Identification of bamboo patches in the lower Gangetic plains using very high resolution WorldView 2 imagery

Aniruddha Ghosh; P. K. Joshi

doi:10.1117/12.2029192

24 October 2013 Identification of bamboo patches in the lower Gangetic plains using very high resolution WorldView 2 imagery

Aniruddha Ghosh, P. K. Joshi

Author Affiliations +

Proceedings Volume 8893, Earth Resources and Environmental Remote Sensing/GIS Applications IV; 889310 (2013) https://doi.org/10.1117/12.2029192
Event: SPIE Remote Sensing, 2013, Dresden, Germany

Abstract

Realizing the potential usage, the country has identified National Bamboo mission for addressing the issues relating to the development of bamboo in the country. Therefore knowledge of spatial distribution of bamboo patches becomes necessary for the evaluation and monitoring of this resource. Medium resolution satellite imagery is ineffective for accurate classification when bamboo occurs in mosaic of small and mixed patches. The present study attempts to address this problem using Very High Resolution (VHR) multispectral (MS) WorldView 2 (WV 2) imagery in South 24 Parganas, West Bengal. West Bengal, a part of Bengal Bay ecological region is one of the areas where bamboo grows naturally. The classification was carried out using additional features namely second order texture components of the first 3 principal components (PCs) of pan-sharpened 8 MS bands. Supervised kernel based (Support Vector Machine, SVM) and ensemble based (Random Forest, RF) machine learning algorithms were applied on the dataset. Moreover, RF based variable importance was analyzed to find the most informative input predictors for classification of bamboo. Altogether seven land use land cover classes were mapped which include agricultural land, built-up, bamboo patches, two canopy classes, fallow land and water body. Variable importance analysis indicated that mean texture measure of PC1 and PC3, and spectral information of MS band 8 (NIR 2) were the most important predictor layers for bamboo mapping. Unlike, RF (accuracy 72%) higher overall accuracy of 80% was achieved using SVM classifier for bamboo mapping. Intermixing of bamboo was seen with other canopies.

Citation Download Citation

Aniruddha Ghosh and P. K. Joshi "Identification of bamboo patches in the lower Gangetic plains using very high resolution WorldView 2 imagery", Proc. SPIE 8893, Earth Resources and Environmental Remote Sensing/GIS Applications IV, 889310 (24 October 2013); https://doi.org/10.1117/12.2029192

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
9 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Image resolution

Image classification

Agriculture

Earth observing sensors

Associative arrays

Spatial resolution

Near infrared

Show All Keywords

Keywords/Phrases

Search In:

Publication Years