Satellite image classification using predictive analytics is a significant task with numerous applications in environmental monitoring, urban planning, and disaster management. This project aims to develop accurate models for classifying satellite images into distinct categories such as 'cloudy', 'desert', 'green area', and 'water'. By leveraging predictive analytics techniques, the project addresses the challenge of automatically analysing satellite images and assigning them to the correct class.
The methodology employed in this project involves a combination of image processing techniques, feature extraction, and machine learning algorithms. Initially, the satellite images undergo pre-processing to enhance their quality and eliminate any noise or artifacts. Relevant features are then extracted from the images to capture the distinctive characteristics of each land cover class. These features serve as inputs to various machine learning algorithms, including logistic regression, decision tree classifier, support vector machines, KNN classifier, Naive Bayes classifier, and random forest classifier.
The project consists of several experiments to evaluate the performance of the classification algorithms. A dataset of satellite images, organized into subfolders representing different land cover classes, is divided into training and testing sets. The training set is utilized to train the models, while the testing set is used to assess their performance. Each algorithm is trained using the training set, and predictions are made on the testing set. The accuracy of each algorithm is measured by comparing the predicted labels with the true labels from the testing set.