Home Articles Plantation type differentiation from fused IRS data – An analysis

Plantation type differentiation from fused IRS data – An analysis

Plantation type differentiation from fused IRS data – An analysis

S. P. S. Kushwaha, P. Rocky and S. Leima
Indian Institute of Remote Sensing, Dehradun – 248 001

The methodology
developed to assess biodiversity over the Western Ghats using remote sensing,
involves a synthesis of information derived at several spatial
scales.




This study deals with the plantation type differentiation and mapping from fused IRS LISS-III and PAN data in Pathri Reserve Forest, Hardwar. A false colour composite was generated from above data using intensity, hue and saturation transformation and replacement of intensity image by PAN image. Pathri Reserve Forest is an entirely planted taungya forest having large scale mixed plantations of sissoo, khair, eucalyptus, mulberry, amaltas and teak trees. Both LISS-III and LISS-III+PAN images were selected for comparative plantation types discrimination through visual interpretation (onscreen digitization). This analysis resulted in two maps showing different plantations on 1:75,000 scale. It was observed that, in spite of inherent differences, the two images facilitated in differentiation and mapping of exactly same number of classes. However, significant plantation boundary refinement was noticed on fused image for all categories. This was very conspicuous in case of smaller patches and strip plantations of eucalyptus, sissoo-teak-amaltas, miscellaneous species, grasslands and settlements. The ponds/tanks and swamps figured more accurately on fused imagery than on LISS-III false colour composite. The individual tree canopies, their linear pattern and the tree shadows were noticed on fused data, which proved to be helpful in identification of various plantation types. The exercise demonstrated good potential of fused imagery in differentiation and mapping of various types of plantations.