Abstract: Machine vision is a key technology in intelligent manufacturing, but the course of optoelectronic image processing has problems such as outdated knowledge system and poor practical teaching effectiveness. The optoelectronic image processing course at Anhui University of Technology focuses on the teaching of machine vision inspection technology in industrial sites. By reconstructing the knowledge system and fully exploring the functions of experimental instruments, real-time images of cooperative enterprise production sites are introduced into laboratories and classrooms to improve practical teaching conditions. Students are guided to participate in various competitions to promote learning and effectively improve the teaching effectiveness of the course.