摘要
针对红外图像去噪领域中,传统小波阈值选取不当造成的图像模糊、峰值信噪比低等问题,提出采用以“3σ”准则确定小波阈值的方法。该方法以高斯噪声的统计特性为基础,若分解后的小波系数幅值大于3σ则保留,小于3σ则置零。将保留的小波系数采用小波逆变换重构,可得去噪后红外图像。去噪实验表明,该方法有效可行,与传统硬、软阈值法、SVD分解法相比,该方法还原红外图像效果更佳。
To solve the problems of image blur and low peak signal-to-noise ratio of the traditional infrared imagery denoising method caused by improper wavelet threshold,a method of determining wavelet threshold by“3σ”criterion was proposed.With the statistical characteristics of Gaussian noise as the foundation,if the amplitude of wavelet coefficient after decomposition was greater than 3σ,it would be retained;otherwise,it would be set to zero.The reserved wavelet coefficients were reconstructed by inverse wavelet transform,thus obtaining the denoised infrared image.The experiment results showed that the method was effective and feasible.The effect of the proposed method was better than those of the hard and soft threshold method and the SVD decomposition method.
作者
孙婷婷
崔少华
SUN Ting-ting;CUI Shao-hua(Department of Computer Science,Huaibei Vocational and Technical College,Huaibei 235000,China;College of Physics and Electronic Information,Huaibei Normal University,Huaibei 235000,China)
出处
《辽东学院学报（自然科学版）》
CAS
2020年第3期201-204,共4页
Journal of Eastern Liaoning University:Natural Science Edition
基金
安徽省大学生创客实验室项目(2016ckjh182)
淮北职业技术学院自然科学研究重点项目(2018-A-2)。
