For the relay platen state identification problem, whether it is a camera or a designated track driving robot inspection, will be affected by the glass door reflection, foreign object obstruction, angle blind area and other platen can not be correctly identified, manual inspection is inevitable. In addition, the current platen image recognition requires a huge background structure, and the inspectors cannot immediately obtain the verification information at the operation site, which greatly affects the work efficiency. In this paper, in response to the above problems, an embedded-based handheld terminal solution is proposed, which uses a miniature intelligent HD camera to achieve multi-angle photography of the pressure plate screen cabinet, effectively solving the problem of obstructing photography due to glass door reflection, foreign object obstruction, angle, etc. In addition, this paper proposes a relay platen status field identification method based on embedded and YOLO Nano algorithm, and selects a high-performance computing core and a special vision recognition module as the hardware platform to avoid a huge background architecture and realize the operation field identification and verification. The test results show that the system has a 100% accuracy rate of relay platen status identification, which effectively solves the operation site identification problem of relay platen status and has strong practicality and applicability.