Improvement of road crack detection processes based on image processing technologies in korea

초록

This paper presents a framework for detecting road cracks using image processing techniques. Currently, the road crack detection process in Korea is manual and labor-intensive. A special vehicle equipped with a dedicated camera acquires road image data. Workers then manually examine the image data to identify road cracks and classify the crack types. We propose an automated framework for analyzing image data in order to minimize costs and errors due to the manual process. This research explores the applicability of existing methodologies in the aforementioned domestic environments. When the methodology was applied to Korea, limitations caused by environmental differences have been identified. In this regard, we propose an improved and extended framework to complement them. The proposed framework contributes to minimizing time, cost and errors by replacing the manual process. © 2019 ICIC International.

키워드