CAJ | 학술논문

提出一种基于加速鲁棒性特征算法（Speed-Up Robust Features，SURF）的抗几何攻击彩色图像水印算法。首先，在彩色图像中的红色分量中检测SURF特征点，并且生成SURF特征描述算子。然后，利用二级离散小波变换（Discrete Wave-let Transform，DWT）得到图像蓝色以及绿色分量的低频信息，并对其进行奇异值分解（Singular Value Decomposition，SVD）。最后将经过一级小波变换后的水印图像对应嵌入彩色图像蓝色分量和绿色分量的低频区域之中。实验结果表明，嵌入水印后的图像在经过几何攻击后，其提取的水印与原水印仍具有较高的相似度，该算法具有较好的鲁棒性。
제출일충기우가속로봉성특정산법（Speed-Up Robust Features，SURF）적항궤하공격채색도상수인산법。수선，재채색도상중적홍색분량중검측SURF특정점，병차생성SURF특정묘술산자。연후，이용이급리산소파변환（Discrete Wave-let Transform，DWT）득도도상람색이급록색분량적저빈신식，병대기진행기이치분해（Singular Value Decomposition，SVD）。최후장경과일급소파변환후적수인도상대응감입채색도상람색분량화록색분량적저빈구역지중。실험결과표명，감입수인후적도상재경과궤하공격후，기제취적수인여원수인잉구유교고적상사도，해산법구유교호적로봉성。
An color image watermarking algorithm that resistant to geometric attack was proposed in this paper, which is based on Speed-Up Robust Features algorithm (SURF). First of all, detected the SURF feature points in the red component of the color image, and generated the SURF feature description operator. Then, two-level Discrete Wavelet Transform (DWT) was used to get the low frequency information of the blue and green components of the image, and carried on the Singular Value Decomposition (SVD). Finally let the watermark image which transformed from one-level wavelet transform embedded to low frequency region of the blue and green components of the color image accordingly. Experimental results showed that the extraction of the watermark from the embedded watermark image was still in high similarity with the original watermark after geometric attack , and the algo-rithm had good robustness.