Research On Low-light Image Enhancement Processing Technology Based On Retina And Cerebral Cortex Theory

Due to the uneven illumination distribution or lack of light source,the image is seriously degraded in brightness,contrast and detail performance.These degraded images make it difficult for people to obtain effective information,thus affecting the observation effect.Therefore,in the night environment,it is inevitable to obtain low-illumination images with poor image quality in the fields of monitoring system,intelligent transportation and all-weather operations.In order to solve this problem,this paper studies the enhancement technology of low-light images.This paper proposed an improved image enhancement algorithm based on the theory of the Retinex,combined with an embedded development platform to enhance processing of low-light images or videos.This paper is divided into the following four chapters: the first chapter studies the background,significance and research status of low-light images enhancement technology,and summarizes the main work of this paper;In the second chapter,we build a low-light image enhancement system based on DE1?SoC,selected an ordinary USB camera as the front-end acquisition device,DE1?SoC as a mid-range processing device,and VGA monitor as a backend display device to display the enhanced image;The third chapter mainly studies the improvement methods of low-light image enhancement algorithms,a weighted guided image filtering algorithm based on Retinex and YUV color space is proposed to overcome the shortcomings of existing algorithms.The improved algorithm can effectively reduce the probability of the occurrence of halo artifacts and have better edge retention effect.According to the proposed improved algorithm,the image under low illumination environment was simulated and the experimental results were compared and evaluated;Chapter 4 introduces the method of implementing the proposed low-light image enhancement algorithm using the DE1?SoC,including the design of embedded systems,the construction of Linux development environment,and the transplantation of enhanced algorithms.We have implemented multiple experiments of the system,and the results can prove the system's high robustness.