Localization and Pixel-Confidence Network for Surface Defect Segmentation

Surface defect segmentation based on deep learning has been widely applied in industrial inspection. However, two major challenges persist in specific application scenarios: first, the imbalanced area distribution between defects and the background leads to degraded segmentation performance; second,...

Full description

Saved in:
Bibliographic Details
Main Authors: Yueyou Wang, Zixuan Xu, Li Mei, Ruiqing Guo, Jing Zhang, Tingbo Zhang, Hongqi Liu
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/15/4548
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Surface defect segmentation based on deep learning has been widely applied in industrial inspection. However, two major challenges persist in specific application scenarios: first, the imbalanced area distribution between defects and the background leads to degraded segmentation performance; second, fine gaps within defects are prone to over-segmentation. To address these issues, this study proposes a two-stage image segmentation network that integrates a Defect Localization Module and a Pixel Confidence Module. In the first stage, the Defect Localization Module performs a coarse localization of defect regions and embeds the resulting feature vectors into the backbone of the second stage. In the second stage, the Pixel Confidence Module captures the probabilistic distribution of neighboring pixels, thereby refining the initial predictions. Experimental results demonstrate that the improved network achieves gains of <inline-formula><math display="inline"><semantics><mrow><mn>1.58</mn><mo>%</mo><mo>±</mo><mn>0.80</mn><mo>%</mo></mrow></semantics></math></inline-formula> in <inline-formula><math display="inline"><semantics><mrow><mi>m</mi><mi>P</mi><mi>A</mi></mrow></semantics></math></inline-formula>, <inline-formula><math display="inline"><semantics><mrow><mn>1.35</mn><mo>%</mo><mo>±</mo><mn>0.77</mn><mo>%</mo></mrow></semantics></math></inline-formula> in <inline-formula><math display="inline"><semantics><mrow><mi>m</mi><mi>I</mi><mi>o</mi><mi>U</mi></mrow></semantics></math></inline-formula> on the self-built Carbon Fabric Defect Dataset and <inline-formula><math display="inline"><semantics><mrow><mn>2.66</mn><mo>%</mo><mo>±</mo><mn>1.12</mn><mo>%</mo></mrow></semantics></math></inline-formula> in <inline-formula><math display="inline"><semantics><mrow><mi>m</mi><mi>P</mi><mi>A</mi></mrow></semantics></math></inline-formula>, <inline-formula><math display="inline"><semantics><mrow><mn>1.44</mn><mo>%</mo><mo>±</mo><mn>0.79</mn><mo>%</mo></mrow></semantics></math></inline-formula> in <inline-formula><math display="inline"><semantics><mrow><mi>m</mi><mi>I</mi><mi>o</mi><mi>U</mi></mrow></semantics></math></inline-formula> on the public Magnetic Tile Defect Dataset compared to the other network. These enhancements translate to more reliable automated quality assurance in industrial production environments.
ISSN:1424-8220