PhotoOCR: Reading Text in Uncontrolled Conditions

Abstract

We describe PhotoOCR, a system for text extraction from images. Our particular focus is reliable text extraction from smartphone imagery, with the goal of text recognition as a user input modality similar to speech recognition. Commercially available OCR performs poorly on this task. Recent progress in machine learning has substantially improved isolated character classification; we build on this progress by demonstrating a complete OCR system using these techniques. We also incorporate modern datacenter-scale distributed language modelling. Our approach is capable of recognizing text in a variety of challenging imaging conditions where traditional OCR systems fail, notably in the presence of substantial blur, low resolution, low contrast, high image noise and other distortions. It also operates with low latency; mean processing time is 600 ms per image. We evaluate our system on public benchmark datasets for text extraction and outperform all previously reported results, more than halving the error rate on multiple benchmarks. The system is currently in use in many applications at Google, and is available as a user input modality in Google Translate for Android.

A Text Recognition and Retrieval System for e-Business Image Management

“…Conventional OCR methods are designed primarily for black-white scanned documents and typically rely on brittle techniques such as binarization [1] [21]. Strong assumptions are usually made in the processing such as the presence of horizontal text lines and that characters have the same size, therefore these methods perform poorly on general images.…”

“…Such approaches are usually referred to as PhotoOCR [1]. Some of these methods address sub-tasks such as text detection [8] [13] or text recognition [20], while others attempt to combine both to have an end-toend solution [10] [19].…”

“…Sliding window based methods process text as a standard task of object detection and recognition [1][11]. Wang et al [22] detected characters with Random Ferns and grouped them into words with a pictorial structure model.…”

“…As a result, photo OCR techniques are required [1] [12]. However, most photo OCR methods were developed for text recognition in generic scenes, and have not been specifically designed for e-business image management.…”

Scene Text Recognition and Retrieval for Large Lexicons

“…In light of this, many attempts have been made to recognize scene text [1][2][3][4][5][6]. Scene text recognition is a challenging problem and its recent success is mostly limited to the small lexicon setting, where an image-specific lexicon containing the ground truth word is provided.…”