Health Detection of Railway Fastener Based on Data Mining Technology

In order to improve the health detection results of railway fastener, a salient re-gional feature model of railway fastener based on data mining technology is de-signed to solve the problems of large error and poor detection efficiency. Railway fasteners are key components of railway tracks, and their condition directly de-termines the safety of the trains passing over them. In this paper, a railway fas-tener state detection algorithm based on a high-speed digital camera is presented. First, we use a spectral residual algorithm to determine the significant region of the railway fastener, and then extract features using the SIFT algorithm. After converting the extracted features to a Fisher vector, the LIBLINEAR classifier is applied to determine the state of each railway fastener. Experimental results show that the proposed algorithm achieves a high recognition rate and is robust to var-iations in weather and lighting conditions.