Research On Recognition System Of Elevator Car Congestion Based On Image Processing

With the continuous development of intelligent high building,elevators has become an indispensable vertical transportation.At the same time,people are demanding for the elevator operating efficiency and comfort higher.As for optimization for elevators efficiency,whether elevators is fully loaded is an essential index.The traditional elevators can only through single weighing system to estimate whether it is fully loaded,therefor there are some mistake stop such as the elevator is not fully loaded but elevator car can t accommodate more passengers.Aiming at the above problems,this paper have researched an elevator car congestion degree recognition system based on image processing,applying the image processing technology to elevator control system,to recognize elevator car internal occupation in real time,making up for the inadequacy of the traditional elevator single weighing system,further improve the elevator operating efficiency and comfort.The main work and achievements are as follows:(1)Three classic moving object detection algorithms are researched: optical flow method,inter-frame difference method,and background subtraction method.Their advantages and disadvantages are analyzed.Mainly test and contrast for four background model in background subtraction.Combined with the operating characteristics of the elevator,put forward an improved algorithm based on adaptive Gaussian model.It has been improved from three aspects: background modeling,binary difference and background updating strategy.(2)The system overall plan is designed,the overall hardware structure of the system is analyzed,and the main chip selection and the hardware design of the peripheral related circuits are completed.Includes the memory and FLASH expansion module,LCD interface circuit,USB expansion circuit,Ethernet network interface circuit,Port communication circuits,and power circuits,etc.Finally,the hardware PCB circuit board is manufactured and the relevant tests is completed.(3)Combined with software functional requirements of the system,the embedded Linux system software platform is transplanted,the system software program is finished and debugged on the platform,and the function modules such as image collection,image processing,calculation of car occupancy rate,serial port communication and so on are realized.(4)The simulate elevator car experimental platform is built,and the system car occupancy recognition stability,car full-load standard and anti-jamming ability test experiment are completed.The experimental results are analyzed,and the system plan is shown to be feasible.