Context pruned histogram of oriented gradients for pedestrian detection

The contextual information has been found to assist the object detection in many ways. The contextual information in the detection tasks can serve to disambiguate the recognition process when the object to be detected is cluttered in the environment. Also, it can help speeding up the detection process itself. In this paper, we have shown that the use of spatial contextual information makes the process of detecting of pedestrian in still images efficient. We have tested our approach in the INRIA dataset consisting of 999 images and the results are found to be encouraging.