Research On Technology Of Visual Perception Enhancement In Weather Of Rainy And Foggy

In the bad weather of rainy and foggy, video captured is often blurred, unclear, that affects the use of subsequent video processing and analysis. So there is great theoretical value and practical application value to remove the interference caused by external factors such as rain and fog and to improve the image clarity. In recent years, visual perception enhancement to remove the interference of rainy and foggy and to improve the image clarity is a hot topic in image processing.This paper takes aim at improving image visual in bad weather, and analies the cause and mechanism of image degradation in depth. We research the techniques of enhancement from image enhancement and physical model. The main tasks are:1. Through a lot of experiment, the inverted foggy image is similar with low-light image. A new algorithm based on Retinex is used to making fog image clear. The algorithm works well for particular images, like mist.2. Atmospheric scattering model explain the principle of fog formation. A clear image is calculated by solving mathematical equations. Transmission image is quickly estimated using Retinex and time is efficient. The algorithm works well in weather of smog.3. Lightness of raindrop is higher than background. Raindrop does not appear at the same position of the neighbor frames. An improved algorithm is presented.This paper focuses on the research of the mechanism of the mechanism. Image defogging is researched from two ways:image enhancement and physical model. Image enhancement is considered in time domain and spatial domain, processing the area we are interested in. Solving the mathematical equations, physical model is to remove the fog from the mechanism of the fog formation. Aimed at image clarity of rainy, an improved algorithm is presented. A lot of experiment shows algorithms work well in the database of He, Tarel and our.