Abstract:Defect detection is essential for the quality control and repair decision-making of various products. Due to collisions, uneven stress, welding parameters and other factors, cracks form on the surface or inside of products, which affect the product appearance and mechanism strength and may even cause huge safety accidents. Nondestructive testing (NDT) is an effective and practical method for accurate defect detection, but it still faces various challenges against complex factors, such as complex backgrounds, poor contrast, weak texture, and class imbalance issues. Recently, deep learning has rapidly improved the performance of automatic defect detection with the strong feature expression ability of deep convolutional neural networks (DCNNs). However, various limitations remain due to the insufficient processing of local contextual features, which affects the detection precision. To address this issue, with the encoder–decoder network structure, a novel nondestructive defect detection network, namely, NDD-Net, is proposed in this paper to construct an end-to-end nondestructive defect segmentation scheme. To make the segmentation network better emphasize the defect areas, an attention fusion block (AFB) is proposed to replace the raw skip connections to acquire more discriminative features and enhance the segmentation performance on microdefects. Meanwhile, by fusing a dense connection convolution network and a residual network, a residual dense connection convolution block (RDCCB) is also proposed to be embedded into the proposed segmentation network to acquire richer information about the local feature maps. Two public datasets with severe class imbalance issues are adopted for model evaluation: the Grima X-ray (GDXray) database and the rail surface discrete defects (RSSDs) dataset. Experimental results show that the proposed segmentation network outperforms other related segmentation models.

Region and Edge-Aware Network for Rail Surface Defect Segmentation

Hybrid deep learning architecture for rail surface segmentation and surface defect detection

RSDNet: A New Multiscale Rail Surface Defect Detection Model

Research on deep learning method for rail surface defect detection

A Decoder-Free Reconstruction Method for Semi-Supervised Surface Defect Detection

Rail Surface Defect Detection Using A Transformer-Based Network

Segmentation of Track Surface Defects Based on Machine Vision and Neural Networks

A Sub-Region Unet for Weak Defects Segmentation with Global Information and Mask-Aware Loss.

Automatic Rail Surface Defects Inspection Based on Mask R-CNN

ISRM:Introspective Self-supervised Reconstruction Model for Rail Surface Defect Detection and Segmentation

RBNet: An Ultra Fast Rendering-based Architecture for Railway Defects Segmentation

Method for Rail Surface Defect Detection Based on Neural Network Architecture Search

Rail surface defect detection based on improved Mask R-CNN

A nondestructive automatic defect detection method with pixelwise segmentation

Uncertainty inspired domain adaptation network for rail surface defect segmentation

Two Deep Learning Networks for Rail Surface Defect Inspection of Limited Samples With Line-Level Label

FS-RSDD: Few-Shot Rail Surface Defect Detection with Prototype Learning

A Rail Surface Defect Detection Method Based on Pyramid Feature and Lightweight Convolutional Neural Network

YOLOv8n-RSDD: A High-Performance Low-Complexity Rail Surface Defect Detection Network

Normal-Knowledge-Based Pavement Defect Segmentation Using Relevance-Aware and Cross-Reasoning Mechanisms

Mcnet: Multiple Context Information Segmentation Network Of No-Service Rail Surface Defects