基于二维经验小波变换的金属断口图像消噪方法研究  

Research on Denoising Method of Metal Fracture Image Based on Two-Dimensional Empirical Wavelet Transform

作　　者：谷士鹏[1] 常志远 马曼曼 李志农[2] 龙盛蓉 程娟[1] Gu Shipeng;Chang Zhiyuan;Ma Manman;Li Zhinong;Long Shengrong;Cheng Juan(China Flight Test and Research Institute,Xi’an 710089,China;Key Laboratory of Nondestructive Testing Ministry of Education,Nanchang Hangkong University,Nanchang 330063,China;Key Laboratory of Jiangxi Province for Image Processing and Pattern Recognition,Nanchang Hangkong University,Nanchang 330063,China)

机构地区：[1]中国飞行试验研究院,陕西西安710089 [2]南昌航空大学无损检测教育部重点实验室,江西南昌330063 [3]南昌航空大学江西省图像处理与模式识别重点实验室,江西南昌330063

出　　处：《航空科学技术》2022年第10期66-73,共8页Aeronautical Science & Technology

基　　金：航空科学基金(20194603001);江西省图像处理与模式识别重点实验室开放基金(ET202008414);江西省教育厅科学与技术项目(GJJ180525)。

摘　　要：针对传统小波变换在金属断口图像分析中存在的不足,结合二维经验模态分解和小波分析,本文提出了一种基于二维经验小波的断口图像消噪算法。该方法通过自适应构造滤波器组构造了经验小波函数,实现了在频域上分割傅里叶频谱的同时又分离出了信号的不同模式,从而提取了具有紧支撑傅立叶频谱的AM-FM成分。将所提方法与传统小波变换方法进行对比,结果表明,不论是从峰值信噪比还是从结构相似指数来看,二维经验小波算法的图像消噪效果都明显优于小波变换方法。最后,将二维经验小波算法应用到金属断口图像消噪中,进一步验证了所提方法的有效性。Based on the shortcomings of traditional wavelet transform in metal fracture image analysis, a fracture image denoising algorithm is proposed while being combined with two-dimensional empirical mode decomposition and wavelet analysis, i.e., 2D empirical wavelet transform. This method constructs the empirical wavelet function by adaptively constructing the filter bank, realizes the segmentation of the Fourier spectrum in the frequency domain, and separates the different modes of the signal, so as to extract the AM-FM component with tightly supported Fourier spectrum. The proposed method is compared with the traditional wavelet transform. The results show that the image denoising effect of 2D empirical wavelet algorithm is obviously better than that of wavelet transform, whether in terms of peak signal-to-noise ratio or structural similarity index. Finally, the 2D empirical wavelet algorithm is applied to metal fracture image denoising to further verify the effectiveness of the proposed method.

关 键 词：二维经验小波 金属断口 图像消噪 小波变换 二维经验模态分解 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]