AN ALGORITHM FOR THE DETECTION AND MEASUREMENT OF RAIL SURFACE-DEFECTS

Defects on the surface of railroad tracks have been the cause of growi ng concern over the past three decades. The automated detection and cl assification of rail surface defects would be of great assistance to r ail maintenance planners, who develop grinding strategies to prevent t he development of potentially dangerous deterioration. Videotaped imag es of the surface of rail have been obtained, but they are subject to distortions due to the acquisition process as well as physical phenome na on the track itself. In this analysis, an algorithm is presented fo r the simultaneous restoration and segmentation of objects in a two-di mensional image. The algorithm relies on distributions that model the relationships between sites and neighbors in order to restore a distor ted image to an estimate of its ideal form, and also obtain detailed i nformation about the objects located in the image. The foundation of t he algorithm is the Iterated Conditional Modes procedure for image res toration. The resulting extension is capable of providing detailed mea surements of the geometric features of objects detected in an image. T he extended algorithm is applied to an image distorted by simulated no ise, and also to an image taken from a videotape of a rail surface. Th e results of the analysis demonstrate the potential for accurate detec tion, measurement, and classification of rail surface defects.