Moving Chessboard Feature Recognition Base On CCD

An iterative method of forward position solution for 6-PTRT parallel mechanism is proposed. The method takes advantage of mechanism characteristics of the parallel mechanism, and establishes Jocabian equations to solve forward position of active platform. Based on the positions during the motion process of the platform, the corresponding motion trajectory is reconstructed.In experiments, a high-precision chessboard is fixed onto the platform and is assumed to stand for the platform. During the motion process of active platform, visual measurement system acquires sequential images, and then extract features with the help of image processing techniques. According to pinhole imaging model, the system figures out each position of the platform, and furthermore, reconstructs the motion trajectory. Experimental results show that absolute error of two motion trajectories recovered from forward position solution method and visual measurement system is within 2mm, and mean error is merely 0.3mm.Visual measurement system is developed using Visual C++ programming environtment and OpenCV library. The system includes functions of motion control, real-time motion trajectory reconstruction and display, sequential image acquisition, camera calibration, feature extraction.