Open AccessQuantification of green areas
Author(s) -
Ramon Octaviano de Castro Matoso,
Lucas do Vale Souza,
Cézar Henrique Barra Rocha
Publication year - 2021
Publication title -
revista nacional de gerenciamento de cidades
Language(s) - English
Resource type - Journals
ISSN - 2318-8472
DOI - 10.17271/2318847297420213022
Subject(s) - computer science , remote sensing , normalized difference vegetation index , software , aerial survey , artificial intelligence , data mining , geography , geology , oceanography , climate change , programming language
The monitoring of green areas has been researched and analyzed by several surveys, however it is not trivial to find data with reliable accuracy and precision. There are well-defined and widely used methodologies for the classification of large area images, but in the case of small areas some authors recommend the analog analysis of aerial photos for classification. However, the acquisition of high definition aerial images is not inexpensive. In addition to being a fully manual and labor intensive classification. The present work proposes a methodology for the collection of a quantitative historical database using remote sensing techniques and Digital images, using the NDVI as a comparison criterion. In this work, free images of the Landsat-5, Landsat-8 and Sentinel-2 satellites were used to verify if it is possible to extract reliable information from areas considered small to use the classification supervised by the maximum likelihood method using the ENVI 5.1 software.