.; the preset size can be set to a fixed value in advance, for example, based on the size of the first filtering area. The larger the first filtering area, the larger the preset size. The preset position and preset size are predetermined, and a sub-region of the preset position and preset size at the boundary of the first filtering area is determined. This sub-region is used as the first sub-region. For example, a sub-region of size x1*y1 to the left of the boundary of the first filtering area is used as the first sub-region.

[0089] The partitioning method can be preset, and can be set according to the type of image data, the size of the image, and the information of the adaptive loop filter corresponding to the image.

[0090] In some embodiments, the partitioning method includes at least one of the following:

[0091] Classification is performed based on the directionality factor of the luminance pixel blocks in the adaptive loop filter;

[0092] Classification is based on the activity factor of the luminance pixel blocks in the adaptive loop filter;

[0093] Classification is based on the sideband information of adaptive loop filtering;

[0094] Classification is based on the residuals of adaptive loop filtering.

[0095] In some embodiments, filtering is performed on the first sub-region based on the filtering control information corresponding to the first adjacent filtering region to obtain the second reconstructed value, including:

[0096] If it is determined that the first adjacent filtering region does not enable adaptive loop filtering based on the filtering control information corresponding to the first adjacent filtering region, the reconstructed value corresponding to the first sub-region in the first reconstructed image is used as the second reconstructed value.

[0097] If the adaptive loop filtering of the first adjacent filtering region is enabled based on the filtering control information corresponding to the first adjacent filtering region, the filtering subset used by the first adjacent filtering region is determined, and the first sub-region is filtered according to the filtering subset to obtain the second reconstructed value.

[0098] Analyze the filter control information corresponding to the first adjacent filter region to determine whether adaptive loop filtering is enabled in the first adjacent filter region. If adaptive loop filtering is not enabled in the first adjacent filter region, the reconstructed value R0 corresponding to the first sub-region in the first reconstructed image is directly used as the second reconstructed value. If adaptive loop filtering is enabled in the first adjacent filter region, the filter control information corresponding to the first adjacent filter region indicates the subset used for filtering. That is, the filter subset used can be determined based on at least one of the fixed subset index, the adaptive parameter set index, and the optional subset index in the adaptive parameter set. By analyzing the filter control information corresponding to the first adjacent filter region, the filter subset used by the first adjacent filter region is determined. Filtering is performed on the first sub-region based on this filter subset to obtain the second reconstructed value. In this case, filtering is performed on the reconstructed value R0 corresponding to the first sub-region.

[0099] In some embodiments, when the number of first adjacent filtered regions is at least two, filtering the first sub-region according to the filtered subset to obtain the second reconstructed value includes:

[0100] Select a filter subset from each of the first adjacent filter regions to determine the target filter subset. Then, filter the first sub-region based on the target filter subset to obtain the second reconstructed value; or...

[0101] The first sub-region is filtered according to the filtering subset of each first adjacent filtering region to obtain at least two second reconstructed values.

[0102] The target filter subset can be understood as a set selected from multiple filter subsets. When the number of first adjacent filter regions is at least two, the number of selectable filter subsets is also at least two. The filter subsets for each first adjacent filter region are selected, choosing one or more filter subsets from multiple filter subsets as the target filter subset. This selection can be random or based on certain rules, priorities, etc., to obtain the target filter subset. The parameters used for filtering are determined based on the target filter subset, and then the first sub-region is filtered to obtain the second reconstructed value. Alternatively, for each filter subset of the first adjacent filter region, the first sub-region is filtered based on this filter subset to obtain the second reconstructed value. When the number of first adjacent filter regions is at least two, the number of filter subsets is also at least two, thus at least two second reconstructed values ​​can be obtained.

[0103] In some embodiments, filtering adjustment is performed based on a first reconstructed value and a second reconstructed value of a first sub-region to determine the filtering result of the first filtering region, including:

[0104] Compare the first reconstructed value with the original image corresponding to the first reconstructed image to determine the first distortion;

[0105] Compare the second reconstructed value with the original image corresponding to the first reconstructed image to determine the second distortion;

[0106] If the second distortion is less than the first distortion, adjust the first sub-region according to the second reconstruction value;

[0107] The filtering result of the first filtering region is determined based on the adjusted reconstruction value of the first sub-region.

[0108] In this embodiment, both the first and second distortions are obtained by comparing two images. The first distortion is calculated by comparing the first reconstructed value with the original image corresponding to the first reconstructed image; the second distortion is calculated by comparing the second reconstructed value with the original image corresponding to the first reconstructed image. The distortion calculated in this embodiment refers to the distortion between the filtered data and the original data. Distortion can be calculated using methods such as mean squared error, peak signal-to-noise ratio, structural similarity index, visual signal-to-noise ratio, and feature similarity index. The first and second distortions are compared. If the second distortion is less than the first distortion, it indicates that the image obtained after filtering is more efficient, and the final reconstructed value of this first sub-region is optimized. The first sub-region is then filtered based on the second reconstructed value. For example, the second and first reconstructed values ​​are weighted, or the first reconstructed value is replaced with a second key value, etc. After filtering, the adjusted reconstructed value of the first sub-region is obtained. Based on the adjusted reconstructed value of the first sub-region, combined with the reconstructed value of the third sub-region in the first filtered region that has not undergone filtering adjustment, the filtering result of the first filtered region is determined.

[0109] In some embodiments, adjusting the first sub-region based on the second reconstructed value includes:

[0110] The second reconstructed value and the first reconstructed value corresponding to the first sub-region are weighted according to the corresponding weight matrix to obtain the adjusted reconstructed value of the first sub-region.

[0111] Determine the weight matrix corresponding to the second reconstructed value and the first reconstructed value corresponding to the first sub-region. This weight matrix can be pre-generated and directly obtained during weighting. The adjusted reconstructed value of the first sub-region is obtained through weighted calculation, which involves weighting the filtered first reconstructed value corresponding to the first sub-region with the filtered and adjusted second reconstructed value.

[0112] In some embodiments, the weight matrix includes at least one of the following:

[0113] The elements in the weight matrix corresponding to the second reconstructed value, and the elements in the weight matrix corresponding to the first reconstructed value of the first sub-region, are fixed values;

[0114] The elements of the weight matrix corresponding to the second reconstructed value and the elements of the weight matrix corresponding to the first reconstructed value of the first sub-region are determined based on the distance between the pixels in the first sub-region and the boundary.

[0115] In this case, the elements in the weight matrix corresponding to the second reconstructed value are fixed values, and the elements in the weight matrix corresponding to the first reconstructed value of the first sub-region are also fixed values. Fixed weights are used for weighting in this scenario. The elements of the weight matrix corresponding to the second reconstructed value and the weight matrix corresponding to the first reconstructed value of the first sub-region are determined based on the distance between the pixels within the first sub-region and the boundary. The closer to the boundary, the larger the value of the element in the weight matrix corresponding to the second reconstructed value, and the smaller the value of the element in the weight matrix corresponding to the first reconstructed value of the first sub-region. In this case, a gradual weighting is used to simulate edge transition; the closer to the boundary, the larger the value of the element in the weight matrix corresponding to the second reconstructed value, i.e., the higher the weight.

[0116] In some embodiments, determining the filtering result of the first filtering region based on the adjusted reconstructed value of the first sub-region includes:

[0117] Cost analysis is performed on the adjusted reconstruction value of the first sub-region and the first reconstruction value corresponding to the first sub-region.

[0118] Based on the cost analysis results, the adjusted reconstruction value of the first sub-region or the first reconstruction value corresponding to the first sub-region is taken as the target reconstruction value of the first sub-region.

[0119] The filtering result of the first filtering region is determined based on the target reconstruction value of the first sub-region.

[0120] Here, the target reconstructed value can be understood as the final reconstructed value of the image. Cost analysis is performed on the adjusted reconstructed value of the first sub-region to determine the cost of the filtered reconstructed value, where the cost can be distortion, bit rate, etc. Cost analysis is also performed on the first reconstructed value corresponding to the first sub-region to determine the cost of the filtered reconstructed value. Comparing the two costs, and combining them with a pre-set selection criterion, a suitable reconstructed value is selected as the target reconstructed value from the adjusted reconstructed value of the first sub-region and the first reconstructed value corresponding to the first sub-region. For example, the reconstructed value with the lowest cost is selected as the target reconstructed value of the first sub-region. The filtering result of the first filtering region is determined based on the target reconstructed value of the first sub-region and the reconstructed value corresponding to the third region.

[0121] This application embodiment selects a suitable reconstruction value through cost analysis, that is, it selects whether to perform filtering adjustment. Finally, it selects a suitable reconstruction value from the first reconstruction value without filtering adjustment and the adjusted reconstruction value of the first sub-region to determine the filtering result of the first filtering region. Alternatively, cost analysis can be omitted, and the filtering result of the first filtering region can be determined directly based on the adjusted reconstruction value of the first sub-region and the corresponding reconstruction value of the third sub-region.

[0122] In some embodiments, after acquiring the first reconstructed image to be processed, the method further includes:

[0123] Based on the encoding information of the first reconstructed image, determine whether filtering adjustment is enabled for the first reconstructed image;

[0124] The encoded information includes at least one of the following:

[0125] The number of adaptive coefficient sets used in the first reconstructed image during the adaptive loop filtering process;

[0126] The number of adaptive coefficient sets used for the current strip brightness component corresponding to the first reconstructed image;

[0127] Temporal hierarchical information of the first reconstructed image;

[0128] The temporal hierarchical information of the current strip corresponding to the first reconstructed image.

[0129] Encoding information can be understood as information describing the image, or it can include information such as the image size, format, and the set of coefficients used for image filtering. After acquiring the first reconstructed image, the encoding information corresponding to the first reconstructed image is determined, further determining the image information corresponding to the original image. Based on the encoding information, it is determined whether filtering adjustment is enabled in the first reconstructed image. Filtering enabling conditions can be preset; if the conditions are met, filtering adjustment is enabled in the first reconstructed image; otherwise, filtering adjustment is disabled. If filtering adjustment is enabled in the first reconstructed image, it can be performed after filtering is complete. If filtering adjustment is disabled in the first reconstructed image, it will not be adjusted even if filtering has been performed.

