A Novel Binarization Method to Remove Verdigris from Ancient Metal Image

Abstract

Nowadays, preserving the knowledge from the ancient metal image has become a tedious process. In ancient temples and old Illams, many important scripts are written on the metal images, due to aging, verdigris, dirt, color change and variation in atmospheric temperature, pressure and humidity, where the images degrade the visual impact of the letters present in them. These types of degradation challenges can be obtained by using image binarization. The proposed novel method effectively removes the degradation from metal images. The degraded metal image is passed to Fast NL means denoising method, which converts the resultant image as gray scale and increases the contrast enhancement, wherein the contrast enhanced image is placed for edge detection and applied an adaptive thresholding. The thresholded image is passed to a morphological open in post enhancement stage to finally obtain the output image by which all types of degradations are removed. Proposed method is giving the exceptional accuracy of about 95.98%.