Image-based search engines analyze a picture and compare it to other images found online. Those pictures that are similar are then sent back to the user using the service. The software that finds the picture varies according to the company that uses it, such as TinEye and Google.
Google's image search engine works by using the picture uploaded into the search bar and creating a mathematical model based on its shapes, colors and proportions. Using this information, it matches it to images already found in Google's index, which results in pictures that are sometimes identical, but also similar in both size and design.
TinEye is one of the most popular reverse image search engines, and is often used to help find images that are illegally being produced. The search engine works by reading an image URL or file format, and then comparing it with its own database of images. Results then link the users to other websites where the image is being hosted. TinEye has several features from its service, such as Mobile image recognition, match engine and TinEye API. The match engine looks for duplicate photos, while the mobile engine is allows users to take pictures of items and search them online.