摘要
提出了一种新的基于多个小波基的图像融合去噪方法。首先利用多个不同的小波基对含噪图像进行阈值去噪,得到多幅恢复图像。然后对这些图像采用小波融合方法进行融合。对于低频系数采用基于边缘的融合算法,在多幅恢复图像中选择最有可能是边缘的点加以保留;对于高频系数,采用了平均的融合算法。最后得到一幅去噪图像。实验结果表明,无论是在视觉效果上还是在峰值信噪比定量指标上该方法去噪效果均明显优于单一小波基去噪。
A novel image denoising method using multiple wavelet bases and fusion technology was proposed. First, every wavelet base was employed to threshold denoise, and recover the image, and then a set of the denoised images were obtained. Next, these denoised images were fused using wavelet transform. While choosing the low frequency coefficients, we selected the pixels that might be the edges from these denoised images. When choosing the high frequency coefficients, we used average method. Finally, the denoised image was reconstructed by the inverse wavelet transform using the new wavelet coefficients. Experimental results showed that the new method presented here was much effective in removing white noise, and gave better Peak Signal Noise Ratio （PSNR） gains than single wavelet base denoising.
出处
《光电工程》
EI
CAS
CSCD
北大核心
2007年第11期103-107,共5页
Opto-Electronic Engineering
基金
国家自然科学基金资助项目(60374053)
河海大学科技创新基金资助项目(06B002-02)
常州国家高新区科技项目(XE120060408)
关键词
小波变换
小波基
图像融合
图像去噪
wavelet transform
wavelet base
image fusion
image denoising
