Multi-Pedestrian Detection

As part of the U.S. Department of Transportations Intelligent Vehicle Initiative (IVI) program, the Federal Highway Administration (FHWA) is conducting R&D in vehicle safety and driver information systems. There is an increasing number of applications where pedestrian monitoring is of high importance. Traffic control, security monitoring, pedestrian flow analysis, and pedestrian counting are some applications which rely heavily on pedestrian tracking. We have developed a new multi-pedestrian detection system. Our system comprises a number of key modules including (1) disparity layered thresholding method that can be used to extract the potential pedestrian(s) layer by layer; (2) pedestrian detection using 3D convex curvature feature that can be used to window out the pedestrian(s) in each disparity layer; (3) continuous pedestrian association that correlates the same pedestrian in subsequent frames; and (4) pedestrian motion estimation of speed and location. We have tested out system using actual images taken with our stereo vision system at some street intersections. Test results have shown that our system has the great potential being used in IVI systems.