[0130] For example, when the number of adaptive coefficient sets used in the adaptive loop filtering process of the first reconstructed image meets the preset rules, it is determined that the first reconstructed image is enabled for filtering adjustment, and filtering adjustment is allowed. When the number of adaptive coefficient sets used in the current strip brightness component corresponding to the first reconstructed image meets the preset rules, it is determined that the first reconstructed image is enabled for filtering adjustment, and filtering adjustment is allowed. When the temporal hierarchical information of the first reconstructed image meets the preset rules, it is determined that the first reconstructed image is enabled for filtering adjustment, and filtering adjustment is allowed. When the temporal hierarchical information of the current strip corresponding to the first reconstructed image meets the preset rules, it is determined that the first reconstructed image is enabled for filtering adjustment, and filtering adjustment is allowed. The adaptive coefficient set can be a historical filtering coefficient set.

[0131] In some embodiments, the filter adjustment conditions include at least one of the following:

[0132] The filter control information corresponding to the first filter region is different from the filter control information corresponding to the first adjacent filter region.

[0133] The width and / or height of the first filtering region meet the preset conditions.

[0134] The preset conditions can be a width threshold, a height threshold, the width-to-height ratio, the ratio of the width of the first filtering region to the width of the image, the ratio of the height of the first filtering region to the height of the image, etc. When the filtering control information corresponding to the first filtering region is different from the filtering control information corresponding to the first adjacent filtering region, and the width and / or height of the first filtering region meets at least one of the preset conditions, the first filtering region is determined to be adjusted for filtering.

[0135] In this embodiment, the first filtering region in the first reconstructed image is processed using the method described above to determine its corresponding filtering result, and finally an encoded image is generated.

[0136] In some embodiments, the method further includes:

[0137] Traverse each first filter region in the encoded image to determine whether there is a first filter region in the encoded image that needs to be adjusted for filtering.

[0138] If so, the filtering parameter information of the encoded image is generated based on the adjustment control information corresponding to each first filtering region.

[0139] In this embodiment, the adjustment control information can be understood as information describing whether and how the first filtering region has undergone filtering adjustment. For example, the adjustment control information could be that the first sub-region to the left of the boundary of the first filtering region has undergone filtering adjustment based on the filtering control information of the adjacent first neighboring filtering region. The filtering parameter information can be understood as information describing the filtering method and parameters used in the encoded image.

[0140] After filtering is completed in each first filtering region of the encoded image, information such as whether filtering adjustment was performed in this first filtering region and how the filtering adjustment was performed is recorded, and this information is used as the adjustment control information for this first filtering region. After generating the encoded image, each first filtering region in the encoded image is traversed to determine whether filtering adjustment has been performed in each first filtering region. If a first filtering region that has undergone filtering adjustment exists, the adjustment control information corresponding to this first filtering region is determined. The adjustment control information corresponding to the first filtering region that has undergone filtering adjustment is written into the image information to generate the filtering parameter information of the encoded image.

[0141] In some embodiments, the filter parameter information includes at least one of the following:

[0142] Index information of the filtering region;

[0143] Control information for filter adjustment.

[0144] The index information of the filtering region is used to indicate which first filtering region's filtering control information should be used for filtering adjustment in the first sub-region. For example, if the first sub-region has three adjacent regions, the first sub-region selects one of them for filtering adjustment, and the index information of the selected first adjacent region is written into the filtering parameter information when generating the filtering parameter information. The filtering adjustment control information is used to indicate whether filtering adjustment should be initiated for the filtering region, the sub-regions of the filtering region, etc.

[0145] In some embodiments, the control information for filter adjustment includes at least one of the following:

[0146] Control information for the luminance component;

[0147] Control information for chromaticity components;

[0148] Control information across components.

[0149] In some embodiments, the control information includes at least one of the following:

[0150] The first identifier, stored in the image header, is used to indicate whether adaptive loop filtering in the filtered region of the image is enabled for filtering adjustment.

[0151] The second identifier, stored in the strip header, is used to indicate whether adaptive loop filtering in the filtering region of the strip is enabled for filter adjustment.

[0152] The third identifier is stored in the syntax structure corresponding to the filtering region. The third identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether each sub-region of the filtering region is enabled for filtering adjustment.

[0153] The fourth identifier is stored in the syntax structure corresponding to the filtering region. The fourth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate the sub-regions in the filtering region that are enabled for filtering adjustment.

[0154] The fifth identifier is stored in the syntax structure corresponding to the filtering region. The fifth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether all sub-regions of the filtering region are enabled for filtering adjustment.

[0155] The first, second, third, fourth, and fifth identifiers are all identification information used to indicate whether filter adjustment is enabled in the filter region, and can be stored in different layers of image information. Taking the first reconstructed image as an example, the filter region indicated in each identifier refers to the first filter region, and the sub-region refers to the first sub-region. These identifiers can indicate whether filter adjustment is enabled in the filter region and its sub-regions. For example, the fourth identifier indicates that adaptive loop filtering in the filter region is enabled, and also indicates the sub-regions in the filter region where filter adjustment is enabled. The decoding end can determine whether filter adjustment is enabled in the encoded image based on the control information, so as to perform corresponding decoding processing.

[0156] In some embodiments, the first filtering region is a coding tree unit.

[0157] The coding unit is called CTU. Each CTU in the image can be used as a first filtering region for corresponding filtering processing. That is, in the process of encoding and filtering the image, each CTU in the image can be filtered using the method provided in the embodiments of this application.

[0158] This application provides an image processing method that solves the problem of poor filtering effect. A first filtering region in a first reconstructed image is filtered using filtering control information to obtain a first reconstructed value. Under any filtering adjustment condition, such as the filtering control information corresponding to the first filtering region being different from the filtering control information corresponding to the first adjacent filtering region, or the width and / or height of the first filtering region meeting preset conditions, a first sub-region within the first filtering region is filtered using the filtering control information of the first adjacent filtering region. Boundary filtering adjustment is performed based on the first and second reconstructed values ​​to achieve local optimization of the first filtering region. During the filtering adjustment process, factors such as distortion and cost after filtering using different methods are considered to comprehensively determine whether to perform filtering adjustment, resulting in a final filtering result for the first filtering region with good filtering effect. Finally, an encoded image is generated based on the filtering result of each first filtering region in the first reconstructed image, improving the image filtering accuracy, reducing local distortion, thereby improving coding performance and image processing effect.

[0159] For example, Figure 4 An example diagram illustrating the implementation of filtering adjustment is provided. The current image contains a total of 9 CTUs, with CTU4 and CTU6 being the adjacent regions above and to the left of CTU7, respectively. When the filter coefficient sets used by CTU4 and CTU6 during ALF filtering differ from those of the current CTU7, this characteristic can be utilized to fine-tune the sub-regions above and to the left of CTU7. For example, the filter coefficients of the current CTU can be omitted, and the filter coefficients of the adjacent CTUs can be used directly to filter the sub-regions. Alternatively, the filter coefficients of the current CTU and the filter coefficients of the adjacent CTUs can be weighted and fused to obtain the final filtering result. In the diagram, the small rectangles in CTU7 represent sub-regions, where the upper sub-region can be optimized using the ALF parameters of CTU4, and the left sub-region can be optimized using the ALF parameters of CTU6.

[0160] Figure 5 A flowchart of another image processing method provided in one embodiment, such as Figure 5 As shown, the image processing method of this application embodiment is applied to the decoding end, and the method includes S210-S230:

[0161] S210. Obtain the encoded image, decode the encoded image to obtain the second reconstructed image, and filter the second reconstructed image to obtain the third reconstructed value.

[0162] In this embodiment, the second reconstructed image can be understood as the image obtained after decoding the encoded image. The third reconstructed value can be understood as the information obtained after filtering the second reconstructed image.

[0163] After encoding the image, the encoding end sends the generated encoded image to the decoding end. When sending the encoded image, the encoding end can generate a video stream based on it. The video stream always includes one or more video frames, each of which is an encoded image. The decoding end receives the video stream and determines the encoded image. After obtaining the encoded image, the decoding end performs decoding processing on the encoded image according to a decoding algorithm to obtain a second reconstructed image. The second reconstructed image is then filtered according to a pre-set filtering method to obtain a third reconstructed value. The filtering process can be at least one of DBF, SAO, and ALF.

[0164] S220. When it is determined that the coded image is enabled for filtering adjustment, the second sub-region of the second filtering region in the second reconstructed image and the second adjacent filtering region corresponding to the second sub-region are determined. The second sub-region is filtered according to the filtering control information corresponding to the second adjacent filtering region to obtain the fourth reconstructed value.

[0165] The second filtering region can be understood as a portion of the second reconstructed image. For example, the second reconstructed image can be divided into regions according to certain rules, and each resulting region can serve as the second filtering region. For instance, the second reconstructed image can be divided using preset sizes, CTUs, CUs, etc., as dividing units to obtain the second filtering regions. The second sub-region can be understood as the portion of the second filtering region that requires filtering adjustment. It can be a region at a specified location within the second filtering region, or it can be any region at any location within the second filtering region. The second adjacent filtering region can be understood as a second filtering region adjacent to the second sub-region.

[0166] To determine whether filtering adjustment is enabled in the encoded image, the encoding end generates indication information when generating the encoded image. This information can be included in the encoded image and sent to the decoding end, or it can be sent separately. Correspondingly, the decoding end determines whether filtering adjustment is enabled based on the indication information in the encoded image. If filtering adjustment is enabled, each filtering region in the second reconstructed image is analyzed to determine the filtering region requiring adjustment, which is denoted as the second filtering region. The second filtering region can be determined based on information from the encoded image or through analysis of the encoded image itself. The second filtering region can be divided into different sub-regions. The sub-regions requiring filtering adjustment within the second filtering region are determined based on information from the encoded image or through analysis of the encoded image itself, and are denoted as the second sub-regions. Other sub-regions are denoted as the fourth sub-regions. After determining the second sub-region, the corresponding second adjacent filtering region is determined based on its position and other information. The filter control information corresponding to the second adjacent filter region is determined. Since the filter control information indicates whether filtering is enabled in the second adjacent filter region and the filter parameters used when filtering is enabled, the filter control information can be used to determine how to filter the second sub-region. The corresponding filter processing method is used to filter the second sub-region to obtain the fourth reconstructed value.

