Application of Convolutional Neural Network for Optical Character Recognition Designed for Kazakhstan Identity Cards

2017

Closed access Metadata (CSL-JSON)
DOI 10.1109/icaict.2017.8686945
Data confidence 85 high
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

The paper describes simple model for printed character recognition designed for Kazakhstan identity cards. In our work we used Convolutional Neural Network for recognition of specific symbols such as digits, English and Kazakh alphabet letters. Proposed solution can be directly implemented as part of the text extraction for identity cards.