A Low Distortion Image Defogging Method Based on Histogram Equalization in Wavelet Domain

Abstract

With the development of computer vision, the requirement of image sharpness in production is increasing. Fog, as a common weather phenomenon, will affect the image quality, so it is necessary to defog the image. In the image defogging technology, the image enhancement method based on non-physical model represented by histogram equalization has the advantage of its relatively mature and simple algorithm, but the disadvantage is that, because of the enhancement effect on the image on the global scale, defogging the detail information will also affect the contour information, resulting in distortion.In this paper, an image defogging method with low distortion based on the histogram equalization in the wavelet domain is proposed, which uses the nondestructive and reversible properties of the integer wavelet transform to separate the detail information and contour information of the image. We equalize the histogram generated by the low-frequency coefficient of the image extracted by the integer wavelet transform, then carry out the inverse integer wavelet transform, and fuse the processed low-frequency coefficient with the high-frequency coefficient of the original image to generate the defogging image. Experimental results show that, compared with the common physical model and non-physical model algorithms, the characteristic of our method is that it has low distortion, on the basis of obvious fog removal effect, it retains the contour completely and meets the application requirements.