Analysis of Different Font Types and Sizes for High Quality Watermarked, Extracted Watermark and Medical Images Using HS Based Reversible Watermarking

Abstract

This paper presents a reversible watermarking technique using gray scale medical images based on histogram shifting. The watermark image used is the medical report of the patient. To utilize the maximum embedding capacity provided by histogram shift technique, the font size and type of the medical report text can be varied. This paper explores the effect of varying font types and sizes to observe the impact on the quality of watermarked image, retrieved watermark and extracted cover image. The performance assessment is done on the basis of MSE, SSIM and PSNR. This paper compares four most commonly used fonts to discover the best suited font to achieve highest PSNR for embedded watermarked image, extracted host medical image and extracted watermark. Also the effect of salt and pepper noise on extracted medical image for different fonts is discussed.