An Effective and Efficient Approach for Single Image Dehazing and Defogging

Images captured by outdoor systems are highly degraded under bad weather conditions, which directly affects the performance of driving assistance, navigation and tracking systems. Thus, recovering the scene from hazy and foggy image is consequential. In this paper, a fast single image deweathering technique is proposed based on morphological operation. In foggy and hazy images, the unequal distribution of depth map is the primary cause of degradation. This unequal variation of depth is compensated by proper estimation of atmospheric light and transmission depth. To accurately estimate the atmospheric light and transmission depth, morphological operation is used. The obtained experimental results in terms of qualitative and quantitative comparisons with the state-of-the-art show the efficacy of the proposed method. Furthermore, the proposed method takes less execution time with little or no halo artifact.