An Improved Method of Image Fast Defogging

Abstract

In order to improve degraded image clarity and color fidelity in foggy weather,an improved defogging method based on dark-channel prior is proposed.According to the more computation time,and on the premise of defogging effect,fast bilateral filtering is used instead of soft matting to refine transmission map that reduces the complexity of the algorithm greatly.The processing time is 5% of original method,which can meet the real-time requirements of the general engineering.Since the brightness of image after haze removal is lower than the actual scene and the color is dim,a simple and effective method of image contrast and brightness enhancement is proposed.It adaptively enhances the image brightness,and modifies the local contrast which is due to the thick fog and haze.The experimental results indicates that this method can restore the fog scene and color rapidly and effectively,and guaranteed the stability and reliability of outdoor imaging system in bad weather in a certain extent.