Google Photos can now search for text that appears in user images with its optical character recognition filter
Abner Li / 9to5Google :
Context & Ripple Effects
Google Lens began in 2017 as an in-photo computer-vision search promised for Google Photos and Google Assistant, then gained real-time camera search and smart text selection at Google I/O 2018. This OCR filter is the next step in that arc: instead of pointing the camera at the world, Google is indexing the text already sitting inside users' photo libraries.
It matters because it converts Photos from a chronological archive into a searchable database — the foundation Google later builds on by bringing Lens OCR to Photos on the web and expanding multisearch, which combines an image with a text query, to more than 70 languages.
First-order effects
- Users can now locate screenshots, signs, and documents inside Google Photos by searching for the words visible in them, without manual albums or captions.
- Google extends its Lens technology from a live viewfinder tool into library-wide indexing, making OCR a core Photos capability rather than a camera feature.
Second-order effects
- The OCR index becomes the substrate for Google's later moves: copying text from images on the web and multisearch queries that pair a photo with text, both of which depend on text being extractable from stored images.
- Photo-library search shifts from keyword-tagged metadata toward content-level understanding, raising the baseline competitors' gallery apps must match on search quality.
Third-order effects
- If the pattern holds, personal photo libraries become queryable multimodal databases — searchable by objects, text, and combined image-plus-text queries — with Google's Lens lineage positioning it as the default interface for that shift.
The trend: Visual search is evolving from pointing a camera at objects to indexing everything inside the photo library itself, culminating in multimodal queries that mix images and text.