Image processing method, processing device and storage medium
By optimizing the inter-frame prediction of image blocks and adjusting the order of the candidate table and index set, the problem of high signaling overhead caused by adding inter-frame prediction tools independently was solved, thus improving the efficiency of video encoding and decoding.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- SHENZHEN TRANSSION HLDG CO LTD
- Filing Date
- 2026-04-17
- Publication Date
- 2026-07-30
AI Technical Summary
In high-efficiency video coding standard protocols, the independent addition of multiple coding tools during inter-frame prediction leads to high signaling overhead, increasing costs and reducing efficiency.
By performing inter-frame prediction on image patches based on the first parameter and/or the second parameter, the order of the candidate table and index set is optimized, avoiding repeated access to unnecessary inter-frame prediction tools and reducing signaling overhead.
It improves the efficiency of inter-frame prediction, supporting the enhancement of video encoding and decoding efficiency.
Smart Images
Figure CN2026091405_30072026_PF_FP_ABST
Abstract
Claims
1. An image processing method, wherein, Including the following steps: S1, perform inter-frame prediction of image patches based on the first parameter and / or the second parameter.
2. The image processing method of claim 1, wherein, It also includes at least one of the following: The first parameter is the candidate table set and / or the candidate tree structure; Adjust the priority and / or order of at least one element in the first parameter based on at least one third parameter; The second parameter is the index set.
3. The image processing method of claim 2, wherein, The candidate set includes at least one of the following: a first candidate set, a multi-level candidate set, and a flattened candidate set; and / or, at least one of the following: At least one candidate table in the multi-layer candidate table set includes at least one of the following: tool type of inter-frame prediction tool, inter-frame prediction tool, inter-frame prediction mode, inter-frame prediction mode parameters, motion vector-based prediction mode, tool type of intra-frame prediction tool, intra-frame prediction tool, intra-frame prediction mode, and intra-frame prediction mode parameters. At least one candidate table in the multi-layer candidate table set is at least one of the following: a second candidate table containing at least one inter-frame prediction tool tool type, a third candidate table containing at least one inter-frame prediction tool, a fourth candidate table containing at least one inter-frame prediction mode, a fifth candidate table containing at least one inter-frame prediction mode parameter, a sixth candidate table containing at least one motion vector-based prediction mode, a seventh candidate table containing at least two inter-frame prediction tools of the same type, an eighth candidate table containing at least two inter-frame prediction modes of the same type, a ninth candidate table containing at least one intra-frame prediction tool tool type, a tenth candidate table containing at least one intra-frame prediction tool, an eleventh candidate table containing at least one intra-frame prediction mode, and a twelfth candidate table containing at least one intra-frame prediction mode parameter; At least one candidate table in the multi-level candidate table set is a candidate table whose priority and / or sorting order are adjusted based on at least one third parameter; At least one candidate table in the multi-level candidate table set is a candidate table that has not undergone priority and / or sorting order adjustment; At least one candidate table in the multi-level candidate table set is a pre-defined candidate table; In a multi-level candidate table set, at least one first candidate table element of the current candidate table has a strong correlation with the next candidate table; In a multi-level candidate table set, at least one second candidate table element in the current candidate table has a non-strongly correlated correspondence with the next candidate table. The priority and / or sorting order of at least one element in the flattened candidate list and / or the first candidate list are adjusted according to the third parameter; At least one element in the flattened candidate list and / or the first candidate list has not been adjusted in priority and / or sorting order; At least one element in the flattened candidate table and / or the first candidate table includes at least one of the following: tool type of inter-frame prediction tool, inter-frame prediction tool, inter-frame prediction mode, inter-frame prediction mode parameters, motion vector-based prediction mode, tool type of intra-frame prediction tool, intra-frame prediction tool, intra-frame prediction mode, and intra-frame prediction mode parameters. The first parameter is the set of most likely candidate tables that includes at least one most likely candidate table; The first parameter is a set of non-most likely candidate tables that includes at least one non-most likely candidate table; The first parameter is a set of non-most likely candidate tables that includes at least one most likely candidate table and a non-most likely candidate table; The first parameter is the most likely candidate tree structure and / or the non-most likely candidate tree structure; The index set includes at least one of the following: a first index representing the tool type of the inter-frame prediction tool, a second index representing the inter-frame prediction tool, a third index representing the inter-frame prediction mode, a fourth index representing the parameters of the inter-frame prediction mode, a fifth index representing the motion vector-based prediction mode, a sixth index representing at least one intermediate element, a seventh index representing at least one candidate table, an eighth index representing the tool type of the intra-frame prediction tool, a ninth index representing the intra-frame prediction tool, a tenth index representing the intra-frame prediction mode, and an eleventh index representing the parameters of the intra-frame prediction mode.
