Using Deep Learning on Satellite Images to Identify Deforestation/Afforestation
| Autor: | Mahadevan Narayanan, Apurva Mhatre, Ajun Nair, Navin Kumar Mudaliar, Aaditya Gurav, Akash Nair |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
education.field_of_study
Computer science business.industry Deep learning Supervised learning Population 02 engineering and technology Convolutional neural network Natural resource 0202 electrical engineering electronic engineering information engineering Leverage (statistics) Afforestation 020201 artificial intelligence & image processing Segmentation Artificial intelligence business education Remote sensing |
| Zdroj: | Computational Vision and Bio-Inspired Computing ISBN: 9783030372170 |
| DOI: | 10.1007/978-3-030-37218-7_113 |
| Popis: | As of 2018 the area covered by the forests was determined to be 30.6% of the world’s land surface which is roughly below 3.8 billion hectares. With the exponential increase in population, the pressure on the natural resources increases which results in cutting the trees for agriculture and for industrial purposes thereby leading to deforestation. In this paper, we investigate the spatial distribution of vegetation cover by using Convolutional Neural Network (CNN) and satellite images. The features of the object identified by the CNN is extremely complex because of the number of filters used to identify various patterns. Supervised learning (learning with annotated data) is used to train the CNN model as it results in better accuracy while determining the forest cover. We leverage this technology of instance segmentation to correctly determine the forest cover in a particular satellite image that radically improves the accuracy. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|