Class-Wised Image Enhancement for Moving Object Detection at Maritime Boat Ramps

In the context of marine boat ramps traffic surveillance, we propose in this paper a novel image enhancement method for interpreting the traffic of boats passing across the boat ramps. As the background dynamics of land and water scenes differ markedly, this new approach classifies areas in each image as either land or water, so that different strategies can be adopted to enhance image on land and on the water, respectively. In particular, the use of the dynamic sharpening size and adaptive sharpening strength significantly increases the robustness of this enhancement method. Experimental results demonstrate that our method is much more able to cope with the highly dynamic land and water composition scenes compared with the state-of-the-art methods.