Video image motion vector encoding and decoding method and encoder/decoder

Abstract

The invention provides a motion vector encoding and decoding method of a video image, an encoder and a decoder. The motion vector encoding method of the video image comprises the following steps of: selecting the motion vector precision with the minimum rate distortion cost as the first motion vector precision; carrying out coding condition judgment on the motion vector difference value; adjustingthe first motion vector precision in a first motion vector precision set range corresponding to the first motion vector precision according to a judgment result; and determining second motion vectorprecision, encoding the motion vector difference according to the second motion vector precision to obtain motion vector information, writing the motion vector information and an identifier corresponding to the first motion vector precision into a code stream, and sending the code stream to a decoding end. Through the scheme, the precision of the motion vector needing to be identified can be reduced, and the code rate and the coding time required for coding the motion vector are reduced while the prediction performance is maintained, so that the coding efficiency is improved, and the coding gain is brought.

Description

technical field [0001] The present application relates to the field of digital signal processing, and in particular to a motion vector encoding and decoding method for video images and an encoding / decoding device. Background technique [0002] Predictive coding is a basic coding tool in video compression, and common predictive coding is inter-frame prediction and intra-frame prediction. Inter-frame prediction refers to the use of video time domain correlation to predict the pixels of the current image using the previously encoded image pixels to achieve the purpose of effectively removing video time domain redundancy. The basic principle is: based on the video sequence usually has a strong Time-domain correlation, for each current encoding block of the current image, find a best matching reference block in the previous encoded image, and use the pixel value of the pixel in the reference block as the pixel value of the pixel in the current encoding block prediction value, an...

AI Technical Summary

Problems solved by technology

On the one hand, the multiple adaptive motion vector accuracy has high complexity, on the other hand, some motion vector predictions are inefficient and bring extra bit costs, which makes the proportion of motion vector information in the code stream constantly rising. Therefore, how to effectively Compressing motion vector information is the key to further improving video compression efficiency

Images