The Research And Implementation Of AGV Visual Guide Algorithm Based On FPGA

The demand of automation equipment is increasing with the development of the society. Because of the high flexibility, Automatic Guided Vehicles are more and more commonly used. The FPGA-based hardware implementation of image processing algorithms is studied in this paper, which could improve image processing speed and precision of the vision guided AGV in order to improve their accuracy and stability.At first, varieties of guide methods of AGV are introduced, and the advantages of visual guide AGV are presented. Image processing algorithms on different platforms are also analyzed. FPGA and DSP as controllers of the image processing system in this paper's are selected based on comparison. In this way the advantages of both FPGA and DSP could perform well, so the speed and accuracy can be improved.The realization of image pre-processing algorithms in the FPGA hardware is researched, which includes data acquisition and control module programming, communication modules designing, interface modules and image processing algorithms developing. Image processing algorithms include improved median filter, edge detection and Hough transform algorithm to detect straight lines. FPGA implementation of the algorithms can complete image processing faster and the precision can be improved. After completing the designs, Matlab, ISE Simulator and ModelSim are used to simulate the image processing algorithms and the control modules. The correctness of the algorithms is proved by the results, with increased processing speed and accuracy. After the image processing system is installed in the AGV, field running tests are also performed. Actual results are matched with simulation results. AGV running stability and precession have been improved accordingly.Finally, the paper is summarized and the prospects for future research are raised.