摘要
小波变换不能够最优地表示图像的边缘,而Curvelet变换硬阈值去噪后的图像过于平滑。二代离散Cur-velet变换运算速度非常快,而且基于Bayes原理的自适应阈值选择是子带变化的,具有最小的Bayesian风险。提出了一种基于二代Curvelet变换同Bayes原理相结合的自适应图像去噪算法,实验结果表明,该算法不仅能够有效地去除了噪声、较好地保留了图像的边缘信息,而且运算快速。
Wavelet transform cant excellently express image edges, and images after Curvelet transform denoising based on hard threshold is too smoothing. Second generation discrete curvelet transform operates rapidly and adaptive threshold selection based on Bayes theory varies along with sub-bands, so it has the least Bayesian risk. The paper proposed an adaptive image denoising algorithm based on second generation curvelet transform and Bayes theory. Experiment results show that the algorithm could not only remove the noise while preserving image edges, but also operate rapidly.
出处
《测绘科学》
CSCD
北大核心
2009年第5期84-86,共3页
Science of Surveying and Mapping
