基于中值预滤波的航空图像小波去噪算法研究  

Aerial image wavelet transformation denoising based on medium pre-filtering

作　　者：匡海鹏[1] 王德江[1] 张景国[1] 陈志超[1] 张雪菲[1] 刘志明[1] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所,吉林长春130033

出　　处：《应用光学》2010年第2期221-224,共4页Journal of Applied Optics

基　　金：中国科学院长春光学精密机械与物理研究所创新工程资助项目(No.ZJ99130B)

摘　　要：结合航空图像的噪声来源与图像特性,提出一种基于中值预滤波的图像小波去噪算法。图像首先经中值滤波器进行预滤波,滤除随机的脉冲式噪声,然后对处理后的图像进行小波变换,与给定阈值相比,对可明显判为信号或噪声的系数进行相应处理;对不确定为信号或噪声的系数进行多尺度上的相关性追溯,判别其归属后进行处理。实验结果表明:该方法客观上提高了图像的信噪比,主观上使去噪后的图像纹理分明,能更好地适合人眼的视觉特性,有利于航空图像的分析、判读。According to noise characteristics of aerial image, this paper proposes a kind of wavelet transformation denosing method based on medium filtering for aerial image. The medium filter is applied to reduce image random noise, and wavelet transformation is made on filtered image. Compared to given threshold, the wavelet coefficients that could be judged as signal or noise obviously are processed, other coefficients are processed before running down in multi scale of wavelet. Experiments prove that this method improves image signal to noise ratio objectivley, and makes denoised image more clearly and suitable for human visual observation, which is good for analysis and judgement of aerial images.

关 键 词：图像去噪 小波变换 中值滤波 

分 类 号：TN911.7[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]