| China became a member of the WTO in 2001. With the improvement of people's living standard,a growing number of international and domestic large-scale shopping malls are opened by which a large number of customers are attracted because of their low price and the numerous products.However, "Customer focus" is the foundation for a future developmentthe. So the information of passenger flow is becoming increasingly important in the business competition. As the passenger flow is the key factor of determining the effectiveness of the shopping malls, and also the basis for mall operators to establish business strategy, so it is undoubtedly a good idea for merchants to devise a intelligent passenger information monitoring system, which analyse the information of passenger flow in a comprehensive way using passenger flow information in combination with other information, to improve their competitiveness in the market.In this paper, a neural network passenger flow counting system is developed in the basis of the radial basis function (RBF), using infrared photoelectric sensor technology.Compared with the Conventional photoelectric sensor count method,this counting method improve the accuracy of real-time passenger flow count largely. And for collecting for each successive time series data on the characteristics of passenger data, it greatly improved the existing segmentation methods and feature extraction method in the actual situation for the passenger flow. At last, it shows the results through the intensive and sparse passenger flow simulation tests. The model presented in this paper, which not only can identify the people side by side, and has relatively low error rate, has a strong theoretical and practical significance.This system consists of four sets of infrared photoelectric sensors, installed in the both sides of mall entrance at ankle height, which can be used to collect customer data. RBF Neural Networks Classification and Feature Extraction Method are the keystones of infrared passenger flow acquisition system. This method not only can reduce the noise among the data, but also can settle the problem when several customers are in a near position.The system is accurate, intelligent and robust, the experimental results are given in the last. Experimental results show that RBF-based passenger flow acquisition system is very effective for the passenger flow. |