Removal of Fog Effect from Highly Foggy Images Using Depth Estimation and Fuzzy Contrast Enhancement Method

摘要

Due to effect of fog visibility of objects decreases and it will be very difficult to recognize any object in hazy atmosphere. So there is always a need of a method which can detect the effect of haze and can reduce the effect of haze to enhance the quality of image. This paper presents depth estimation and Fuzzy Contrast Enhancement based model to remove effect of haze from image by making a linear model to estimate scene depth of the hazy image. underneath this model the depth data is well recovered. Fuzzy logic system is developed for enhancement of contrast. The aim is to get rid of haze from a hazy image and it may be achieving by generate a picture of high contrast than the first image by applying more weights on input gray levels so that it can match with mean gray value. Experimental results ensure that our technique is extremely good. While comparing our result with existing methods on the basis of MSE, PSNR and time and result shows average PSNR of our method is 16.56 which is much better than the existing methods.