Data Transmission Based on a Unified Digital Recognition Technology

In this paper, a low-input and fast-deployment application framework based on the video recognition technology is proposed for the data transmission for cross-system or isolated equipment without proper interfaces. The technical work of this paper mainly includes two parts. Firstly, a YOLO-based deep learning extraction and recognition algorithm is designed to solve a unified digital recognition for both printed style digital and tube style ones. Secondly, aiming at the problem of deviation of recognition area caused by video displace or rotation, we propose a correction technology based on image registration. Experimental results and practical application results show the effectiveness of the technology.