Fog Image Enhancement Method

Fog is a usual natural phenomenon. There are amounts of suspended particles in foggy weather, because these particles scatter light, the contrast of image is relatively low and image is got blurred, so it has a very important value to recover image as soon as possible.Currently there are two kinds of image defogging techniques, one is based on physical models of atmospheric degradation, the model parameters of this method require some priori conditions and physical devices, its application is limited. The other is based on image enhancement method, this method doesn't need to establish the atmospheric model and directly enhance fog-degraded image details. This paper uses the second method, which mainly includes the following contents:First of all, we introduce the development of some traditional enhancement methods based on image processing, and then the wavelet transform methods is focused on.Wavelet analysis is developed in recent years and quickly applied to image processing and voice analysis, it can achieve both image enhancement and denoising.This paper presents a method of wavelet image enhancement based on wavelet transfom. This method has not only good detail enhancement but also good de-noising result. Firstly, the image is processed by wavelet transform, then the image is enhanced by wavelet transfom based on semi-soft-thresholding denoising.Lastly, the enhanced image is performed by inverse wavelet. Because of the multi resolution analysis of wavelet transfom, this method can well suppress noise, but wavelet transfom is located in the frequency domain, the enhanced image is not sensitive to the scene depth information. Image is further stretched contrast in time domain so as to enhance the contrast of the image.Experimental results demonstrate that this methods are effective for contrast enhancement of fog-degraded images and can be applied to the practical situations.