Abstract: With the excessive consumption of non-renewable resources such as coal and oil, human beings are facing increasingly seriously energy shortage crises and environmental problems. In this case, solar energy, as clean and renewable energy, is receiving increasing attention and application. Solar cells are the key parts of solar power generation, and their quality directly affects energy conversion efficiency, service life and device stability. In order to help students understand the relationship between solar cell fabrication process, defects and solar cell performance and to find ways to improve the performance of solar cells, this paper designs a solar cells defect detection experimental system based on machine vision. The experiment introduces the detection method of electroluminescence defects, the structure of the direction system and the interaction and interconnection between the modules. Taking monocrystalline silicon solar cells as an example, the various defects of solar cells were tested and some problems in the testing process were discussed. The practical results show that students’ understanding of the theoretical knowledge of solar cells is deepened. At the same time, in the course or the experiment, students have developed the practical ability to find problems and solve problems, and achieve good teaching results.