时空域自适应滤波非均匀性校正算法

郭玉婷; 贾晓洪; 李丽娟; 刘俊明

时空域自适应滤波非均匀性校正算法

郭玉婷^1,,
贾晓洪^{1, 2},
李丽娟^{1, 2},
刘俊明^{1, 2}

1.
中国空空导弹研究院, 河南洛阳 471009
2.
航空制导武器航空科技重点实验室, 河南洛阳 471009

详细信息

作者简介:
郭玉婷（1998-）女，河南洛阳人，硕士研究生，主要研究方向为红外导引头信息处理技术。E-mail：641536183@qq.com

中图分类号: TJ761.3
计量
- 文章访问数: 181
- HTML全文浏览量: 101
- PDF下载量: 62
出版历程
- 收稿日期: 2022-04-18
- 修回日期: 2022-06-01
- 刊出日期: 2023-05-19

Space-Time Domain Adaptive Filtering Non-uniformity Correction Algorithm

GUO Yuting^1,,
JIA Xiaohong^{1, 2},
LI Lijuan^{1, 2},
LIU Junming^{1, 2}

1.
China Air-to-Air Missile Research Institute, Luoyang 471009, China
2.
Aviation Key Laboratory of Science and Technology on Airborne Guide Weapons, Luoyang 471009, China

摘要

摘要: 由于红外焦平面探测器受到制造工艺等限制，图像不可避免地会存在非均匀性。传统神经网络算法会留下“鬼影”的问题，本文改进传统神经网络算法，利用引导滤波图像作为期望模板，防止图像的边缘被滤波器平滑。当场景运动时，通过时域迭代的策略来不断进行非均匀性校正参数的更新。为了抑制算法中常见的鬼影现象，设计了基于空域局部方差和时域场景变化率相结合的自适应学习率，利用前后的校正参数自适应调整阈值。实验仿真表明，本文所提的算法相比于传统算法均方根误差下降45.45%左右，可以在校正图像非均匀性的同时很好地抑制“鬼影”现象。
- 红外焦平面 /
- 非均匀性校正 /
- 基于场景 /
- 神经网络 /
- 引导滤波
Abstract: Because the infrared focal plane detector is limited by manufacturing technology, the image is inevitably nonuniform. The traditional neural network algorithm solves the "ghost" problem using the guided filtering image as the expected template to prevent image edge smoothing by the filter. When the scene is moving, the nonuniformity correction parameters are continuously updated using the time-domain iteration strategy. To suppress the common ghosting phenomenon in the algorithm, an adaptive learning rate was designed based on a combination of the spatial local variance and the time-domain scene change rate, and the threshold was adjusted adaptively using the correction parameters before and after. Simulation results show that the root mean square error of the proposed algorithm is reduced by 45.45% compared with that of the traditional algorithm, and the proposed algorithm can suppress the "ghost" phenomenon well while correcting image nonuniformity.
- infrared focal plane /
- non-uniformity correction /
- scene-based /
- neural network /
- guided filter

HTML全文

图 1 神经网络法原理

Figure 1. Neural network method

下载: 全尺寸图片幻灯片

图 2 室内场景校正过程图

Figure 2. Indoor scene correction process diagram

下载: 全尺寸图片幻灯片

图 3 室内场景校正前后的图像粗糙度对比

Figure 3. Roughness comparison of indoor scene before and after correction

下载: 全尺寸图片幻灯片

图 4 室内场景校正后均方根误差对比

Figure 4. RMSE comparison of indoor scenes after correction

下载: 全尺寸图片幻灯片

图 5 室外场景校正过程图（每组图中左侧为传统神经网络算法，右侧是本文改进算法）

Figure 5. Outdoor scene correction process diagram(In each set of figures, the left is the traditional neural network algorithm, and the right is the improved algorithm in this paper)