[0167] S230. Determine the filtering result of the second filtering region in the second reconstructed image based on the third reconstruction value and the fourth reconstruction value of the second sub-region, and generate a decoded image based on the filtering result of each second filtering region in the second reconstructed image.

[0168] For the second sub-region, there are two corresponding reconstructed values: a third reconstructed value and a fourth reconstructed value. These values ​​are obtained by filtering using two different filtering control information. The third and fourth reconstructed values ​​for the second sub-region are then processed comprehensively. For example, one of the three values ​​is selected as the final filtering result for the second sub-region; alternatively, the third and fourth values ​​are weighted, and operations such as maximizing or minimizing are performed to obtain the final filtering result for the second sub-region. For the second filtering region, the filtering result is determined based on the final filtering result for each second sub-region and the third reconstructed value for the fourth sub-region. A decoded image is then generated based on the filtering results for each second filtering region in the second reconstructed image and the third reconstructed values ​​for other filtering regions.

[0169] This application provides an image processing method that solves the problem of poor filtering effect. A second reconstructed image is obtained by decoding an encoded image, and a third reconstructed value is obtained by filtering the second reconstructed image. With filtering adjustment enabled in the encoded image, a second sub-region of the second filtering region and its corresponding second adjacent filtering region are determined in the second reconstructed image. The second sub-region of the second filtering region is filtered using the filtering control information of the second adjacent filtering region. Then, the filtering result of the second filtering region is determined based on the third and fourth reconstructed values, achieving local optimization of the second filtering region and obtaining the final filtering result of the second filtering region. Finally, an encoded image is generated based on the filtering result of each second filtering region of the second reconstructed image, improving the image filtering accuracy.

[0170] In some embodiments, determining whether filtering adjustment is enabled for the encoded image includes:

[0171] Parse the image header or strip header of the encoded image to obtain the filtering parameter information;

[0172] Determine whether filtering adjustment is enabled for the encoded image based on the filtering parameter information.

[0173] The image header or strip header of the encoded image can carry filtering parameter information. Parsing the image header or strip header yields the filtering parameter information. Based on the information included in the filtering parameter information, it can be analyzed to determine whether filtering adjustment is enabled in the encoded image. For example, if the filtering parameter information includes an indication that adaptive loop filtering is enabled in the filtering region, it can be determined that filtering adjustment is enabled in the encoded image.

[0174] In some embodiments, determining a second sub-region of the second filtering region in the second reconstructed image and a second adjacent filtering region corresponding to the second sub-region includes:

[0175] The syntax structure of each second filtering region in the encoded image is analyzed to obtain the filtering parameter information;

[0176] The second sub-region and the corresponding second adjacent filtering region are determined based on the filtering parameter information.

[0177] The syntax structure of the filtered regions in the encoded image can carry filtering parameter information. Parsing the syntax structure of each second filtered region yields the filtering parameter information. Based on the information included in the filtering parameter information, the second sub-region requiring filtering adjustment is determined, along with its corresponding second adjacent filtered region. For example, the filtering parameter information includes identifiers indicating the second sub-region within the second filtered region and identifiers indicating the second adjacent filtered region. These identifiers allow for the identification of the corresponding second sub-region and the second adjacent filtered region.

[0178] In some embodiments, the filter parameter information includes at least one of the following:

[0179] Index information of the filtering region;

[0180] Control information for filter adjustment.

[0181] In some embodiments, the control information for filter adjustment includes at least one of the following:

[0182] Control information for the luminance component;

[0183] Control information for chromaticity components;

[0184] Control information across components.

[0185] In some embodiments, the control information includes at least one of the following:

[0186] The first identifier, stored in the image header, is used to indicate whether adaptive loop filtering in the filtered region of the image is enabled for filtering adjustment.

[0187] The second identifier, stored in the strip header, is used to indicate whether adaptive loop filtering in the filtering region of the strip is enabled for filter adjustment.

[0188] The third identifier is stored in the syntax structure corresponding to the filtering region. The third identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether each sub-region of the filtering region is enabled for filtering adjustment.

[0189] The fourth identifier is stored in the syntax structure corresponding to the filtering region. The fourth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate the sub-regions in the filtering region that are enabled for filtering adjustment.

[0190] The fifth identifier is stored in the syntax structure corresponding to the filtering region. The fifth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether all sub-regions of the filtering region are enabled for filtering adjustment.

[0191] Since the filtering parameter information includes one or more of the above information, the second sub-region and the second adjacent filtering region corresponding to the second sub-region can be determined by parsing the filtering parameter information.

[0192] In some embodiments, filtering is performed on the second sub-region based on the filtering control information corresponding to the second adjacent filtering region to obtain the fourth reconstructed value, including:

[0193] The filter subset used in the second adjacent filter region is determined based on the filter control information of the second adjacent filter region.

[0194] The second sub-region is filtered based on the filter subset used in the second adjacent filter region to obtain the fourth reconstructed value.

[0195] The filtering control information of the second adjacent filtering region is analyzed. This information includes whether filtering is enabled in the second adjacent filtering region and the subset used for filtering. Based on the analysis results, the filtering subset used in the second adjacent filtering region is determined. The second sub-region is then filtered using the filtering subset used in the second adjacent filtering region to obtain the fourth reconstructed value.

[0196] In some embodiments, determining the filtering result of the second filtered region in the second reconstructed image based on the third reconstructed value and the fourth reconstructed value of the second sub-region includes:

[0197] The fourth reconstruction value and the third reconstruction value corresponding to the second sub-region are weighted according to the corresponding weight matrix to obtain the adjusted reconstruction value of the second sub-region.

[0198] The filtering result of the second filtering region in the second reconstructed image is determined based on the adjusted reconstruction value of the second sub-region.

[0199] Determine the weight matrix corresponding to the fourth reconstructed value and the third reconstructed value corresponding to the second sub-region. This weight matrix can be pre-generated and directly obtained during weighting. The adjusted reconstructed value of the second sub-region is obtained through weighted calculation, which involves weighting the filtered third reconstructed value corresponding to the second sub-region with the fourth reconstructed value obtained after filtering adjustment. Based on the adjusted reconstructed value of the second sub-region, combined with the reconstructed value of the fourth sub-region in the second filtered region that has not undergone filtering adjustment, the filtering result of the second filtered region is determined.

[0200] In this embodiment, the second filtering region in the second reconstructed image is processed using the method described above to determine the corresponding filtering result, and finally a decoded image is generated.

[0201] In some embodiments, the weight matrix includes at least one of the following:

[0202] The elements in the weight matrix corresponding to the fourth reconstructed value, and the elements in the weight matrix corresponding to the third reconstructed value of the second sub-region, are fixed values;

[0203] The elements of the weight matrix corresponding to the fourth reconstructed value and the elements of the weight matrix corresponding to the third reconstructed value in the second sub-region are determined based on the distance between the pixels in the second sub-region and the boundary.

[0204] In some embodiments, the filter control information includes at least one of the following:

[0205] Adaptive loop filter control information for the luminance component;

[0206] Adaptive loop filter control information for chrominance components;

[0207] Cross-component adaptive loop filter control information.

[0208] In some embodiments, the adaptive loop filter control information includes at least one of the following:

[0209] Whether adaptive loop filtering is enabled in the filtering region or for the components of the filtering region;

[0210] The index of the fixed subset of filters corresponding to the filtering region;

[0211] Index of the adaptive parameter set corresponding to the filtering region;

[0212] The index of the optional subset in the adaptive parameter set corresponding to the filtering region.

[0213] In some embodiments, the second sub-region includes at least one of the following:

[0214] A sub-region of a preset position and size located at the boundary of the second filtering region;

[0215] Sub-regions obtained using a preset partitioning method.

[0216] In some embodiments, the partitioning method includes at least one of the following:

[0217] Classification is performed based on the directionality factor of the luminance pixel blocks in the adaptive loop filter;

[0218] Classification is based on the activity factor of the luminance pixel blocks in the adaptive loop filter;

[0219] Classification is based on the sideband information of adaptive loop filtering;

[0220] Classification is based on the residuals of adaptive loop filtering.

[0221] In some embodiments, the filter control information includes at least one of the following:

[0222] Adaptive loop filter control information for the luminance component;

[0223] Adaptive loop filter control information for chrominance components;

[0224] Cross-component adaptive loop filter control information.

[0225] In some embodiments, the adaptive loop filter control information includes at least one of the following:

[0226] Whether adaptive loop filtering is enabled in the filtering region or for the components of the filtering region;

[0227] The index of the fixed subset of filters corresponding to the filtering region;

[0228] Index of the adaptive parameter set corresponding to the filtering region;

[0229] The index of the optional subset in the adaptive parameter set corresponding to the filtering region.

[0230] In some embodiments, the second filtering region is a coding tree unit.

[0231] This application provides an image processing method that solves the problem of poor filtering effect. The method filters the second reconstructed image using filtering control information to obtain a third reconstructed value. By parsing the image header or strip header of the encoded image, filtering parameter information is obtained. Based on the filtering parameter information, when filtering adjustment of the encoded image is enabled, the second sub-region of the second filtering region in the second reconstructed image and the corresponding second adjacent filtering region are determined. The second sub-region of the second filtering region is filtered using the filtering control information of the second adjacent filtering region to obtain a fourth reconstructed value. Boundary filtering adjustment is performed based on the third and fourth reconstructed values ​​to achieve local optimization of the second filtering region, resulting in a better filtering effect. Finally, a decoded image is generated based on the filtering results of each second filtering region of the second reconstructed image, improving the image filtering accuracy, reducing local distortion, thereby improving coding performance and image processing effect.

[0232] The image processing process is illustrated through the following examples:

[0233] Example 1

[0234] This embodiment provides an image processing method that fine-tunes the boundary regions of an image through loop filtering. Applied to the encoding end, the image processing process is illustrated using a first filtering region (CTU) as an example. The encoder obtains the reconstructed values ​​of the original image to be processed, i.e., the first reconstructed image. The first reconstructed image undergoes filtering processing, where the luminance component, chrominance component, and cross component undergo ALF filtering decisions at the slice level and CTU level. For the sub-regions at the boundary positions within each CTU, optimization is performed based on the filtering parameter information of adjacent CTUs, thereby reducing local distortion and improving encoding performance. Specific steps include:

[0235] Step 1: The encoder obtains the reconstructed value R0 of the current image to be processed based on the existing encoding information, performs filtering processing on the first reconstructed image, including DBF, SAO, and ALF, and uses the RDO rate distortion optimization method to make decisions for each CTU to determine the filtering control information. The filtering control information is used to select whether to enable ALF, and if ALF is enabled, the subset used by the current CTU.

