摘要
介绍了提升方法(Lifting Scheme)的基本原理,给出了用提升方法构造传统小波的实现方法。在提升小波分解变换的基础上,研究一种自适应阈值的图像去噪方法——AdaptThr Shrink去噪法。这种方法是基于Bayes框架,在不同子带和不同方向上选择不同的最佳阈值。结合软阈值法对图像进行去噪,与传统方法相比,此种方法提高了去噪后图像的峰值信噪比(PSNR),而且使图像更加清晰。基于提升小波的自适应阈值图像去噪法实现简单、计算速度快、去噪效果好。
Introduces basic principle of lifting scheme and presents method of construction of traditional wavelets via lifting scheme. An adaptive threshold based on lifting wavelet transform for image denoising is studied. This method is derived in a Bayesian framework and threshold is chosen according to different subbands and orientations. Comparing with traditional denoising methods, this method combined with soft threshold algorithm, can improve the PSNR more effectively and also makes denoised image more clearly. Adaptive threshold based on lifting wavelet transform,which can be computed fast with a simple implementation,has a good effect for image denoising.
出处
《计算机技术与发展》
2008年第2期42-45,共4页
Computer Technology and Development
关键词
小波变换
提升方法
阈值
图像去噪
wavelet transform
lifting scheme
threshold
image denoising
