Binarization of Document Images Using Image Dependent Model

Binarization of document images with poor contrast, strong noise complex patterns and variable modalities in the gray-scale histograms is a challenging problem. In this paper we present a binarization algorithm based on image dependen t model to address this problem for the c heque processing application. The proposed algorithm seeks an optimal threshold that w ould eliminate the background noise, while preserving as much character stroke data as possible. The strategy is based on the use of information extracted from one clean part of the image, referred to as the "model" sub-image, to optimize the binarization in another problematic part of the image, referred to as the "target" sub-image. Experiments with 4,200 cheque images, provided by our industrial partner, showed signi cant improvement in the binarization quality in comparison with other well-established algorithms.