A Hybrid Image Restoration Algorithm for Digitized Degraded Historical Documents

The historical documents have quite importance in our life because historical documents hold the important information about the old civilizations. However, because of many factors these documents are found in degraded form. Historical documents contain important information so we need to restore them. There are hundreds of thousands historical documents are available. If these documents are restored manually then it is a time consuming task. Now days, the historical documents are stored in the computer memories in the form of digital historical documents. The digitized historical documents are easy to restore than the manual restoration process on the historical documents. In this paper, the hybrid method is used to restore the historical documents. The proposed algorithm is capable of restoring most of the digitized historical documents accurately. This algorithm is capable of removing the most of the noises and spots from the digitized historical documents. This algorithm is the hybridization of morphological operations, image de-noising filters (wiener filter [10], median filter, etc), image dilation, OTSU thresholding, and Sauvola thresholding techniques. The normalized absolute error (NAE), Mean Square Error (MSE), and peak signal to noise ratio (PSNR) is taken as the performance parameters. The performance of the proposed algorithm is based on these parameters is better and the results are higher than the existing algorithms, when compared.