OCR

Introduction

Optical Character Recognition (OCR) is a technology of detecting and recognizing letters or printed text in image and converting it into machine-readable text data through computer vision. It is a key element of document digitization and a significant area of research in AI, pattern recognition, and computer vision.

Usage

It is commonly used in data entry automation from printed paper data records, digitizing typed texts so that they can be electronically edited, translated, searched, indexed and stored effectively, as well as assisting blind and visually impaired persons. Besides the detection of text from document images, our OCR model can also recognize text from printed product names, traffic warning signs, street name signs, billboard text, and even the vehicle licence plate from a car.

Using the pre-built model for OCR on EdgeAI Studio

To use OCR block for a simple optical character recognition, you can set up a pipeline as below:

TIP

The following languages are supported by our OCR engine: English, Traditional Chinese, Simplified Chinese, French, Italian, Germany, Spanish, Portuguese. You can specified the language(s) in the "OCR" block for recognition.

Examples

Road sign detection and OCR

You may find a demo pipeline here (opens new window).

Automatic car plate detection and OCR

You may find a demo pipeline here (opens new window).