With the rapid development of the industrial sector in the era, special equipment plays a crucial role in various industries. The nameplate of special equipment is an important component that identifies the device's identity and performance, holding significant significance for equipment management and maintenance. However, in the face of the inefficiency and high labor costs associated with traditional manual identification methods, which cannot meet the growing demands for recognition, this design proposes an embedded OCR text recognition system capable of achieving pattern selection, real-time monitoring, nameplate recognition, cloud data uploading, cloud monitoring, and warning tracking functions. It relies on modules such as STM32103ZET6, Alibaba Cloud IoT platform, Raspberry Pi 4B+ and a series of related modules, including the CMOS high-definition camera, are utilized in the system. A super lightweight OCR recognition system specifically designed for nameplate text recognition is employed. The overall model size of the system is below 5MB, and it achieves an overall recognition rate of over 84%.