中文摘要：

为了使数字水印综合性能更好，根据图像邻域像素之间具有很强的相关性这一特点，提出了一种基于支持向量机的图像水印算法。该算法将支持向量机的思想用于数字水印，并取得了较好的效果。由于支持向量机在有限训练样本的情况下具有很好的学习和泛化能力，因此，可以首先利用回归型支持向量机较好地建立图像邻域像素之间的关系模型，然后，通过调整模型的输出值与中心像素值之间的大小关系来嵌入或提取水印。实验表明，用该技术嵌入水印后的图像不仅具有很好的图像感知质量和较强的鲁棒性，对图像增强、JPEG压缩、噪声、几何剪切等抵抗强，而且安全性好、实用性较强。

英文摘要：

Considering the coherence among neighborhood pixels in an image, a kind of spatial domain watermarking scheme based on support vector machine is proposed. It uses support vector machine to embed the watermark and gains satisfied results. Due to the good learning ability and generalization ability of SVM with limited training samples, it can learn the relationship between the selected pixel and its neighboring pixels well with support vector regression. Then, a bit of the watermark is embedded or extracted by adjusting the value between the selected pixel（i, e. desired output） and the actual output of the trained SVR. Experimental results show that the proposed algorithm has good image perceptual quality and high watermark robustness to common image processing operation and the JPEG compression, which also possesses good security and practicability.