Quality Inspection Algorithm of Printed Character Based on OCR Technology

This electronic document is a “livUsing OCR technology to automatically recognize identification characters will greatly reduce people’s workload and improve work efficiency. However, in the case of low contrast between background and characters, traditional OCR technology has poor anti-interference ability and low recognition accuracy. This article proposes a printing character quality inspection algorithm based on AlexNet. The quality inspection process includes multiple steps such as image acquisition, image preprocessing, feature extraction, classification recognition, and quality inspection. In the design of fuzzy character recognition software, a graphical interaction interface is developed using the Python Tkinter module, and a character recognition system is designed, including a character recognition module and an information storage and query module, to enhance the human-machine interaction ability of the system and facilitate product sourcing and management. The experimental results show that the model in this paper can achieve recognition accuracy of 96.1%, and the recognition time for individual item characters is 0.29 seconds. The AlexNet model proposed in this article has better performance of printed character recognition than that of CNN and RNN models. The experimental conclusion of this article can be used for character recognition and quality inspection in the tobacco industry.