Images & photos · Image Metadata Privacy Tool
Camera Serial Numbers in EXIF: How Photos Get Linked Back to You
· Why it matters
exif image-privacy privacy
Body and lens serial numbers are written into EXIF by many cameras, and often again inside the MakerNote. Two photos with the same serial are provably from the same camera, which can tie an anonymous account to a named one.
Two accounts, one camera — the concrete way a serial number connects photos you never meant to associate
Two accounts can avoid names and locations yet publish files carrying the same BodySerialNumber or LensSerialNumber. A repeated identifier provides a technical link between those files even when it does not identify the owner by itself. ToolAcre classifies standard body, lens, owner and image identifiers as device-and-identity fields so they appear before ordinary exposure settings.
Where serials live — the BodySerialNumber and LensSerialNumber tags added in Exif 2.3, plus manufacturer-specific fields inside the MakerNote
Readable serials can live in standard Exif sub-directory fields such as BodySerialNumber and LensSerialNumber. CameraOwnerName and ImageUniqueID provide other linking material. MakerNote is also classified as device-related because vendor data can contain serial and focus information, but this parser does not decode the private structure inside it field by field.
Which files carry them — interchangeable-lens cameras often do; phones vary; exported edits may keep or drop them depending on the software
JPEG, PNG and WebP can all carry the same TIFF-shaped EXIF directories in their respective metadata blocks. Whether a particular camera, phone or editor writes a serial is an input fact, not a guarantee of the format. Inspect each exported file; do not assume a conversion preserved the field or assume a missing readable serial means no opaque device data exists.
How a repeated identifier can link files
A stable value repeated across photographs lets an observer group those files as likely sharing a device or lens. The workbook called that link provable ownership, which is too strong: files can be copied, metadata can be edited and equipment can change hands. The defensible claim is correlation based on a repeated value, not authentication of a person.
Legitimate identification and privacy trade-offs
Serial fields support inventory, support and provenance workflows as well as privacy risks. The repository does not document ownership registries or stolen-camera databases, so this section avoids those external claims. Decide whether the publishing copy needs technical identifiers; if it does not, remove the containing block while retaining a private master for legitimate records.
Worked example: inspect readable serial fields and the MakerNote marker
Inspect a camera JPEG and read the Device and identity table for owner, body, lens and image identifiers. MakerNote may appear only as an opaque numeric summary because UNDEFINED values are bounded and its internals are not chased. Removal does not need to understand the blob: the entire APP1 Exif segment leaves as one unit.
What this does not cover — sensor-noise fingerprinting from the pixels, which survives metadata removal, and visible watermarks
Metadata is not the only way photographs can be correlated. The repository names steganography as out of scope and makes no sensor-fingerprint claim. Pixel content, repeated scenes and publishing behaviour remain separate clues, while colour profiles are deliberately kept and may contain descriptions. Do not turn block removal into a promise of unlinkability.
Takeaway: check for serials before publishing — the Image Metadata Privacy Tool shows the standard tags and removes them along with the rest of the EXIF block
Check for readable serials before publishing and treat opaque MakerNote content conservatively. Save any technical details you need from the private original, then clean a derivative and verify that no field readable by this parser remains. The result removes the whole EXIF block without recompressing the image, but it does not erase identifying information visible in the pixels.