The Application of Computer Vision in Part Recognition
Industrial part recognition, localization, and sorting are essential components of intelligent manufacturing in the Industry 4.0 era. With the rapid development of deep learning, the YOLO series of detectors has become the mainstream solution for industrial vision due to their balance of real-time performance and accuracy. Recent studies have not only improved YOLO architectures—through lightweight design, attention mechanisms, and optimized loss functions—but also explored hybrid strategies that combine deep learning with classical geometric methods to enhance robustness in complex environments. In parallel, complete pipelines that integrate recognition, localization, and robotic grasping have demonstrated feasibility in real-world production scenarios. Experimental results report high accuracy (mAP exceeding 95%), inference speeds above 30 FPS, and localization precision within the sub-millimeter range. Nonetheless, challenges such as limited domain generalization, difficulty with reflective or occluded parts, and the lack of standardized 3D pose benchmarks remain unresolved. Future progress is expected to be driven by synthetic data and domain adaptation, multi-modal sensing, unified multi-task architectures, and industrial-grade deployment pipelines. This review provides a comprehensive synthesis of state-of-the-art approaches and highlights promising directions for advancing intelligent robotic manufacturing.
References
- Ma J., Liu X., Yang Y., et al. Application of computer vision technology in parts recognition [J]. China Equipment Engineering, 2020, (21):244-246.
- Mi Z., Lian Z. A review of YOLO methods for general object detection [J]. Computer Engineering and Applications, 2024, 60(21):38-54.ZX. Liu, Lightweight YOLOv5 for Part Recognition, 2024.
- Li R., Fu J., Zhai F., et al. Recognition and Pose Estimation Method for Stacked Sheet Metal Parts [J]. Applied Sciences, 2023, 13(7).
- Wang L. Research on industrial parts recognition method based on deep learning [D]. Lanzhou University of Technology, 2023. DOI:10.27206/d.cnki.ggsgu.2023.000415.
- Wang Q. Research on industrial stacked parts target recognition based on machine vision [D]. Anhui University of Science and Technology, 2024. DOI:10.26918/d.cnki.ghngc.2024.000790.J. Wang, Small-Target Part Recognition with Improved YOLOv8, 2024.
- Redmon, Joseph and Ali Farhadi.“YOLOv3:An Incremental Improvement.” ArXiv abs/1804.02767(2018).
- Feng D., Fang Z., Deng L. Real-time multi-object detection of assembly parts based on YOLOv3 [J]. Internal Combustion Engine & Parts, 2022, (12):91-93. DOI:10.19475/j.cnki.issn1674-957x.2022.12.039.
- Teng C., Chen M., Cai W., et al. Intelligent recognition method of mechanical parts based on YOLOv3 algorithm [J]. Information & Computer (Theory Edition), 2023, 35(21):53-55.
- Zhang J., Liu F., Wang R. Research on industrial parts recognition algorithm based on YOLOv3 in intelligent assembly [J]. Journal of Optoelectronics·Laser, 2020, 31(10):1054-1061. DOI:10.16136/j.joel.2020.10.0264.B. Guo, Stereo Vision-Based Recognition and Localization of Stacked Parts, 2024.
- Vadim P., Ekaterina P. Solving the part identification problem using their STL models [J]. Journal of Physics: Conference Series, 2022, 2373(7).M. Gao, Stacked Part Recognition and Robotic Grasping System, 2025.