A photograph is often the densest piece of evidence in a case and the least used. It carries a place, a time, objects, text and people — and most investigations extract one of those, by eye, and move on.
Vision is the image side of the workspace. It reads images with AI, runs reverse search, and works out where a photograph was taken.
Three questions about one image
What is in it? Objects, text, signage, vehicles, clothing, documents held up to a camera. A description that can be searched later is worth more than an image you have to remember.
Where else does it appear? Reverse search places the same or a similar image elsewhere on the internet — often revealing the original posting, an earlier date, or a different name attached to the same face or product.
Where was it taken? Geolocation reasons from what is visible — terrain, architecture, signage, layout — toward a place, and presents the reasoning rather than a bare coordinate.
Metadata is a bonus, not the method
Most images that matter have been stripped of their metadata by the platform that published them. An approach that depends on EXIF works on holiday photos and fails on evidence. Vision treats what is visible in the frame as the primary material.
A location is a finding, and findings get reviewed
A geolocation is a claim about the physical world, and it is presented as one — with the visual elements it rests on. A confident-looking pin with no reasoning behind it is worse than useless in a report, because it survives until someone checks it.
Accepted results join the case file with their support; uncertain ones stay marked as uncertain. See accepted and rejected evidence.
Where it fits in a case
Results flow into the rest of the workspace: locations reach Engramite Map, entities and links reach Engramite Graph, and admitted findings reach the report through Engramite Writer.
Frequently asked questions
Does it work without EXIF data? Yes. That is the normal case for published images.
Is a geolocation a certainty? No. It is a supported finding, shown with its reasoning, for a human to accept or reject.
Can it analyse images I already hold in the case? Yes — documents, downloads and material provided by a client all live in the same case file.
Read next: Engramite Map or the Archive.