Analysis of Tsunami-Affected and Reconstructed Areas in Nagapattinam Using Remote Sensing

Abstract

The recent experiences in the 2004 Indian Ocean tsunami showed the efficiency of remote sensing techniques in quick damage mapping and recovery efforts. This study develops an object-based image analysis method for mapping the tsunami-affected areas. The proposed method is carried out in four steps. In the first step, data preprocessing operations, data registration, and cloud removal are carried out based on the multi-source data. In the second step, object-based classification operation is performed to separate the images into segments. In the third step, infrastructural damages are identified. In the fourth step, developments after the tsunami are found out. The results showed that the proposed method has given a higher accuracy than the earlier methods. The proposed combination of techniques gives a high accuracy of 99% for the detection of changes in vegetation. The change detection result found 39% change in the building in 2005, 32% of soil area and 8.7% of water region had been increased within 2 years. The building elevation method estimated newly developed one story and two story buildings as 11,112 and 142, respectively.