Research And Application Of Video Image Enhancement Algorithm For Closed-Circuit Television Monitoring System Of Rail Transit

Nowadays,the closed-circuit television(CCTV)monitoring system is widely used in the rail transit,including the rail obstacle detection,the appearance inspection of the rail equipment,and the important information extraction,etc.However,the captured images are generally low contrast.Enhancement technology for improving image quality has become a research hotspot.In this paper,the advantages and disadvantages of several common image enhancement techniques are discussed and compared.Then an improved algorithm for CCTV monitoring system is proposed.This paper includes the following three aspects:1.When dark channel prior fog removal algorithm is applied to the rail traffic image,the algorithm will generate the error in estimating atmospheric light value process.As such,the thesis proposes the improved dark channel prior fog removal algorithm,whose key idea is to take the weighted average value of the maximum gray value and the average gray value as the threshold to get the estimation of atmospheric light.The experimental results show that the algorithm achieves the better performance.2.On the one hand,high and low frequency separation method is used to adjust the brightness of the low illuminative image.The adaptive gamma correction method is used for the low frequency part.High frequency part adopts Fourier transform frequency domain enhancement method instead of the filter methods.At last,the high and low frequency are fused.If the values of the original image are smaller than that of the low frequency component,then them will be set as zero in the previous algorithm.However,the proposed method preserves the missing signal to obtain the complete detail information.On the other hand,we choose the Global Histogram Equalization(GHE)to address the shortcomings of the Multi-scale Retinex(MSR)algorithm after analyzing and comparing various spatial enhancement algorithms.The optimal weighted average value is selected according to a large number of experiments,which improves the information entropy,contrast,clarity,and average brightness of the image.3.For the proposed fog removal algorithm,the brightness adjustment algorithm,and the contrast enhancement algorithm,the software platform applied to the CCTV monitoring system is developed independently,and the enhancement function of the image and video isrealized.