The curvelet transform for image denoising
The curvelet transform for image denoising
复制标题
DOI：
10.1109/icip.2001.958937
复制
发表时间：
2001-10
期刊：
影响因子：
--
通讯作者：
Jean-Luc Starck;E. Candès;D. Donoho
中科院分区：
文献类型：
--
作者：
Jean-Luc Starck;E. Candès;D. Donoho
仅给出摘要形式，如下所示。我们提出了两种新数学变换的近似数字实现，即脊波变换和曲波变换。我们的实现提供了精确的重建，稳定的扰动，易于实施，计算复杂性低。我们应用这些数字变换去噪的一些标准图像嵌入在白色噪声。在这里报道的测试中，简单的曲波系数阈值是非常有竞争力的“最先进的”技术的基础上小波，包括阈值的抽取或未抽取小波变换，也包括基于树的贝叶斯后验均值方法。此外，曲波重建表现出比基于小波的重建更高的感知质量，提供视觉上更清晰的图像，特别是更高质量的边缘和微弱的线性和曲线特征的恢复。
Summary form only given, as follows. We present approximate digital implementations of two new mathematical transforms, namely, the ridgelet transform and the curvelet transform. Our implementations offer exact reconstruction, stability against perturbations, ease of implementation, and low computational complexity. We apply these digital transforms to the denoising of some standard images embedded in white noise. In the tests reported here, simple thresholding of the curvelet coefficients is very competitive with 'state of the art' techniques based on wavelets, including thresholding of decimated or undecimated wavelet transforms and also including tree-based Bayesian posterior mean methods. Moreover, the curvelet reconstructions exhibit higher perceptual quality than wavelet-based reconstructions, offering visually sharper images and, in particular, higher quality recovery of edges and of faint linear and curvilinear features.