摘要: 基于人眼视觉系统特性,根据图像几何结构特征,提出一种多尺度脊波字典的构造方法。构造出的多尺度脊波字典具有多分辨率、多尺度、各向异性和多方向等特性,能够更为稀疏地表示图像,从而得到一种基于多尺度脊波字典的稀疏性约束图像去噪算法。实验结果表明,该去噪算法能有效去除噪声,更好地保留图像的边缘等细节。
关键词:
稀疏表示,
脊波,
图像去噪,
匹配追踪
Abstract: A multiscale ridgelet dictionary is constructed in terms of the perception character of human visual system and geometric properties of the local structures. The multiscale ridgelet dictionary is multiresolution, multiscale, anisotropy and multidirectional. It can provide sparser representation for image. Furthermore, a sparsity constrained image denoising algorithm based on multiscale ridgelet dictionary is proposed. Experimental results indicate that the proposed denoising algorithm can effectively reduce noise, and preserve the image edge details.
Key words:
sparse representation,
ridgelet,
image denoising,
matching pursuit
中图分类号:
邓承志. 基于多尺度脊波字典的图像去噪算法[J]. 计算机工程, 2010, 36(23): 207-208,211.
DENG Cheng-Zhi. Image Denoising Algorithm Based on Multiscale Ridgelet Dictionary[J]. Computer Engineering, 2010, 36(23): 207-208,211.