摘要
毫米波被动成像具有全天时工作能力。与红外、可见光成像相比,其不足之处是分辨率较低,不能完全反映场景与目标的细节情况。采用小波域正则化方法,首先对毫米波图像进行小波域局部噪声方差估计,然后用自适应正则化方法重构超分辨率毫米波图像。毫米波图像处理的实验证明,该方法消噪效果明显,能锐化图像,保持图像细节。
Millimeter wave passive imaging has the ability to work day. But compared with infrared and visible light imaging, one of most important drawback in millimeter-wave passive imaging is the lower resolution,cannot fully reflect the scene and the target's detail. It is most important to improve millimeter-wave image resolution for better detecting and identifying the target. This paper, we use wavelet domain regularization method.Conduct millimeter-wave image wavelet domain local noise variance estimation, the latter with adaptive regularization method for reconstruction of high-resolution millimeter wave images. The experimental results show that this method significantly noise cancellation, sharpen the image, while preserving image detail.
引文
[1]李兴国,李跃华.毫米波近感技术基础[M].北京:北京理工大学出版社,2009.
[2]程正兴,杨守志,冯晓霞.小波分析的理论、算法、进展和应用[M].北京:国防工业出版社,2007.
[3]张彬,杨风暴.小波分析方法及其应用[M].北京:国防工业出版社,2011.
[4]王知强.一种基于新阈值函数的小波图像去噪算法[J].哈尔滨理工大学学报,2011,16(4):56-58.
[5]矫媛,黄斌文.一种新的小波图像去噪方法[J].科技信息学报,2010,21(3):86-87.
[6]吕慧显,杨斌,李京.基于局部化自适应阈值的小波图像降噪[J].青岛理工大学学报,2010,31(2):118-122.
[7]刘成云,陈振学,马于涛.自适应阈值的小波图像去噪[J].光电工程,2007,34(6):77-81.
[8]钱苏敏,张琳絮,张云,等.基于小波阈值改进的去噪算法研究[J].光电工程,2012,31(5):49-51.
[9]周光祥,杨丹丹.基于自适应阈值的提升小波去噪方法[J].桂林电子科技大学学报,2012,32(2):150-154.
[10]刘毅文,李玲玲,李翠华,等.基于小波多阈值和子带增强的图像去噪[J].厦门大学学报,2012,51(3):342-347.
[11]王艳,胡章芳.图像复原中正则化参数的自适应选择方法[J].重庆邮电大学学报,2007,12,19(6):45-47.
[12]Chen W F,Chen M,Zhou J.Adaptively regularized constrained total least-squares image restoration[J].IEEE Transactions on Image Processing,2000,9(4):588-596.
[13]沈焕锋,李平湘.张良培.一种基于正则化技术的超分辨影像重建方法[J].中国图像图形学报,2005,10(4):436-440.
[14]苗晴,杨灵娥.基于空域正则化的模糊图像复原方法[J].佛山科学技术学院学报,2008,26(2):5-8.