Document Analysis and Recognition

The cornerstone of this project is to create a DAR system that will convert its functionality into applications in the areas of research and commercial systems. The system will allow the user to convert printed text into an editable digital format, making recognition easier and more efficient. The system will rely on well-known methods, such as image processing, pattern recognition and artificial intelligence, to which will be added a unique technique, namely natural image processing. Effective use of the CNN algorithm would help reduce image blur while effectively recognizing text. The entire system will be a threestep process of creating a digital document from a hard copy, scanning the digital document, and recognizing templates in the document for correct, error-free results.