Features How it works Model Start Analysis
CNN-powered · Live camera or photo upload

Spot the dent
before the buyer does.

Point a camera at any vehicle, or upload a photo, and DentScan AI tells you in seconds whether it's damaged or clean — trained on real inspection images, running on a convolutional neural network.

Start Analysis See how it works
camera-01 · live inspection feed
🚗
rear-quarter panel
Confidence
91%
Damage detected — scratch / dent identified on the rear-quarter panel
Live camera capture
Photo upload mode
CNN classifier — trained from scratch
Instant confidence score
Downloadable PDF report
Runs locally — no cloud upload
Live camera capture
Photo upload mode
CNN classifier — trained from scratch
Instant confidence score
Downloadable PDF report
Runs locally — no cloud upload

Everything a quick
inspection needs

No manual checklist. Point, capture, and know.

Photo Upload

Drop in an existing photo of a vehicle and get an instant verdict — damaged or clean.

Confidence Score

Every result comes with a model confidence percentage, not just a yes/no guess.

Scan History

Every capture during a session is logged so you can review past results at a glance.

PDF Report

Download a clean, one-page report of any scan — image, verdict, confidence, and timestamp.

Runs Locally

The model runs on your own machine — no images are sent to a third-party cloud service.

From camera to
verdict, in seconds

Three steps, no manual inspection checklist.

01
Choose your input

Upload a photo, or open your live camera feed.

02
Capture & analyze

The CNN classifies the image as damaged or normal.

03
Review & download

See the confidence score and export a PDF report.

VerdictDamaged
Confidence91%
Scan Time0.4s
Front bumper — scan 3Normal
Rear panel — scan 4Damaged

Stop guessing
if it's damaged.

Run your first scan in under a minute — free, no account needed.

Start Analysis
A university project demonstrating CNN-based image classification