DEVELOPMENT OF AN OPTICAL CHARACTER RECOGNITION SYSTEM FOR THE AUTOMATION OF SECURITY CHECKS

Rusakov Alexey Mikhailovich, Filatov Vyacheslav Valeryevich, Dolzhenkov Sergey Sergeevich, Antonov Emil Petrovich, Andreev Dmitry Igorevich, Kulikov Daniil Mikhailovich

This article deals with different approaches for optical character recognition (OCR) systems on images in order to automate security checks. A review and analysis of current methods and algorithms for character recognition in streaming images is given. The basic technology for optical information processing is considered, and the scope of this technology in the field of information security is determined. The features of CAPTCHA solution automation implementation are shown. OpenCV and Tesseract are considered to be the main libraries. The article develops CAPTCHA analysis software architecture in order to detect the characters depicted on it. The main points in realization of this software solution are given and it is shown that it is possible to apply this software solution in practice thus showing vulnerabilities of CAPTCHA security checking systems.

Keywords:Image character recognition, image mining, CAPTCHA recognition, computer vision