[0236] The encoder uses the Rate-Distortion Optimization (RDO) method to make decisions for each CTU, including decisions for each component of the CTU, to obtain filter control information. This filter control information includes:

[0237] 1) ALF control information for the luminance component;

[0238] 2) ALF control information for chromaticity components, including at least one of the following: control information for the two chromaticity components, Cb and Cr;

[0239] 3) ALF control information across components, including at least one of the following: control information for two components, CCALF-Cb and CCALF-Cr.

[0240] The ALF control information currently used by CTU includes at least one of the following:

[0241] 1) Whether ALF is enabled for the current CTU or its components;

[0242] 2) The current fixed subset index of the filter used by CTU;

[0243] 3) The index of the adaptive parameter set currently used by the CTU (e.g., the index of the historical adaptive parameter set currently used by the CTU);

[0244] 4) Optional subset indexes of the adaptive parameter set currently used by the CTU (e.g., optional subset indexes of the historical adaptive parameter set currently used by the CTU).

[0245] Step 2: Based on the decision obtained in Step 1, perform ALF filtering on the current CTU of the current image to obtain the first reconstructed value R1 after filtering;

[0246] The encoding end performs ALF filtering on the current CTU in at least one of the following ways:

[0247] 1) Based on the ALF filtering control information of the luminance component in step 1, perform ALF filtering on the luminance component of the current CTU to obtain the reconstructed value of the luminance component.

[0248] 2) Based on the ALF filtering control information of the chromaticity component and the cross-component CC-ALF filtering control information in step 1, perform ALF filtering on the chromaticity component of the current CTU to obtain the reconstructed value of the chromaticity component.

[0249] Step 3: Determine the sub-region located at the boundary in the current CTU (i.e., the first sub-region), and obtain the ALF control information used by the CTUs at adjacent positions on each boundary (i.e., the first adjacent filtering region);

[0250] The sub-regions that can be fine-tuned at the encoding end refer to the regions at pre-specified boundary positions. During the fine-tuning process of the filtering results, the encoding end traverses one or more sub-regions according to a preset method, and the decoding end uses the same preset method. Relevant information does not need to be transmitted in the bitstream. The position of the boundary, the size of the sub-region, and the number of sub-regions are determined based on the actual effect during implementation, and can also be selected according to different service types and different requirements. This embodiment does not impose any restrictions. Figure 6 A schematic diagram of sub-region division is provided, in which the boundary sub-region is set to the upper part according to the encoding order (e.g., ...). Figure 6 The boundary sub-region 1 shown and the left side (as shown) Figure 6 The boundary sub-region shown is 0) two sub-regions. Figure 7 Another diagram illustrating sub-region division is provided, defining three boundary sub-regions, including the left side (e.g., Figure 7 The boundary sub-region shown is 0), and the upper left (e.g.) Figure 7 The boundary sub-region 1 shown above and above (as shown) Figure 7 The boundary sub-region shown is 2). The boundary sub-region can be used as the first sub-region for boundary filtering.

[0251] In addition to the four pre-defined fixed sub-regions at the boundaries, the CTU can be categorized to generate multiple sub-regions. Among these sub-regions, the sub-regions near the boundaries can be optimized using the boundary fine-tuning method described in this scheme. The internal categorization of the CTU mentioned here can be obtained by categorizing according to the pre-defined categorization method of the encoding and decoding ends. The categorization method can include various methods, such as the ALF luminance pixel block classification method in VVC (based on directional factors and activity factors), or the extended ALF classification method in ECM (based on sideband information or residual classification). This embodiment does not limit the specific methods.

[0252] Step 4: When the ALF control information used by the CTU at an adjacent location is different from that of the current CTU, for each sub-region, obtain the reconstructed value R0 before ALF, and the encoder uses the ALF control information used by the CTU adjacent to the sub-region to re-filter the sub-region to obtain the filtered reconstructed value R2.

[0253] Since for the luminance component of each CTU, after the RDO decision, there are three possible decision results: not enabling ALF, filtering using the filtering coefficients of a fixed subset, and filtering using the filtering coefficients of a historical APS subset (including the latest APS subset calculated in the current frame), the encoder may also re-filter the sub-region using the ALF control information adopted by the CTUs adjacent to the sub-region, including one of the following:

[0254] 1) When the adjacent CTU does not enable ALF, the reconstructed value R0 before ALF in this sub-region is directly used as the filtered reconstructed value R2;

[0255] 2) When adjacent CTUs use a fixed subset or a historical APS subset for filtering (i.e., ALF is enabled), the same fixed subset or historical APS subset is used to filter the sub-region to obtain the reconstructed value R2.

[0256] Step 5: Compare the pixel values ​​of the original image. Compare the reconstruction values ​​R1 obtained in Step 2 and R2 obtained in Step 4 with the distortion of the original image. If the distortion corresponding to the reconstruction value R2 is smaller, use R2 to fine-tune the reconstruction value of the sub-region.

[0257] When the distortion is reduced after filtering the current CTU sub-region using the ALF filtering coefficients of the neighboring CTU, the encoder can optimize the final reconstructed value of the sub-region. For example, the reconstructed value R2 and the reconstructed value R1 can be weighted and fused to obtain the adjusted reconstructed value R.

[0258] R = (R1°W1 + R2°W2)

[0259] Where ° represents the operation of element-wise multiplication, and the sum of the two weights corresponding to each pixel in W1 and W2 is 1.

[0260] Weighted fusion includes at least the following three methods:

[0261] 1) The reconstructed value R2 is selected as the fusion result, that is, W2 is a matrix in which all elements are 1;

[0262] 2) The reconstructed values ​​R1 and R2 are weighted with constant fixed weights, that is, the elements in W1 and W2 are fixed values;

[0263] 3) The reconstructed values ​​R1 and R2 are weighted with a gradual weight to simulate edge transition. The closer to the boundary, the higher the corresponding weight value in W2. That is, the elements in W1 and W2 are determined based on the distance between the pixels in the sub-region and the boundary.

[0264] The above formula corresponds to a scenario where the sub-region includes only one adjacent CTU. When the number of adjacent CTUs in the sub-region is greater than 1, the above formula can be extended to a weighted fusion of multiple reconstructed values. This embodiment does not limit this.

[0265] Step 6: The encoding end performs RDO (Reconstruction-Oriented Decision) decisions for both the scheme with and without boundary fine-tuning. This involves cost analysis of the reconstructed values ​​R and R1 after sub-region adjustment, selecting the scheme with the lowest cost as the final filtering method. Based on the cost analysis results, one of R and R1 is chosen as the target reconstructed value.

[0266] It should be noted that when the number of adjacent CTUs in a sub-region is 1, the encoder only needs to decide whether to use the boundary fine-tuning scheme. When the number of adjacent CTUs in a sub-region is greater than 1 (for example, in a sub-region at the vertex, the set of comparable filter coefficients can include up to 4, namely the filter coefficients of the adjacent CTU above, the adjacent CTU to the left, the adjacent CTU to the upper left, and the filter coefficient set of this CTU), the encoder can also use RDO to make a choice among the schemes corresponding to multiple adjacent CTU coefficients. Accordingly, the index information of the selected adjacent CTUs can be written into the video encoded bitstream for the decoder to refer to for decoding.

[0267] Step 7: Perform the above steps sequentially on all CTUs in the current image, process them using the filtering scheme obtained in Step 6, and obtain the filtered reconstructed value R2 of the current image; and process the reconstructed value R2 accordingly to obtain the target reconstructed value.

[0268] After each CTU in the first reconstructed image is filtered in the manner described above, an encoded image is generated.

[0269] Step 8: The encoding end traverses all CTUs in the current encoded image. If there is a CTU that uses boundary fine-tuning in this image, it is determined that the ALF filtering boundary fine-tuning method is enabled for this image, and the adjustment control information related to ALF boundary fine-tuning is written into the video encoding bitstream to generate filtering parameter information.

[0270] Example 2: Fine-tuning loop filtering method for defined boundary regions

[0271] This embodiment provides a further coding end decision-making method for the loop filtering method for boundary region fine-tuning described in Embodiment 1. In addition to using RDO to decide whether to use the boundary fine-tuning filtering method, the coding end can also determine whether to turn this method off or on in the current image based on information such as the number of adaptive coefficient sets used in the current image (e.g., the number of historical filter coefficient sets) and the temporal level, thereby improving coding efficiency.

[0272] Optional Implementation Method 1:

[0273] Steps 1-2: Same as in Example 1, obtain the reconstructed value of the current CTU in the current first reconstructed image after ALF filtering;

[0274] Step 3: (Optional) Obtain the number of adaptive coefficient sets (sh_num_alf_aps_ids_luma or ph_num_alf_aps_ids_luma) used by the first reconstructed image or the current strip luminance component. If the number of adaptive coefficient sets used by the luminance component meets the preset rules, the boundary fine-tuning method is allowed to be enabled; otherwise, it is not allowed to be enabled.

[0275] For example, when the number of sh_num_alf_aps_ids_luma or ph_num_alf_aps_ids_luma is greater than the threshold T, the first reconstructed image is allowed to enable the boundary fine-tuning method (i.e., enable filter adjustment), otherwise it is not allowed to enable it;

[0276] Step 4: If the first reconstructed image allows the boundary fine-tuning method to be enabled, perform steps 3-8 of Example 1.

[0277] Optional Implementation Method Two:

[0278] Steps 1-2: Same as in Example 1, obtain the reconstructed value of the current CTU in the current first reconstructed image after ALF filtering;

[0279] Step 3: (Optional) Obtain the temporal layer information (temporal_layer_id) of the first reconstructed image or the current strip. If the temporal layer meets the preset rules, enable the boundary fine-tuning method; otherwise, disable it.

[0280] For example, when the number of time domain levels is greater than a threshold T, the first reconstructed image is allowed to enable the boundary fine-tuning method; otherwise, it is not allowed.

