Self-adaptive bottom video mining method based on contrast resolution compensation

Abstract

The invention provides a self-adaptive bottom video mining method based on contrast resolution compensation, comprising the steps of obtaining the original average gray of a video frame; obtaining the best compensation factor of the contrast resolution of the video frame according to the average gray press-down type; compensating the contrast resolution of the video frame by using the best compensation factor; displaying and memorizing the compensated video stream, thus completing the bottom video mining. In the invention, the self-adaptive bottom video mining on the video photographed under low luminance can be realized by the vision contrast resolution compensation method. The essential of the method is as follows: the contrast is replaced by sacrificing the information quantity that can not be sensed by human vision so as to mine the main information that is arranged in photos and can be sensed by human vision conveniently; and the method in the invention can be used for realizing the machine vision for bottom video mining.

Description

technical field [0001] The invention belongs to digital video processing technology, in particular, relates to a mining processing method for video obtained under low illumination and difficult for human vision to distinguish. Background technique [0002] Human vision is divided into scotopic vision under low illumination, photopic vision under normal lighting, and intermediate vision in between. Videos obtained in low-light environments (such as nighttime or other special environments) are difficult for human vision to distinguish due to low contrast. Due to the limitation of the contrast resolution of human vision, human vision cannot distinguish the target images taken under dark vision (low light) conditions, which brings great difficulties to monitoring, criminal investigation, detection and other work in low-light environments . Under scotopic conditions, the contrast resolution of human vision is very low. If it can be compensated, the contrast resolution of human ...

AI Technical Summary

Problems solved by technology

Videos obtained in low-light environments (such as nighttime or other special environments) are difficult for human vision to distinguish due to low contrast

Due to the limitation of the contrast resolution of human vision, human vision cannot distinguish the target images taken under dark vision (low light) conditions, which brings great difficulties to monitoring, criminal investigation, detection and other work in low-light environments

Under scotopic conditions, the contrast resolution of human vision is very low. If it can be compensated, the contrast resolution of human beings under scotopic conditions can be greatly improved. Visually indistinguishable video images become clearly visible video images, just like cats, canine mammals, and birds such as nightingales have night vision functions. This technology is called bottom-level video mining, and there is no bottom-level video mining at present. technology related reports

[0003] Although the human sensory system has a good adaptive adjustment function, there are still some defects, and the visual system is no exception. Human visual contrast resolution limitation is one of many human visual defects

Images