eduzhai > Applied Sciences > Transportation >

A Review of Vision-Based Pedestrian Detection Techniques

  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: The problem of pedestrian detection has been considered as a vital one in the domain of Computer Vision and Pattern Recognition. It is formulated as the problem of automatically identifying and locating the pedestrians in images or videos. In the past many techniques have been proposed to solve this problem accurately and efficiently. However, the variations in images such as body attire and pose, occlusion, different illumination parameters in different scenarios and clutter present in the background poses challenges in attaining high accuracy. This paper provides a review of some well-known existing techniques for pedestrian detection in images. Few benchmarked datasets are discussed, performance evaluation metrics are also presented. Furthermore, this paper also discusses the interesting observations of the experiments and potential research direction for the pedestrian detection problem.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×