How krishgarg344/crop_monitoring_sih works
CropVision is a satellite-based crop-health monitor. Its main user workflow is selecting a farm in the Streamlit dashboard, loading processed vegetation-index rasters, normalizing and stacking them, running patch-wise CNN inference, and viewing health maps and summaries. The advanced dashboard also presents simulated soil metrics and estimates pest risk from crop stress and simulated weather. A separate Flask API currently predicts on a random dummy patch, rather than on uploaded satellite imagery. Notebook and data-preparation details are documented but not sampled, so those relationships are not expanded.
Main entry points: dashboard_1.py Multi-tab dashboard · dashboard.py Basic dashboard · crop_health_model.keras Crop health CNN · app.py Flask test API
Components 15
Farm dashboard
- Multi-tab dashboard
dashboard_1.py - Basic dashboard
dashboard.py - Processed indices
data/processed
Model inference
- Index normalization
dashboard_1.py - Patch prediction
dashboard_1.py - Crop health CNN
model/crop_health_model.keras - Health map
dashboard_1.py
Farm insights
- Health distribution
dashboard_1.py - Simulated soil
dashboard_1.py - Pest risk rules
dashboard_1.py - Farm summary
dashboard_1.py
Test API
- Flask test API
app.py - Crop health CNN
model/crop_health_model.keras
Other components
- Farm user
- API caller
Connections 23
- Farm user → Multi-tab dashboard: selects farm
Readme.md - Multi-tab dashboard → Processed indices: reads indices
dashboard_1.py - Multi-tab dashboard → Crop health CNN: loads model
dashboard_1.py - Multi-tab dashboard → Index normalization: normalizes indices
dashboard_1.py - Index normalization → Patch prediction: provides stack
dashboard_1.py - Multi-tab dashboard → Patch prediction: requests map
dashboard_1.py - Patch prediction → Crop health CNN: predicts patches
dashboard_1.py - Patch prediction → Health map: reconstructs map
dashboard_1.py - Multi-tab dashboard → Health distribution: calculates percentages
dashboard_1.py - Health map → Health distribution: counts classes
dashboard_1.py - Multi-tab dashboard → Simulated soil: generates metrics
dashboard_1.py - Multi-tab dashboard → Pest risk rules: assesses risk
dashboard_1.py - Health distribution → Pest risk rules: supplies stress
dashboard_1.py - Multi-tab dashboard → Farm summary: renders summary
dashboard_1.py - Health distribution → Farm summary: supplies percentages
dashboard_1.py - Farm user → Basic dashboard: selects farm
dashboard.py - Basic dashboard → Processed indices: reads indices
dashboard.py - Basic dashboard → Crop health CNN: loads model
dashboard.py - Basic dashboard → Index normalization: normalizes indices
dashboard.py - Basic dashboard → Patch prediction: requests map
dashboard.py - Patch prediction → Basic dashboard: returns map
dashboard.py - API caller → Flask test API: requests prediction
app.py - Flask test API → Crop health CNN: loads and predicts
app.py