Image Defogging Combined with Compensation of High Lighted Areas

This paper proposes an improved dark channel prior defogging algorithm to improve the color distortion in the sky and other high lighted areas. Firstly, the dark channel prior is used to estimate the transmittance and the atmospheric light value. Then, guided filter is used to smooth the transmittance, and L
0
gradient minimization filter is used for further refinement. Finally, the constructor method is used to compensate the transmittance to improve the color distortion in the high lighted area. Experimental results show that the method can accurately estimate the transmittance and enhance the contrast of haze removal images, the color distortion of high lighted areas is improved. The haze removal results obtained after haze removing are clearer and more adaptive.