Detecting Probable Regions of Humans in Still Images Using Raw Edges

Human detection remains a challenge in computer vision due to highly articulated body postures, viewpoints changes, varying illumination conditions and cluttered background. Because of these difficulties, most of the previous publications often focus only on low-articulated postures, e.g. pedestrians, in still images. In this paper, we propose a new method to detect a human region from still images using raw edges. Not exhaustively detecting all of people occurrences in images; nevertheless; our approach can perform significantly on many types of images, typically, sports images with various poses. Instead of sliding window-style approaches for detecting, we rely on characteristics of boundaries and interest points by combining several image-processing techniques such as image filter, image segmentation, edge detection…Afterward, we use K-mean algorithm and probability for choosing a human region. Especially, we do not need a training phase. Despite not being the same purpose on detecting domain to previous works, in certain degrees, we also try to compete to typical works. Two challenging datasets are involved in discovering interesting facts needed to be concerned when designing proposed method for detecting people.