[0281] Step 4: If the first reconstructed image allows the boundary fine-tuning method to be enabled, perform steps 3-8 of Example 1.

[0282] Optional Implementation Method 3:

[0283] Steps 1-2: Same as in Example 1, obtain the reconstructed value of the current CTU in the first reconstructed image after ALF filtering;

[0284] Step 3: The encoder determines whether the first reconstructed image allows the ALF boundary fine-tuning method to be enabled;

[0285] Step 4: If the first reconstructed image allows the boundary fine-tuning method to be enabled, for the current CTU, obtain the CTU width and height. For CTUs whose width or height does not meet the preset conditions, the ALF boundary fine-tuning method is not allowed to be enabled.

[0286] Step 5: If the ALF boundary fine-tuning method is enabled, execute steps 4-8 of Example 1 to fine-tune the current CTU boundary according to the filtering control information of adjacent CTUs.

[0287] Example 3: Fine-tuning loop filtering method for decoding side boundary region

[0288] This embodiment provides an image processing method that fine-tunes the boundary regions of an image through loop filtering. Applied to the decoding end, the image processing process is illustrated using the second filtering region as a CTU as an example. After obtaining the second reconstructed value corresponding to the encoded image, the decoder performs filtering processing on the second reconstructed image. Based on the ALF filtering control signaling and boundary fine-tuning related signaling parsed from the bitstream, after ALF filtering at the CTU level, for the sub-region (i.e., the second sub-region) at the boundary position in each CTU, optimization is performed based on the filtering parameter information of adjacent CTUs. Specific steps include:

[0289] Step 1: The decoder receives the video stream, which includes one or more video frames and corresponding ALF filter control signaling and ALF filter boundary fine-tuning control signaling.

[0290] Step 2: The decoder decodes the encoded image in the video stream to obtain the second reconstructed image, and performs filtering processing on the second reconstructed image, including DBF, SAO, and ALF.

[0291] Step 3: The decoder parses the image header or strip header to obtain the filtering parameter information (the filtering parameter information may include ALF filtering boundary fine-tuning control signaling), and determines whether the ALF filtering boundary fine-tuning method is enabled for the encoded image;

[0292] Step 4: If the encoded image uses the ALF filtering boundary fine-tuning method, the decoder further parses the CTU syntax structure to obtain the filtering parameter information. The filtering parameter information in the syntax structure may include ALF filtering boundary fine-tuning control signaling to determine whether the current CTU uses the ALF filtering boundary fine-tuning method.

[0293] Step 5: If the current CTU uses the ALF filtering boundary fine-tuning method, determine the CTU sub-region that uses boundary fine-tuning, and determine the ALF filtering control information used by the adjacent CTUs in the sub-region;

[0294] Step 6: For sub-regions using boundary fine-tuning, two sets of ALF filtering control information are used to perform ALF filtering on the reconstructed image of the sub-region, and then weighted and fused. The two sets of ALF filtering information include the ALF filtering control information used by the current CTU and the ALF filtering control information used by the adjacent CTUs of the sub-region. For pixels in other regions of the CTU, the ALF filtering control information of the current CTU is directly used for filtering to obtain the reconstructed image.

[0295] Example 4: Grammar

[0296] This embodiment provides a method for indicating ALF boundary fine-tuning control information in the bitstream. The syntax and semantic information can be applied to combinations of various cases described in Embodiments 1 to 3. It should be noted that the following syntax is only an example applied to the ALF filter boundary fine-tuning method for the luminance component. For control signaling of other components such as the chrominance component and CC-ALF, the control signaling settings for the luminance component can be referenced.

[0297] Image header syntax: A flag indicating whether CTU boundary fine-tuning is enabled for the brightness ALF, for example, the first flag. Table 1 shows the image header syntax structure.

[0298] Table 1 Image Header Syntax Structure

[0299]

[0300] Semantics:

[0301] For example, setting ph_alf_boundary_merge_enabled_flag in the image header control signaling indicates whether the ALF boundary fine-tuning method is enabled for the current image; 1 indicates enabled, and 0 indicates disabled.

[0302] Strip header syntax: A flag indicating whether CTU boundary fine-tuning is enabled in the luminance ALF, for example, a second flag. Table 2 shows the strip header syntax structure.

[0303] Table 2. List of heading syntax structures

[0304]

[0305] Semantics:

[0306] For example, setting sh_alf_boundary_merge_enabled_flag in the slice header control signaling indicates whether the ALF boundary fine-tuning method is enabled for the current slice, with 1 indicating enabled and 0 indicating disabled.

[0307] CTU syntax:

[0308] Option 1: Each CTU structure includes an identifier indicating whether the luminance ALF of that CTU is enabled for CTU boundary fine-tuning, and an identifier indicating whether boundary fine-tuning is enabled for each sub-region, such as a third identifier. For example, when the CTU contains preset sub-regions in the upper and left directions, the CTU-level ALF filter boundary fine-tuning control signaling can be set as follows, as shown in Table 3, which is a CTU syntax structure table:

[0309] Table 3 CTU Syntax Structure Table

[0310]

[0311]

[0312] Semantics:

[0313] alf_boundary_merge_flag is used to indicate whether the ALF boundary fine-tuning method is enabled in the current CTU. 1 indicates that it is enabled and 0 indicates that it is not enabled.

[0314] alf_boundary_merge_flag_b0 is used to indicate whether the ALF boundary fine-tuning method is enabled in the left subregion of the current CTU. 1 indicates enabled, and 0 indicates disabled.

[0315] alf_boundary_merge_flag_b1 is used to indicate whether the ALF boundary fine-tuning method is enabled in the sub-region above the current CTU. 1 indicates enabled, and 0 indicates disabled.

[0316] Option 2: Each CTU structure includes an identifier indicating whether the CTU's luminance ALF is enabled for CTU boundary fine-tuning, and the ID of the sub-region where boundary fine-tuning is enabled, for example, the fourth identifier. For example, when the CTU contains preset sub-regions in the upper, upper left, and upper directions, and only one sub-region is enabled for ALF filter fine-tuning at a time, the CTU-level ALF filter boundary fine-tuning control signaling can be set as follows. Table 4 shows another CTU syntax structure table:

[0317] Table 4 CTU Syntax Structure Table

[0318]

[0319] Semantics:

[0320] alf_boundary_merge_flag is used to indicate whether the ALF boundary fine-tuning method is enabled in the current CTU. 1 indicates that it is enabled and 0 indicates that it is not enabled.

[0321] alf_boundary_subregion_idx is used to indicate the subregion ID where boundary fine-tuning is currently enabled in the current CTU. 0 indicates the left subregion, 1 indicates the upper left subregion, and 2 indicates the upper subregion.

[0322] Option 1: Each CTU structure includes an identifier indicating whether the luminance ALF of that CTU enables CTU boundary fine-tuning, controlling whether boundary fine-tuning is enabled in all sub-regions, for example, the fifth identifier. Table 5 shows another CTU syntax structure table.

[0323] Table 5 CTU Syntax Structure

[0324]

[0325] Semantics:

[0326] For example, the alf_boundary_merge_flag can be set in the control signaling to indicate whether the ALF boundary fine-tuning method is enabled in the current CTU's preset sub-regions. 1 indicates that it is enabled and 0 indicates that it is not enabled.

[0327] It should be noted that the boundary fine-tuning mentioned in the above embodiments refers to filter adjustment.

[0328] Figure 8 This is a schematic diagram of the structure of an image processing apparatus provided in one embodiment. This apparatus is applied to the encoding end, such as... Figure 8 As shown, the device includes: a first reconstructed image acquisition module 310, a second reconstructed value determination module 320, and a first filtering adjustment module 330.

[0329] The first reconstructed image acquisition module 310 is used to acquire the current first reconstructed image to be processed, make decisions for each first filtering region of the first reconstructed image, determine the filtering control information corresponding to each first filtering region, and perform adaptive loop filtering on the first filtering region according to the filtering control information to obtain the first reconstructed value.

[0330] The second reconstruction value determination module 320 is used to determine the first sub-region in the first filtering region and the first adjacent filtering region corresponding to the first sub-region. When the first filtering region meets the filtering adjustment conditions, the first sub-region is filtered according to the filtering control information corresponding to the first adjacent filtering region to obtain the second reconstruction value.

[0331] The first filtering adjustment module 330 is used to perform filtering adjustment based on the first reconstructed value and the second reconstructed value of the first sub-region, determine the filtering result of the first filtering region, and generate an encoded image based on the filtering result of each first filtering region of the first reconstructed image.

[0332] This application provides an image processing apparatus that solves the problem of poor filtering effect. It filters a first filtering region in a first reconstructed image using filtering control information to obtain a first reconstructed value. When the first filtering region meets the filtering adjustment conditions, it filters a first sub-region within the first filtering region using filtering control information from a first adjacent filtering region. Then, it adjusts the filtering based on the first and second reconstructed values ​​to achieve local optimization of the first filtering region, obtaining the final filtering result of the first filtering region. Finally, it generates an encoded image based on the filtering results of each first filtering region of the first reconstructed image, thereby improving the image filtering accuracy.

[0333] In some embodiments, the filter control information includes at least one of the following:

[0334] Adaptive loop filter control information for the luminance component;

[0335] Adaptive loop filter control information for chrominance components;

[0336] Cross-component adaptive loop filter control information.

[0337] In some embodiments, the adaptive loop filter control information includes at least one of the following:

[0338] Whether adaptive loop filtering is enabled in the filtering region or for the components of the filtering region;

[0339] The index of the fixed subset of filters corresponding to the filtering region;

[0340] Index of the adaptive parameter set corresponding to the filtering region;

[0341] The index of the optional subset in the adaptive parameter set corresponding to the filtering region.

[0342] In some embodiments, the adaptive loop filtering of the first filtering region includes at least one of the following methods:

[0343] Adaptive loop filtering is performed on the luminance components of the first filtering region based on the adaptive loop filtering control information of the luminance components.

[0344] Based on the adaptive loop filtering control information of the chrominance components and the adaptive loop filtering control information across components, adaptive loop filtering is performed on the chrominance components of the first filtering region.

[0345] In some embodiments, the first sub-region includes at least one of the following:

[0346] A sub-region of a preset position and size located at the boundary of the first filtering region;

[0347] Sub-regions obtained using a preset partitioning method.

