Certification & Online Trace Collection · service active
WACZ · ISO 28500/ eIDAS timestamping/ Client area
C.E.R.T.O.
Sign in Register free
IT EN
C.E.R.T.O. / Modules / Image Evidence Collection
04 · IMG

Image Evidence Collection

Forensically acquire and analyse a user-supplied image: full metadata, geolocation, perceptual fingerprints and tampering indicators, sealed into a verifiable bundle.

C.E.R.T.O. Desktop “Image Evidence Collection” screen: preview of the user-supplied image, optional qualified eIDAS InfoCert timestamp and start of the certification with in-depth forensic analysis.

C.E.R.T.O.'s Images module performs the forensic evidence collection of an image supplied by the operator and runs an in-depth analysis on it: it extracts EXIF/IPTC/XMP metadata and GPS data, computes cryptographic and perceptual hashes and runs ELA, histogram, Luminance Gradient, Level Sweep and the examination of the JPEG quantization tables. Everything is sealed as digital evidence with a double RFC 3161 timestamp and an Ed25519 signature. It is the tool for court-appointed and party experts, lawyers and law enforcement who must certify a photo and document its authenticity and traces of re-processing.

Key features

What this module does.

  • Extended EXIF/IPTC/XMP extraction: hardware/software/filesystem dates.
  • Full GPS data with geolocation map of the shot.
  • Forensic analysis: ELA, interactive histogram, JPEG quantization tables, Luminance Gradient and Level Sweep.
  • Perceptual hashes (aHash, dHash, pHash, wHash) at multiple depths for near-duplicate matching.
  • BagIt 1.0 bundle signed with Ed25519, double RFC 3161 timestamp and CASE/UCO.
Final forensic report: the interactive dashboard of the image bundle — summary, image, EXIF/IPTC/XMP, forensic analysis, perceptual hashes, chain of custody, hash inventory and integrity check.
Forensic pipeline

How the module operates.

A repeatable, documented procedure: from the cryptographic fingerprint to the in-depth pixel analysis, every examination of the image leaves a verifiable trace inside the bundle.

01 · NTP

Synchronised time

Multi-source NTP sync with documented offset and roundtrip: the moment of evidence collection is anchored.

02 · HASH

Original fingerprint

Computation of MD5/SHA-1/SHA-256/SHA-512 on the supplied original image, as is, before any processing: the integrity reference.

03 · META

EXIF/IPTC/XMP metadata

Full extraction of embedded metadata — camera, lens, shooting parameters, dates, GPS, author, copyright and C2PA provenance — formatted and human-readable.

04 · PHASH

Perceptual hashes

Four algorithms (aHash, dHash, pHash, wHash) at three bit depths: fingerprints of the visual appearance, useful to find the same image or detect modified versions.

05 · ANALYSIS

In-depth analysis

Error Level Analysis (ELA), histogram, pixel statistics, Luminance Gradient, Level Sweep and JPEG quantization tables with quality estimation.

06 · HISTORY

Editing history

Reconstruction, from XMP metadata, of the software and tools used, the derivation chain and the Content Credentials (C2PA), distinguishing development from tampering.

07 · MARK

Rendering & watermark

Generation of a watermarked version (for review) alongside the authoritative original, with their respective hashes side by side.

08 · SEAL

Signature & double timestamp

manifest.json signed with Ed25519 + double RFC 3161 timestamp, packaging into a BagIt 1.0 bundle with a CASE/UCO description and verify.sh / verify.bat verifiers.

Advanced analysis

More than a hash: in-depth forensic analysis.

C.E.R.T.O. does not just certify the file: it studies its pixels and metadata with image-analysis tools, to document its authenticity, history and any traces of re-processing.

The report's “In-depth forensic analysis” tab: JPEG compression and quality, quantization tables, interactive histogram, pixel statistics, XMP editing history, embedded EXIF preview, Error Level Analysis, Luminance Gradient and Level Sweep. Click to see it in full.

Error Level Analysis (ELA)

Recompresses the JPEG and amplifies the differences: re-processed areas tend to show a different error level. An indicative, not conclusive, technique.

Histogram & statistics

Tone distribution per channel (R/G/B/Luminance) with clipping, entropy, sharpness and dominant colour: an objective baseline for comparison.

JPEG quantization tables

The 8×8 matrices that encoded the JPEG: they are the “encoder fingerprint” and, if non-standard, indicate the saving program or a recompression.

Luminance Gradient

Maps the direction of light on surfaces: edges or areas with lighting inconsistent with their surroundings are a classic copy-paste cue.

Level Sweep

Highlights pixels at a given tonal level, linked to the histogram: a pasted region often has a slightly different tonal distribution.

