A kemelized Probabilistic Neural network approach for counting pedestrians

Abstract

An improved, intelligent pedestrian counting system, using images obtained from a single video camera, is described in this paper. This system is capable of detecting and counting a group of pedestrians in the region of interest. Groups can be extracted by using the image processing method, and a Kernel-induced Probabilistic Neural Network (KPNN) employed to perform the classification, and estimate the number of pedestrians in a group. We validated the pedestrian-counting system on a pedestrian dataset, and this analysis indicates th at the proposed KPNN-type classifier provides good results.