Study on Three-dimensional Image Inspection System for Wheelset Treads
As the only contact surface between the wheel and the rail,the automatic detection of wheel size and tread condition is one of the important prerequisites to ensure the safety of train operation.Aiming at the problems of low detection accuracy,high false alarm rate,high environmental dependence and low manual efficiency of the two-dimensional image detection system of wheelset treads and manual measurement of wheelset dimensions,this paper develops a three-dimensional image detection system of treads with the accompanying function of wheelset dimensions detection.The system determines the depth information of the measured object based on the laser triangulation method,accurately extracts each tread curve through the improved Gaussian optical center extraction algorithm,and ultimately restores the three-dimensional condition information of the tread of the wheelset from 1,800 tread curves.Through a large number of fault simulation tests,it is found that the system is not only able to accurately identify the simulated faults,but also able to exclude the interference of the unchanged depth information such as stains and water stains on the wheelset,with recognition rate reaching more than 95％,which provides a set of more applicable three-dimensional detection system for the condition detection of the tread as well as wheelset dimension detection.
万方数据
维普