[0348] In some embodiments, the division method includes at least one of the following:

[0349] Classification is performed based on the directionality factor of the luminance pixel blocks in the adaptive loop filter;

[0350] Classification is based on the activity factor of the luminance pixel blocks in the adaptive loop filter;

[0351] Classification is based on the sideband information of adaptive loop filtering;

[0352] Classification is based on the residuals of adaptive loop filtering.

[0353] In some embodiments, the step of filtering the first sub-region according to the filtering control information corresponding to the first adjacent filtering region to obtain the second reconstructed value includes:

[0354] If it is determined that the first adjacent filtering region does not enable adaptive loop filtering based on the filtering control information corresponding to the first adjacent filtering region, the reconstruction value corresponding to the first sub-region in the first reconstructed image is taken as the second reconstruction value.

[0355] If it is determined that the first adjacent filtering region is enabled by adaptive loop filtering based on the filtering control information corresponding to the first adjacent filtering region, the filtering subset used by the first adjacent filtering region is determined, and the first sub-region is filtered according to the filtering subset to obtain the second reconstructed value.

[0356] In some embodiments, when the number of the first adjacent filtered regions is at least two, the step of filtering the first sub-region according to the filtered subset to obtain the second reconstructed value includes:

[0357] Select a filter subset from each of the first adjacent filter regions to determine a target filter subset, and then filter the first sub-region according to the target filter subset to obtain a second reconstructed value; or...

[0358] The first sub-region is filtered according to the filtering subset of each of the first adjacent filtering regions to obtain at least two second reconstructed values.

[0359] In some embodiments, the step of performing filtering adjustment based on the first reconstructed value and the second reconstructed value of the first sub-region to determine the filtering result of the first filtering region includes:

[0360] Compare the first reconstructed value with the original image corresponding to the first reconstructed image to determine the first distortion;

[0361] The second reconstructed value is compared with the original image corresponding to the first reconstructed image to determine the second distortion;

[0362] If the second distortion is less than the first distortion, the first sub-region is adjusted according to the second reconstruction value;

[0363] The filtering result of the first filtering region is determined based on the adjusted reconstruction value of the first sub-region.

[0364] In some embodiments, adjusting the first sub-region based on the second reconstructed value includes:

[0365] The second reconstructed value and the first reconstructed value corresponding to the first sub-region are weighted according to the corresponding weight matrix to obtain the adjusted reconstructed value of the first sub-region.

[0366] In some embodiments, the weight matrix includes at least one of the following:

[0367] The elements in the weight matrix corresponding to the second reconstructed value and the elements in the weight matrix corresponding to the first reconstructed value of the first sub-region are fixed values;

[0368] The elements of the weight matrix corresponding to the second reconstructed value and the elements of the weight matrix corresponding to the first reconstructed value of the first sub-region are determined based on the distance between the pixels in the first sub-region and the boundary.

[0369] In some embodiments, determining the filtering result of the first filtering region based on the adjusted reconstructed value of the first sub-region includes:

[0370] Cost analysis is performed on the adjusted reconstruction value of the first sub-region and the first reconstruction value corresponding to the first sub-region.

[0371] Based on the cost analysis results, the adjusted reconstruction value of the first sub-region or the first reconstruction value corresponding to the first sub-region is taken as the target reconstruction value of the first sub-region.

[0372] The filtering result of the first filtering region is determined based on the target reconstruction value of the first sub-region.

[0373] In some embodiments, after acquiring the first reconstructed image to be processed, the method further includes:

[0374] Based on the encoding information of the first reconstructed image, determine whether filtering adjustment is enabled for the first reconstructed image;

[0375] The encoded information includes at least one of the following:

[0376] The number of adaptive coefficient sets used in the adaptive loop filtering process of the first reconstructed image;

[0377] The number of adaptive coefficient sets used for the current strip brightness component corresponding to the first reconstructed image;

[0378] The temporal hierarchical information of the first reconstructed image;

[0379] The temporal hierarchical information of the current strip corresponding to the first reconstructed image.

[0380] In some embodiments, the filter adjustment conditions include at least one of the following:

[0381] The filter control information corresponding to the first filter region is different from the filter control information corresponding to the first adjacent filter region.

[0382] The width and / or height of the first filtering region meet preset conditions.

[0383] In some embodiments, the device further includes:

[0384] The filter parameter information generation module is used to traverse each first filter region in the encoded image and determine whether there is a first filter region in the encoded image that needs to be adjusted; if so, it generates filter parameter information of the encoded image based on the adjustment control information corresponding to each first filter region.

[0385] In some embodiments, the filtering parameter information includes at least one of the following:

[0386] Index information of the filtering region;

[0387] Control information for filter adjustment.

[0388] In some embodiments, the control information for filter adjustment includes at least one of the following:

[0389] Control information for the luminance component;

[0390] Control information for chromaticity components;

[0391] Control information across components.

[0392] In some embodiments, the control information includes at least one of the following:

[0393] The first identifier, stored in the image header, is used to indicate whether adaptive loop filtering in the filtered region of the image is enabled for filtering adjustment.

[0394] The second identifier, stored in the strip header, is used to indicate whether adaptive loop filtering in the filtering region of the strip is enabled for filter adjustment.

[0395] The third identifier is stored in the syntax structure corresponding to the filtering region. The third identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether each sub-region of the filtering region is enabled for filtering adjustment.

[0396] The fourth identifier is stored in the syntax structure corresponding to the filtering region. The fourth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate the sub-regions in the filtering region that are enabled for filtering adjustment.

[0397] The fifth identifier is stored in the syntax structure corresponding to the filtering region. The fifth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether all sub-regions of the filtering region are enabled for filtering adjustment.

[0398] In some embodiments, the first filtering region is a coding tree unit.

[0399] The image processing apparatus proposed in this embodiment belongs to the same inventive concept as the image processing method proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in any of the above embodiments. Furthermore, this embodiment has the same beneficial effects as performing the image processing method.

[0400] Figure 9 This is a schematic diagram of another image processing apparatus provided in one embodiment, which is applied to the decoding end, such as... Figure 9 As shown, the device includes: an encoded image acquisition module 410, a second filtering adjustment module 420, and a decoded image generation module 430.

[0401] The encoded image acquisition module 410 is used to acquire an encoded image, decode the encoded image to obtain a second reconstructed image, and filter the second reconstructed image to obtain a third reconstructed value.

[0402] The second filtering adjustment module 420 is used to determine, when the coded image is determined to enable filtering adjustment, a second sub-region of the second filtering region in the second reconstructed image and a second adjacent filtering region corresponding to the second sub-region, and to perform filtering processing on the second sub-region according to the filtering control information corresponding to the second adjacent filtering region to obtain a fourth reconstructed value.

[0403] The decoding image generation module 430 is used to determine the filtering result of the second filtering region in the second reconstructed image based on the third reconstruction value and the fourth reconstruction value of the second sub-region, and to generate a decoding image based on the filtering result of each second filtering region of the second reconstructed image.

[0404] This application provides an image processing method that solves the problem of poor filtering effect. A second reconstructed image is obtained by decoding an encoded image, and a third reconstructed value is obtained by filtering the second reconstructed image. With filtering adjustment enabled in the encoded image, a second sub-region of the second filtering region and its corresponding second adjacent filtering region are determined in the second reconstructed image. The second sub-region of the second filtering region is filtered using the filtering control information of the second adjacent filtering region. Then, the filtering result of the second filtering region is determined based on the third and fourth reconstructed values, achieving local optimization of the second filtering region and obtaining the final filtering result of the second filtering region. Finally, an encoded image is generated based on the filtering result of each second filtering region of the second reconstructed image, improving the image filtering accuracy.

[0405] In some embodiments, determining whether the encoded image has filtering adjustment enabled includes:

[0406] The image header or strip header of the encoded image is parsed to obtain the filtering parameter information;

[0407] Based on the filtering parameter information, determine whether the encoded image should be filtered for adjustment.

[0408] In some embodiments, determining the second sub-region of the second filter region in the second reconstructed image and the second adjacent filter region corresponding to the second sub-region includes:

[0409] The syntax structure of each second filtering region of the encoded image is analyzed to obtain filtering parameter information;

[0410] The second sub-region and the second adjacent filtering region corresponding to the second sub-region are determined based on the filtering parameter information.

[0411] In some embodiments, the filtering parameter information includes at least one of the following:

[0412] Index information of the filtering region;

[0413] Control information for filter adjustment.

[0414] In some embodiments, the control information for filter adjustment includes at least one of the following:

[0415] Control information for the luminance component;

[0416] Control information for chromaticity components;

[0417] Control information across components.

[0418] In some embodiments, the control information includes at least one of the following:

[0419] The first identifier, stored in the image header, is used to indicate whether adaptive loop filtering in the filtered region of the image is enabled for filtering adjustment.

[0420] The second identifier, stored in the strip header, is used to indicate whether adaptive loop filtering in the filtering region of the strip is enabled for filter adjustment.

[0421] The third identifier is stored in the syntax structure corresponding to the filtering region. The third identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether each sub-region of the filtering region is enabled for filtering adjustment.

[0422] The fourth identifier is stored in the syntax structure corresponding to the filtering region. The fourth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate the sub-regions in the filtering region that are enabled for filtering adjustment.

[0423] The fifth identifier is stored in the syntax structure corresponding to the filtering region. The fifth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether all sub-regions of the filtering region are enabled for filtering adjustment.

[0424] In some embodiments, the step of filtering the second sub-region according to the filtering control information corresponding to the second adjacent filtering region to obtain the fourth reconstructed value includes:

[0425] The filter subset used by the second adjacent filter region is determined based on the filter control information of the second adjacent filter region.

[0426] The second sub-region is filtered based on the filter subset used in the second adjacent filter region to obtain the fourth reconstructed value.

[0427] In some embodiments, determining the filtering result of the second filtered region in the second reconstructed image based on the third reconstructed value and the fourth reconstructed value of the second sub-region includes:

[0428] The fourth reconstructed value and the third reconstructed value corresponding to the second sub-region are weighted according to the corresponding weight matrix to obtain the adjusted reconstructed value of the second sub-region.

[0429] The filtering result of the second filtering region in the second reconstructed image is determined based on the adjusted reconstruction value of the second sub-region.

