The aim of this paper is to resolve what exactly lies behind the optical character recognition process, what its components are, and how it looks in practice. An optical character recognition, in simple words is translating an analog document into a searchable digital version and the concept of digitization is explained as the root part of the theoretical part of the paper, and what are the projects and initiatives of digitization that would not exist without optical character recognition. This process is interwoven with several related disciplines, of which the pattern recognition, computer vision, and machine learning are briefly explained in the paper. The theoretical part of the paper is formulated around a description of eight major components of the optical character recognition process: optical scanning, segmentation of image regions, preprocessing, normalization, segmentation, representation, feature extraction, and teaching and recognition. Although most of the literature is mathematically based, the components of optical character recognition are mathematically illustrated to the limit understandable to all. Optical Character Recognition applications were then introduced and a comparison study was performed on ABBYY FineReader 15, Free OCR, Google Drive OCR, I2OCR and Convertio applications. The methodology used in the research is a qualitative comparison of five different optical character recognition applications and a quantitative comparison of how many errors were made and how many languages were included in the architecture of each application. After the research, a discussion of the results was conducted and conclusions were drawn based on the research and the theory analyzed.