Computer Vision Methods for Recognition Data of Financial Organization

This graduation project is devoted to the research study conducted for automation of activities of internal audit services in financial organization. Within the audit checks there was an attempt to develop a text recognition application which will deal with scanned copies of documents, which confirms client’s income. The requirement in such instrument stems from the need to analyze big amount of data in order to validate the correctness of client’s income calculated by bank employees. This graduation work set out the aims and purposes of the research study. The first chapter deals with the theory of internal audit and the audit objectives. The second chapter contains the overview of computer vision methods and the business-logic of developed solution. The third chapter of the project gives details about the results of the research and the technical implementation of the text recognition algorithm.