AN ATTEMPT TO IDENTIFY THE ROCKFALL DANGER SECTION OF ROAD USING ARTIFICIAL INTELLIGENCE AND DRIVE RECORDER

Abstract

Since the road disaster prevention inspection in 1996, inspection areas for rockfall and collapse have been set, and rockfall accidents have been greatly reduced due to intensive disaster prevention measures. However, small-scale rockfalls occur frequently on mountain roads, and property damage accidents and road closures still occur even if they do not lead to serious accidents.

In this research, in order to propose the effectiveness of the AI method in road maintenance, the section where rockfall progresses from the cumulative tendency by counting the rockfalls that remain along the road or on the rockfall protection net from the video of the drive recorder.