This topic mainly studies the navigation parameters obtained by image processing technology to achieve omnidirectional mobile AGV autonomous navigation. The camera is mounted on the bottom of the body of the AGV and captures the black tape path on the ground. Image preprocessing is performed, including image graying, improved CANNY algorithm edge detection, morphological processing, and so on. Then the Hough transform is used to detect the path of the preprocessed image. Finally, a straight line is selected to obtain the effective edge line, and the navigation deviation parameters are extracted. Experimental results show that the effectiveness of the guidance techniques in this paper has achieved the expected results.