Research Of The Vision Navigation For Wheelchair Robot

With the development of robot technology and the market's push, the intelligent robot becomes one research focus. The robot study for old people and disable person has favorable economic value and society significance, the key technology of the robot is its moveable performance and navigation technology. According to this actuality, a kind of vision navigation experiment platform based on intelligent wheelchair robot is developed, in the system, the USB camera is adopted as input device of surrounding's informations, the efficient digital image processing arithmetic is adopted to identify route, the multidimensional plan method with forecasting and feedback parameters based on fuzzy control model is performed to plan the going route. The contents of the paper include these parts:Firstly, the hardware platform of vision navigation of the intelligent wheelchair robot is established. The scheme of upper computer and lower computer is used to construct the vision navigation control system, the input of system is image collected by USB camera, the output of system is PWM waves amplified by drive module from lower computer and the USB communication module is used to connect the upper computer and lower computer.Secondly, the digital image processing system is established. The related digital image processing arithmetics are realized by C language in the VC++ platform, in these arithmetics, the local area optimized threshold, the blackest dot informations of gray image and the edge of trace are together used to divide the gray image to black and white one, which makes the image processing system quick and efficient.Thirdly, the routine plan arithmetic based on vision navigation is designed. After analysing and comparing many routine plan of vision navigation, an innovative multidimensional routine plan method with forecasting and feedback parameters is designed, which bases on fuzzy control module. Considering the present follow states of intelligent wheelchair, the direction of ahead routine in visual filed and the corrective result feedback of last action, the navigation module gives control signal to lead the wheelchair robot next motion. Finally, the vision navigation trackings of routines with different directions shape are carried out. The experiments indicate that the vision navigation system has made good tracking effect of these routines in structure surroundings.