What Is Reverse Image Search?

The short definition

A reverse image search uses an image as the query. You upload a photo or paste an image's web address, and the engine returns pages that contain the same image, edited or cropped copies of it, and pictures that look visually similar.

It answers questions plain text search cannot: Where did this photo come from? Who else is using it? Is there a higher-resolution version? Is this profile picture stolen from someone else?

How it works under the hood

Search engines do not compare your image pixel by pixel against the whole web. Instead they convert the picture into a compact numerical fingerprint that captures shapes, edges, colors and textures. That fingerprint is compared against the fingerprints of billions of already-indexed images.

Because each engine builds its own index and uses its own matching model, the same image can return very different results on Google, Bing, Yandex and TinEye. That is exactly why searching more than one engine is worth the extra few seconds.

What people use it for

Common, everyday uses include:

• Finding the original source or photographer of an image
• Checking whether a dating or social profile photo is stolen (catfish check)
• Spotting fake news and out-of-context photos
• Finding a higher-resolution or unwatermarked version
• Identifying products, plants, animals, landmarks and artwork
• Seeing where your own photos have been reposted