一种改进的小波去噪方法在红外图像中应用  

Improved wavelet de-noising method for infrared image application

作　　者：易清明[1] 陈明敏[1] 石敏[1] 

机构地区：[1]暨南大学信息科学技术学院,广州510632

出　　处：《计算机工程与应用》2016年第1期173-177,共5页Computer Engineering and Applications

基　　金：广东高校产学研结合示范基地及科技成果转化重大项目(No.cgzhzd1103);广东省教育部产学研结合项目(No.2012B091000133);广东省工程技术研究中心项目(No.2012gczx A003)

摘　　要：针对小波软阈值去噪函数会产生恒定误差导致图像边缘模糊的缺点,提出了一种改进阈值函数的去噪算法。该算法中当小波系数较大时,阈值函数趋向于硬阈值函数;当小波系数较小时,趋向于软阈值函数,具有自适应性。采用维纳滤波消除图像小波变换中低频频带中残留的噪声。实验结果表明,改进后的阈值函数结合贝叶斯阈值的方法与传统小波软阈值去噪相比,能够有效去除红外图像中的噪声,同时保持红外图像热差细节,具有较高的峰值信噪比,非常适用于去除红外图像中的噪声。The soft threshold function will produce a constant deviation which causes image edges blur. An improved wavelet threshold function method is proposed. When thewavelet coefficient is large, the threshold function is a hard threshold. The threshold tends to be soft threshold when the wavelet coefficient is small. The low-frequency bands use Wiener filter. The improved wavelet threshold function method with Bayes shrink threshold has a good de-nosing effect.The MATLAB simulation results show that compared with the traditional soft threshold wavelet de-nosing method, this method can effectively remove the infrared image noise and maintain the differential thermal infrared image detail with high signal to noise ratio. It is very useful for removing noise in the infrared image.

关 键 词：阈值函数 小波变换 图像去噪 红外图像 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]