What this page does
This page does not try to find an exact address from a photo. Note fixed clues such as mountain skylines, roads, signs, vegetation, and shadows. Use them to narrow the possible region and state your confidence.
Back to workflows
GEOINT workflow
Narrow candidate regions from a photo.
Notes fixed clues in a photo to narrow possible regions; does not claim an exact address and keeps confidence visible.
Passive and safe research workflow
When to use this
You have a photo or video frame with a location claim but you are not sure yet. Use this flow for news images, rural scenes, mountain terrain, or visible fixed clues like signs.
Rural photos, terrain clues, news images, and visual geolocation practice.
PhotoVideo frameClaimed locationVisible clues
Do not overclaim
Similar appearance alone is not proof. Read the boundaries below and keep uncertainty visible in your report.
Do not treat one visual similarity as proof of exact location.
- Do not treat one visual similarity as proof of exact location.
- Do not expose private homes, vulnerable people, or live location detail.
- Do not expand sensitive facility or operational detail unnecessarily.
- Do not turn people in the image into targets.
Step by step
Work in order and take notes at each step.
List fixed clues in the image
What to look at
Mountain skyline, valley, road, coastline, sign, building type, vegetation, shadow direction, and weather.
What to write down
Write each clue on its own line: what you see, fixed or temporary, and how sure you are.
Mistake to avoid
Fixating on one detail and reading the whole photo through that single clue.
Are fixed clues, temporary details, and guesses kept separate?
Create broad candidate regions
What to look at
Whether climate, terrain, language, architecture, and vegetation fit the photo.
What to write down
Two to five possible regions, with a short note on which clues support each.
Mistake to avoid
Writing the first place that comes to mind as the only option.
Do you have a short comparable list, not one guess?
Try 3D terrain and mountain silhouette
What to look at
Skyline, valley direction, and viewing angle in Google Earth or PeakVisor.
What to write down
Which candidate fits better, which was ruled out, and why.
Mistake to avoid
Confirming a place because the silhouette looks similar.
Is the 3D match used as one supporting signal, not final proof?
Cross-check maps and imagery
What to look at
The same fixed clues in OpenStreetMap, Mapillary, satellite imagery, or archives.
What to write down
Source name, date, and what you found; supporting and conflicting points.
Mistake to avoid
Reaching a conclusion from one source alone.
Do at least two independent sources support the same candidate?
Check time context
What to look at
Shadow direction with SunCalc, season, snow line, weather, and image date.
What to write down
Whether the date claim fits, is unclear, or conflicts.
Mistake to avoid
Skipping image date and focusing only on place.
Do the place and date claims conflict with the visible clues?
Report with a confidence level
What to look at
All your notes: supporting clues, ruled-out candidates, open questions.
What to write down
One of verified, strong possibility, possible, unverified, or refuted, with a short reason.
Mistake to avoid
Writing an uncertain result as a definite sentence.
Do supporting evidence, conflicts, and limits appear in the report?
Short glossary
- Fixed clue
- Something in the photo that does not change easily: skyline, road layout, building type, sign.
- Candidate region
- A possible area still being tested. Keep more than one candidate.
- 3D terrain
- Elevation and viewing-angle models in tools like Google Earth, used for silhouette comparison.
- Confidence level
- A label for how sure you are: verified, possible, unverified, and similar.
- GEOINT
- Geospatial intelligence from imagery. Here it means comparing maps and photos, not a magic locator.
Before you report
- Fixed clues are written on separate lines
- At least two candidate regions are listed
- 3D or map comparison was done
- A second independent source was used
- Confidence level and reason are included
- Sensitive location detail is reduced when needed
What you should have
Candidate region list, eliminated options, supporting clues, limits, and confidence.
Sources I open
What I can and cannot say
The photo supports a candidate area only when several fixed clues align; exact location should not be overstated.
Move to evidence notesSources
Start with these sources.
AFP Fact Check
Verification and fact-checking
Agence France-Presse fact-checking site covering viral claims, images, videos, and misinformation.
Open sourceBellingcat Online Investigation Toolkit
Verification and fact-checking
Curated toolkit and guides for open-source investigation methods.
Open sourceBing Visual Search
Image and reverse-image search
Microsoft visual search tool for images, products, landmarks, and similar visuals.
Open sourceCopernicus Browser
Satellite imagery
European Copernicus Data Space browser for Sentinel imagery and products.
Open sourceExifTool
Image and reverse-image search
Command-line utility for reading, writing, and comparing image and media metadata.
Open sourceFotoForensics
Image and reverse-image search
Educational image analysis site with error level analysis and metadata tools.
Open source