<p>The surge of digital infrastructures demands robust and efficient image encryption schemes to ensure privacy and security. This paper presents a lightweight image encryption scheme that integrates the Walsh Hadamard Transform (WHT) with chaos-driven confusion diffusion for grayscale and RGB images. The novelty lies in performing binary chaotic masking directly in the WHT domain, coupled with an explicitly defined composite key space from logistic and tent maps with optional Arnold iterations, which strengthens resistance to CPA/KPA while preserving low arithmetic complexity. The contribution includes per-channel WHT decorrelation, chaos-guided coefficient permutation, transform-domain XOR diffusion, and a transparent security evaluation using entropy, H/V/D correlation, NPCR, UACI, and runtime. Experiments on standard images resized to powers of two yield cipher entropies of 7.996–7.999 bits/pixel, near-zero adjacent-pixel correlations, NPCR = 99.6%, UACI = 33%, and execution times of 10–15 ms for 256 × 256 and 512 × 512 images, demonstrating strong randomness, differential robustness, and suitability for real-time, resource-constrained deployments.</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

Image encryption technique using walsh transform and chaotic systems

  • Adesh Kumar,
  • Jaishree

摘要

The surge of digital infrastructures demands robust and efficient image encryption schemes to ensure privacy and security. This paper presents a lightweight image encryption scheme that integrates the Walsh Hadamard Transform (WHT) with chaos-driven confusion diffusion for grayscale and RGB images. The novelty lies in performing binary chaotic masking directly in the WHT domain, coupled with an explicitly defined composite key space from logistic and tent maps with optional Arnold iterations, which strengthens resistance to CPA/KPA while preserving low arithmetic complexity. The contribution includes per-channel WHT decorrelation, chaos-guided coefficient permutation, transform-domain XOR diffusion, and a transparent security evaluation using entropy, H/V/D correlation, NPCR, UACI, and runtime. Experiments on standard images resized to powers of two yield cipher entropies of 7.996–7.999 bits/pixel, near-zero adjacent-pixel correlations, NPCR = 99.6%, UACI = 33%, and execution times of 10–15 ms for 256 × 256 and 512 × 512 images, demonstrating strong randomness, differential robustness, and suitability for real-time, resource-constrained deployments.