Counting people by using a single camera without calibration

The counting of people is considered important for air-conditioning system in the intelligent building. Based on image processing, an automatic simple bi-directional counting method of pedestrians through a gate is proposed. In this method, only one common camera is used which is hung from the ceiling of the gate with a directly downward view. An experimental formula is used to estimate the number of people in the foreground blobs. Then, a simple-fast tracking algorithm is used to track blobs. The moving direction is determined by using the track information and the path classification. In addition, the method also has merge-split counting strategy to solve the merging or splitting cases. The method has been tested in a 3.2GHz Intel Core machine with 40fps on 1080×720 images without code optimization. Average accuracy rate of 98% and 95% are achieved on videos with normal traffic flow and videos with many cases of merges and splits, respectively.