A practical approach to validate the authenticity of identity documents

There is a need for a system capable of determining the authenticity of identity documents in a simple way due to the increase over the years of crimes such as fraud and identity theft, this in the specific case of Costa Rica; the reason for this situation is that many processes and transactions can be carried out virtually, which opens the door for this type of crime to take place. For the above purpose, the present work consists of the research, design, and development of a prototype system to determine whether an identification document is authentic. This authenticity estimator consists of a composition of algorithms applied to images, such as filters and transformations; in order to perform a process called Optical Character Recognition (OCR), which consists of the recognition of symbols or characters from a given image. This proposal uses classical computer vision techniques, although it is important to mention that machine learning techniques can also be used for this purpose. Nevertheless, with this approach, 80 percent of accuracy was obtained in the recognition of identity documents.