A People Counting Method Based on Multiple Cameras Information Fusion

Abstract

As people counting is becoming a research hotspot, a method for people counting based on multiple cameras information fusion is proposed. First, an adaptive sliding window algorithm is designed for people counting in single camera. Second, based on the homography theory, the common area between multiple cameras is calculated. Finally, the object matching strategy considering the factor of occlusion, is designed for improving the people counting results of single camera and achieving the overall people counting of multiple cameras. Experiments results have shown that, this method can perform well in complex scenes.