Editing history (XMP)

Software, tools, derivation chain and Content Credentials (C2PA) from XMP metadata: distinguishes normal photographic development from tampering.

Bundle contents

Everything that gets generated.

Each evidence collection produces a coordinated set of artefacts, each with a precise forensic role, organised into clearly-named folders inside data/.

Original image

The file supplied by the operator, kept as is with its own hash: it is the authoritative media of the bundle, never re-processed.

evidence/<immagine>

Watermarked version

A C.E.R.T.O.-watermarked copy for review and sharing, with its own hash distinct from the original's.

evidence/<immagine>_watermarked

EXIF · IPTC · XMP

The embedded metadata extracted and formatted: shooting parameters, dates, GPS, author, copyright, software and C2PA provenance.

reports/exif-report.txt · metadata/

Perceptual hashes

The fingerprints of the visual appearance (aHash/dHash/pHash/wHash at 64/144/256 bits), to find the image or detect modified versions.

hashes/perceptual-hashes.json

Forensic analysis

The analysis outputs: ELA, histogram and statistics, Luminance Gradient, Level Sweep, quantization tables and the embedded EXIF preview.

reports/analysis/

Forensic report & hashes

The report in PDF and TXT with its own RFC 3161 timestamp and the complete hash inventory (MD5/SHA-1/SHA-256/SHA-512) of all artefacts.

reports/report.pdf · hashes/file-hashes.json

Self-validation

A bundle that proves itself.

The bundle does not need C.E.R.T.O. to be validated: anyone, even years from now, can verify its authenticity with standard tools. The BagIt 1.0 structure and the interactive dashboard make it self-explanatory.

  • interactive.html — the navigable offline dashboard: image and comparison, EXIF/IPTC/XMP, forensic analysis, perceptual hashes, hash inventory and client-side integrity check.
  • manifest.json signed with Ed25519 (RFC 8032), bound to the identity of the device registered at first launch.
  • Double RFC 3161 timestamp: inner anchor on data/tsa.tsr and outer seal on tagmanifest-sha256.txt.tsr. Free cascade Sectigo→DigiCert→GlobalSign; optional qualified eIDAS InfoCert.
  • manifest-sha256.txt and tagmanifest-sha256.txt (RFC 8493): fixity of the payload and of the control files; no file can be added or altered without the check failing.
  • metadata/evidence.case.jsonldCASE 1.3 / UCO 1.4 description of the evidence, and tsa-ca.pem for verifying the timestamp even offline.
  • verify.sh / verify.bat — standalone verifiers: they recompute the hashes, check the double timestamp and the signature, and declare “VALID BUNDLE”.
FAQ

Frequently asked questions

Forensic image evidence, in-depth analysis, photo authenticity and bundle verification: the most common questions.

What is forensic image evidence collection?
It is the certification of a user-supplied image as digital evidence: its hashes are computed, all embedded metadata is extracted and an in-depth forensic analysis is performed, sealing everything into a bundle verifiable by third parties.
What forensic analysis does C.E.R.T.O. run on the image?
An in-depth analysis: Error Level Analysis (ELA), an interactive histogram, pixel statistics, Luminance Gradient, Level Sweep, JPEG quantization tables and quality estimation, the embedded EXIF preview, and the editing history reconstructed from XMP metadata (software, tools, derivation).
Does the analysis prove whether an image has been tampered with?
The techniques (ELA, Luminance Gradient, Level Sweep, quantization tables) are indicative, not conclusive: they flag anomalies and traces of re-processing to be assessed. It is important to distinguish normal photographic development (e.g. RAW conversion in Lightroom) from tampering: C.E.R.T.O. provides the technical elements, the assessment remains with the expert.
What are perceptual hashes?
They are fingerprints of the image's visual appearance (not its bytes): unlike cryptographic hashes, visually similar images produce similar perceptual hashes. C.E.R.T.O. computes four (aHash, dHash, pHash, wHash) at three bit depths, useful to find the same image or detect modified versions.
What metadata is extracted?
All embedded metadata: EXIF (camera, lens, shooting parameters, dates, GPS), IPTC (captions, author, copyright), XMP (software, editing history, C2PA/Content Credentials provenance) and the file's technical properties (dimensions, colour space, ICC profile, subsampling, JPEG tables).
Is the certified image valid as evidence in court?
The bundle follows recognised standards (ISO/IEC 27037, BagIt RFC 8493, RFC 3161, CASE/UCO) with an Ed25519 signature and a double timestamp; authenticity and integrity can be verified by anyone, even offline. The analysis documents the state of the file; the final assessment rests with the adjudicating authority.

Collect evidence with the Image Evidence Collection module.

Register for free and download C.E.R.T.O. Desktop for Windows and macOS from your client area.