Presentation of an effective method of text recognition in the image, based on the criteria of similarity of structural models of symbols similarity criteria

Abstract

The object of research is the methods of handwriting recognition. The article reviews the basic principles of optical recognition. The aim of the work is to present an algorithm for text recognition in the image, which has high accuracy with a small training sample. To achieve this goal, an original similarity criterion is proposed, which can be used to effectively compare the formed structural models of symbols. The reader is offered a detailed consideration of the algorithm of the mentioned similarity criterion. Ref. 5, pic. 4

Keywords

• character recognition
• structural model
• similarity criterion
• bipartite graph