Image Enhancement Algorithm Research And Its Application To Images Defogging

As an important part of image processing, image enhancement technology aims to emphasize the useful information of images by improving the image contrast, and get better image quality and visual effects. According to the processing of different spaces, image enhancement can be classified into two types, spatial domain algorithm and frequency domain algorithm. Al-though both two algorithms can effectively expand the contrast, the algorithm based on frequency domain has the large amount of computation that is not suitable for real-time processing. So the spatial domain algorithm is applied more widely. There-fore, this paper focuses on the spatial domain algorithm based on histogram and Retinex theory.Image enhancement technology is widely used in image processing software, industrial, medical, military and other fields. By carrying on enhancement processing to all kinds of pictures, we can obtain the ideal picture effect to meet the requirement of image information extraction in our life. In foggy weather conditions, Image acquisition device get blurred and low recognition images which seriously affects us to extract the useful information of image and also cause unpredictable in-fluence on the subsequent image processing. Therefore, the research of image enhancement algorithm and its application to image defogging has important practical significance.This paper firstly elaborated the background and significance of image enhancement research, and discussed the current situation to the image enhancement defogging algorithm at home and abroad, and also introduced some basic enhancement algorithm from two angles of the spatial domain and the frequency domain. Then, it mainly focuses on the analysis of image enhancement defogging method based on histogram and Retinex theory, and get the algorithm effect picture through the simulation. what’s more, it has given a new kind of improved algorithm which is based on linear stretch, the algorithm enhanced the image contrast gradient and optimized the picture quality. Finally, this paper analyzes the advantages and disadvantages of each of these algorithms, and present a new enhanced defogging algorithm which combined contrast limited histogram equalization with multi-scale Retinex. It analyzes the algorithm effect picture by image quality evaluation criterion and shows the new algorithm effectively improve the quality of the image after defogging.