IMAGE CLASSIFICATION USING MACHINE LEARNING FOR ENVIRONMENTAL MONITORING: AN ANALYSIS OF SATELLITE IMAGERY

The monitoring of the environment is essential for addressing the effects of human activity on ecosystems. Satellite imagery can provide us with data on a massive scale as it allows us to observe changes in the vegetation growth, the land cover, and more. Due to the complexity of satellite imagery, it can be hard to accurately classify and analyze it. This is why it is important that we use machine learning techniques to improve the efficiency of our environmental monitoring. Some of these include the use of neural networks and CNN. The goal of this study is to analyze the effectiveness of CNN, RNN, and ANN in properly classifying satellite images collected by the MODIS dataset. It also explores the applications of this technology in monitoring the changes in the environment over time. The process involved preprocessing the images and creating labels based on the different land cover classifications. After training the models, they were evaluated for their accuracy. The CNN model was able to achieve an accuracy of 93.5% in properly classifying satellite images. It was also more accurate than the other two models. The study showed that machine learning can be used to improve the efficiency of environmental monitoring. The study demonstrates how machine learning can be used for environmental monitoring. It also calls for further research to understand its full potential