Hardware Implementation of Minimum Filter based on 2-D Cumulative Histogram

Recently, advance technology such as automated driving or surveillance requires the clear scene of input image or video, therein lies the reason for the growing importance of haze removal algorithm. In order to remove fog, atmospheric light estimation is required. Since white objects can be mistaken as fog, minimum filter is needed to prevent this problem by reducing the effect of these objects in the image's local patches. This paper proposes hardware design of minimum filter based on cumulative histogram. The synthesis is done using a SoC evaluation board.