Research of the Handwriting Recognition Methods

Abstract

This paper examines the methods for handwriting recognition. In the first part, the basic concept of the text recognition system is considered and the features of the operation of these systems are described, it also talks about possible problems in recognition that arise when working with text. The second part examines the difficulties in recognizing handwritten characters associated with the peculiarities of a language, style and handwriting recognition. The third part contains an analysis of methods for preliminary processing of images to
improve the quality of further segmentation and text recognition. The next section describes approaches to segmenting text in a document. The fifth section contains an analysis of some of the methods for classifying handwritten characters. This section talks about the principles of such methods as using templates, using outline models, and also attribute-based methods. The mathematical models and algorithms for each method are described. The conclusion summarizes the state of the handwriting recognition systems and provides a general algorithm for the system's operation in order of its further implementation based on the above analyses.