Research on the Identification and Positioning Method of Assembly Unit for Personalization

Abstract Within the framework of Industry 4.0 and the Made in China 2025 initiative, the trend of personalized customization is rapidly expanding. However, the process of identifying and positioning diverse assembly units is hindered by issues like low accuracy and reduced efficiency. This study introduces the GE YOLOv5 algorithm, which refines the YOLOv5 algorithm by incorporating innovative lightweight techniques and attention mechanisms. The results from experiments indicate that this algorithm markedly enhances the performance of personalized customization production lines, achieving a real-time detection speed of 97.59 FPS, an average precision of 0.966 at a 0.5 threshold, and an accuracy rate of 98.6%. Compared to the original YOLOv5 algorithm, it demonstrates clear advantages and significantly advances the shift of personalized customization production towards more sophisticated production philosophies.