Real-time landslide monitoring using close-range stereo image sequences analysis

Abstract

Landslide is one of the major natural hazards, threatening human lives and properties. Real-time landslide monitoring is very important to disaster relief. In this paper, the effectiveness of real-time landslide monitoring method based on close-range photogrammetry is investigated in an on-campus landslide test site at Tongji University. Feature point tracking method using landslide image sequences is tested first, and then the elevation change detection method on the landslide surface is also tested. According to the experimental results, it is shown that both of these two methods can effectively monitor the landslide movements, which can provide important pre-experimental results for the real-time landslide monitoring in Hongkou Town, Sichun, China.