Vision-Based Passenger Flow Estimation Method Based on Regional Disparity Extraction

YU Hai-bin,LIU Ji-lin
DOI: https://doi.org/10.3969/j.issn.1004-1699.2007.07.023
2007-01-01
Abstract:To improve the accuracy rate of vision-based passenger flow estimation system by using stereo,a novel method based on regional disparity extraction to detect passenger's head was presented.It was the combination of monocular image processing and stereo.The monocular image processing was employed to extract candidate head areas in which the head disparity would be attained with the aid of the regional disparity histogram.The corresponding relation between disparity and body height and the perspective relation between disparity and head size were applied to eliminate the false candidate head areas and then the passenger flow information was acquired accurately.As the candidate head areas will only be the 1/5~1/4 of entire view,the amount of stereo matching computation can be so largely reduced that the method can be used in real-time embedded system.The experiment results show that the accuracy rate achieves a great improvement when the method is applied to the vision-based passenger flow estimation system.
What problem does this paper attempt to address?