Scholarlink

<p>This paper proposes a novel Hadamard Transformation-based image defense methodology tailored for rice grain images, targeting enhanced security and computational efficiency. The pipeline includes halftoning, share generation, secret sharing via Hadamard transformation, and quality-enhanced reconstruction. Privacy-sensitive regions are isolated and resized to reduce processing overhead. The method is evaluated using metrics such as pixel-wise difference, entropy, histogram analysis, radar charts (PSNR/SSIM), and featural integrity, and compared with benchmark techniques. The proposed model achieves 97.01% reconstruction accuracy, 45&#xa0;ms processing time, 60&#xa0;KB memory usage, and demonstrates strong robustness (score: 0.89), scalability (60&#xa0;images/s), and energy efficiency (0.035&#xa0;J/image), outperforming existing approaches in security and performance.</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

Hadamard defense mechanisms for rice grain image privacy

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月1日

查看全文
  • D. Suma,
  • V. G. Narendra,
  • M. Raviraja Holla,
  • M. Darshan Holla

摘要

This paper proposes a novel Hadamard Transformation-based image defense methodology tailored for rice grain images, targeting enhanced security and computational efficiency. The pipeline includes halftoning, share generation, secret sharing via Hadamard transformation, and quality-enhanced reconstruction. Privacy-sensitive regions are isolated and resized to reduce processing overhead. The method is evaluated using metrics such as pixel-wise difference, entropy, histogram analysis, radar charts (PSNR/SSIM), and featural integrity, and compared with benchmark techniques. The proposed model achieves 97.01% reconstruction accuracy, 45 ms processing time, 60 KB memory usage, and demonstrates strong robustness (score: 0.89), scalability (60 images/s), and energy efficiency (0.035 J/image), outperforming existing approaches in security and performance.