多幅相同场景的图像的降噪  

Multiple Copy Image Denoising

作　　者：贺卫国[1] 黎绍发[1] 

机构地区：[1]华南理工大学计算机科学与工程学院,广州510641

出　　处：《计算机工程与应用》2005年第12期15-17,共3页Computer Engineering and Applications

基　　金：国家自然科学基金资助项目(编号:60273064)

摘　　要：随着数字图像采集设备的发展,利用多幅相同场景的图像以获得更高质量的图像的方法已具有现实意义。文章分析了多幅图像降噪中面临的问题,并以高斯白噪声污染图像为例,论证了多幅图像的最佳线性组合方法,采用自适应维纳滤波与均值滤波相结合的降噪方法,试验证明了多幅图像降噪效果比单幅图像降噪效果明显要好。With the development of digital image acquisition equipment,the method to make use of multiple images with same objects to improve image quality is of usefulness.This paper analyzes the possible existed questions in multiple copy image denoising.An example is presented by using multiple images corrupted by white Gaussian noise.It proposes a method of linear combination and proves it is the optimal linear combination.Adaptive Wiener filter cascading mean filter is chosen as the denoising method.Simulation proves multiple copy image decosing is better than single image denoising.

关 键 词：维纳滤波 高斯白噪声 降噪 图像融舍 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]