Open Access Open Access  Restricted Access Subscription Access

Detection of Coastal Landforms in a Deltaic Area Using a Multi-Scale Object-Based Classification Method


Affiliations
1 Geosciences Group, National Remote Sensing Centre, Indian Space Research Organization, Hyderabad 500 037, India
2 Department of Geo-Engineering, Andhra University College of Engineering (A), Vishakhapatnam 530 003, India
3 Department of Physical Sciences, Mahatma Gandhi Chitrakoot Gramodaya Vishwavidyalaya, Chitrakoot 485 780, India
 

Coastal landforms play an important role in protecting deltaic areas from erosion due to the action of waves. However, landforms in the deltas are dynamic and vulnerable to changes due to the effect of natural disasters like floods and cyclones. Automatic detection of dynamic landforms from satellite data can provide important inputs for effective coastal zone management. In this study, we developed an Object-Based Image Analysis (OBIA) technique to identify and map landforms in the Krishna delta, east coast of India using Resourcesat-2 LISS-IV multispectral image (5.8 m) and digital elevation model (DEM) (4 m). Since landforms are represented at multiple scales, the plateau objective function method was used to select appropriate scales during multiresolution segmentation. Knowledge-based rules in OBIA, using the parameters tone, texture, shape and context derived from satellite images and height from DEM were developed for classification of landforms. A total of 11 landforms (beach, beach ridge, swale, tidal creek, marsh, spit, barrier bar, mangrove, natural levee, channel island and channel bar) were mapped using this approach. High detection accuracy of these landforms indicates that the method developed has the potential for geomorphological mapping of dynamic landforms in low lying deltaic areas.

Keywords

Beach, Cyclone, DEM, Image Segmentation, Mangrove, OBIA, Resourcesat-2.
User
Notifications
Font Size


  • Detection of Coastal Landforms in a Deltaic Area Using a Multi-Scale Object-Based Classification Method

Abstract Views: 457  |  PDF Views: 183

Authors

Tapas R. Martha
Geosciences Group, National Remote Sensing Centre, Indian Space Research Organization, Hyderabad 500 037, India
A. Mohan Vamsee
Department of Geo-Engineering, Andhra University College of Engineering (A), Vishakhapatnam 530 003, India
Vikas Tripathi
Department of Physical Sciences, Mahatma Gandhi Chitrakoot Gramodaya Vishwavidyalaya, Chitrakoot 485 780, India
K. Vinod Kumar
Geosciences Group, National Remote Sensing Centre, Indian Space Research Organization, Hyderabad 500 037, India

Abstract


Coastal landforms play an important role in protecting deltaic areas from erosion due to the action of waves. However, landforms in the deltas are dynamic and vulnerable to changes due to the effect of natural disasters like floods and cyclones. Automatic detection of dynamic landforms from satellite data can provide important inputs for effective coastal zone management. In this study, we developed an Object-Based Image Analysis (OBIA) technique to identify and map landforms in the Krishna delta, east coast of India using Resourcesat-2 LISS-IV multispectral image (5.8 m) and digital elevation model (DEM) (4 m). Since landforms are represented at multiple scales, the plateau objective function method was used to select appropriate scales during multiresolution segmentation. Knowledge-based rules in OBIA, using the parameters tone, texture, shape and context derived from satellite images and height from DEM were developed for classification of landforms. A total of 11 landforms (beach, beach ridge, swale, tidal creek, marsh, spit, barrier bar, mangrove, natural levee, channel island and channel bar) were mapped using this approach. High detection accuracy of these landforms indicates that the method developed has the potential for geomorphological mapping of dynamic landforms in low lying deltaic areas.

Keywords


Beach, Cyclone, DEM, Image Segmentation, Mangrove, OBIA, Resourcesat-2.

References





DOI: https://doi.org/10.18520/cs%2Fv114%2Fi06%2F1338-1345