激光主动成像图像噪声分析与抑制  

Noise analyzing and denoising of intensity image for laser active imaging system

作　　者：李晓峰[1] 徐军[1] 罗积军[1] 曹立佳[1] 张胜修[1] 

机构地区：[1]第二炮兵工程学院,陕西西安710025

出　　处：《红外与激光工程》2011年第2期332-337,共6页Infrared and Laser Engineering

基　　金：国家自然科学基金资助项目(60874093)

摘　　要：分析了激光主动成像图像的噪声机理,针对强度像噪声抑制问题,提出了一种基于同态滤波和小波域SURE(Stein′s Unbiased Risk Estimate)的激光主动成像图像降噪算法。该方法首先用同态变换将乘性散斑噪声转换为加性噪声;然后在小波域,没有将小波系数看作随机变量,而是以最小化均方误差MSE为目的,将图像降噪过程看作是一个小波系数的加权和,通过SURE获得近似最优的小波系数的权值;最后再作相应的小波逆变换和同态逆变换,得到降噪后的图像。实验结果表明:该方法具有较好的噪声抑制效果,且计算量极小。The noise mechanism of laser active imaging system was analyzed. In order to suppress the noise of intensity image, a new image denoising algorithm for laser active imaging system which based on homomorphic filtering and wavelet domain Stein＇s Unbiased Risk Estimate （SURE） was proposed. First, it made the speckle noise convert to the additive noise by homomorphic transform, and then minimized the estimate of mean square error between the clean image and the denoised one in wavelet domain. Contrary to the custom methods, wavelet coefficients were not considered as random variable anymore, the denoising process were parameterized directly as a sum of elementary nonlinear processes with unknown weights. The SURE was introduced to obtain the near-optimal weights. Finally, inverse wavelet transform and inverse homomorpgic transform were carried out and the denoised intensity image was obtained. Experimental results show that the proposed algorithm has advanced denoising performance and the computation time is less than others.

关 键 词：激光主动成像 图像降噪 小波变换 SURE 

分 类 号：TN958.98[电子电信—信号与信息处理]