Self-adaptive CU splitting decision-making method based on deep learning and multi-feature fusion

Abstract

The invention provides an adaptive CU splitting decision method based on deep learning and multi-feature fusion. The method comprises the steps: firstly, calculating the texture complexity SD of a current CU through standard deviation, building a threshold model through a quantization parameter function and a depth function, and dividing the current CU into a complex CU and a uniform CU; secondly,if the complex CU belongs to the edge CU, judging whether the complex CU is split or not by utilizing a CNN structure based on multi-feature fusion; otherwise, judging whether the complex CU is splitor not by utilizing the self-adaptive CNN structure. According to the method, deep learning and multi-feature fusion are combined, and the problem of coding complexity is solved. The CNN structure based on multi-feature fusion and the CNN structure based on self-adaptation can successfully process the training samples, calculation of rate distortion RDO of all CU and complex CU is avoided, and therefore the calculation complexity is reduced, and the coding time is saved.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an adaptive CU split decision method based on deep learning and multi-feature fusion. Background technique [0002] With higher demands on video compression, it becomes more important to develop more efficient video coding standards. JVET has developed the next generation video coding standard - H.266 / VVC. The H.266 / VVC Test Model (VTM) implements many novel techniques that can significantly improve coding efficiency. H.266 / VVC uses a quadtree-nested multi-type tree (QTMT) coding block architecture for block partitioning, which shows better coding performance, but results in enormous computational complexity, possibly 5 times that of HEVC , and H.266 / VVC also contains 67 intra-frame prediction modes for intra-frame prediction. Among them, the planar mode and DC mode are the same as H.265 / HEVC. The prediction mode becomes denser, so more accurate prediction can ...

AI Technical Summary

Problems solved by technology

[0007] In view of the deficiencies in the above-mentioned background technology, the present invention proposes an adaptive CU split decision method based on deep learning and multi-feature fusion, which combines deep learning and multi-feature fusion to solve the technical problem of coding complexity

Images