The Analyzing Method And Device Of Video-based Bus Passengers Crowding Level

Transportation is the base of city development. As the fast development of domestic economy and increase of car ownership, the bus has become the main transport vehicle for the public. However, the dense passengers and uncomfortable experience on bus are suffered much criticism by the citizens. Consequently, the bus enterprise need to cover the real-time passenger flow circumstance and release the bus passenger congestion level information to the public. In recent years, each of city is promoting the construction of intelligent bus system vigorously, and intelligent vehicle terminal has become the standard equipment for the bus which realized the function of vehicle positioning, wireless communication, automatic reporting station, violation detection, audio and video surveillance storage. However, it lacks effective methods of passengers flow detection. Traditional radio frequency card, infrared and binocular video and other passenger flow collection equipment have obvious defects. Consequently, it contributes much significance to research and develop vehicle terminal with the function of analyzing the passenger congestion level in the bus.This paper aims at promoting a video analyzing methods and designing the relevant vehicle terminal to realize the function analyzing degree of congestion in the bus with surveillance camera. Firstly, the paper introduces the research background and current circumstances and the purpose and significance of the project. Secondly, it illustrates the main function of the vehicle terminal and introduces the video analyzing technology theory including video standard, color space and interface standard. The hardware and software design of vehicle terminal especially the core video module design is described in the third part. Fourthly, it indicate the two proposal based on the video congestion analyzing, which are passenger flow detection and density analysis. The density analysis is chosen by comparison. The next part is core algorithm including total process design, video analyzing area choice and division, application and algorithm optimization of frame difference and gradient two video analysis methods and brief introduction of programming. The last part is the test and evaluation of the vehicle terminal including hardware type experiment, system resource test, congestion analysis function evaluation. The experiment results satisfied the design objects which accurately reflect the real time passenger flow status during the bus operating progress.The research results of the project, vehicle terminal, which supports congestion analysis has been promoted to the market and be applied in a number of cities in China. The real time congestion degree information provided by the vehicle terminal provides the data support for the transportation choice for the public and transport capacity arrangement, line network planning, dynamic scheduling for the public bus.