University of Twente Student Theses

Login

Object-oriented method for postevent Landslide mapping: the utility of Texture

Harris, Melanie Tramaine Marian (2010) Object-oriented method for postevent Landslide mapping: the utility of Texture.

[img] PDF
2MB
Abstract:In the immediate aftermath of an event, the most pressing issue is how to collect and disseminate accurate information to aid disaster managers in organising rescue, emergency and damage assessment operations in the field. Post-event hazard information is important and it needs to be produced as quickly and as accurately as possible; therefore, inventories need to be generated rapidly from available satellite imagery and other data. Object-oriented analysis (OOA) has been proven useful for identifying landslides, however, a gap in knowledge surrounding the utility and systematic application of texture measures for landslide detection. Thus this research sought to evaluate and assess the utility of object-based texture analysis for identifying landslides. To do this a series of tests were set up to compare actual values of pixel vs. object-based texture calculations, determine how a single texture value is arrived at for each object, examine the factors affecting texture calculations (band, scale factor, directionality) and determine which texture measure is most useful for landslide mapping. Finally, an object-based procedure was developed and applied in a test site (Wenchuan County, Sichuan Province, China) to test what was learnt in the evaluation of object-based texture. The results show that texture cannot be used in isolation to distinguish landslides from its false positives such as bare ground and settlement areas. If however it is used in conjunction with spectral and morphological data identification and ultimately classification results can be enhanced. Keywords: GLCM Texture, OOA, landslides
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/92529
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page