Research on Surface Defect Detection Based on Semantic Segmentation

Yu-ting LIU, Ya-ning YANG, Chao WANG, Xiang-yu XU, Tao ZHANG

Abstract


Surface defect detection plays an important role in ensuring product quality. In view of the problem of surface defect detection in industrial production, a surface defect detection method based on semantic segmentation is introduced, which uses the idea of transfer learning. A better network model can be trained by using fewer defect samples. In addition, the defect can be classified by this method, and the defect type can be labeled, and the defect area can be obtained. In order to verify the effectiveness of the proposed method, the performance of the method is analyzed by DAGM 2007 dataset. The experimental results show that the defect detection accuracy of this method is more than 99.6% and meets the practical requirements of industrial production.

Keywords


Defect Detection, Semantic Segmentation, Fully Convolutional Networks


DOI
10.12783/dtcse/aicae2019/31504

Full Text:

PDF

Refbacks

  • There are currently no refbacks.