A Novel System for Authentication of Identity Cards Using YOLOv5 and YOLOv8
摘要
This paper introduces an advanced authentication system leveraging the You Only Look Once (YOLO) object detection technique, specifically YOLOv5 and YOLOv8. The system identifies individuals within an institution based on the detection of their ID cards. Upon detection, the ID cards are isolated from the image frame for further processing. The isolated card images are then subjected to another round of YOLO detection, targeting specific text fields. After the text is localized, Optical Character Recognition (OCR) is utilized for precise extraction of the textual information for identity verification. This streamlined process boosts security of the organization, facilitating the prompt identification of non-compliant individuals and efficient verification of others. The system’s promising results indicate broader potential applications in various settings that demand reliable identity verification protocols.