一种基于卡通纹理分解的X光图像恢复方法*

针对X光图像去噪时在抑制噪声的同时会模糊图像边缘的情况,提出采用图像卡通纹理分解和基于全变分的增广拉格朗日算法进行图像恢复.图像可以分解为卡通部分和纹理部分,噪声信息及图像的快变信息被分离到图像的纹理部分.通过基于增广拉格朗日算法的全变分去噪模型对纹理图像进行去噪处理,将卡通图像与处理后的纹理图像加权合成得到恢复图像.仿真实验结果表明,该方法不仅可以对图像进行快速处理,而且能够较好地保持图像的边缘信息,获得较高的输出信噪比.

X-ray Image Restoration Method Based on Cartoon Texture Decomposition

While the X-ray image denoised, there is a conflict between noise suppression and edge blurring, and a kind of approach is proposed by using the cartoon texture image decomposition and the ALM to realize image restoration. The image can be decomposed into the cartoon part and texture one, and the noise information and the quick change information of the image can also be pulled in the texture of the image. It is thought to select the ALM to denoise the texture image, and then compose the cartoon part and the restoration of the texture part to achieve restoration image. It is well shown by the results that this method can not only shorten the time of image processing, but also preserve the edge information of the image, and obtain higher output PSNR than the traditional TV method.