下载: 全尺寸图片幻灯片

图 6 室外场景校正前后的图像粗糙度对比

Figure 6. Roughness comparison of outdoor scene before and after correction

下载: 全尺寸图片幻灯片

图 7 室外场景校正后均方根误差对比

Figure 7. RMSE comparison of outdoor scenes after correction

下载: 全尺寸图片幻灯片

参考文献(17)

[1]	马晓平, 赵良玉. 红外导引头关键技术国内外研究现状综述[J]. 航空兵器, 2018(3): 3-10. https://www.cnki.com.cn/Article/CJFDTOTAL-HKBQ201803001.htm MA Xiaoping, ZHAO Liangyu. An overview of infrared seeker key technologies at home and abroad[J]. Aero Weaponry, 2018(3): 3-10. https://www.cnki.com.cn/Article/CJFDTOTAL-HKBQ201803001.htm
[2]	周永康, 朱尤攀, 赵德利, 等. 基于场景的红外焦平面非均匀校正算法综述[J]. 红外技术, 2018, 40(10): 952-960. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS201810005.htm ZHOU Yongkang, ZHU Youpan, ZHAO Deli, et al. Overview of scene-based non-uniform correction algorithms for infrared focal plane[J]. Infrared Technology, 2018, 40(10): 952-960. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS201810005.htm
[3]	李旭, 杨虎. 基于两点的红外图像非均匀性校正算法应用[J]. 红外与激光工程, 2008, 37: 608-610. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ2008S2064.htm LI Xu, YANG Hu. Application of a non-uniformity correction algorithm for IRFPAs based on two points[J]. Infrared and Laser Engineering, 2008, 37: 608-610. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ2008S2064.htm
[4]	关同辉, 张同贺. 一种新型实时两点非均匀性校正方法[J]. 航空兵器, 2021, 28(4): 112-117. https://www.cnki.com.cn/Article/CJFDTOTAL-HKBQ202104018.htm GUAN Tonghui, ZHANG Tonghe. A new real-time two-point non-uniformity correction method[J]. Aero Weaponry, 2021, 28(4): 112-117. https://www.cnki.com.cn/Article/CJFDTOTAL-HKBQ202104018.htm
[5]	陈芳林. 基于场景的红外焦平面非均匀性校正算法及FPGA实现[D]. 南京: 南京理工大学, 2017. CHEN Fanglin. Scene-based Infrared Focal Plane Non-Uniformity Correction Algorithm and FPGA Implementation[D]. Nanjing: Nanjing University of Science and Technology, 2017.
[6]	钱润达, 赵东, 周慧鑫, 等. 基于加权引导滤波与时域高通滤波的非均匀性校正算法[J]. 红外与激光工程, 2018, 47(12): 1204001-1-1204001-6. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ201812022.htm QIAN Runda, ZHAO Dong, ZHOU Huixin, et al. Non-uniformity correction algorithm based on weighted guided filter and temporal high-pass filter[J]. Infrared and Laser Engineering, 2018, 47(12): 1204001 -1-1204001-6. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ201812022.htm
[7]	徐泽林, 路东明, 王利平, 等. 利用灰度差估计的条纹非均匀性校正方法[J]. 光学学报, 2021, 41(5): 511001-1-511001-6. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB202105004.htm XU Zelin, LU Dongming, WANG Liping, et al. Fringe non-uniformity correction method based on gray different estimation[J]. Acta Optica Sinic, 2021, 41(5): 511001-1-511001-6. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB202105004.htm
[8]	CHAO Zuo, QIAN Chen, GU Guohua, et al. Improved interframe registration based nonuniformity correction for focal plane arrays[J]. Infrared Physics & Technology, 2012, 55(4): 263-269. . http://www.onacademic.com/detail/journal_1000035395320710_83c7.html