[0430] In some embodiments, the weight matrix includes at least one of the following:

[0431] The elements in the weight matrix corresponding to the fourth reconstructed value and the elements in the weight matrix corresponding to the third reconstructed value of the second sub-region are fixed values;

[0432] The elements of the weight matrix corresponding to the fourth reconstructed value and the elements of the weight matrix corresponding to the third reconstructed value of the second sub-region are determined based on the distance between the pixels in the second sub-region and the boundary.

[0433] In some embodiments, the filter control information includes at least one of the following:

[0434] Adaptive loop filter control information for the luminance component;

[0435] Adaptive loop filter control information for chrominance components;

[0436] Cross-component adaptive loop filter control information.

[0437] In some embodiments, the adaptive loop filter control information includes at least one of the following:

[0438] Whether adaptive loop filtering is enabled in the filtering region or for the components of the filtering region;

[0439] The index of the fixed subset of filters corresponding to the filtering region;

[0440] Index of the adaptive parameter set corresponding to the filtering region;

[0441] The index of the optional subset in the adaptive parameter set corresponding to the filtering region.

[0442] In some embodiments, the second sub-region includes at least one of the following:

[0443] A sub-region of a preset position and size located at the boundary of the second filtering region;

[0444] Sub-regions obtained using a preset partitioning method.

[0445] In some embodiments, the division method includes at least one of the following:

[0446] Classification is performed based on the directionality factor of the luminance pixel blocks in the adaptive loop filter;

[0447] Classification is based on the activity factor of the luminance pixel blocks in the adaptive loop filter;

[0448] Classification is based on the sideband information of adaptive loop filtering;

[0449] Classification is based on the residuals of adaptive loop filtering.

[0450] In some embodiments, the second filtering region is a coding tree unit.

[0451] The image processing apparatus proposed in this embodiment belongs to the same inventive concept as the image processing method proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in any of the above embodiments. Furthermore, this embodiment has the same beneficial effects as performing the image processing method.

[0452] This application also provides a communication node. Figure 10 A schematic diagram of the structure of a communication node is provided as an embodiment, such as... Figure 10 As shown, the communication node provided in this application includes a processor 510, a memory 520, and a computer program stored in the memory and executable on the processor. When the processor 510 executes the program, it implements the image processing method described above.

[0453] The communication node may also include a memory 520; the processor 510 in the communication node may be one or more. Figure 10 Taking a processor 510 as an example; memory 520 is used to store one or more programs; the one or more programs are executed by the one or more processors 510, so that the one or more processors 510 implement the image processing method as described in the embodiments of this application.

[0454] The communication node also includes: a communication device 530, an input device 540, and an output device 550.

[0455] The processor 510, memory 520, communication device 530, input device 540, and output device 550 in the communication node can be connected via a bus or other means. Figure 10 Taking the example of a connection between China and Israel via a bus.

[0456] Input device 540 can be used to receive input digital or character information, and to generate key signal inputs related to user settings and function control of the communication node. Output device 550 may include display devices such as a display screen.

[0457] The communication device 530 may include a receiver and a transmitter. The communication device 530 is configured to perform information transmission and reception communication under the control of the processor 510.

[0458] The memory 520, as a computer-readable storage medium, can be configured to store software programs, computer-executable programs, and modules, such as program instructions / modules corresponding to the image processing method described in the embodiments of this application (e.g., the first reconstructed image acquisition module 310, the second reconstructed value determination module 320, and the first filter adjustment module 330 in an image processing device, or the encoded image acquisition module 410, the second filter adjustment module 420, and the decoded image generation module 430 in an image processing device). The memory 520 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created according to the use of the communication node, etc. In addition, the memory 520 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 520 may further include memory remotely located relative to the processor 510, and these remote memories can be connected to the communication node via a network. Examples of the above-mentioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0459] The communication node in this application embodiment can also be an electronic device, such as a desktop computer, a smart terminal, etc.

[0460] This application also provides a storage medium storing a computer program, which, when executed by a processor, implements any of the image processing methods described in this application.

[0461] Optionally, this image processing method, applied at the encoding end, includes: acquiring a first reconstructed image to be processed; making decisions for each first filtering region of the first reconstructed image; determining filtering control information corresponding to each first filtering region; performing adaptive loop filtering on the first filtering region according to the filtering control information to obtain a first reconstructed value; determining a first sub-region within the first filtering region and a first adjacent filtering region corresponding to the first sub-region; when the first filtering region meets the filtering adjustment conditions, performing filtering processing on the first sub-region according to the filtering control information corresponding to the first adjacent filtering region to obtain a second reconstructed value; performing filtering adjustment based on the first reconstructed value and the second reconstructed value of the first sub-region; determining the filtering result of the first filtering region; and generating an encoded image based on the filtering result of each first filtering region of the first reconstructed image.

[0462] Optionally, this image processing method, applied at the decoding end, includes: acquiring an encoded image; decoding the encoded image to obtain a second reconstructed image; and filtering the second reconstructed image to obtain a third reconstructed value; when it is determined that the encoded image has filter adjustment enabled, determining a second sub-region of the second filter region in the second reconstructed image and a second adjacent filter region corresponding to the second sub-region; filtering the second sub-region according to the filter control information corresponding to the second adjacent filter region to obtain a fourth reconstructed value; determining the filtering result of the second filter region in the second reconstructed image according to the third reconstructed value and the fourth reconstructed value of the second sub-region; and generating a decoded image according to the filtering result of each second filter region in the second reconstructed image.

[0463] The computer storage medium in this application embodiment can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. For example, a computer-readable storage medium can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable CD-ROM, optical storage device, magnetic storage device, or any suitable combination thereof. The computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0464] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit programs for use by or in connection with an instruction execution system, apparatus, or device.

[0465] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, radio frequency (RF), etc., or any suitable combination thereof.

[0466] This application provides a computer program product, which includes a computer program that, when executed by a processor, implements the image processing method described in any one of the embodiments of this application.

[0467] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0468] The above description is merely an exemplary embodiment of this application and is not intended to limit the scope of protection of this application.

[0469] Those skilled in the art will understand that the term user terminal encompasses any suitable type of wireless user equipment, such as mobile phones, portable data processing devices, portable web browsers, or vehicle-mounted mobile stations.

[0470] Generally, the various embodiments of this application can be implemented in hardware or dedicated circuitry, software, logic, or any combination thereof. For example, some aspects can be implemented in hardware, while others can be implemented in firmware or software that can be executed by a controller, microprocessor, or other computing device, although this application is not limited thereto.

[0471] Embodiments of this application can be implemented by executing computer program instructions through the data processor of a mobile device, for example, in a processor entity, or through hardware, or through a combination of software and hardware. The computer program instructions can be assembly instructions, Instruction Set Architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages.

[0472] Any block diagram of logical flow in the accompanying drawings of this application may represent program steps, or may represent interconnected logic circuits, modules, and functions, or may represent a combination of program steps and logic circuits, modules, and functions. The computer program may be stored on memory. Memory may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as, but not limited to, read-only memory (ROM), random access memory (RAM), optical storage devices and systems (Digital Video Disc (DVD) or Compact Disk (CD), etc.). Computer-readable media may include non-transitory storage media. Data processors may be of any type suitable to the local technical environment, such as, but not limited to, general-purpose computers, special-purpose computers, microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), and processors based on multi-core processor architectures.

[0473] A detailed description of exemplary embodiments of this application has been provided above through exemplary and non-limiting examples. However, various modifications and adjustments to the above embodiments will be apparent to those skilled in the art when considered in conjunction with the accompanying drawings and claims, without departing from the scope of this application. Therefore, the proper scope of this application will be determined by the claims.

Claims

1. An image processing method, characterized in that, Applied to the encoding end, including: The first reconstructed image to be processed is obtained. Decisions are made for each first filtering region of the first reconstructed image to determine the filtering control information corresponding to each first filtering region. Adaptive loop filtering is performed on the first filtering region according to the filtering control information to obtain the first reconstructed value. A first sub-region and a first adjacent filter region corresponding to the first sub-region are determined. When the first filter region meets the filter adjustment conditions, the first sub-region is filtered according to the filter control information corresponding to the first adjacent filter region to obtain the second reconstructed value. Based on the first reconstructed value and the second reconstructed value of the first sub-region, a filtering adjustment is performed to determine the filtering result of the first filtering region, and an encoded image is generated based on the filtering result of each first filtering region of the first reconstructed image.

2. The image processing method according to claim 1, characterized in that, The filter control information includes at least one of the following: Adaptive loop filter control information for the luminance component; Adaptive loop filter control information for chrominance components; Cross-component adaptive loop filter control information.

3. The image processing method according to claim 2, characterized in that, The adaptive loop filter control information includes at least one of the following: Whether adaptive loop filtering is enabled in the filtering region or for the components of the filtering region; The index of the fixed subset of filters corresponding to the filtering region; Index of the adaptive parameter set corresponding to the filtering region; The index of the optional subset in the adaptive parameter set corresponding to the filtering region.

4. The image processing method according to claim 1, characterized in that, The adaptive loop filtering of the first filtering region includes at least one of the following methods: Adaptive loop filtering is performed on the luminance components of the first filtering region based on the adaptive loop filtering control information of the luminance components. Based on the adaptive loop filtering control information of the chrominance components and the adaptive loop filtering control information across components, adaptive loop filtering is performed on the chrominance components of the first filtering region.

5. The image processing method according to claim 1, characterized in that, The first sub-region includes at least one of the following: A sub-region of a preset position and size located at the boundary of the first filtering region; Sub-regions obtained using a preset partitioning method.

6. The image processing method according to claim 5, characterized in that, The division method includes at least one of the following: Classification is performed based on the directionality factor of the luminance pixel blocks in the adaptive loop filter; Classification is based on the activity factor of the luminance pixel blocks in the adaptive loop filter; Classification is based on the sideband information of adaptive loop filtering; Classification is based on the residuals of adaptive loop filtering.

7. The image processing method according to claim 1, characterized in that, The step of filtering the first sub-region according to the filtering control information corresponding to the first adjacent filtering region to obtain the second reconstructed value includes: If it is determined that the first adjacent filtering region does not enable adaptive loop filtering based on the filtering control information corresponding to the first adjacent filtering region, the reconstruction value corresponding to the first sub-region in the first reconstructed image is taken as the second reconstruction value. If it is determined that the first adjacent filtering region is enabled by adaptive loop filtering based on the filtering control information corresponding to the first adjacent filtering region, the filtering subset used by the first adjacent filtering region is determined, and the first sub-region is filtered according to the filtering subset to obtain the second reconstructed value.

