Study On The Method Of Denoising Low SNR Video

With the application and popularization of photoelectric imaging system,people put forward higher requirements,they are gradually applied to the worse environment,such as night,underground,tunnel,etc,under these circumstances due to insufficient light,uneven illumination,sensor sensitivity limit,tiny particles of light refraction,the acquisition of the video showed the following characteristics:small gathering information,noise intensity distribution is wide.This greatly increases the remove noise,restore the difficulty of the video information.Therefore,the significance of low SNR noise suppression technique of digital video sequence is significant.Aiming at the low SNR video with high-intensity noise,according to the video sequence image information of each frame correlation in the time domain,and the noise has no correlation in the time domain,this paper proposed a method of temporal and spatial joint video denoising based motion estimation and compensation of gray.The algorithm will combine background subtraction and coterminous frame subtraction method up as motion estimation method,which can effectively extract the prospect information in the case of high intensity noise interference,and non-local means prospect of denoising,background partly for the mean of time domain denoising.With the defects of reducing about constract by mean of time domain filtering,this paper proposed a method for compensating gray denoising results after treatment.We finally simulation experiment on the algorithm proposed in this paper.The experiment results show that:add high-intensity white Gaussian noise for a low SNR video,this method can not only effectively suppress noise and significantly improve the video signal to noise ratio,but also the details are effectively retained and restored,and the background frame in the process of background subtraction directly as a result of the mean time domain,which greatly makes the algorithm can be further optimized.