TV image processing to determine the position of a robot vehicle

摘要

An algorithm for determining the position of a robot from only one TV image has been developed. The image processing algorithm consists of three steps; the first involves transforming the standard square pattern of the input image into its skeleton, in the second step the positions of the four vertices of the square on the TV is determined by applying the method of least squares, in the third step the position of the robot is computed.