摘要
在小波变换域中去除图像中的噪声是近年来的研究热点之一。目前在小波域中对加性噪声的去除已经有了许多研究结果,比如Donoho等的处理方法都得到了很好的应用。但是由于超声图像噪声情况的复杂性,其对去噪的方法提出了更高的要求。为了在去除噪声的同时能够更好的保护边缘及有用的细节信息,本研究结合Birg-éMassart等提出的非参数自适应估计理论,提出一种在平稳小波变换域中对超声图像去噪的方法。实验证明,这种基于非参数自适应估计理论的超声图像去噪方法,与Donoho阈值去噪方法相比,去噪效果有所提高。
It is one of the research hotspots that de-noising images in wavelet transform domain recently. There are many methods for additive noise reduction have been presented in wavelet domain now. For example the thresholding method proposed by Donoho, which has got wide applications in signal processing field. But it makes a higher request for de-noising method for the ultrasonic images, since the noises situation is complex in ultrasonic images. To remove noise＇s effects and preserve more useful edges and details in the mean time, we proposed a new de-noising method by using the threshold method based on nonparametric adaptive estimation which is presented by Birgé-Massart in stationary wavelet transform domain in this paper. The experiments show that our de-noising method represents better characteristic than the de-noising method based on Donoho strategy.
出处
《中国生物医学工程学报》
CAS
CSCD
北大核心
2006年第4期400-403,416,共5页
Chinese Journal of Biomedical Engineering
基金
教育部留学回国人员科研启动基金资助项目(2004.176.4)
山东省自然科学基金资助项目(Z2004G01
2004ZRC03016)
关键词
非参数
自适应估计
二维平稳小波变换
超声图像去噪
nonparametric estimation
adaptive estimation
2-D stationary wavelet transform
ultrasonic image de-noising
