Landuse Landcover Classification And Accuracy Assessment For Vijayawada Using Remote Sensing And Gis

Authors

  • T. Shalini , Dr. Dinesh Singh

Abstract

Remote sensing is one of the important tools within the creation of Land use Land cover maps are
administered via an image classification method. To make the image classification operation
successful, a combination of factors should consider which include the availability of high-quality
satellite images and second-hand images data, error-free classification process, and user's experience
and competence program. This study aims to classify and produce the Land use Land cover maps
using remote sensing and GIS techniques administered in Vijayawada for the years 2009, 2016, and
2019. For this study, Landsat 4-5 TM, Landsat OLI8, Sentinel 2 satellite data were downloaded and
processed. During this study, the maximum likelihood supervised classification method was used and
therefore the LULC was classified into 5 categories: Agriculture, Barren Land, Waterbody,
Vegetation, and Settlements. These areas are being visited; the collection of GPS data is done so that
areas can be approved for classification. The accurate assessment of the land cover classification
requires some random areas within the region has been selected to represent the different kinds of
classes on the map. To generate the overall accuracy, the classified areas were compared with the
land cover map. Kappa coefficient is used to assess the quality of classification. The overall accuracy
of LULC obtained for 2009, 2016, and 2019 was 79%, 80%, and 86%, and the kappa coefficient was
0.70, 0.71, and 0.80. The kappa coefficient is appraised as considerable and therefore the classified
image is determined to be suited for enhanced investigation. This work suggests a fundamental
supply of facts where the organizers and decision-makers can employ to moderately plan the
circumstances.

Published

2020-11-01

Issue

Section

Articles