Deforestation Analysis using Unsupervised Clustering and Satellite Images
K. Pradeep Mohan Kumar1, Rushan Mukherjee2, Mayank Dewangan3
1K. Pradeep Mohan Kumar*, Department of Computer Science and Engineering, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
2Rushan Mukherjee, Department of Computer Science and Engineering, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
3Mayank Dewangan, Department of Computer Science and Engineering, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
Manuscript received on March 15, 2020. | Revised Manuscript received on April 01, 2020. | Manuscript published on April 10, 2020. | PP: 1651-1656 | Volume-9 Issue-6, April 2020. | Retrieval Number: F4335049620/2020©BEIESP | DOI: 10.35940/ijitee.F4335.049620
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Expansion of farmland and unplanned land encroachments have increased as the earth’s population boomed, which has led to uncontrolled deforestation across the world. Deforestation and industrialization have given rise to global warming, causing mayhem in the current ecosystem. The weather patterns are disrupted, and natural calamities are occurring more frequently. The after-effects of these events have to lead to dramatic losses in flora and fauna. Even though a large part of India’s land is urbanized, there are many protected areas in specific parts of the country that represent significant vegetation that has been affected if we observe from the scale of a subcontinent. In this paper, we aim to trace the deforestation in the Sundarbans from 1988 to 2019. This period is essential as a lot of industrialization and population boom happened during this time. We have selected the Sundarbans because mangroves are a natural defense to cyclones and also provide shelter to a plethora of living organisms. Our study area covers more than 7000 square kilometers of the Indian Sundarbans. The satellite images are from Landsat-5, Landsat-7, and Landsat-8, which are orthorectified. We make use of open-source software like Quantum GIS (QGIS), Google Earth Engine (GEE), and Google Colab, which is a Python IDE in this project. We make use of the K-means clustering, which is an unsupervised learning algorithm. Here, we have described a method to analyze deforestation accurately using low-cost techniques, which can be used by underdeveloped nations and private organizations to help in the fight against illegal deforestation.
Keywords: Deforestation, K-means Clustering, Machine learning, NDVI.
Scope of the Article: Machine learning,