The Technology Research And Implementation Of Bus Passenger Flow Statistics Based On Video Analysis

With the rapid development of China's economy, the number of population is rapid growth in large cities. Many cities propose "Low carbon environmental protection, green travel" slogan to encourage people to choose public transport travel. In order to meet the travel needs of residents, many cities are expanding the construction of public transport facilities. However, when it is the peak of travel using car at work and tourist that still can not meet the resource requirements, and when it is not travel peak that will cause waste of resources. Therefore, how to optimize the allocation of public transportation resources has become a problem faced by managers and the bus company that.In recent years, as people continuous raise awareness of the importance of data, bus passenger flow statistics also become an important data as optimizing the allocation of public transportation resources.However, the passenger flow data acquisition means lags behind and a low statistical accuracy can not to provide accurate and effective reference data for operational decisions and integrated management of public transportation enterprises issues.Against the bus passenger flow statistics current faced proplem, we design a set bus passenger flow statistics system to suitable the bus operating environment based on video analysis. This system sovles the drawback of traditional data collection method, a preliminary discusses the system architecture and software development of bus passenger flow statistics. When carrying out the bus traffic statistics algorithm, we investigate the current popular object detection and object tracking algorithm, analyze the advantages and disadvantages of each method, and proposed a bus passenger flow statistics algorithm to be used for the embedded devices under retaining it advantage basis.In this paper proposed algorithm, we employ the histogram of oriented gradient aglorithm as the object detection and the tracking-learning-detection aglorithm as the object tracking. In addition, we employ the baseline judge method as the analysis of passenger behavior. When developing the bus passenger flow statistics software, we utilize OpenCV2.3.1 visual library in Visual Studio 2010 platform to achieve the bus passenger flow statistics algorithm based on video analysis, and set up the simulation environment on a PC to test the performance of the bus passenger flow statistics algorithm.The bus passenger flow statistics algorithm based on video analysis proposed in this paper can real-time understand the peope traffic changes of the various bus lines and timely optimiz the bus layout and vehicle scheduling through continuous collection and analysis of the bus passenger flow data. So that it can provide accurate basis for road planning manager of the city and the bus company operation and management, and provide power for the construction and development of intelligent transport.