Drag & drop scrap images here
or click to browse — multiple images supported
—
Score-CAM visualization: which image regions drove the model's prediction
Generating attention map…
Original Image
AI Attention Map
Red regions indicate areas that most influenced the model's prediction. Bright red on metal surfaces typically indicates zinc-coated (galvanized) texture patterns detected by the EfficientNet-B0 backbone.
Upload a PCB or e-waste image to get an AI-powered assessment and compliance report.
Drop PCB or e-waste image
Motherboards, PCBs, mixed e-waste
Count ICs, capacitors, connectors. Estimate board density and recovery potential.
Manual vs. mechanical paths per board type. Links to India's authorized recyclers.
India generates ~3.2 million tonnes of e-waste annually (MoEF 2023) and ranks 3rd globally. Less than 22% is processed through formal channels. ScrapScan Phase 2 targets bridging this gap by putting AI-driven e-waste intelligence in the hands of registered recyclers and dismantlers.
| Supplier | Loads | Avg Grade | Zinc Risk | Compliance |
|---|---|---|---|---|
| Sharma Scrap Pvt Ltd | 48 | HMS-1 | LOW | ✔ IS 2314 |
| Gupta Metal Works | 31 | HMS-2 | MEDIUM | ⚠ Partial |
| Rajasthan Scrap Corp | 22 | Galv. | HIGH | ✖ Non-compliant |
| Delhi Iron Traders | 17 | HMS-1 | LOW | ✔ IS 2314 |
| Mumbai Metals Ltd | 39 | Mixed | MEDIUM | ⚠ Partial |
No scans yet. Upload an image to begin.