Automatic Mall Traffic Statistics Based on Head Target Detection
Jia Shi-ji · Journal of Dalian Jiaotong University · 2015
Mall traffic automatic statistics has great significance in security management,staff scheduling and commodity procurement. For the counting difficulty in the case of overlapping,automatic mall traffic statistical algorithms is put forward based on the human head target detection. Firstly,Haar features are extracted to train Adaboost head target classifier. Secondly,Camshift algorithm is employed to track the target,and Kalman algorithm is used to narrow the search scope. Finally,head template is used to match the pedestrian target. Experiment results show that the average accuracy of the proposed method is 98. 2%,and the counting time of each pedestrian target is only 19 ms.