Video Enhancement Under Bad Light

Video and image information not only becomes more and more important in people's lives, but also is the basis of all kinds of video surveillance, visual analysis and processing. However, the proportion of all kinds of bad image acquisition environment is increasing in real life, which result in low contrast image. Moreover, the image is vague even important content blocked, which directly affects the quality of image and video. In order to obtain the ideal in a variety of adverse circumstances of video and image, laying the foundation for further application, it is necessary to study the various factors of adverse environment on image quality, building imaging model, recovering contaminated video and image, and ultimately eliminate the adverse effects of the environment.This paper studies the effect of image quality based on a large number of previous studies, and obtains some results:1) The illumination condition of the scene is poor in the night environment, which leads to the decrease of the overall brightness and contrast of the video and the image. In addition, the target cannot distinguish from the surrounding. In this paper, we study the imaging characteristics of the night environment, analyze the classic image enhancement algorithm for night environment, and propose a novel image and video enhancement algorithm. This algorithm in this paper introduces characteristics of root function and phagocytosis of details, not only enhances the night image rapidly, but also maintains the true color of image at the same time.2) Due to a large amount of water vapor and particulate matter in the air of ambient light and sun light for a variety of scattering and absorption in foggy environment, the video and the overall image is pale and fuzzy,the overall brightness of the image is enhanced but the contrast is reduced.After studying the atmospheric scattering model and the Dr.He's theory of dark channel, a fast haze removal algorithm is proposed based on light and dark channel. The algorithm put forward new light channel thoughts for the estimation of the atmospheric environment value, introduce fast guided filtering for the refinement of the atmospheric environment and the transmission value, enhancing the fog image and video, and improve the processing speed.3) LED pixel brightness enhancement when rain or snow appear on the target occlusion, which make a big difference with the surrounding environment. This paper studies the physical characteristics, and visual model, and analyzes some existing video and image based on snow removal algorithm. Then this paper puts forward the temporal and spatial characteristics of snow removal algorithm. This algorithm uses the spatial and temporal characteristics of the snow and snow model to separate the moving objects, and then to get the pixel classification for repair.At last, this paper designs and implements a video image enhancement system. A large number of experiments show that the algorithm designed in this paper can effectively enhance the image and video, and has better real-time performance.