Journal of Natural Environment

Journal of Natural Environment

Application of the Weighted Indexes Using Training Data and Genetic Algorithm on High Resolution Images for Vegetation Detection in Urban Areas

Document Type : Research Paper

Authors
Millad Janalipour 1
Ali Mohammadzadeh 2
Mohammad Javad Valadan Zoej 3
1 PhD Candidate, Department of Remote Sensing and Photogrammetry, Geomatics Engineering Faculty, K. N. Toosi University of Technology, Tehran, Iran
2 Assistant Professor, Department of Remote Sensing and Photogrammetry, Geomatics Engineering Faculty, K. N. Toosi University of Technology, Tehran, Iran
3 Associate Professor, Department of Remote Sensing and Photogrammetry, Geomatics Engineering Faculty, K. N. Toosi University of Technology, Tehran, Iran
10.22059/jne.2014.51296
Abstract
Various indexes such as RVI, NDVI, SAVI and OSAVI have been proposed for vegetation detection using satellite images. These indexes have been obtained based on high reflectance of the vegetation in near infrared band and its high absorption in red band. Basic defect of these indexes are using them in various regions without any changes in index structure. In other words, these indexes have not capability of adaption to various regions and in some of the researches have tried to reduce this defect using the empirical coefficients. In this article, all of the bands are used to produce in vegetation index. For used all of the bands in the proposed indexes, each band is assigned a weight. These weights are estimated using training data and the proposed algorithms. The study areas were Shiraz, Bam and New Brunswick, that high resolution images are used from these areas. Results in study areas show the high capability of the proposed indexes to vegetation detection.
Keywords
Genetic Algorithm
remote sensing
threshold
vegetation
weighted vegetation index
Journal of Natural Environment
Volume 67, Issue 2 - Serial Number 2
Summer 2014
Pages 145-155