차량용 시스템의 안개제거 영상처리를 위한 선예도기반 성능평가 방법

This paper presents a new comparison method of haze-removal algorithms for the next generation automotive. Compared to the previous PSNR-based comparison, which measures the similarity, the proposed MTF-based method checks the sharpness for selecting more suitable haze-removal algorithm for the following lane detection problem. Among the practical filtering schemes used for haze-removal algorithm, experimental results show that Gaussian filtering effectively preserves the sharpness of road images, enhancing the lane detection accuracy.