摘要
对小波阈值去噪中的常用阈值和阈值函数进行分析,提出一种自适应的模糊阈值去噪算法,该算法在BayesShrink阈值基础上,通过增加一个修正因子,并结合模糊理论,自适应地对图像进行模糊阈值函数处理。实验表明该算法与BayesShrink软阈值函数去噪算法相比,去噪后图像的峰值信噪比PSNR和最小均方误差MSE均有所提高,并且图像也更清晰,具有较好的去噪效果。
Common thresholds and threshold functions in wavelet shrinkage image denoising are analyzed in this paper. A new adaptive fuzzy threshold filter algorithm based on wavelet transform is proposed in this paper. The improved algorithm modifies BayesShrink by increasing an amendatory factor , and combines fuzzy theory to denoise . It can be adaptive to denoise image by fuzzy threshold function. Compared with soft- threshold algorithm based on BayesShrink and improved algorithm, experimental results show that improved algorithm is not only improves the PSNR and MSE, but also makes denoised image more clear. It is effective to image denoising.
出处
《计算技术与自动化》
2008年第1期79-81,共3页
Computing Technology and Automation
基金
湖南省高等学校科学研究项目(07C087)
关键词
图像去噪
小波系数
阈值
阈值函数
模糊理论
image denoising
wavelet coefficient
threshold
threshold function
fuzzy theory