Convolutional Neural network based Online Rail surface Crack Detection
Chitturi Akhila, Cheemala Astha Diamond, A. Mary Posonia
Abstract
The proposed system utilizes CNN to detect the faults in the railway tracks with the help of images. Their characteristics are obtained to extract the default railway track. This model helps to reduce the manual inspection work. While performing the task, it lowers the time and cost utilization and increases the safety of passengers. To train and validate the CNN algorithm, the ground truth databases on images of masks are utilized. To this end, this CNN maintains the both space and time coherence of the defects in the tracks, and the final output of the false prediction effectively.
Topics & Concepts
Convolutional neural networkComputer scienceGround truthTask (project management)Artificial intelligenceTrack (disk drive)Computer visionReal-time computingPattern recognition (psychology)EngineeringSystems engineeringOperating systemInfrastructure Maintenance and MonitoringRailway Engineering and DynamicsSurface Roughness and Optical Measurements