An adaptive approach for visibility enhancement in aerial imagery

A novel spatial domain image enhancement algorithm, in which dynamic range of the scene illumination is compressed from the human visual perspective to improve the visual quality and visibility in digital images captured under degraded visual conditions, is proposed. The proposed algorithm employs an adaptive approach so that local image statics, namely the local standard deviation and the local mean in the image are modified simultaneously utilizing an intensity transform based on a “S” shape curve, the curvature parameter of which is determined adaptively followed by a local contrast enhancement process. The performance of the algorithm is evaluated by a statistical visual measure, along with visual comparisons of the proposed method with state-of-the-art enhancement algorithms are given.