19 December 2008 Pattern geomorphologic analysis for soil erosion study
Author Affiliations +
Abstract
The geologist try to understand relationship between soil erosion observed and natural landscape structure. Erosion can effectively appears in the vicinity of linear or planar structures of soil (lines, faults or materials change). Once eroded areas are mapped, an inventory of relief linear shapes is done. The crossing geomorphological analysis with other environmental parameters allows to predict the becoming eroded areas. Lineaments detection is usually made by photointerpretation. DEM (Digital Elevation Model) visual analysis is another alternative but not sufficient, so it uses the derived models from DEM called hillshade images. The DEM is lighted up by a virtual source with a direction and height incidence. A good study require a complete lightings visual interpretation which is very slow and subjective. This paper propose an automatic process that help geologist to detect and analyse the geomorphological structures present in the landscape by using image analysis methods. This study focus on lines and catchments basins structures. First a new watershed and catchments basins segmentation method is developed it defines an attractive structure between pixels (based on path of steepest slope). After these lines are automatically extracted by Hough transform and their preferential direction is analysed by a technique called directions rose. Some results are given on DEM and Hillshade images for a particular areas of the main New Caledonia island where soil erosion is a serious problem mainly due to tropical weather (violent rains) and human activities (mining, bush fire) on the weathered rocks (laterites) in mountain.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Nazha Selmaoui, Nazha Selmaoui, Isabelle Rouet, Isabelle Rouet, Mélanie Mahot, Mélanie Mahot, } "Pattern geomorphologic analysis for soil erosion study", Proc. SPIE 7150, Remote Sensing of Inland, Coastal, and Oceanic Waters, 71501L (19 December 2008); doi: 10.1117/12.804884; https://doi.org/10.1117/12.804884
PROCEEDINGS
9 PAGES


SHARE
Back to Top