Release: Land Cover Classification using Sentinel-2 Imagery

Release: Land Cover Classification using Sentinel-2 Imagery
Overview
I am excited to announce the release of my project, “Land Cover Classification using Sentinel-2 Imagery,” which leverages the power of Google Earth Engine to classify land cover types for the year 2023. This project demonstrates the application of machine learning techniques to large-scale environmental data, providing valuable insights into land cover patterns.

Key Features
Sentinel-2 Image Collection: Utilizes the Harmonized Sentinel-2 MSI: MultiSpectral Instrument, Level-2A dataset for comprehensive land cover analysis.
Random Forest Classifier: Implements a Random Forest classifier to accurately classify different land cover types.
Spectral Indices: Incorporates the Normalized Difference Vegetation Index (NDVI) to enhance classification accuracy.
Training and Validation: Includes detailed steps for merging, sampling, and splitting training data, as well as evaluating classifier performance.
Visualization: Provides clear visualization of classified land cover types using a color-coded map.
Export Functionality: Allows exporting of training data for further analysis and use.
How to Use
Load and Preprocess Data: Load the Sentinel-2 image collection and preprocess the data by selecting relevant bands and adding spectral indices.
Merge and Sample Training Data: Combine different land cover classes and sample training data for feature extraction.
Train the Classifier: Train a Random Forest classifier on the sampled data.
Classify and Display Results: Apply the trained classifier to classify the image and display the results on the map.
Evaluate Performance: Assess the classifier’s performance using confusion matrices and accuracy metrics.
Export Data: Export the training data for further analysis.
Outputs
Land Cover Classification Map: A visual representation of classified land cover types.
Accuracy Metrics: Training and validation accuracy metrics to evaluate classifier performance.
Exported Training Data: Training data exported in SHP format for further use.
Getting Started
Open the script in Google Earth Engine.
Modify the region of interest (ddn) as needed.
Run the script to classify land cover and export the results.
I hope this project serves as a valuable resource for those interested in applying machine learning to environmental data. Your feedback and contributions are welcome!

Release by:
Anish K M