Binarization of Color Historical Document Images Using Local Image Equalization and XDoG

This paper presents a new binarization method for color images of degraded historical document. The proposed method makes use of local image equalization based on color constancy, and an extension to the standard difference of Gaussians edge detection operator, XDoG. The binarization is achieved after three main steps: the first step removes undesirable degradation artifacts using a local image equalization and Otsu binarization algorithm. The second step uses global image equalization and XDoG edge detection operator to binarize the text. The final step combines the two previous steps, performing a cleanup to remove remaining degradations artifacts and fix possible missing text or area, to produce the final result. The proposed approach proved to achieve more satisfactory results than other state-of-the-art methods.