Abstract: The number of dismantled energy meters is increasing year by year, and manual sorting and breaking is time-consuming and laborious. Therefore, an automatic sorting and breaking line system for dismantled energy meters is designed in the form of production line. The system hardware part focuses on three kinds of devices: high-definition cameras, image capture cards, and vision sensors. The software part first obtains the image of the dismantled energy meter through image acquisition, and then extracts the key features in the image by using machine vision algorithm. Finally, according to the similarity system, the sorting results are automatically output to realize the classification of the dismantled watt-hour meter. The test results show that the system's recognition accuracy of a number of key information is close to 100%, and the batch label printing and pasting tasks achieve zero errors, fully demonstrating the system's high reliability, accuracy and efficient automatic processing capabilities.