Signature image improvement with gradient adaptive morphology

Abstract

The paper presents a method for improving signature images, by using a directional field guided morphology. The method adapts a circular or linear structural element and its orientation. There is presented the comparison of the algorithm with the results of binarization process and with the existing post-processing algorithms, such as the morphological opening, closing and median filtering. The experiments show that the authors’ method significantly improves the quality of images by removing unwanted artifacts.