强激光与粒子束, 2019, 31 (10): 103210, 网络出版: 2019-10-14  

宽带电磁图像卷积神经网络盲恢复方法研究

Research on blind recovery method of wideband electromagnetic image convolutional neural network
作者单位
北京航空航天大学 电子信息工程学院 电磁兼容技术研究所, 北京 100191
摘要
在利用抛物反射面对电磁干扰源成像过程中, 由于系统衍射受限及成像频带较宽, 导致干扰源成像模糊, 分辨率低, 难以分辨, 不同频率不同区域干扰源所成图像分辨率不同, 采用已有超分辨算法难以提高分辨率。为了实现宽带电磁图像的盲复原, 应用卷积神经网络的方法。网络训练是直接输入模糊图像, 不假设任何特定的模糊和噪声模型情况下, 重建出高质量图像。实验和仿真结果证明了卷积神经网络盲恢复方法在宽频带不同成像区域下表现了优于其他盲恢复算法的优势。
Abstract
In the process of electromagnetic interference sources imaging testing using parabolic reflection, the diffraction phenomenon of the system leads to blurred and low resolution images. Interference sources have different resolution capabilities in different areas and frequencies, so it’s difficult to enhance image resolution by using existed super-resolution algorithm. In order to realize the blind recovery of wideband electromagnetic images, a method based on convolutional neural network is proposed. Network training is the process which directly inputs a blurred image and reconstructs a high quality image without assuming any particular blur and noise model. Both experiment and simulation results demonstrate that the convolutional neural network blind recovery method outperforms other blind recovery calculations in different imaging regions of wide frequency band.
参考文献

[1] 李良超, 杨建宇, 郑鑫,等.无源毫米波超分辨算法及实时性研究[J].电子科技大学学报, 2009, 38(6): 952-956.(Li Liangchao, Yang Jianyu, Zheng Xin, et al. Research on passive millimeter wave image super-resolution algorithm and real-time implementation. Journal of University of Electronic Science and Technology of China, 2009, 38(6): 952-956)

[2] 杜鑫,谢树果,郝旭春,等.一种电磁干扰源成像多分辨率分区算法[J].强激光与粒子束, 2015, 27: 103223.(Du Xin, Xie Shuguo, Hao Xuchun, et al. An electromagnetic interference source imaging algorithm of multi-resolution partitions. High Power Laser and Particle Beams, 2015, 27: 103223)

[3] 李垣江, 李跃华, 张光锋.一种改进的毫米波图像超分辨复原方法[J].现代雷达, 2011, 33(8): 43-46.(Li Yuanjiang, Li Yuehua, Zhang Guangfeng. A modified method of super-resolution restoration for passive millimeter wave images. Modern Radar, 2011, 33(8): 43-46)

[4] Chan T F, Wong C. Total variation blind deconvolution diffraction[J]. IEEE Trans Image Processing, 1998, 7(3): 370-375.

[5] Krishnan D, Tay T, Fergus R. Blind deconvolution using a normalized sparsity measure[C]//IEEE Conference on Computer Vision and Pattern Recognition. 2011: 233-240.

[6] Levin A, Weiss Y, Durand F, et al. Efficient marginal likelihood optimization in blind deconvolution[C]//IEEE Conference on Computer Vision and Pattern Recognition. 2011: 2657-2664.

[7] Xu L, Zheng S, Jia J. Unnatural L0 sparse representation for natural image deblurring[C]//IEEE Conference on Computer Vision and Pattern Recognition. 2013: 1107-1114.

[8] Comaniciu D, Meer P. Mean shift analysis and applications[C]//Proceedings of IEEE 7th International Conference on Computer Vision. 1999, 2: 1197-1203.

[9] Sun L, Cho S, Wang J, et al. Edge-based blur kernel estimation using patch priors[C]//IEEE Conference on Computational Photography. 2013: 1-8.

[10] Michal H, Jan K, Pavel Z, et al. Convolutional neural networks for direct text deblurring [C]//British Machine Vision Conference. 2015: 1-12.

[11] Oliver W, Josef S, Andrew Z, et al. Non-uniform deblurring for shaken images[C]//2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 2010.

朱艳菊, 谢树果, 李元豪, 张娴. 宽带电磁图像卷积神经网络盲恢复方法研究[J]. 强激光与粒子束, 2019, 31(10): 103210. Zhu Yanju, Xie Shuguo, Li Yuanhao, Zhang Xian. Research on blind recovery method of wideband electromagnetic image convolutional neural network[J]. High Power Laser and Particle Beams, 2019, 31(10): 103210.

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!