Program for Handwriting Imitation

This paper is dedicated to handwriting imitation. In this paper, the areas of potential use of automatic handwriting generation are described as well as the examples of the existing methods of handwriting synthesis and their advantages and disadvantages. The proposed handwriting generation algorithm is described, which includes such steps as recognition of user-provided handwriting sample which is a pre-written and scanned text, then, generation a database of characters based on the location of each character and its neighbours, and, finally, the creation the required text using letters from database. The main goal of this work is the development of handwriting imitation program - Desktop-application, running on the Windows operating system. To create the application, the language Python and numpy, opencv, scikit-learn were used. The paper contains 61 pages, 3 chapters, 50 images, 1 tables, 30 references, 4 appendices. Key Words: optical character recognition, handwriting imitation, computer vision