Scholarlink

<p>This paper proposes a Histogram of Oriented Gradients (HOG) based high capacity and blind image steganography technique with tamper detection and content authentication capabilities to verify the integrity and authenticity of the secret and confidential data. In this approach, an HOG feature vector is computed for each block of the cover image to determine the dominant gradient direction. The gradient magnitude corresponding to the dominant gradient direction is thresholded to generate Blocks of Interest (BOIs), and the message bits are embedded in the BOIs according to the dominant gradient direction. The simulation results demonstrate that the proposed technique improves the payload by up to 27.2% and raises the Peak Signal-to-Noise Ratio (PSNR) value by up to 3.4 dB compared to other state-of-the-art methods. By adjusting the values of the embedding parameters, the proposed scheme provides a good balance between capacity and image quality. Authentication analysis confirms that the proposed method exhibits high fragility to various attacks, with the Number of Blocks Tampered (NBT) reaching up to 75.7%, highlighting its effectiveness in tamper detection and content authentication. Additionally, the proposed technique is computationally efficient with reduced embedding and extraction times compared to the state-of-the-art approaches.</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

Hog based high capacity and blind image steganography technique with tamper detection and content authentication

  • 期刊论文

  • 发表日期：

    2025年12月29日

查看全文
  • Iffat Rehman Ansari,
  • Mohd Ayyub Khan,
  • Ekram Khan

摘要

This paper proposes a Histogram of Oriented Gradients (HOG) based high capacity and blind image steganography technique with tamper detection and content authentication capabilities to verify the integrity and authenticity of the secret and confidential data. In this approach, an HOG feature vector is computed for each block of the cover image to determine the dominant gradient direction. The gradient magnitude corresponding to the dominant gradient direction is thresholded to generate Blocks of Interest (BOIs), and the message bits are embedded in the BOIs according to the dominant gradient direction. The simulation results demonstrate that the proposed technique improves the payload by up to 27.2% and raises the Peak Signal-to-Noise Ratio (PSNR) value by up to 3.4 dB compared to other state-of-the-art methods. By adjusting the values of the embedding parameters, the proposed scheme provides a good balance between capacity and image quality. Authentication analysis confirms that the proposed method exhibits high fragility to various attacks, with the Number of Blocks Tampered (NBT) reaching up to 75.7%, highlighting its effectiveness in tamper detection and content authentication. Additionally, the proposed technique is computationally efficient with reduced embedding and extraction times compared to the state-of-the-art approaches.