Identification of vegetated landslides using only a LIDAR-based Terrain Model and derivatives in an object oriented environment

Light Detection and Ranging (LiDAR) and its derivative products have become a powerful tool in landslide research, particularly for landslide identification and landslide inventory mapping. In contrast to the many studies that use expert-based analysis of LiDAR derivatives to identify landslides, only few studies, all pixel-based, have attempted to develop computer-aided methods for extracting landslides from LiDAR. It has not been tested whether object-oriented analysis (OOA) could be an alternative, even though OOA has gained increased attention for (semi-)automated landslide identification from monotemporal and multitemporal passive optical airborne and satellite sensor data, and using either multispectral or panchromatic imagery. Elevation data has been used before in OOA-based landslide detection work (e.g. Martha et al., 2010), but only to support the removal of false positives and the discrimination of different landslide types. The objective of this study is to exploit OOA for landslide inventory mapping using single pulse LiDAR data for both the segmentation and classification steps, without the support of any spectral information. We focus on (1) cognitive perception of landslides as geomorphic features consisting of parts with different morphometric characteristics, (2) testing existing scale optimization methods, and (3) selection of morphometric indicators for the identification of different landslide features. We further discuss the possibilities and limitations of LiDAR-based identification of landslides covered by vegetation. The Flemish Ardennes (Belgium), a hilly region characterised by loose tertiary lithology and affected by more than 200 landslides is selected as study area.status: publishe