摘要： 为更好地对图像进行稀疏表示，以改善去噪效果，提出一种傅里叶变换与非下采样轮廓波变换(NSCT)相结合的自适应阈值去噪算法。在傅里叶域中对含噪图像去噪，在NSCT域中利用分层噪声估计的贝叶斯阈值算法，结合多尺度多方向的能量阈值修正方案自适应地滤除剩余噪声。实验结果表明，该算法的去噪性能较好。
关键词:
图像去噪,
傅里叶变换,
非下采样轮廓波变换,
自适应阈值,
贝叶斯框架
Abstract: In order to express a sparse image better and achieve better de-noising effect, an adaptive threshold de-noising algorithm combined with Fourier Transform(FT) and Non-subsampled Contourlet Transform(NSCT) is proposed. The noisy image is de-noised in Fourier domain, and the remaining noise is filtered out in NSCT domain. It is based on stratified noise estimation and adaptive Bayes threshold, combined with a flexible multi-scale and multi-directional energy correction threshold scheme. Experimental results show that the proposed algorithm can improve de-noising performance.
Key words:
image de-noising,
Fourier Transform(FT),
Non-subsampled Contourlet Transform(NSCT),
adaptive threshold,
Bayes frame
中图分类号:
李利, 杨恢先, 何雅丽, 冷爱莲, 岳许要. 结合FT与NSCT的自适应阈值去噪算法[J]. 计算机工程, 2012, 38(17): 238-241.
LI Li, YANG Hui-Xian, HE Ya-Li, LENG Ai-Lian, YUE Hu-Yao. Adaptive Threshold De-noising Algorithm Combined with Fourier Transform and Non-subsampled Contourlet Transform[J]. Computer Engineering, 2012, 38(17): 238-241.