中文摘要：

针对图像传感数据传输时的版权保护问题,提出了一种基于压缩感知的双域水印算法。该算法利用离散余弦变换作为稀疏基,在完成压缩感知图像传感数据采集之后,通过矩阵乘法、加法等简单运算将水印嵌入到数据的压缩感知域,其中,带水印的数据在公开信道上传输,水印嵌入密钥在私密信道中传输。接收端在稀疏域进行水印的提取和验证,并重构恢复出图像数据。理论分析表明,该算法可以很好地对所采集的图像传感数据提供版权保护,安全性好。实验结果表明,在提取水印之后,可以得到较高精度的数据;同时,该算法具备很好的鲁棒性,能够有效地保证水印提取质量。

英文摘要：

Aiming at the copyright protection problem of image sensing data transmission, a double-domain watermarking algorithm based on compressive sensing was proposed. In this algorithm, Discrete Cosine Transform （DCT） was used as a sparsity basis, and the watermark was embedded in the compressive sensing domain through the simple matrix multiplication and addition after compressive image sensing. The watermarked image data was transmitted over the public channel and the watermark embedding key was transmitted over the private channel. At the receiver site, the watermark is extracted and verified in the sparse domain, and then the image data was reconstructed. Theoretical analysis shows that the proposed algorithm can provide copyright protection to the image sensing data, and has good security. The experimental results show that it can get highly accurate data after the watermark was extracted. At the same time, the proposed algorithm has a good robustness, which can ensure the quality of the extracted watermark.