摘要
图像的平移不变量小波去噪法是小波收缩阈值去噪法的改进,能够有效抑制小波阈值去噪过程中出现的伪吉布斯现象.分析了图像的平移不变量小波去噪法的原理及其算法,并将该方法应用于颗粒显微图像的去噪.对采用不同去噪方法的图像按照分水岭算法进行分割,然后进行对比.实验结果表明图像的平移不变量小波去噪法的去噪效果较好。
Translation invariant (TI) wavelet de-noising method is an improved method of wavelet thresholding shrinks; it can suppress Pseudo-Gibbs artifacts effectively. The theory and algorithm of TI wavelet de-noising method are analyzed in this paper. The same microscopic particle image is de-noised by the proposed method and two conventional methedsand the de-noised images are segmented with watershed algorithm. Comparison tests show that the method of TI wavelet de-noising method is better than others.
出处
《河海大学常州分校学报》
2006年第4期17-20,共4页
Journal of Hohai University Changzhou
基金
河海大学常州校区科技创新基金资助项目(05B002-04)
关键词
平移不变量
小波收缩
伪吉布斯现象
颗粒图像
translation invariant
wavelet shrink
Pseudo-Gibbs artifacts
particle image