Bobin Jérôme, Starck Jean-Luc, Fadili Jalal M, Moudden Yassir, Donoho David L
DAPNIA/SEDI-SAP, Service d'Astrophysique, CEA/Saclay, 91191 Gif sur Yvette, France.
IEEE Trans Image Process. 2007 Nov;16(11):2675-81. doi: 10.1109/tip.2007.907073.
In a recent paper, a method called morphological component analysis (MCA) has been proposed to separate the texture from the natural part in images. MCA relies on an iterative thresholding algorithm, using a threshold which decreases linearly towards zero along the iterations. This paper shows how the MCA convergence can be drastically improved using the mutual incoherence of the dictionaries associated to the different components. This modified MCA algorithm is then compared to basis pursuit, and experiments show that MCA and BP solutions are similar in terms of sparsity, as measured by the l1 norm, but MCA is much faster and gives us the possibility of handling large scale data sets.
在最近的一篇论文中，提出了一种称为形态成分分析（MCA）的方法，用于从图像的自然部分中分离纹理。MCA依赖于一种迭代阈值算法，使用一个在迭代过程中线性递减至零的阈值。本文展示了如何利用与不同成分相关联的字典的互不相干性来大幅提高MCA的收敛速度。然后将这种改进的MCA算法与基追踪进行比较，实验表明，就由l1范数衡量的稀疏性而言，MCA和BP的解相似，但MCA速度更快，使我们有能力处理大规模数据集。
