Design and Development of a Fastener Identification and Location System Based on YOLOv8

Abstract

Small and medium - sized auto and motorcycle parts enterprises still use traditional manual sorting methods for sorting small - sized fasteners, which are characterized by high sorting error rates, slow sorting speeds, and high labor costs. This paper designs and develops a fastener identification and location system based on YOLOv8. By building an image acquisition platform and designing a visual operation interface, it can real - time identify camera video streams, dynamic videos, and static photos. Methods such as histogram equalization are used for image enhancement and pre - processing, and the optimized YOLOv8 algorithm is adopted as the core detection method. 500 images are collected, and 4000 images of fasteners of different batches and models are obtained through image enhancement to train the model. The results show that the detection accuracy of this algorithm reaches 92%, the recall rate is 88%, and the detection speed is 60FPS. This system can meet the accuracy and real - time requirements of fastener detection in industrial production, optimize and replace traditional sorting operations, and has economic and practical value.