Crop classification using NDVI
summary
Uses Sentinel-2 10 m imagery from 2019–2021 to build multi‑temporal composites and compute NDVI time series for each field, then fits a harmonic analysis (amplitude + phase) to capture seasonal signatures and applies unsupervised clustering to separate banana, potato, corn/bean classes. Produces a crop map and shows clusters work without labeled training data but notes validation gaps and prediction errors because ground truth is limited; suggests testing EVI and SAVI next.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:35Challenges
- 3:18Clustering
- 7:13Method
- 9:10Satellite images
- 15:10Goals
- 15:47Research Area
- 16:29Mentality
- 18:21Cluster
- 18:56Limitations
- 19:25Other spectral indices
- 20:34Conclusion
- 20:59References
- 21:13Thank you
related talks
21mUnderstanding World food economy with satellite
36mClassification of Satellite Imagery
36mDeep Learning - Segmentation, Identification and Classification
32mMeasuring Vegetation Health to Predict Natural Hazards
28mGeoAI: Evolution of AI-Based Tree/Vegetation Detection Models Using Imagery
31mBridging IoT and Machine Learning with Python for Sustainable Agriculture
19mPlant Factory: Sensor, Data, Machine Learning
27mSmart Forestry: AI in the Service of Nature