Scholarlink

<p>Security, privacy, and authenticity of electronic patient records (EPR) are the major concerns in tele-health services. Watermarking is one of the techniques to address these issues. For an accurate diagnosis, the medical data’s region of interest (ROI) is essential. So, these regions must have high visual similarity and resilience after watermarking. The proposed ROI-based watermarking technique is used to protect the integrity ROI of medical data. The medical image is divided into ROI and region of non-interest (RONI) using a maximally stable extremal regions&#xa0;(MSER) feature extraction algorithm. To improve the security of input host image, a hybrid watermark image is generated in redundant domain using patient’s Aadhar card and the fingerprint images. Then the integrity of ROI of medical image is protected using hybrid watermark image. To get the optimum scaling factor Gaussian quantum-behaved particle swarm optimization (GQ-PSO) algorithm is used. Hash values of patient identity is generated using SHA-384 and transmitted on ThingSpeak IoT server for real-time identity verification. To guarantee that the ROI properties remain unchanged, binary robust invariant scalable keypoints (BRISK), are employed. The average percentage improvement in imperceptibility is 57.4232%, robustness is 11.10%, and total execution time is 25.54%. Thus, compared with other existing techniques, the proposed method shows higher robustness, better imperceptibility, and lesser computational complexity. </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

ROI-based optimized image watermarking with real-time authentication

  • 期刊论文

  • 发表日期：

    2025年7月31日

查看全文
  • Divyanshu Awasthi,
  • Vinay Kumar Srivastava

摘要

Security, privacy, and authenticity of electronic patient records (EPR) are the major concerns in tele-health services. Watermarking is one of the techniques to address these issues. For an accurate diagnosis, the medical data’s region of interest (ROI) is essential. So, these regions must have high visual similarity and resilience after watermarking. The proposed ROI-based watermarking technique is used to protect the integrity ROI of medical data. The medical image is divided into ROI and region of non-interest (RONI) using a maximally stable extremal regions (MSER) feature extraction algorithm. To improve the security of input host image, a hybrid watermark image is generated in redundant domain using patient’s Aadhar card and the fingerprint images. Then the integrity of ROI of medical image is protected using hybrid watermark image. To get the optimum scaling factor Gaussian quantum-behaved particle swarm optimization (GQ-PSO) algorithm is used. Hash values of patient identity is generated using SHA-384 and transmitted on ThingSpeak IoT server for real-time identity verification. To guarantee that the ROI properties remain unchanged, binary robust invariant scalable keypoints (BRISK), are employed. The average percentage improvement in imperceptibility is 57.4232%, robustness is 11.10%, and total execution time is 25.54%. Thus, compared with other existing techniques, the proposed method shows higher robustness, better imperceptibility, and lesser computational complexity.