Question 1

How accurate is the NDVI output?

Accepted Answer

Our model achieves SSIM 0.8060 across 440 held-out validation images — 16.8% above the published benchmark for this task.

Question 2

How was the model validated, and has it been independently recognised?

Accepted Answer

We validated on 440 real Sentinel-2 field pairs held out from training. The work received recognition from MSME (Ministry of Micro, Small & Medium Enterprises) and was presented at an IEEE student-branch symposium. SSIM and Pearson r metrics are logged per inference run for auditability.

Question 3

Do I need a multispectral drone?

Accepted Answer

No. Nabhya works with any standard RGB drone or satellite image. No special hardware required.

Question 4

What image types are supported?

Accepted Answer

PNG, JPG, JPEG, and TIFF. Maximum file size 10MB.

Question 5

How long does analysis take?

Accepted Answer

Typically under 2 seconds. Tested on images up to 10MB.

Question 6

Can I integrate this into my existing software?

Accepted Answer

Yes. Nabhya is a REST API — one POST request returns a full NDVI heatmap and JSON statistics. Most integrations take under a day.

Question 7

What happens after my 50 free analyses?

Accepted Answer

You will receive an email before you run out. Upgrade to a paid plan to continue. No automatic charges.

Question 8

Is Nabhya's SSIM validation methodology auditable?

Accepted Answer

Yes. Every inference run logs SSIM and Pearson r scores against ground-truth. On Enterprise plans we can share the full validation dataset provenance and model card on request.