Digital Image Classification of Mango and Coconut for Natham Taluk, Dindigul District using Sentinel-2a Optical Data
S. Parthipan
S. Parthipan, M.Tech, Geoinformatics, Department of Geography, University of Madras, Guindy, Chennai (Tamil Nadu), India.
Manuscript received on 27 November 2020 | Revised Manuscript received on 03 December 2020 | Manuscript Accepted on 15 December 2020 | Manuscript published on 30 December 2020 | PP: 33-39 | Volume-1 Issue-1, December 2020 | Retrieval Number: A1009021121/2021©LSP
Open Access | Ethics and Policies | Indexing and Abstracting
© The Authors. Published by Lattice Science Publication (LSP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Remote sensing and GIS have been widely applied in agriculture. Several methods exist for mango classification of satellite data which can be utilized by the agricultural sector. This study focuses on using supervised classification approaches to classify mango and coconut plantations Natham taluk, Dindigul district Tamil Nadu. Sentinel 2A acquired on 3 rd February2018wasusedforimageclassification.Groundtruthdatacol lectionwasperformed through the taluk. The land use and land cover of the study area were distinguished into five classes viz., coconut, mango, cropland, settlements and waterbody. Supervised image classification technique such as Mahalanob is Distance, Maximum likelihood Classifier, Spectral angle mapper and Spectral correlation mapper methods were applied over the image. The accuracy measures, suchasproducer’ sacc uracy, user’ saccuracy, overall accuracy and kappa coefficient were estimated. The results showed that maximum likelihood supervised classifier had the highest overall accuracy of 51.4% while other supervised classifier such as Maha lanob is Distance (32.4%), Minimum Distance classifier (42.86%), Spectral Angle Mapper (42.85%), Spectral Angle Mapper (42.85%) and Spectral Correlation Mapper (34.53%) had lower accuracy. It is suggested to utilize multi-date data for classification for crop discrimination utilising the unique phenology of various crops for better accuracy.
Keywords: Mahalanobis, Coconut, Mango, Cropland, Settlements