With respect to the positioning navigation for indoor robots' automatic recognition of specific color targets in specified areas, a robotic system development method based on distributed vision technology is proposed. The image of the robot working site is collected by the camera outside the venue, and the machine vision processing is performed by the computer to obtain the position information of the robot, and then the information is used to control the robot to follow the route. After calibration test in the area of 3m×5m, it is proved that the maximum position error of the robot is: 15.6cm, the average error is: 8.1cm; the azimuth error is: 11.3°, and the average error is 8.05°.