Automated Identification of Geomorphological and Topographical Features of the Seabed by Object Based Image Analysis of Digital Terrain Models

作者：Siakavara K^*; Argialas D

来源：Zeitschrift fur Geomorphologie, 2013, 57(3): 63-74.

DOI：10.1127/0372-8854/2013/S-00144

• seabed
• bedforms
• object based image analysis
• multibeam echo sounder

Seabed feature identification is of great importance in many engineering and geomorphological projects. Multibeam echo sounder datasets are often used in those projects for tasks such as producing Digital Terrain Models (DTM). An effort in automating the interpretation and delineation of seabed geomorphological and topographical features from a DTM and its derivatives using an object based image analysis approach was developed. The approach consists of four basic steps. First objects were created through multiresolution segmentation of the DTM. Then the classes of geomorphological features were initially defined and were successively refined through the multiresolution analysis procedure. The final classes were smooth ground, channels, scars, mounds, ridges and landslide scars. Subsequently each object was assigned to one of the classes representing those bedforms by fuzzy classification. Finally in the fourth step a visual qualitative evaluation of the results was performed.

• 出版日期2013-9