1. Study on Zenithal Pedestrian Detection Algorithm
- Author
-
D.Y. Shen and M.H. Xu
- Subjects
InformationSystems_INFORMATIONINTERFACESANDPRESENTATION(e.g.,HCI) ,Computer science ,business.industry ,Pedestrian detection ,ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION ,ComputerApplications_COMPUTERSINOTHERSYSTEMS ,Pedestrian ,Edge detection ,Hough transform ,law.invention ,Object-class detection ,law ,Feature (computer vision) ,Computer vision ,Enhanced Data Rates for GSM Evolution ,Artificial intelligence ,business ,Algorithm ,ComputingMethodologies_COMPUTERGRAPHICS - Abstract
Pedestrian detection is a hot and difficulty topic in computer vision. This paper presents a zenithal pedestrian detection algorithm designed with the circular feature of pedestrian head in case of top view. It is a reliable pedestrian detection strategy for intelligent illumination control system. The active edge detection algorithm mextracts the pedestrian edge, and grads Hough transform detects the circular feature of pedestrian head. Matlab simulation experiments show that the detection method can effectively detect pedestrian and eliminate the interference. Keywords-pedestrian detection; hough transform; edge
- Published
- 2015
- Full Text
- View/download PDF