8. The image processing method according to claim 7, characterized in that, When the number of the first adjacent filtered regions is at least two, the step of filtering the first sub-region according to the filtered subset to obtain the second reconstructed value includes: Select a filter subset from each of the first adjacent filter regions to determine a target filter subset, and then filter the first sub-region according to the target filter subset to obtain a second reconstructed value; or... The first sub-region is filtered according to the filtering subset of each of the first adjacent filtering regions to obtain at least two second reconstructed values.

9. The image processing method according to claim 1, characterized in that, The step of adjusting the filter based on the first reconstructed value and the second reconstructed value of the first sub-region to determine the filtering result of the first filtering region includes: Compare the first reconstructed value with the original image corresponding to the first reconstructed image to determine the first distortion; The second reconstructed value is compared with the original image corresponding to the first reconstructed image to determine the second distortion; If the second distortion is less than the first distortion, the first sub-region is adjusted according to the second reconstruction value; The filtering result of the first filtering region is determined based on the adjusted reconstruction value of the first sub-region.

10. The image processing method according to claim 9, characterized in that, The adjustment of the first sub-region based on the second reconstructed value includes: The second reconstructed value and the first reconstructed value corresponding to the first sub-region are weighted according to the corresponding weight matrix to obtain the adjusted reconstructed value of the first sub-region.

11. The image processing method according to claim 10, characterized in that, The weight matrix includes at least one of the following: The elements in the weight matrix corresponding to the second reconstructed value and the elements in the weight matrix corresponding to the first reconstructed value of the first sub-region are fixed values; The elements of the weight matrix corresponding to the second reconstructed value and the elements of the weight matrix corresponding to the first reconstructed value of the first sub-region are determined based on the distance between the pixels in the first sub-region and the boundary.

12. The image processing method according to claim 9, characterized in that, The step of determining the filtering result of the first filtering region based on the adjusted reconstructed value of the first sub-region includes: Cost analysis is performed on the adjusted reconstruction value of the first sub-region and the first reconstruction value corresponding to the first sub-region. Based on the cost analysis results, the adjusted reconstruction value of the first sub-region or the first reconstruction value corresponding to the first sub-region is taken as the target reconstruction value of the first sub-region. The filtering result of the first filtering region is determined based on the target reconstruction value of the first sub-region.

13. The image processing method according to claim 1, characterized in that, After obtaining the first reconstructed image to be processed, the process also includes: Based on the encoding information of the first reconstructed image, determine whether filtering adjustment is enabled for the first reconstructed image; The encoded information includes at least one of the following: The number of adaptive coefficient sets used in the adaptive loop filtering process of the first reconstructed image; The number of adaptive coefficient sets used for the current strip brightness component corresponding to the first reconstructed image; Temporal hierarchical information of the first reconstructed image; The temporal hierarchical information of the current strip corresponding to the first reconstructed image.

14. The image processing method according to claim 1, characterized in that, The filter adjustment conditions include at least one of the following: The filter control information corresponding to the first filter region is different from the filter control information corresponding to the first adjacent filter region. The width and / or height of the first filtering region meet preset conditions.

15. The image processing method according to claim 1, characterized in that, Also includes: Traverse each first filter region in the encoded image to determine whether there is a first filter region in the encoded image that needs to be adjusted for filtering. If so, the filtering parameter information of the encoded image is generated based on the adjustment control information corresponding to each first filtering region.

16. The image processing method according to claim 15, characterized in that, The filtering parameter information includes at least one of the following: Index information of the filtering region; Control information for filter adjustment.

17. The image processing method according to claim 16, characterized in that, The control information for the filter adjustment includes at least one of the following: Control information for the luminance component; Control information for chromaticity components; Control information across components.

18. The image processing method according to claim 16, characterized in that, The control information includes at least one of the following: The first identifier, stored in the image header, is used to indicate whether adaptive loop filtering in the filtered region of the image is enabled for filtering adjustment. The second identifier, stored in the strip header, is used to indicate whether adaptive loop filtering in the filtering region of the strip is enabled for filter adjustment. The third identifier is stored in the syntax structure corresponding to the filtering region. The third identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether each sub-region of the filtering region is enabled for filtering adjustment. The fourth identifier is stored in the syntax structure corresponding to the filtering region. The fourth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate the sub-regions in the filtering region that are enabled for filtering adjustment. The fifth identifier is stored in the syntax structure corresponding to the filtering region. The fifth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether all sub-regions of the filtering region are enabled for filtering adjustment.

19. The image processing method according to any one of claims 1-18, characterized in that, The first filtering region is a coding tree unit.

20. An image processing method, characterized in that, Applied to the decoding end, including: The encoded image is acquired, the encoded image is decoded to obtain a second reconstructed image, and the second reconstructed image is filtered to obtain a third reconstructed value; When it is determined that the coded image is enabled for filtering adjustment, the second sub-region of the second filtering region in the second reconstructed image and the second adjacent filtering region corresponding to the second sub-region are determined. The second sub-region is filtered according to the filtering control information corresponding to the second adjacent filtering region to obtain the fourth reconstructed value. The filtering result of the second filtering region in the second reconstructed image is determined based on the third reconstructed value and the fourth reconstructed value of the second sub-region, and a decoded image is generated based on the filtering result of each second filtering region in the second reconstructed image.

21. The image processing method according to claim 20, characterized in that, Determining whether filtering adjustment is enabled for the encoded image includes: The image header or strip header of the encoded image is parsed to obtain the filtering parameter information; Based on the filtering parameter information, determine whether the encoded image should be filtered for adjustment.

22. The image processing method according to claim 20, characterized in that, Determining the second sub-region of the second filtering region in the second reconstructed image and the second adjacent filtering region corresponding to the second sub-region includes: The syntax structure of each second filtering region of the encoded image is analyzed to obtain filtering parameter information; The second sub-region and the second adjacent filtering region corresponding to the second sub-region are determined based on the filtering parameter information.

23. The image processing method according to claim 21 or 22, characterized in that, The filtering parameter information includes at least one of the following: Index information of the filtering region; Control information for filter adjustment.

24. The image processing method according to claim 23, characterized in that, The control information for the filter adjustment includes at least one of the following: Control information for the luminance component; Control information for chromaticity components; Control information across components.

25. The image processing method according to claim 24, characterized in that, The control information includes at least one of the following: The first identifier, stored in the image header, is used to indicate whether adaptive loop filtering in the filtered region of the image is enabled for filtering adjustment. The second identifier, stored in the strip header, is used to indicate whether adaptive loop filtering in the filtering region of the strip is enabled for filter adjustment. The third identifier is stored in the syntax structure corresponding to the filtering region. The third identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether each sub-region of the filtering region is enabled for filtering adjustment. The fourth identifier is stored in the syntax structure corresponding to the filtering region. The fourth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate the sub-regions in the filtering region that are enabled for filtering adjustment. The fifth identifier is stored in the syntax structure corresponding to the filtering region. The fifth identifier is used to indicate whether the adaptive loop filtering of the filtering region is enabled for filtering adjustment, and to indicate whether all sub-regions of the filtering region are enabled for filtering adjustment.

26. The image processing method according to claim 20, characterized in that, The step of filtering the second sub-region according to the filtering control information corresponding to the second adjacent filtering region to obtain the fourth reconstructed value includes: The filter subset used by the second adjacent filter region is determined based on the filter control information of the second adjacent filter region. The second sub-region is filtered based on the filter subset used in the second adjacent filter region to obtain the fourth reconstructed value.

27. The image processing method according to claim 20, characterized in that, The step of determining the filtering result of the second filtering region in the second reconstructed image based on the third reconstructed value and the fourth reconstructed value of the second sub-region includes: The fourth reconstructed value and the third reconstructed value corresponding to the second sub-region are weighted according to the corresponding weight matrix to obtain the adjusted reconstructed value of the second sub-region. The filtering result of the second filtering region in the second reconstructed image is determined based on the adjusted reconstruction value of the second sub-region.

28. The image processing method according to claim 27, characterized in that, The weight matrix includes at least one of the following: The elements in the weight matrix corresponding to the fourth reconstructed value and the elements in the weight matrix corresponding to the third reconstructed value of the second sub-region are fixed values; The elements of the weight matrix corresponding to the fourth reconstructed value and the elements of the weight matrix corresponding to the third reconstructed value of the second sub-region are determined based on the distance between the pixels in the second sub-region and the boundary.

29. The image processing method according to claim 20, characterized in that, The filter control information includes at least one of the following: Adaptive loop filter control information for the luminance component; Adaptive loop filter control information for chrominance components; Cross-component adaptive loop filter control information.

30. The image processing method according to claim 29, characterized in that, The adaptive loop filter control information includes at least one of the following: Whether adaptive loop filtering is enabled in the filtering region or for the components of the filtering region; The index of the fixed subset of filters corresponding to the filtering region; Index of the adaptive parameter set corresponding to the filtering region; The index of the optional subset in the adaptive parameter set corresponding to the filtering region.

31. The image processing method according to claim 20, characterized in that, The second sub-region includes at least one of the following: A sub-region of a preset position and size located at the boundary of the second filtering region; Sub-regions obtained using a preset partitioning method.

32. The image processing method according to claim 31, characterized in that, The division method includes at least one of the following: Classification is performed based on the directionality factor of the luminance pixel blocks in the adaptive loop filter; Classification is based on the activity factor of the luminance pixel blocks in the adaptive loop filter; Classification is based on the sideband information of adaptive loop filtering; Classification is based on the residuals of adaptive loop filtering.

33. The image processing method according to claim 20, characterized in that, The second filtering region is a coding tree unit.

34. A communication node, characterized in that, include: The image processing method comprises a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for establishing communication between the processor and the memory, wherein the program, when executed by the processor, implements the steps of the image processing method as described in any one of claims 1-33.

35. A storage medium for computer-readable storage, characterized in that, The storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the image processing method according to any one of claims 1-33.

36. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the image processing method according to any one of claims 1-33.

Images