Overloading of trucks has serious negative impacts on traffic safety, road infrastructure and the environment. The current axle type detector cannot accurately categorize truck models, and the vehicle axle type recognition system lacks the relevant devices and traceability counts. Therefore, this paper develops a convenient vehicle axle type recognition system based on machine vision. Taking the truck axle as the research object, this paper proposes a machine vision-based vehicle axle type detection and identification system by analyzing and researching the vehicle axle type detection and identification technology. In-depth research is carried out on the structural design of the tracked carrier platform, the embedded design of the system, the improved YOLOv7 target detection algorithm and other key technologies, which realizes the efficient and intelligent measurement of the vehicle axle type, and is of great theoretical significance and practical value for the traceability counting of the truck axle and the classification and identification of the vehicle type.