摘要
文中针对现有去噪算法存在的问题,提出了一种基于双正交小波和边缘加权的新的图像去噪算法。该算法对图像进行基于图像移位相关性的自适应二叉分解,研究了白高斯噪声在双正交小波分解下的功率谱,并结合图像的边缘信息,对不同区域的去噪阈值以不同权重加权。实验结果表明,文中算法去噪所得图像的MSE优于小波变换全局阈值去噪,视觉效果明显优于维纳滤波去噪。
To eliminate some defects of temporary image denoising algorithms, an image denoising algorithm based on biorthogonal wavelet and edge detection is developed. The algorithm takes advantage of the shift correlation of image and subbands to perform adaptive binary-tree decomposition of images. This paper analyzes the PSD (Power Spectrum Density) model of the Additive White Gaussian Noise (AWGN) in subbands under biorthogonal wavelet decomposition, and the edge of image is utilized to weigh different coefficie...
出处
《通信技术》
2008年第5期145-148,共4页
Communications Technology
关键词
图像去噪
边缘检测
双正交小波
小波变换
image denoising
edge detection
biorthogonal wavelet
wavelet transform