[9]	ZUO C, ZHANG Y, CHEN Q, et al. A two-frame approach for scene-based nonuniformity correction in array sensors[J]. Infrared Physics & Technology, 2013, 60(1): 190-196. http://www.zuochao.org/uploads/1/1/0/7/11076000/2013_irt_tfnuc.pdf
[10]	李谦, 杨波, 粟宇路, 等. 基于神经网络的红外焦平面光学非均匀性校正改进算法[J]. 红外技术, 2019, 41(3): 251-255. http://hwjs.nvir.cn/article/id/hwjs201903009 LI Qian, YAN Bo, SU Yulu, et al. An improved algorithm for irfpa optical nonuniformity correction based on neural networks[J]. Infrared Technology, 2019, 41(3): 251-255. http://hwjs.nvir.cn/article/id/hwjs201903009
[11]	RONG Shenghui, ZHOU Huixin, QIN Hanlin, et al. Guided filter and adaptive learning rate based non-uniformity correction algorithm for infrared focal plane array[J]. Infrared Physics & Technology, 2016, 76: 691-697. http://www.sciencedirect.com/science?_ob=ShoppingCartURL&_method=add&_eid=1-s2.0-S1350449515300529&originContentFamily=serial&_origin=article&_ts=1480172376&md5=3e335208352d8436eca4420b9fdcff9b
[12]	刘会通, 马红伟. 自适应非均匀性校正中"鬼影"问题的分析[J]. 红外技术, 2003, 23(5): 30-36. DOI: 10.3969/j.issn.1001-8891.2003.05.008 LIU Huitong, MA Hongwei. An analysis of the ghosting artifact in adaptive nonuniformity correction[J]. Infrared Technology, 2003, 23(5): 30-36. DOI: 10.3969/j.issn.1001-8891.2003.05.008
[13]	陈芳林, 张宝辉, 汪贵华, 等. 改进的神经网络非均匀性校正算法研究[J]. 科学技术与工程, 2016, 33(16): 215-220. https://www.cnki.com.cn/Article/CJFDTOTAL-KXJS201633039.htm CHEN Fanglin, ZHANG Baohui, WANG Guihua, et al. Research on improved neural network non-uniformity correction algorithm[J]. Science Technology and Engineering, 2016, 33(16): 215-220. https://www.cnki.com.cn/Article/CJFDTOTAL-KXJS201633039.htm
[14]	Scribner D A, Caulfield J T. Nonuniformity correction for staring IR focal plane arrays using scene - based techniques[C]//Proceedings of SPIE the International Society for Optical Engineering, 1990, 12: 21730.
[15]	张红辉, 罗海波, 余新荣, 等. 改进的神经网络红外图像非均匀性校正方法[J]. 红外技术, 2013, 35(4): 232-237. http://hwjs.nvir.cn/article/id/hwjs201304011 ZHANG Honghui, LUO Haibo, YU Xinrong, et al. Improved algorithm of neural network used in IR image non-uniformity correction[J]. Infrared Technology, 2013, 35(4): 232-237. http://hwjs.nvir.cn/article/id/hwjs201304011
[16]	张菲菲, 王文龙, 马国锐, 等. 基于非局部均值滤波与神经网络的红外焦平面阵列非均匀性校正算法[J]. 红外技术, 2015, 37(4): 265-271. http://hwjs.nvir.cn/article/id/hwjs201504001 ZHANG Feifei, WANG Wenlong, MA Guorui, et al. Neural network nonuniformity correction algorithm for infrared focal plane array based on non-local means filter[J]. Infrared Technology, 2015, 37(4): 265-271. http://hwjs.nvir.cn/article/id/hwjs201504001
[17]	陆余洋君. 基于神经网络的红外焦平面非均匀性校正算法研究与FPGA实现[D]. 合肥: 中国科学技术大学, 2021. LU Yuyangjun. Research on Algorithms Nonuniformity Correction on Infrared Focal Plane Array Based on Neural Net Works and FPGA Implement[D]. Hefei: University of Science and Technology of China, 2021.