Mangrove Mapping of Ratnagiri Coast using Different Classification Techniques

Abstract

Mangrove coverage of Ratnagiri coast was mapped by using different techniques such as supervised classification, supervised classification of principal components, unsupervised classification and unsu- pervised classification of principal components and overall classification accuracy ranged from 79.46- 86.19, 82-89, 84.52-89 and 89-93% respectively. The kappa co-efficient for supervised classification, supervised classification of principal components, unsupervised classification and unsupervised clas- sification of principal components were 0.74-0.82, 0.78-0.86, 0.81-0.87 and 0.87-0.90 respectively. Over- all classification accuracy achieved by unsupervised classification of principal components technique was comparatively better than overall classification accuracy achieved by other techniques. Thus this technique is found appropriate for mapping the mangrove coverage in the Ratnagiri block.