Study And Implementation On Defog Method For The Camears And Surveillance System

As the leading edge of computer vision engineering field, video supervision is playing a more important part of security system. With the development of networks, image processing and automatic control, intelligent supervision system will definitely meet its application prospect, and it will be the development trend of video surveillance industry.Surveillance system equipped with defogging module has become the development of mainstream. But in the process of application research, the requirement of real time is the key point, and countries around the world are all carrying out research to improve the defogging performance of surveillance system.With the independently developed video surveillance platform, the research of image defogging is implemented, and the main contents of the research include the following aspects:1. Research and design of defogging algorithm:by studying the comparative analysis of Dark channel prior theory, histogram equalization and defogging method based on the depth of field under the background of Atmospheric scattering physical model, defogging methods and methods selectionstrategywhich is suitable for video monitoring system is provided.2. Design and analysis of defogging module in the surveillance system:through the analysis of video stream processing, a real-time and high applicable system design schema is provided by utilizing advanced NVR of the current mainstream.Moreover, the image quality evaluation is also carried on to demonstrate the feasibility and rationality, average defogging time of video frames with different dimensions is also tested to verify engineering practicability of defogging methods. Through the test above, the function and performance of the defogging algorithm and system design schema is confirmed.