Enhancement of Under-Exposed Image for Object Tracking Algorithm Through Homomorphic Filtering and Mean Histogram Matching

Object tracking through video or image becomes more popular in recent years. Indeed, clear and high contrast images are essential to attain good tracking results. The problem arises when the object in an image or video is under-exposed, resulting it to be hardly visible and differentiated from the background. Existing methods are able to solve some of the aforementioned problems, but produce other problems such as over-enhanced effect and color distortion. Thus, the object of interest may become untraceable due to these distortions. This paper proposes an image enhancement method to improve under-exposed images or videos through homomorphic filtering and mean histogram matching, in order to produce more visible and traceable objects. This method integrates homomorphic filtering method and histogram modification technique which consists of histogram matching and dual-histogram stretching. The proposed method is designed to reduce non-uniform illumination while increasing image/video contrast and visibility. The experiment results show that the proposed method outperforms some state-of-the-art methods in terms of visibility and contrast level on some standard benchmark database.