4. The image processing method as described in claim 3, wherein, It also includes at least one of the following: The most likely candidate table is at least one of the following: the first candidate table and / or the flattened candidate table after priority and / or order adjustment based on at least one third parameter, and at least one candidate table after priority and / or order adjustment based on at least one third parameter in the multi-level candidate table set; The non-most likely candidate table is at least one of the following: the first candidate table and / or the flattened candidate table that have not been adjusted in priority and / or order, and at least one candidate table in the multi-level candidate table set that has not been adjusted in priority and / or order. The architecture and / or branching structure in the most likely candidate tree structure is the architecture and / or branching structure after priority and / or ordering adjustment based on at least one third parameter; Some or all of the architectures in the non-most likely candidate tree structure have not been prioritized and / or ordered. Some or all of the branch structures in the non-most likely candidate tree structure have not been prioritized and / or ordered.
5. The image processing method as described in claim 2, wherein, Candidate tree structures include at least one of the following: An architecture comprising at least one of the following: tool type of at least one inter-frame prediction tool, inter-frame prediction tool, inter-frame prediction mode, inter-frame prediction mode parameters, motion vector-based prediction mode, tool type of intra-frame prediction tool, intra-frame prediction tool, intra-frame prediction mode, and intra-frame prediction mode parameters. At least one layer of architecture, and at least one first node and / or first element of the at least one layer of architecture has a branching structure; At least one layer of architecture, and at least one second node and / or second element of the at least one layer of architecture has a non-branching structure; At least one layer of architecture and / or at least one branch structure that adjusts priority and / or order of arrangement based on at least one third parameter; The structure comprises at least one branch structure including at least one of the following: tool type of at least one inter-frame prediction tool, inter-frame prediction tool, inter-frame prediction mode, inter-frame prediction mode parameters, motion vector-based prediction mode, tool type of intra-frame prediction tool, intra-frame prediction tool, intra-frame prediction mode, and intra-frame prediction mode parameters. At least one layer of architecture and / or at least one layer of branching structure, wherein at least one layer of architecture and / or at least one layer of branching structure includes at least one of the following: a thirteenth candidate table containing at least one inter-frame prediction tool, a fourteenth candidate table containing at least one inter-frame prediction tool, a fifteenth candidate table containing at least one inter-frame prediction mode, a sixteenth candidate table containing at least one inter-frame prediction mode parameter, a seventeenth candidate table containing at least one motion vector-based prediction mode, an eighteenth candidate table containing at least two inter-frame prediction tools of the same type, a nineteenth candidate table containing at least two inter-frame prediction modes of the same type, a twentieth candidate table containing at least one intra-frame prediction tool, a twenty-first candidate table containing at least one intra-frame prediction tool, a twenty-second candidate table containing at least one intra-frame prediction mode, and a twenty-third candidate table containing at least one intra-frame prediction mode parameter.
