A two-stage binarization approach for document images

Binarization of a gray scale document image is one of the most important steps for automatic document processing. The paper presents a two-stage document image binarization approach. The approach applies a region based binarization technique first to the whole image and utilizes a neural network based binarization technique to those text blocks in which a good character segmentation cannot be achieved at the first stage. Experimental results on a number of document images show that our two-stage binarization approach performs better than other binarization techniques in terms of character segmentation quality and computing time