Video Surveillance System Image Clarity Research

Video images are the main carrier for people to obtain information,so the quality of images is getting more and more attention.There are many reasons why the video image is unclear,and some can be avoided,such as poor focus adjustment,dirt on the lens,and so on.But the most common cause is mainly the weather,which causes the image itself to deteriorate.To this end,this paper focuses on the clear development of video images in foggy days,low light and rainy environments,and the overall effect is good.The main research work and results of the thesis are as follows:(1)For the smog environment,the histogram equalization and Retinex theory algorithm are used to clear the fog haze image,and the Retinex theory algorithm uses single scale(SSR)and multi-scale(MSR)algorithms.The experimental results show that the multi-scale algorithm has the best processing effect,the image details are prominent,clear and identifiable,the overall color is relatively natural,and the visual effect is good.(2)For the weak illumination environment,this paper first analyzes the characteristics of the image,and then adopts several algorithms in the foggy environment.Based on this,a multi-scale Retinex theoretical algorithm with color restoration and histogram equalization are proposed.The algorithm combined with each other accounts for 1/2,which makes the processed image effects retain their respective advantages,and the effect is good,which has certain engineering application value.(3)For the rainy environment,the characteristics of raindrops are analyzed firstly.Then,based on the deficiency of the inter-frame difference method,a background difference method based on single Gaussian model is proposed for rain removal,and the influence of learning rate on raindrops is explored.The experimental results show that the image processing effect is the best when the learning rate is 0.6,and it has certain robustness,and the algorithm can be applied to the medium rain,heavy rain and snowy environment at the same time.Finally,based on MATLAB,a video surveillance image clarity processing system is constructed.The system can realize the clearing of video surveillance images in foggy days,weak illumination and rainy environments.Different algorithms are intuitive and have certain engineering application value.