Article: Robust watermarking of medical images using SVM and hybrid DWT-SVD Journal: International Journal of Information and Computer Security (IJICS) 2024 Vol.24 No.3/4 pp.214 - 235 Abstract: In the present scenario, the security of medical images is an important aspect in the field of image processing. Support vector machines (SVMs) are a supervised machine learning technique used in image classification. The roots of SVM are from statistical learning theory. It has gained excellent significance because of its robust, accurate, and very effective algorithm, even though it was applied to a small set of training samples. SVM can classify data into binary classification or multiple classifications according to the application's needs. Discrete wavelet transform (DWT) and singular value decomposition (SVD) transform techniques are utilised to enhance the image's security. In this paper, the image is first classified using SVM into ROI and RONI, and thereafter, to enhance the images diagnostic capabilities, the DWT-SVD-based hybrid watermarking technique is utilised to embed the watermark in the RONI region. Overall, our work makes a significant contribution to the field of medical image security by presenting a novel and effective solution. The results are evaluated using both perceptual and imperceptibility testing using PSNR and SSIM parameters. Different attacks were introduced to the watermarked image, which shows the efficacy and robustness of the proposed algorithm. Inderscience Publishers - linking academia, business and industry through research

Title: Robust watermarking of medical images using SVM and hybrid DWT-SVD

Authors: Kumari Suniti Singh; Harsh Vikram Singh

Addresses: Department of Electronics Engineering, KNIT, Sultanpur 228118, India ' Department of Electronics Engineering, KNIT, Sultanpur 228118, India

Abstract: In the present scenario, the security of medical images is an important aspect in the field of image processing. Support vector machines (SVMs) are a supervised machine learning technique used in image classification. The roots of SVM are from statistical learning theory. It has gained excellent significance because of its robust, accurate, and very effective algorithm, even though it was applied to a small set of training samples. SVM can classify data into binary classification or multiple classifications according to the application's needs. Discrete wavelet transform (DWT) and singular value decomposition (SVD) transform techniques are utilised to enhance the image's security. In this paper, the image is first classified using SVM into ROI and RONI, and thereafter, to enhance the images diagnostic capabilities, the DWT-SVD-based hybrid watermarking technique is utilised to embed the watermark in the RONI region. Overall, our work makes a significant contribution to the field of medical image security by presenting a novel and effective solution. The results are evaluated using both perceptual and imperceptibility testing using PSNR and SSIM parameters. Different attacks were introduced to the watermarked image, which shows the efficacy and robustness of the proposed algorithm.

Keywords: support vector machine; SVM; discrete wavelet transform; DWT; singular value decomposition; SVD; watermark embedding; image watermarking; data security; robustness; DICOM images.

DOI: 10.1504/IJICS.2024.141598

International Journal of Information and Computer Security, 2024 Vol.24 No.3/4, pp.214 - 235

Received: 22 May 2022
Accepted: 03 May 2023
Published online: 26 Sep 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Robust watermarking of medical images using SVM and hybrid DWT-SVD

Keep up-to-date