Clear Processing Research On Image Of The Snowing Gets

The mostly existing video surveillance systems work in clear weather conditions, but in fact the monitoring system is not always in fine condition, with the video monitoring technology is widely used in various fields, which requires the system to be able to be applied in various occasions. Snowing is a common natural phenomenon in the winter of northern, and also one of the main factors that affect image quality of the outdoor video monitoring system. In order to the control system can work well under the environment of snow, this article shows that the treatment clear research to make the image that takes in snow condition, for the image getting in the snow sharpens process, to optimize the effect of image.This article based on the snow pollution point of the video images as the research object, the image that access under the snow conditions as process research. Firstly, according to sampling requirements of research, the images of diseased samples are collected. No matter which kinds of equipments are used, the collected images often are not satisfying. In order to improve the image quality, the noise in the image needs to be removed by the method of median filter, image enhancement, for follow-up studies provide a clear image. Secondly, get thought the characteristics of snow video images are researched carefully and deeply. Then takes two kinds of effective detection and removal of snow video method, one kind is, according to the traditional frame difference method the basic idea and the combination of snow image time-domain characteristics, puts forward an improved frame difference method; the other is, modified K-means clustering algorithm, the algorithm based on image brightness changes, for every pixel in the image brightness change clustering estimation, so as to obtain the background luminance myopia estimates, as snowflakes brightness value is much larger than the background luminance values, with the background brightness value to replace snow brightness value idea to remove snow orders. Last, it make image clarity recovery by based on Contourlet domain wiener filtering method. In order to gets image that closer to the actual image.Based on the theoretical research, designing and implementing the snow gets a processing system of the clear image through the Visual C++6.0platform,making the mage clear with the collection of video, the results show that, this study can obtain effective treatment to the images, in order to get the purpose of optimizing the image information.