Hybrid Algorithm of Dark Chanel Prior and Guided filter for Single Image Dehazing

Abstract

In this paper, we propose a hybrid algorithm of dark channel prior (DCP) and a guided filter for single image dehazing. First, it takes a haz image as the input, when cleaning the haze in the image with background area and low contrast. Next, in order to change the contrast and intensity of haze removal image with the proposed method. Then, a modified approach is applied to rebuilding the pixel of the resulting image. Reducing the sharpness and the air light expands the whiteness in the image capture. From the weather with fog and haze caused by floating particles, worsen the quality of the image. Haze removal algorithms are more useful for many applications in vision. The researcher has proposed a method to adjust the contrast and intensity of the image in removing the haze from The DCP method by using the guided filter. Which, the method we propose uses a variety of outdoor haze images test. Experimental results shown that the proposed approach outperforms, it was found that the DCP combining with guided filter algorithm haze methods give better performance in terms of mean structural similarity (MSSIM) and peak signal-to-noise ratio (PSNR), which is effective in hazing. Comparative results of the peak signal-to-noise ratio and image quality index demonstrate the robustness of the proposed methods model.