A Multistage Binarization Technique for the Degraded Document Images

Abstract

This paper presents a multistage binarization technique for the degraded document images. The proposed technique can deal with many types of the degradations, where we propose edge detection methods to find the edges of the objects in the image even if this image suffers from inhomogeneities. Then, to find the rest of the objects contents, we propose a combination of Niblack's method, integral image, and a machine learning technique, where Markov random field is applied in an energy minimization framework using graph cut. Also we propose new formulas to measure the binarization efficiently and a new method for the binarization measurement. The results of extensive experiments on many datasets show the robustness of the proposed technique on various types of degradations in the document images where our technique demonstrates superior performance against also many other methods.