Scholarlink

<p>The large volume of medical data that needs to be effectively stored, retrieved, and distributed as well as the rising number of security threats that must be expressly addressed have resulted in new problems for healthcare information management. Multiple watermark based strategies can be utilized simultaneously for various applications of image watermarking, whereas the real-time utilization of a single watermark-based strategy has a number of constraints. In this paper, multiple watermark images are embedded into cover medical image using two different approaches for security of medical image for smart healthcare system. In first approach, MNNIT ‘logo’ watermark is embedded into LL sub-band of Lifting wavelet transform (LWT)-Schur decomposition processed ultra-sound of liver image of patient. Second watermark image ‘Aadhar’ is embedded in HL sub-band of cover image processed using LWT-Discrete cosine transform. In the suggested method watermarked image is encrypted using Arnold-cat map encryption and FAST features of cover image and watermarked image are matched for authentication. The hybrid optimization techniques based on harmony search optimization and gray wolf optimization technique is utilized in the proposed method because it overcomes the parameter selection limitations that most metaheuristic algorithms face. In this work, the watermark is pre-processed in Schur domain to avoid false positive problem. Under no attack, the proposed technique gives a normalized correlation coefficient (NCC) value of 1.000 and a peak signal to noise ratio (PSNR) of around 45&#xa0;dB. The effectiveness of the proposed method is also assessed against hybrid attacks and compared with the existing schemes; in most of the cases this method improves the PSNR value more than 10&#xa0;dB and NCC value more than 10% with low computational time makes it suitable for smart healthcare system.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

MIWHSmark: Multiple Image Watermarking Based on Hybrid Technique in Schur Domain for Smart Healthcare System

  • 期刊论文

  • 发表日期：

    2025年6月11日

查看全文
  • Anurag Tiwari,
  • Vinay Kumar Srivastava

摘要

The large volume of medical data that needs to be effectively stored, retrieved, and distributed as well as the rising number of security threats that must be expressly addressed have resulted in new problems for healthcare information management. Multiple watermark based strategies can be utilized simultaneously for various applications of image watermarking, whereas the real-time utilization of a single watermark-based strategy has a number of constraints. In this paper, multiple watermark images are embedded into cover medical image using two different approaches for security of medical image for smart healthcare system. In first approach, MNNIT ‘logo’ watermark is embedded into LL sub-band of Lifting wavelet transform (LWT)-Schur decomposition processed ultra-sound of liver image of patient. Second watermark image ‘Aadhar’ is embedded in HL sub-band of cover image processed using LWT-Discrete cosine transform. In the suggested method watermarked image is encrypted using Arnold-cat map encryption and FAST features of cover image and watermarked image are matched for authentication. The hybrid optimization techniques based on harmony search optimization and gray wolf optimization technique is utilized in the proposed method because it overcomes the parameter selection limitations that most metaheuristic algorithms face. In this work, the watermark is pre-processed in Schur domain to avoid false positive problem. Under no attack, the proposed technique gives a normalized correlation coefficient (NCC) value of 1.000 and a peak signal to noise ratio (PSNR) of around 45 dB. The effectiveness of the proposed method is also assessed against hybrid attacks and compared with the existing schemes; in most of the cases this method improves the PSNR value more than 10 dB and NCC value more than 10% with low computational time makes it suitable for smart healthcare system.