Image segmentation as a preliminary step for character recognition in ancient printed documents

Bedini L, Tonazzini A

OCR  Image segmentation  Image restoration  [Image Processing 

After analyzing and processing several ancient printed documents, we argued that the joint restoration and segmentation of images is the first fundamental step for character recognition, which usually relies on isolated characters. This step is particularly critic in the case of ancient printed documents where several degradation processes may cause the characters to touch and merge one another. In this paper we propose to to integrate techniques of image restoration with techniques of image segmentation based on Markov Random Field models. Several results of both simulated and real experiments are shown to validate the method.