ID Engine: automatic data extraction from Ukrainian national documents in night vision mode
The id.engine system powered by GreenOCR allows you to scan various types of IDs in more than 96 languages in vivo. The document may contain inscriptions made in the Latin alphabet, Cyrillic, hieroglyphs and numbers. There’s no need for the user to create any special conditions for shooting, look for additional lighting, align the document in front of the camera or aim the camera to make a shot for a long time for data extraction. It allows you to scan and recognize identity documents (passports, ID cards, driver’s license) on mobile or stationary devices equipped by a camera.
Here we show the capabilities of the system to automatically extract data from the fields of Ukrainian national documents: Ukrainian ID, Ukrainian national passport, Ukrainian driving license - in very difficult conditions of insufficient lighting, at different picture skew and rotation. id.engine (https://smartengines.com/ocr-engines/...) can recognize in twilights the documents which are printed both in Cyrillic and in Latin at the same time. The system also extracts data from the page of Ukrainian passport, where the machine-readable zone (MRZ) is located.
All you need is to present the document to the camera, and Smart IDReader will recognize all required data in less than 1 second and then fill them in any file, form or a worksheet.
About Smart Engines:
Smart Engines is leading the way to a new era of mobility, channelling the latest scientific knowledge into world-class document recognition systems that meet the challenges of the digital revolution.
Smart Engines Limited specialises in researching artificial intelligence – with particular expertise in computer vision software. Our class-leading proprietary AI-based recognition systems support public and private organisations worldwide by significantly improving business processes through superior document scanning and analysis. Smart Engines technologies help our clients to solve the problems in recognition and data extraction from many types of documents.
web https://smartengines.com/ocr-engines/...
linkedin / smart-engines
medium / smartengines
twitter / smartenginesllc
The documents demonstrated in the video contain synthetic data made for demonstration purposes and do not present real faces nor private data.