A review of segmentation and contextual analysis techniques for text recognition

摘要

This paper presents a review of the literature on text-processing techniques. The techniques covered are segmentation of text and contextual recognition, both of which are required when considering text recognition of documents. Several different methods of text segmentation are compared for various image data formats. The problem of correct segmentation of joined and broken characters is also considered. Two techniques for contextual recognition are considered: the Markov-based methods and dictionary look-up methods. The various techniques for storing dictionary information are compared. A discussion of the importance of the choice of the correct context is given, together with guidance on which methods are best suited to which applications.