Rail surface defect detection is vital for railway safety. Traditional methods falter with varying defect sizes and complex backgrounds, while two-stage deep learning models, though accurate, lack real-time capabilities. To overcome these challenges, we propose an enhanced one-stage detection model based on CenterNet. We replace ResNet with ResNeXt and implement a multi-branch structure for better low-level feature extraction. Additionally, we integrate SKNet attention mechanism with the C2f structure from YOLOv8, improving the model’s focus on critical image regions and enhancing the detection of minor defects. We also introduce an elliptical Gaussian kernel for size regression loss, better representing the aspect ratio of rail defects. This approach enhances detection accuracy and speeds up training. Our model achieves a mean accuracy (mAP) of 0.952 on the rail defects dataset, outperforming other models with a 6.6% improvement over the original and a 35.5% increase in training speed. These results demonstrate the efficiency and reliability of our method for rail defect detection.

Rail-fastening components are essential for ensuring the safety of urban rail systems by securing rails to sleepers. Traditional inspection methods rely heavily on manual labor and are inefficient. This paper introduces a novel approach to address these inefficiencies and the challenges faced by computer vision-based inspections, such as missed detections due to imbalanced samples and limitations in conventional image segmentation techniques. Our approach transitions the industry’s focus from qualitative to a more precise quantitative analysis of rail-fastening components. We propose Mask-FRCN, an advanced image segmentation network that incorporates three key technological enhancements: the fully refined convolutional network module (FRCN),which refines the segmentation boundaries for SFC-type fasteners; the Channel-WiseKnowledge Distillation (CWD) algorithm, which boosts the model’s inference efficiency; and the FCRM methodology, which enhances the extraction capabilities for features specific to SFC-type fasteners. Furthermore, we introduce a fastener system inspection and quantization method based on the Mask FRCN method (FIQ), a novel technique for quantifying the condition of components by using image features, template matching with random forests, and a clustering calculation method derived from segmentation results. Experimental results validate that our method significantly surpasses existing techniques in accuracy, thereby offering a more efficient solution for inspecting rail-fastening components. The enhanced Mask-FRCN achieves a segmentation accuracy of 96.01% and a reduced network size of 36.1 M. Additionally, the FIQ method improves fault detection accuracy for SFC-type fasteners to 95.13%, demonstrating the efficacy and efficiency of our innovative approach.

Detecting pantographs remains a challenging task due to complex scenes, variable weather conditions, and noise interference. Existing pantograph detection methods struggle to effectively segment the complete shape of the pantograph from intricate backgrounds and adverse weather, and they often exhibit inadequate real-time performance. To address these challenges, we propose a novel pantograph segmentation method that leverages a deep learning multi-scale strip pooling attention mechanism. Our approach utilizes the PidNet semantic segmentation network as the baseline architecture, while we introduce a newly designed multi-scale strip pooling attention mechanism specifically for the detail extraction branch. The multi-scale strip convolution branch effectively extracts the pantograph pixel-level detail features, while the pooling branch effectively extracts the macroscopic features of the pantograph. The unique linear interpolation method effectively mitigates the influence of weather, enhancing segmentation accuracy while maintaining a lightweight structure. In the context aggregation branch, a multi-scale context aggregation module utilizing gated convolution has been developed to replace the original network’s module, which possesses strong pantograph positioning capabilities. In comparison to existing pantograph detection methods, our model demonstrates the ability to accurately segment the pantograph with a clearly defined shape, effectively filter out extraneous background noise, and exhibit high robustness to variations in illumination and weather conditions. In addition, a rich pantograph dataset was created, including various scenarios and weather conditions, which also enhanced the robustness of the model. When the IOU and accuracy are 92.91% and 96.04%, respectively, the inference speed can still exceed 30 FPS on a single 2080Ti GPU.