Study on portable detection system of the bending radius of spacecraft cables based on image recognition
In response to the low accuracy and consistency of current manual visual inspection methods, we propose an efficient detection method based on image recognition and develop a portable automated detection system. By utilizing pixel clustering for foreground region positioning and global threshold combined with sliding windows for cable area segmentation, precise cable identification is achieved. Additionally, through cable centerline extraction and bending radius calculation based on hazardous regions and points, quantitative detection of cable bending radius is realized. Experimental results demonstrate that the detection accuracy of this system is no less than 5%, providing a reference for other cable bending radius detection method.
Authors
- Mingyu Guo (ORCID: https://orcid.org/0000-0002-3478-9201)
- Lei Huang (ORCID: https://orcid.org/0000-0002-8854-6387)
- Guanghui Fu (ORCID: https://orcid.org/0000-0002-6391-5983)
- Yue Yu (ORCID: https://orcid.org/0000-0002-4758-8492)
Institutions
- China Academy of Space Technology (CN)
Publication Details
- Published
- 2025-07-16
- DOI
- https://doi.org/10.1117/12.3070618
- Primary Topic
- Industrial Vision Systems and Defect Detection
- Type
- article
- Field-Weighted Citation Impact
- 0.00