Digital Restoration of Deteriorated Mural Images

Abstract

In this paper, an integrated methodology is proposed to virtually enhance the mural images by taking the weighted average of original image with the mean image. The algorithm consists of four major steps as described in the paper. A new line detection and extraction technique using correlation followed by convolution with different templates is implemented and explained. The synthesis of the templates is also explained in detail. Toggle filter is used to enhance the lines. This step is followed by K-means clustering, averaging pixels and weighted average. An idea on recovery of degraded patch is also presented. The results of our experiment are found to be good and may be used to restore deteriorated digital mural images.