Detection And Removal Of Noise In Video Surveillance

Detection and removal of image noise in video surveillance is an important aspect in image processing. For the purpose of detection of video noise is to avoid the noise’s impaction on the video images, to ensure the normal operation of the monitoring system. Video equipment is running in various environments, and the operation will be subject to environmental impact. For example, rainwater may will enter the camera’s intern; the winds can cause monitoring equipment or shaking. various environmental effects, will lead to video noise, such as little noise, snowflake, stripes and so on when the video surveillance system is subjected to the interference of external environment, resulting in noise appears in the video surveillance, it will lead to customers like blurred, and even to lose important information. When the video monitoring system is under the interference of the external environment, it can cause the emergence of the noise in video surveillance video, and make the images blurred and lose the important information. The detection and removal of noise is mainly used the general method, which can lead to the detection of the effect is not very good. These method also can fail to detect some videos which have many small noise but also can cause the loss of much more information. Noise detection method proposed in this paper is from the point of view of classification noise. We do the classification to the noise, and then test various noise. In this paper, the noise is divided into three categories, respectively, little noise, snow, stripe. This paper uses the TV model value detection method to detect the little noises; the snow noise’s detection is by using the connected domain based on shape’s method; and the stripe noise detection method is an improved Fu Li ye transform method. In this paper, through the classification and detection, we can achieve good results on the detection of noise on video. At the same time we can also get the good result we want on the video detection in noise’s removal.