Study On Image Clearness Technique In Bad Weather

With the development of the video surveillance system techniques, the demands of system's reliability and stability have raised. In order to make video surveillance systems operate normally in bad weather, image clearness technique is proposed in the paper which aims to enhance weak information by software methods.The paper studies on the image clearness technique in fog. Firstly, the classification of fog is introduced. Then the characteristic of atmospheric scattering models and the difficulty caused by these models are analyzed. To solve these problems, the paper presents two classes of ways: One based on image restoration technique and the other based on image enhancement technique. Due to the difference degree of information lost in haze and dense fog, different methods are adopted. Adaptive segmentation and local enhancement methods are used to process images taken in haze. As the result, higher contrast of image is obtained without noise magnification in sky region. Single depth image can easily be processed through models. Fist, black pixel whose radiance is zero is chosen mutually, then image optical depth is computed, at last image information is recovered, two images which* This Work is supported by the Major Project of Science and Technology Research of Ministry of Education of China under Grant No: 01113are taken in different degree of fog are processed by difference image information amplification. Results are satisfactory, besides theoretical proof also has been given. According to the characteristics that limited dynamic range of images taken in dense fog, dynamic range stretch as a global enhancement method is used. In order to reduce the effect of noise it brings, sparse coding is adopted.In this paper, computer techniques and images process techniques are used to enhance the clearness of images taken in fog automatically, which has a great significance in improving the robustness of outdoor surveillance systems and also be helpful to the process of images taken in other bad weather.