ZeroTrace OSINT
Reverse Image Composer
Generate optimised reverse-image-search queries across Google Lens, Yandex, Bing, TinEye, SauceNAO, and more.
The reverse image composer takes a photo (local file or URL) and builds optimised search queries for the major reverse-image-search engines. It pre-processes the image into multiple search variants to maximise hit rate.
The toolkit does not host an image index — it composes deep links into the public engines that do. You run the searches in your browser; the composer makes the searches one-click.
Why a composer
Reverse image search is the single most-asked OSINT capability. Every engine has different strengths:
- Google Lens — best general-purpose, especially for objects and text in images.
- Yandex — best for non-Western scenes, outdoor photos, and faces (often beats Google for European / Asian geolocation cases).
- Bing Visual Search — second-best general purpose, sometimes catches what Google misses.
- TinEye — exact-match focus; finds where the same image has appeared even after re-encoding.
- SauceNAO — anime, illustration, art-focused. Useful for identifying source material in a non-photographic case.
- Karma Decay — Reddit-focused; good for tracking memes and viral images.
A typical reverse-image investigation runs four or five engines and merges the results. The composer makes that one button-press per engine.
Search variants
The composer generates four pre-processed variants of your image:
| Variant | What it does | Best for |
|---|---|---|
| Original | Unchanged | Baseline |
| Centre-cropped square | 50% area, centred | Subjects in the middle of the frame |
| Top half / bottom half | Two sub-images | Subjects in one quadrant; image-of-image cases |
| Greyscale + edge-only | Sketch-style transform | Yandex-style scene matching |
Each variant gets its own deep-link set across all engines. The result is a grid of one-click searches you can fire off in any order.
Multi-variant hashing
The composer also computes:
- pHash, dHash, aHash, wHash — useful for spotting whether two images are near-identical.
- SHA-256 of the bytes — useful for spotting whether two images are byte-identical.
Hashes are a cheap up-front check before a reverse-search. If you already have a candidate match, comparing hashes locally is faster than running another engine.
Yandex in particular often beats Google for non-Western scenes — outdoor photos taken in Europe, Asia, or anywhere with non-English signage. If Google Lens returns nothing useful, run the same image against Yandex before giving up.
Bulk-paste links
The composer offers a "copy all links" affordance — every engine + every variant as a flat list. Useful when you want to:
- Open them all in tabs at once (with a browser extension).
- Send a colleague a complete search-set.
- Save the search set as part of an investigation profile.
What the composer does not do
- It does not run the searches. You click each link; the search engine runs in your browser. The composer constructs the URLs.
- It does not scrape engine results. All result interpretation happens in your browser, manually.
- It does not upload your image to ZeroTrace. The image stays on your machine. The image is uploaded to the search engine only when you click the link — that upload goes from your browser directly to the engine.
Pivots
| Click on... | Pivot to |
|---|---|
| The image | Image metadata, geo clues, sun & shadow solver |
| Image hashes | Photo clustering |
| Search engine link | Opens in browser |
Sources
Every engine the composer constructs links for is named in the result. Click-through is to the named engine; no intermediary, no proxy. The image you upload (when you click) goes from your browser to the engine over the engine's TLS connection.