A video image processing method and device

Abstract

The invention provides a video image processing method and device. The method comprises the following steps: A, partitioning a video into a plurality of scenes according to a pixel gray value distribution characteristic of a video image frame; B, adaptively selecting a filter function corresponding to each scene according to the whole pixel gray value distribution characteristic of the scene; C, building a mapping lookup table between an original image pixel value and a filtered pixel value in each scene according to the selected filter function; and D, performing the same filter operation on all pixels in each frame of image in each scene by using the mapping lookup table corresponding to the scene. The method and the device can achieve a relatively good enhancement effect on the scenes and keep the filter consistency of the scenes and the continuity of watching quality. Therefore, the method and the device can better improve the quality of various video images, including aspects such as video image definition, color richness and watching comfort level.

Description

technical field [0001] The present invention relates to the technical field of video images, in particular to a video image processing method and device. Background technique [0002] With the massive popularity of personal smartphones, video cameras, cameras, etc., more and more users are starting to take videos and images by themselves. The popularity of the Internet has enabled user-generated (UGC) videos and images to be easily shared with more users. [0003] The entire process of video images from source acquisition to terminal display involves acquisition, (pre-processing), encoding, transmission, decoding, (post-processing), rendering and display and other multimedia processing operations. There is a certain correlation between these process operations, which will affect the viewing experience of end users. [0004] However, due to the variety and complexity of video image acquisition equipment and shooting environments, the processing capabilities and strategies o...

AI Technical Summary

Problems solved by technology

But only a single contrast stretch may be even worse for scenes where the overall video image is too dark or too bright

The histogram equalization method is also easy to cause excessive adjustment, and it is image / frame related, so the processing of each image frame is independent, which is not suitable for video, because the parameters of each frame in the scene may not be uniform , it is easy to cause severe fluctuations in video quality, incoherent subjective quality, and affect the viewing effect; and it will reduce the temporal correlation between adjacent image frames in the same scene, reducing video coding performance, (if the same subjective quality is maintained) may be greatly reduced Increase the bit rate required for video encoding

Images