Development and Research of OCR Algorithms

Automation of digitization of paper documents is topical in all public and commercial organizations. The object of the study in this paper are the character recognition algorithms and the subject of the study - an OCR system. The main aim if the study: to develop a high accuracy algorithm of Optical Character Recognition from passport images by combining the new image preprocessing methods with Tesseract OCR library and vocabulary control of the results. The result of the is the prototype of optical character recognition system which is able to to recognize data from passports. THe application which has been elaborated during the study is approbated on set of test data. Testing of application was conducted and there was a successful implementation of this system in the territorial fund of obligatory medical insurance of the Nizhny Novgorod. Its further development and improvement will elevate the quality of recognition and increase functionality.