Hybrid Method for Elimination of Uneven Illumination from Camera-based Document Images

Abstract

In this paper, is addressed the issues of uneven lighting and complex background of the camera-based document image using hybridization of new split and merge method. The input image is divided into four uniform size sub-images, and CLAHE enhancement technique was applied to all four sub-images to rectify noise amplification in each part of input image and then adaptive document binarization operation was used on four sub-images. Then, all four sub section images are merged to get the binarized image with elimination of noises, uneven illumination. Encouraging results are obtained by this method as compared on to other existing methods of Sauvola, Feng and Kasar found in the literature. The comparative analysis is made on collected datasets, as the standard datasets are not available.