Research On RBF Neural Network And Its Application In Infrared People-counting

Under the condition of the global economy's into an organic whole and the knowledge economy's coming, business enterprises have to face the serious challenge. In order to survive and develop, business enterprises should focus on customer which is a certain choice for competition. As the computer technology is universal in commercial areas, the importance of passenger information in commercial competition is growing. It can improve the competitive capacity of businesses by making an intelligent monitoring system of passenger information which centre on passenger information and combine other information for analysis.This paper construct a Customer-counting system based on RBF neural net using technology of infrared sensor. Compared with traditional Customer-counting using sensor, this application has more accuracy. This article improve an adapted method of data segmentation and the feature of customer flow, which based on the feature of Customer-counting data continuous space-time sequence. By using both sparse and dense customers flow, the simulation results show effectiveness of recognizing the situation of customers who entry at the same time, which testify the significance of our system not only in principle research but also in actual application.As a new modeling method, Radial Basis Function (RBF) neural network can learn regulations form history data,auto-cluster and organize network structure according to given issue, it can overcome the shortcoming of partial minimum point problem which occurs in the BP neural network. RBF network is wildly used for its excellent performance in many fields such as system recognition and dynamic forecasting and so on, so we put forward a people counting model based on the RBF neural network in this paper.This system includes four photoelectric transducers of infrared ray, which are fixed in the entry of Market, and the height to the ankle. It collects data for passing people. There are two focuses about the proposed Customer-counting system: the approach of feature extraction and the proposed hybrid RBF network for classification. This approach reduces data noisy. The system can deal with the problem of the passing people walks closely each other. This system has high accuracy, intelligence and robusticity. Some test results are given at last. The presented results show that the proposed hybrid RBF based system is able to deliver promising results in performing people counting in an open public passage.At last, it has constract a systemwhich can output passenger flow.