Enable an excellent UX in your app with the best in class mobile OCR
Real-time OCR allows mobile app developers to solve a range of text recognition problems using a mobile device camera. Our high precision real-time text recognition is based on the latest research in machine learning and it is highly optimized for mobile devices.
Machine learning approach
DeepOCR is designed as a custom-made machine learning system for OCR. Our research team is continuously finding new ways to design state-of-the-art neural networks, and our development team makes sure that the networks are running fast while requiring minimal memory. This enables high accuracy and speed when supporting even the most complex use cases.
Optimized for mobile
Developed and optimized specifically for mobile devices our text recognition technology uses all the computing power of a mobile device for real-time performance on a device. Parallel processing with SIMD and multithreading, as well as our integrated camera management, enable us to give the fastest and most accurate OCR results.
Many possibilities
Mobile OCR technology is the core of all our products to improve user experience in a wide variety of use cases: from mobile payments, onboarding of new users, KYC, expense tracking, ticket validation, check in at airports and hotels, to security and ID scanning for voting registration. We love to work on new challenging use cases that push our OCR development even further!
Deep reads
User-friendly mobile OCR: Real-time auto capture
Read about real-time auto capture for document detection from a live video stream, for when documents are scanned in imperfect conditions.
From OCR to DeepOCR
Let us take you behind the curtain and show you what it took for us to incorporate machine learning in our OCR development.
Have you heard of Photomath?
The app that reinvented how to learn maths.
Photomath scans and solves mathematical problems in real-time by using a camera of a mobile device. In September 2016 it became the world’s first app solving handwritten equations. It’s among top 5 educational apps with 90M+ downloads. Photomath is powered by DeepOCR technology by Microblink.
