An image-level weakly supervised segmentation method for No-service rail surface defect with size prior. (15th February 2022)

• Record Type: Journal Article
• Authors: Zhang, Defu; Song, Kechen; Xu, Jing; Dong, Hongwen; Yan, Yunhui
• Is Part Of: Mechanical systems and signal processing. Volume 165(2022)
• Publication Date: 2022-02-15
• Subjects: Surface defect segmentation -- No-service rail -- Image-level weakly supervised -- Size prior information -- Pooling combination module; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2021.108334
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: A category is decomposed with the size prior for weakly supervised segmentation. A pooling combination module makes full use of the attributes of sub-category. Pseudo labels are generated by classification network to train segmentation network. Our designed method and MCnet obtain the best performance on the NRSD-MN. Abstract: Surface defect segmentation of no-service rail is important for its quality assessment. Several deep learning methods have obtained good performances based on computer vision testing. However, these methods all require a large number of training samples with pixel-level labels. Marking pixel-level labels is a difficult task – time-consuming, labor-consuming, and professionals. To address the issue, a novel image-level weakly supervised segmentation formulation is proposed for no-service rail surface defects. These defects are decomposed into three sub-categories (strip-shaped, spot-shaped, block-shaped) according to the size prior information (area and shape). Then, a method is presented with a pooling combination module. The pooling combination module makes full use of the size attributes of the sub-category by utilizing different pooling functions for different sub-categories. Experimental results demonstrate that our method is effective and outperforms the state-of-the-art methods. Index terms - Surface defect segmentation; no-service rail; image-level weakly supervised; size prior information; pooling combination module.