Research On Image Dehazing And Color Correction Algorithm

Due to the development of computer vision technology and the improvement of computer processing speed, computer vision system enjoys increasing popularity in everyday life. However, images captured in outdoor environments are usually degraded because of the presence of bad weather conditions such as fog and haze. The low contrast and color distortion of images leads to video surveillance and information acquirement problems. Therefore, image dehazing becomes a heated discussed topic in academic fields.Nowadays, the dehazing algorithms can be mainly divided into two categories:non-model dehazing algorithms which based on traditional image enhancement methods; image restoration based modeling ones. In general, the non-model methods enjoy lower complexity and are able to procedure real-time process. However, the non-model based algorithms causes less effective dehazed images. The restoration methods build the hazy image degradation model based on the theory of the atmospheric scattering theory. However, the dehazed results contain false color with high processing complexity.This article focuses on image dehazing based on the haze image degradation model and color correction. We analyze the cause of hazy image distortion and solve the dehazing problems of hazy images and video. Firstly, we briefly review all related work domestic and aboard and then deeply research on the classic algorithms. And then introduced the hazy image degradation model based on the theory of the atmospheric scattering theory, and did further study on the key technologies which are Dark Channel Prior (DCP) based dehazing, transmission refined and the estimation of the atmospheric light. Based on the the image degradation model and key technologies above, we propose some novel dehazing algorithms to achieve clear dehazed images. Experiments’results show the vivid color and the accuracy of the restored edges and gradients. A comparison study is made among all these algorithms.This article makes innovation which mainly includes two items:· a single image dehazing and color correction algorithm based on HSI color space (HSICT) is proposed. Since the degradation of hazy image in color channels of HSI color space is different, the different restoration methods can be applied in different color channels. The initial recovered image usually suffers from color distortion, so color transform method is introduced. The experimental results show that the algorithm can improve the contrast of the image and avoid the color distortion.· a novel approach using image fusion framework (DehazeIF) for single image dehazing is proposed. Firstly, we employ the fast wavelet transform based image fusion method to refine the transmission. With the estimated transmission, the haze-free image can be obtained with high contrast but suffers from color distortion. Secondly, we implement a color correction scheme based on weighted image fusion to modify the color information of the haze-free image.