University students collaborating at a computer on a deep-learning and computer-vision framework for detecting image forgery

Digital image forensics

Is this photograph telling the truth?

iPerfect examines any photo and classifies it as authentic, copy-move forgery, spliced or AI-manipulated — showing the forensic evidence behind every decision.

Start an examination

Analyse your image now

Choose a file or use the camera. The analysis runs on this device and usually takes a few seconds.

100% processed on deviceNothing is uploaded to a server
Error-level & noise analysis SHA-256 integrity hash Signed .ipeg report export

What iPerfect detects

Four verdicts, backed by measurable evidence

Every indicator contributes a score, and the fused result is reported with a confidence level rather than a bare yes or no.

01

Authentic images

Consistent compression, coherent sensor noise and intact capture metadata point to an untouched photograph.

02

Copy-move forgery

A region cloned from elsewhere in the same picture is found through a shift-correlation search over textured areas.

03

Spliced content

Material pasted from another image betrays itself through error-level and noise discontinuities along its edges.

04

AI-generated or edited

Synthetic imagery shows unusually smooth noise residuals, missing camera metadata and atypical entropy.

The process

From capture to a verifiable report

Icon of a phone camera with a cloud upload arrowIcon of a magnifying glass inspecting a gridded photo regionIcon of a bar chart beside a confidence gaugeIcon of a report sheet with a verified shield and download arrow

Step 1 of 4

Acquire the image

Upload a photograph or capture one with the device camera. Nothing leaves your browser — the whole examination runs locally.

Already holding a report?

Check any copy of an image against its signed .ipeg evidence file.

Open the verifier