A 'Divide and Embed' approach in a robust crypto-watermarking technique for enhancing the embedding capacity. (July 2022)

• Record Type: Journal Article
• Authors: Verma, Usha; Sharma, Neelam
• Is Part Of: Biomedical signal processing and control. Volume 76(2022)
• Publication Date: 2022-07
• Subjects: Medical image watermarking -- Hybrid embedding -- Crypto-watermarking -- Region-based watermarking; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.103694
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Address the trade-off between embedding capacity and perceptual quality. Higher embedding capacity without affecting the imperceptibility and robustness. 'Divide and Embed' approach to enhance the embedding capacity. One solution for authentication, preservation, robustness, imperceptibility and embedding capacity. Abstract: Medical Image watermarking techniques have a major concern of addressing the issue of embedding capacity as the electronic health record (EHR) of a patient is large in size. Embedding large size watermark into medical images degrades the perceptual quality of the image which may lead to changes in morphological features of the tissues or organs. The main aim of the proposed work is to address the challenge of the trade-off between embedding capacity and image quality along with maintaining robustness. A 'Divide and Embed' approach is proposed to enhance the embedding capacity without affecting the image perceptibility and robustness. This approach is based on the availability of multiple medical images with a single patient's record. The patient information used as a watermark is divided into several parts equal to the number of medical images available with a patient record and embedded separately using a region-based crypto-watermarking technique with a wavelet-based hybrid embedding process. The projected technique is tested by varying embedding capacity to check for both imperceptibility and robustness against various unintentional and intentional attacks for both CT scan and MR Images. The PSNR achieved is 77.9 dB for the higher embedding capacity of 56, 548 pixels by maintaining the robustness with average BER to 0.0609 and average NC to 0.9406 after applying various attacks. Therefore, the proposed approach has enhanced the embedding capacity by maintaining the perceptual quality of the medical image as well as robustness is also consistent as compared to the other researcher's work.