Low Quality Camera Color Cast And Noise Processing

With the development of the times,the network become more and more popular,and many cheap network camera are used in our daily life and work.In general,the low quality of the video camera creates partial color and noise problem in the video,which to a certain extent,the impact of the video viewing.So in this paper,we mainly deal with the phenomenon of partial color and noise problem in the video which is collected by the low quality camera.The main work of this paper is as follows.For video of partial color phenomena.In this paper,we use the temporal correlation of the vedio to dvide the input video into many units;then,for each video unit,using the equivalent circle ideas to detect whether there is partial color phenomena;finally,for the video with partial color phenomena,we combine the gray world theory and the Retinex theory to create a nonlinear partial color correction model.And the experiment of real data verifies the validity of our algorithm.Aiming at the problem of noise in the video.This paper according to the light conditions of the videos divide the noise into several grades;then,according to the different nature of the noise,we use the median filtering algorithm and noise reduction algorithm based on multi order wavelet transform to processing noise problem.And the experiment of real data verifies the validity of the algorithm.Based on the above algorithm research,a set of software for the color cast and noise processing of the video that collected by the low quality camera is developed,which can be processed in real time to correct the partial color phenomena and noise reduction,and then improve the quality of video.