摘要
为进一步提高图像质量,提出一种基于高密度离散小波变换的改进图像降噪方法。给出二维高密度离散小波变换的分解与重构快速算法,通过该算法对图像进行多尺度分解,利用相邻尺度小波系数相关性对各层小波系数进行双变量收缩阈值处理,重构降噪后的图像。实验结果表明,与其他常用小波降噪方法相比,该方法能进一步提高图像降噪效果,且在降噪过程中较好地保留图像细节。
To improve the quality of the image,this paper presents an improved image denoising method based on high density discrete wavelet transform.The two-dimensional fast decomposition and reconstruction algorithm is given,and it is used to decompose the image in multi-scale.The wavelet coefficients at each level are processed with bivariate shrinkage threshold according to the correlation of wavelet coefficients of adjacent scales.The denosed image is reconstructed.Experiments show that compared with other wavelet denoising method,the method proposed in the paper further enhances the image denoising performance,and still keeps the details of the image.
出处
《计算机工程》
CAS
CSCD
2012年第1期211-214,共4页
Computer Engineering
基金
国家"111"计划基金资助项目(B08036)
关键词
高密度
小波变换
双变量收缩阈值
图像降噪
high density
wavelet transform
bivariate shrinkage threshold
image denoising
