Robust Object Detection and Enhancement in Weather Degraded Images

摘要

Detecting objects of interest and obtaining their clear visual appearances are critical requirements for visual surveillance systems.In this paper we propose a novel algorithm to detect foreground objects from weather degraded image sequences and then enhance their visibilities.First,we propose a novel metric based on the dark channel prior method to measure the image fog property to decide whether the image scene is obscured by fog.Second,if there is heavy fog in the scene,a novel approach for object detection based on an air atmospheric scattering model is proposed.During the object detection step the background depth map is calculated from two different weather condition images.This background depth map can be computed just once for the same scene.Then the depth map for the current image is generated.Finally,the difference between the current depth map and background depth map is measured and the major different regions indicate the foreground regions.Once the foreground object is detected,we enhance its visibility based on a scene radiance recovery method based on its depth values.Our proposed algorithm is tested with some surveillance video recorded under different fog conditions.Experimental results show that the proposed approach is efficient and accurate in foreground object detection and visibility enhancement.