Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGG
Fiber-flaw detection on pill surfaces is a critical yet challenging task in industrial pharmacy due to diverse defect characteristics. To overcome the limitations of traditional methods in accuracy and real-time performance, this study introduces SA-MGhost-DVGG, a novel lightweight network for enhan...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Computers |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-431X/14/5/200 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850127429202345984 |
|---|---|
| author | Jipei Lou Hongyi Wang Haodong Liang Ziwei Wu |
| author_facet | Jipei Lou Hongyi Wang Haodong Liang Ziwei Wu |
| author_sort | Jipei Lou |
| collection | DOAJ |
| description | Fiber-flaw detection on pill surfaces is a critical yet challenging task in industrial pharmacy due to diverse defect characteristics. To overcome the limitations of traditional methods in accuracy and real-time performance, this study introduces SA-MGhost-DVGG, a novel lightweight network for enhanced detection. The proposed network integrates an MGhost module for reducing parameters and computational load, a mixed-channel spatial attention (SA) module to refine features specific to fiber regions, and depthwise separable convolutions (DepSepConv) for efficient dimensionality reduction while preserving feature information. Experimental evaluations demonstrate that SA-MGhost-DVGG achieves a mean detection accuracy of 99.01% with an average inference time of 2.23 ms per pill. The findings confirm that SA-MGhost-DVGG effectively balances high accuracy with computational efficiency, offering a robust solution for industrial applications. |
| format | Article |
| id | doaj-art-f5d899c2f69e454f867e7a5ed4ca2cfd |
| institution | OA Journals |
| issn | 2073-431X |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Computers |
| spelling | doaj-art-f5d899c2f69e454f867e7a5ed4ca2cfd2025-08-20T02:33:42ZengMDPI AGComputers2073-431X2025-05-0114520010.3390/computers14050200Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGGJipei Lou0Hongyi Wang1Haodong Liang2Ziwei Wu3Tianjin Key Laboratory of Intelligent Control of Electrical Equipment, School of Control Science and Engineering, Tiangong University, Tianjin 300387, ChinaTianjin Key Laboratory of Intelligent Control of Electrical Equipment, School of Artificial Intelligence, Tiangong University, Tianjin 300387, ChinaTianjin Key Laboratory of Intelligent Control of Electrical Equipment, School of Artificial Intelligence, Tiangong University, Tianjin 300387, ChinaTianjin Key Laboratory of Intelligent Control of Electrical Equipment, School of Artificial Intelligence, Tiangong University, Tianjin 300387, ChinaFiber-flaw detection on pill surfaces is a critical yet challenging task in industrial pharmacy due to diverse defect characteristics. To overcome the limitations of traditional methods in accuracy and real-time performance, this study introduces SA-MGhost-DVGG, a novel lightweight network for enhanced detection. The proposed network integrates an MGhost module for reducing parameters and computational load, a mixed-channel spatial attention (SA) module to refine features specific to fiber regions, and depthwise separable convolutions (DepSepConv) for efficient dimensionality reduction while preserving feature information. Experimental evaluations demonstrate that SA-MGhost-DVGG achieves a mean detection accuracy of 99.01% with an average inference time of 2.23 ms per pill. The findings confirm that SA-MGhost-DVGG effectively balances high accuracy with computational efficiency, offering a robust solution for industrial applications.https://www.mdpi.com/2073-431X/14/5/200flaw detectionfiber-flawpill defectimage processinglightweight network |
| spellingShingle | Jipei Lou Hongyi Wang Haodong Liang Ziwei Wu Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGG Computers flaw detection fiber-flaw pill defect image processing lightweight network |
| title | Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGG |
| title_full | Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGG |
| title_fullStr | Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGG |
| title_full_unstemmed | Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGG |
| title_short | Detection of Fiber-Flaw on Pill Surface Based on Lightweight Network SA-MGhost-DVGG |
| title_sort | detection of fiber flaw on pill surface based on lightweight network sa mghost dvgg |
| topic | flaw detection fiber-flaw pill defect image processing lightweight network |
| url | https://www.mdpi.com/2073-431X/14/5/200 |
| work_keys_str_mv | AT jipeilou detectionoffiberflawonpillsurfacebasedonlightweightnetworksamghostdvgg AT hongyiwang detectionoffiberflawonpillsurfacebasedonlightweightnetworksamghostdvgg AT haodongliang detectionoffiberflawonpillsurfacebasedonlightweightnetworksamghostdvgg AT ziweiwu detectionoffiberflawonpillsurfacebasedonlightweightnetworksamghostdvgg |