四元数奇异值分解与彩色图像去噪

提出了一种基于四元数的彩色图像去噪算法.该算法对彩色图像进行四元数奇异值分解，得到表征彩色图像的不同分量的奇异值，同时适当地选择和丢弃分别表征图像和噪声的奇异值，可以有效地去除彩色图像的加性噪声.该算法的特点是采用一种新颖的基于彩色图像能量测度模型，自适应地确定去噪图像重构的奇异值数目，因此具有快速去噪和简单可行的优点.实验结果表明，提出的方法针对彩色图像去噪具有较好的效果.

Quaternion singular value decomposition approach to color image de-noising

A color image de-noising algorithms have been proposed based on quaternion. By using corresponding knowledge of quaternion singular value decomposition, the singular values on the diagonal matrix obtained through QSVD represent different components in color image. The additive noise of a color image can be eliminated effectively through selecting the proper singular values that represent signal and discarding the ones that represent noise. By a color image energy model, the singular number for image reconstruction and to eliminate the noise is adaptively determined, so the algorithms can denoise rapidly and it can be also implemented easily in practice. The experiment results show that the algorithms and the model are effective.