🌿
CropScan
AI Precision Agriculture
Checking…
🌱
Diagnose Your Crop
38 diseases · MobileNetV3 · NDVI fusion · CAF
🔧 Background Removal
Removes soil/sky — improves accuracy
Environmental Inputs
Adjust manually or tap Live Weather to auto-fill
Temperature 28°C
Humidity 65%
NDVI 0.45
Moderate vegetation
Model Pipeline
🧠
MobileNetV3-Large
38-class PlantVillage · 224×224 · 2-phase fine-tune
🔥
Grad-CAM (Eq.4–5)
Explainability heatmap on conv_head layer
🌍
NDVI / EVI / SAVI (Eq.10–12)
Vegetation indices · normalised Vn (Eq.13)
☁️
Weather Stress (Eq.14)
Temp + humidity → Ws stress score
⚗️
XGBoost Fusion (Eq.15)
w1=0.45 · w2=0.35 · w3=0.20 (Eq.18)
✨
CAF — Novel Algorithm
Confidence-adaptive weights per sample
85%+
Accuracy
38
Diseases
CAF
Novel Fusion
ONNX
Exported
🌿
CropScan
Analysing your leaf…
Analysing leaf…
Running MobileNetV3 · NDVI fusion · CAF
⬜ Preprocessing image (224×224)
⬜ MobileNetV3 inference
⬜ Computing NDVI + Weather stress (Eq.14)
⬜ XGBoost health fusion (Eq.15)
⬜ CAF adaptive weights (novel)
Scan Result
MobileNetV3 + CAF
Health Score — Eq.15
Environmental Inputs Used