An image processing system for health monitoring of road pavement

点击翻译按钮获取中文摘要

This research presents an image processing system for health monitoring of roadpavements. As the first step, an image filter to detect cracks accurately from variouskinds of images of road pavement. Then, in order to utilize the filter effectively, wepropose two methods as pre-processes for applying the filter. One is a method to classifytype of pavement (concrete or asphalt) by using maximum likelihood classifier. Anotheris a method to recognize the region where cars might mainly pass from the images. Finally,these methods are combined into one system and it enables us to automatically evaluatehealth level of the pavement from the images and indicate points where maintenance is required.