Central Library

Normal view MARC view ISBD view

Real time deforestation detection using ANN and Satellite images [electronic resource] : The Amazon Rainforest study case / by Thiago Nunes Kehl, Viviane Todt, Maur�icio Roberto Veronez, Silvio Cesar Cazella.

By: Nunes Kehl, Thiago [author.].
Contributor(s): Todt, Viviane [author.] | Roberto Veronez, Maur�icio [author.] | Cesar Cazella, Silvio [author.] | SpringerLink (Online service).
Material type: materialTypeLabelBookSeries: SpringerBriefs in Computer Science: Publisher: Cham : Springer International Publishing : Imprint: Springer, 2015Description: X, 67 p. 25 illus., 21 illus. in color. online resource.Content type: text Media type: computer Carrier type: online resourceISBN: 9783319157412.Subject(s): Geography | Artificial intelligence | Remote sensing | Geography | Remote Sensing/Photogrammetry | Artificial Intelligence (incl. Robotics)Additional physical formats: Printed edition:: No titleDDC classification: 910.285 Online resources: Click here to access online
Contents:
1 Introduction -- 2 Literature Review -- 3 Method -- 4 Results and Discussion -- 5 Conclusions and Future Work.
In: Springer eBooksSummary: The foremost aim of the present study was the development of a tool to detect daily deforestation in the Amazon rainforest, using satellite images from the MODIS/TERRA sensor and Artificial Neural Networks. The developed tool provides parameterization of the configuration for the neural network training to enable us to select the best neural architecture to address the problem. The tool makes use of confusion matrices to determine the degree of success of the network. A spectrum-temporal analysis of the study area was done on 57 images from May 20 to July 15, 2003 using the trained neural network. The analysis enabled verification of quality of the implemented neural network classification and also aided in understanding the dynamics of deforestation in the Amazon rainforest, thereby highlighting the vast potential of neural networks for image classification. However, the complex task of detection of predatory actions at the beginning, i.e., generation of consistent alarms, instead of false alarms has not been solved yet. Thus, the present article provides a theoretical basis and elaboration of practical use of neural networks and satellite images to combat illegal deforestation.
    average rating: 0.0 (0 votes)
No physical items for this record

1 Introduction -- 2 Literature Review -- 3 Method -- 4 Results and Discussion -- 5 Conclusions and Future Work.

The foremost aim of the present study was the development of a tool to detect daily deforestation in the Amazon rainforest, using satellite images from the MODIS/TERRA sensor and Artificial Neural Networks. The developed tool provides parameterization of the configuration for the neural network training to enable us to select the best neural architecture to address the problem. The tool makes use of confusion matrices to determine the degree of success of the network. A spectrum-temporal analysis of the study area was done on 57 images from May 20 to July 15, 2003 using the trained neural network. The analysis enabled verification of quality of the implemented neural network classification and also aided in understanding the dynamics of deforestation in the Amazon rainforest, thereby highlighting the vast potential of neural networks for image classification. However, the complex task of detection of predatory actions at the beginning, i.e., generation of consistent alarms, instead of false alarms has not been solved yet. Thus, the present article provides a theoretical basis and elaboration of practical use of neural networks and satellite images to combat illegal deforestation.

There are no comments for this item.

Log in to your account to post a comment.