期刊名称：Journal of Emerging Trends in Computing and Information Sciences
电子版ISSN：2079-8407
出版年度：2014
卷号：5
期号：7
页码：527-531
出版社：ARPN Publishers
摘要：In Korea, damaged pavement like potholes are increasing due to the climate change such as heavy rains and snow, and thus complaints and lawsuits of accidents related to potholes are growing. In order to solve this problem, a new research is being conducted for developing a technology which can detect and recognize potholes based on images, which may contribute to improvement of survey efficiency and pavement quality by prior investigation and immediate action. The purpose of this study is to develop pothole 2D DB based on 2D images and video data as a pre-study of developing a technology detecting and recognizing potholes based on images. In this study, we collect images of potholes through video recording on a moving vehicle, categorize potholes by type of shape and road class based on collected images, and develop a query system for the built pothole 2D DB. The obtained pothole video data and a query system for pothole 2D DB will be utilized for testing algorithms that will be developed to detect and recognize potholes based on images and video data.
关键词：Pothole; 2D images; Video data; DB; Query system