6. The image processing method as described in claim 2, wherein, The third parameter is determined or obtained based on at least one of the following: The inter-frame prediction tool and / or inter-frame prediction mode used in at least one of the following: left neighbor block, top neighbor block, top left neighbor block, top right neighbor block, non-neighbor block, cross component block, co-position block, temporal block, default block, and candidate block; The cost of at least one inter-frame prediction tool and / or inter-frame prediction mode; Size parameter of at least one of the following: image block, left neighbor block, top neighbor block, top left neighbor block, top right neighbor block, non-neighbor block, cross-component block, co-location block, temporal block, default block, and candidate block; The motion vector distribution of at least one of the following: left neighbor block, top neighbor block, top left neighbor block, top right neighbor block, non-neighbor block, cross-component block, co-position block, temporal block, default block, and candidate block; The motion vector height of at least one of the following: left neighbor block, top neighbor block, top left neighbor block, top right neighbor block, non-neighbor block, cross component block, co-position block, temporal block, default block, and candidate block; The motion trend of at least one of the following: left neighbor block, top neighbor block, top left neighbor block, top right neighbor block, non-neighbor block, cross component block, co-position block, temporal block, default block, and candidate block; Boundary features of at least one of the following: left neighbor block, top neighbor block, top left neighbor block, top right neighbor block, non-neighbor block, cross-component block, co-position block, temporal block, default block, and candidate block; Intra-frame prediction results of at least one of the following: left neighbor block, top neighbor block, top left neighbor block, top right neighbor block, non-neighbor block, cross-component block, co-position block, temporal block, default block, and candidate block.
7. The image processing method as described in claim 2, wherein, It also includes at least one of the following: Inter-frame prediction tools include at least one of the following: merge mode with motion vector difference, geometric partition mode, affine motion compensation, advanced temporal motion vector prediction, joint inter-frame and intra-frame prediction, sub-block transform mode, overlapping block motion compensation, local illumination compensation, coding unit-level bidirectional prediction weighting, and merge mode. Step S1 includes: determining or obtaining a fourth parameter based on the first parameter and / or the second parameter, and performing inter-frame prediction on the image patch based on the fourth parameter.
8. The image processing method as described in claim 7, wherein, The fourth parameter includes at least one of the following: Tool types for inter-frame prediction tools; Inter-frame prediction tools; Inter-frame prediction mode; Inter-frame prediction mode parameters; Prediction patterns based on motion vectors; Intra-frame prediction tool types; Intra-frame prediction tools; Intra-frame prediction mode; Intra-frame prediction mode parameters; At least one of the candidate lists from the first to the twenty-third candidate list; The first indication information represents at least one item in the first to twenty-third candidate lists; A second indication information characterizing at least one of the following: tool type of inter-frame prediction tool, inter-frame prediction tool, inter-frame prediction mode, inter-frame prediction mode parameters, motion vector-based prediction mode, tool type of intra-frame prediction tool, intra-frame prediction tool, intra-frame prediction mode, and intra-frame prediction mode parameters. At least one index in the index set.
9. The image processing method as described in claim 1, wherein, Also includes: The first parameter is processed based on the image patch's construction factor.
10. The image processing method as described in claim 9, wherein, It also includes at least one of the following: The build factor is determined or obtained based on at least one of the size parameters of image blocks, left neighbor blocks, top neighbor blocks, top left neighbor blocks, top right neighbor blocks, non-neighbor blocks, cross-component blocks, co-position blocks, temporal blocks, default blocks, and candidate blocks, the cost of at least one inter-frame prediction tool, and at least one of the syntax elements obtained from the bitstream. The number and / or position of at least one element in the candidate table set or candidate tree structure are determined or obtained based on at least one construction factor; The candidate table set and / or candidate tree structure are truncated, and / or adjusted, and / or updated based on at least one construction factor.
11. A processing apparatus, wherein, include: The system includes a memory and a processor, wherein the memory stores an image processing program, and when the image processing program is executed by the processor, it implements the steps of the image processing method as described in claim 1.
12. A storage medium, wherein, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the image processing method as described in claim 1.