From photo to decision
Everything GrainSight does, and exactly how it works — from your first upload to a verified inspection record.
What it does
AI Visual Inspection
Upload a rice sample image and GrainSight estimates five visual quality indicators: whole, broken, chalky, yellow, and damaged kernels.
Image Quality Check
Unsuitable images — overlapping grains, poor lighting, blur — are flagged and rejected instead of producing an unreliable estimate.
Quality Standard Comparison
Define custom thresholds for whole, broken, chalky, yellow, and damaged kernels, and compare every inspection against them automatically.
PASS / REVIEW / FAIL
Every inspection resolves to a clear decision: PASS meets your configured criteria, REVIEW needs a closer look, and FAIL clearly falls outside requirements.
Human Verification
Confirm or correct the AI estimate. The original AI result and the human-verified result both stay visible for transparency.
Inspection History
Every inspection is kept and searchable by sample, lot, supplier, date, and result — so nothing gets lost between screenings.
Reports
Turn any inspection into a shareable report — sample details, the estimate, the comparison, and the final decision in one document.
Taking a good sample photo
Do
- Spread rice in a thin, even layer
- Use bright, even lighting
- Keep the camera directly above the sample
- Fill most of the image with rice
- Make sure the image is sharp
Avoid
- Avoid grain piles
- Avoid strong shadows
- Avoid blurry photos
- Avoid very dark environments
Ready to see it on your own samples?
Start with a rice image. Get an AI-assisted estimate in seconds.