Railway Fastener Anomaly Detection via Multisensor Fusion and Self-Driven Loss Reweighting
Yang Gao, Zhiwei Cao, Yong Qin, Xuanyu Ge, Lirong Lian, Jie Bai, Hang Yu
Abstract
Fasteners are a critical part of the rail and are used to fix the rail, which are important for train operation. Rail vibrations during train operation can cause anomalies in fasteners. At present, the 3-D structured light camera is used to detect anomalies on railway sites, but there is a lack of sufficient mining of 3-D data and effective fusion of multisensor data. To address this issue, this article proposes a novel approach for railway fastener anomaly detection via multisensor fusion and self-driven loss reweighting. First, a pixel-level attention mechanism multisensor fusion method is applied, where the depth map is used as an attention factor to highlight edge contours and enhance abrupt changes in the gray level. Second, a feature fusion-decoupled module is proposed to obtain the dense feature maps and then decouple the detection task to output class, location, and confidence. Finally, given the characteristics of the sample imbalance in the fastener dataset, a dynamic self-driven loss reweighting method is used to improve the detection accuracy of difficult samples. The experimental results show that the proposed method can achieve 86.6% precision and 61.72-FPS detection speed, better than other state-of-the-art algorithms.