Half-grouting sleeve fullness detection using consumer-grade endoscope-based vision method and ensemble soft voting classifier.
Saved in:
| Title: | Half-grouting sleeve fullness detection using consumer-grade endoscope-based vision method and ensemble soft voting classifier. |
|---|---|
| Authors: | Wu, Yanqi1,2 (AUTHOR), Huang, Zhengrong3 (AUTHOR), Zhang, Jian2,4 (AUTHOR) jian@seu.edu.cn, Zhang, Xinyu2 (AUTHOR) |
| Source: | Nondestructive Testing & Evaluation. Jul2026, Vol. 41 Issue 7, p4221-4241. 21p. |
| Subjects: | Image recognition (Computer vision), Quality control, Construction materials, Machine learning, Engineering inspection, Deep learning |
| Abstract: | As a commonly used connection method for prefabricated components, sleeve grouting connections play a crucial role in the quality, safety, and durability of these structures. Given that current grouting quality detection methods are mostly qualitative and that industrial endoscopes for quantitative detection are costly, this paper proposes a half-sleeve grouting fullness detection method based on consumer-grade endoscopes and deep learning. By utilising pre-embedded plastic pipes as channels for endoscope inspection, intelligent recognition of grouting fullness at inspection sites is achieved through an ensemble soft voting classifier (ESVC) trained on collected endoscopic images. This method delivers high detection accuracy while reducing the subjectivity of manual interpretation. Based on accurate localisation of the upper surface of the grout and variations in endoscope insertion depth, the proposed approach can effectively quantify the fullness index under incomplete grouting conditions. This detection process is expected to shift traditional sleeve grouting quality control from 'post-detection' to 'process monitoring', providing essential data support for timely repair decisions and ensuring grouting continuity. The approach establishes an effective closed loop between detection and repair. [ABSTRACT FROM AUTHOR] |
| Copyright of Nondestructive Testing & Evaluation is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | As a commonly used connection method for prefabricated components, sleeve grouting connections play a crucial role in the quality, safety, and durability of these structures. Given that current grouting quality detection methods are mostly qualitative and that industrial endoscopes for quantitative detection are costly, this paper proposes a half-sleeve grouting fullness detection method based on consumer-grade endoscopes and deep learning. By utilising pre-embedded plastic pipes as channels for endoscope inspection, intelligent recognition of grouting fullness at inspection sites is achieved through an ensemble soft voting classifier (ESVC) trained on collected endoscopic images. This method delivers high detection accuracy while reducing the subjectivity of manual interpretation. Based on accurate localisation of the upper surface of the grout and variations in endoscope insertion depth, the proposed approach can effectively quantify the fullness index under incomplete grouting conditions. This detection process is expected to shift traditional sleeve grouting quality control from 'post-detection' to 'process monitoring', providing essential data support for timely repair decisions and ensuring grouting continuity. The approach establishes an effective closed loop between detection and repair. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 10589759 |
| DOI: | 10.1080/10589759.2025.2518593 |
