Randomized redundant DCT: efficient denoising by using random subsampling of DCT patches
Randomized redundant DCT: efficient denoising by using random subsampling of DCT patches
复制标题
DOI：
10.1145/2820903.2820923
复制
发表时间：
2015-11
期刊：
影响因子：
--
通讯作者：
Shu Fujita;Norishige Fukushima;M. Kimura;Y. Ishibashi
中科院分区：
文献类型：
--
作者：
Shu Fujita;Norishige Fukushima;M. Kimura;Y. Ishibashi
J在本文中，我们提出了一种加速方法，用于使用冗余离散余弦变换（R-DCT）进行图像denosing。图像denoising对于图像处理至关重要，其效率对于图形应用很重要。带有限制性或收缩方法的R-DCT可以在保持细节纹理的同时执行脱氧。此外，该方法与最先进的去核方法（例如BM3D）相比在计算上是有效的。但是，计算成本仍然不足以实时处理。因此，我们通过使用DCT斑块的随机亚采样来加速该方法。实验结果表明，我们的方法可以加速处理，而降解性能的降解有点。
j In this paper, we propose an acceleration method for image denoising with a redundant discrete cosine transform (R-DCT). Image denoising is essential for image processing, and its efficiency is important for graphics applications. R-DCT with a hard-thresholding or shrinkage method can perform denoising while keeping detail textures. Moreover, the method is computationally efficient compared with state-of-the-art denoising methods, such as BM3D. The computational cost, however, is still insufficient for real-time processing; hence, we accelerate the method by using randomized subsampling of DCT patches. Experimental results show that our method can accelerate the processing while the degradation of denoising performance is a little.