|
|
|
Abstract
Multi Sensor Change Detection in Shallow Coral Reef Environments using IKONOS & Landsat(TM) Imagery Systems
Nima Pahlevan
M.Sc. student
Khaje-Nasire-Toosi University Of Technology, Faculty of Geodesy & Geomatics, Iran
n_p60@yahoo.com
Dr. A. Alimohamadi
Dean of GIS Department
Khaje-Nasire-Toosi University Of Technology, Iran
alimohamadi@kntu.ac.ir
Dr. M.J. Valadanzouj
Remote sensing professor
Khaje-Nasire-Toosi University Of Technology, Iran
valadanzouj@kntu@ac.ir
Abstract :
Remote sensing applications in coastal and coral reef environments have been developed over the past 3 decades to map characteristics of aquatic ,from the water surface ,to water column constituents and substrate cover type. the successful lunch of IKONOS has extensively overcome the spatial constraints of existing sensors, particularly when the geographical scope of interest located in patchy coral areas.
The study were conducted in the eastern part of Kish Island which are ideally suited for remote sensing due to clear, shallow water(0-7m) containing not so complex number of habitats. Three time series of Landsat(TM),1992, and two images of IKONOS ,2001 & 2004, were investigated. During field survey over 200 points were sampled for defining habitat categories, ground-truthing and accuracy assessment. Pre-processing included Geometric, Radiometric and water column correction. Lacking in situ measurement of atmospheric conditions, removal of Atmospheric effects was implemented by to different methods. The first method used relative atmospheric correction by applying Pseudo-invariant objects and the latter extract atmospheric condition utilizing MODIS Data. After conducting water column correction using Lyzenga's method, Supervised classification was implemented applying Mahalanobis distance classifier(due to subsequence band correlation) and 7 separate classes were discriminated in each time acquisition. At the end, pixel based image differencing method applied to three different images to reveal changes. We concluded that there was a huge loss of corals in the region detected with acceptable level of confidence by means of high & moderate spatial resolution comparison.
|
|
|