Scholarlink

Ensuring the authenticity of LiDAR data is essential to prevent manipulations that could affect decision-making in critical applications. This work presents a watermarking technique that embeds and extracts digital marks directly onto a luminance map generated from the RGB channels of LiDAR point clouds. The method operates in the spatial domain and uses the zero-frequency (DC) coefficient in \(4 \times 4\) pixel blocks to embed watermark bits, avoiding the need to compute the full Fourier transform. A quantization strategy is employed to adjust pixel intensities and embed the information imperceptibly. The effectiveness of the technique is evaluated in terms of imperceptibility and robustness, using the Peak Signal-to-Noise Ratio (PSNR) metric. Simulations demonstrate that it is possible to maintain a balance between visual imperceptibility and resistance to tampering under various quantization step values, offering a lightweight and practical solution for LiDAR data authentication.

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

Authentication in LiDAR Data Using a Spatial Domain Technique

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Valeria López-Rodríguez,
  • Moisés Salinas-Rosales,
  • Manuel Cedillo-Hernández

摘要

Ensuring the authenticity of LiDAR data is essential to prevent manipulations that could affect decision-making in critical applications. This work presents a watermarking technique that embeds and extracts digital marks directly onto a luminance map generated from the RGB channels of LiDAR point clouds. The method operates in the spatial domain and uses the zero-frequency (DC) coefficient in \(4 \times 4\) pixel blocks to embed watermark bits, avoiding the need to compute the full Fourier transform. A quantization strategy is employed to adjust pixel intensities and embed the information imperceptibly. The effectiveness of the technique is evaluated in terms of imperceptibility and robustness, using the Peak Signal-to-Noise Ratio (PSNR) metric. Simulations demonstrate that it is possible to maintain a balance between visual imperceptibility and resistance to tampering under various quantization step values, offering a lightweight and practical solution for LiDAR data authentication.