-
复杂背景下低可见度目标追迹是成像领域的一个重要研究方向, 现有方法难以对运动方向和速率均无规则变化的低可见度目标进行追迹成像. 运动衬度成像可大幅提升目标追迹的灵敏度, 已在X射线成像领域取得重要应用, 但仅局限于固定轨迹或单调背景的成像. 本文将运动衬度成像引入到时变复杂背景中轨迹不确定的目标追迹成像任务, 通过解决现有方法受复杂背景和高频噪声干扰严重的问题, 实现目标的高灵敏追迹成像. 天空和树林中飞鸟的追迹结果表明, 该方法能有效地消除野外自然光和树叶无规则摆动导致背景灰度无规则变化的影响, 实现时变复杂背景中自由运动弱信号小目标的高灵敏追迹成像. 本文发展的方法有望为低可见度目标追迹成像提供一种新的手段.Tracking of low-visibility targets in complex backgrounds is an important research field, where existing methods struggle to image low-visibility targets with irregular changes in moving direction and speed. Move contrast imaging can greatly improve the sensitivity of target tracking, which has achieved important applications in the field of X-ray imaging, including high-resolution imaging to microvessels in living rats with the help of contrast agents, in-situ dynamic observation of ion migration and redox reactions during electrochemical reactions, and water refilling along vessels in willow branch without resorting to agents. However, all these applications are limited to imaging with fixed trajectories or monotonous backgrounds. In principle, move contrast imaging is based on the frequency spectral characteristics of the time-domain grayscale signal and is highly sensitive to moving components, which is wavelength-independent. This paper extends the move contrast imaging to the visible light waveband for tracking free-moving targets in time-varying complex backgrounds. To meet the need for tracking imaging of free-moving targets in complex backgrounds, we develop a move contrast imaging (MCI) method based on continuous wavelet transform (CWT) and Hilbert-Huang transform (HHT) with high discriminatory capability for non-stationary signals. Selecting birds in the sky and forest for the tracking imaging, the irregular grayscale changes caused by natural light intensity in the wild field and random swaying of tree leaves result in complex imaging backgrounds. The tracing results of low-visibility free-moving targets show that FT-MCI method, CWT-MCI method and HHT-MCI method can improve the target tracing imaging sensitivity by 179.9 times, 175.8 times and 214.6 times compared with temporal subtraction imaging, respectively. The results of tracking imaging of free-moving targets in complex backgrounds show that compared with the FT-MCI method and CWT-MCI method, the HHT-MCI method can further effectively suppress the influence of background noise on tracking imaging of targets of interest, thus achieving high sensitivity imaging of free-moving targets in time-varying complex backgrounds. Combining the phase diagram of FT-MCI and the imaging parameters, we can further show the motion direction, the motion speed or the distance from the observation point. Therefore, the HHT-MCI imaging method developed in this paper is expected to provide a novel method for tracking free-moving targets in time-varying complex backgrounds.
-
Keywords:
- high sensitivity tracking imaging /
- move contrast imaging /
- time-varying complex background /
- low visibility target tracking
[1] Cheng Y H, Wang J 2014 AMM 490 1283Google Scholar
[2] 徐艳, 王培光, 杨青, 董江涛 2019 68 164203Google Scholar
Xu Y, Wang P G, Yang Q, Dong J T 2019 Acta Phys. Sin. 68 164203Google Scholar
[3] 沈瑜, 王新新 2017 自动化与仪器仪表 4 122Google Scholar
Shen Y, X. W X 2017 Autom. Instrum. 4 122Google Scholar
[4] Husein A M, Calvin, Halim D, Leo R, William 2019 J. Phys. Conf. Ser. 1230 012017Google Scholar
[5] Markandey V, Reid A, Wang S 1996 IEEE Trans Aerosp. Electron. Syst. 32 866Google Scholar
[6] 崔智高, 王华, 李艾华, 王涛, 李辉 2017 66 084203Google Scholar
Cui Z G, Wang H, Li A H, Wang T, Li H 2017 Acta Phys. Sin. 66 084203Google Scholar
[7] 袁国武, 陈志强, 龚健, 徐丹, 廖仁健, 何俊远 2013 小型微型计算机系统 34 668
Yuan G W, Chen Z Q, Gong J, Xu D, Liao R J, He J Y 2013 J. Chin. Comput. Syst. 34 668
[8] Han X W, Gao Y, Zheng L, Zhang Z M, Niu D 2015 Fifth International Conference on Instrumentation & Measurement, Computer, Communication, and Control (IMCCC) Qinhuangdao, China, September 18–20, 2015 p579
[9] 连可, 严明, 李丹, 王厚军 2011 电讯技术 51 49
Lian K, Yan M, Li D, Wang H J 2011 Telecommun. Eng. 51 49
[10] Dong X B, Huang X S, Zheng Y B, Bai S J, Xu W Y 2014 Infrared Phys. Technol. 65 36Google Scholar
[11] 张弘, 赵保军, 毛二可, 朱梦宇 2001 红外与激光工程 30 96
Zhang H, Zhao B J, Mao E K, Zhu M Y 2001 Infrared Laser Eng. 30 96
[12] Lü P Y, Sun S L, Lin C Q, Liu G R 2018 Infrared Phys. Technol. 91 107Google Scholar
[13] Wu Y, Yang Z, Niu W L, Zheng W 2018 IGARSS Valencia, Spain, July 22−27, 2018, p7066
[14] 郑晓枫 2015 硕士学位论文 (杭州: 杭州电子科技大学)
Zheng X F 2015 M. S. Thesis (Hangzhou: Hangzhou Dianzi University) (in Chinese)
[15] 吕腾蛟, 袁子乔, 雷刚, 任泽宇 2021 火控雷达技术 50 78Google Scholar
Lyu T J, Yuan Z Q, Lei G, Ren Z Y 2021 Fire Contrl Radar Technol. 50 78Google Scholar
[16] Liu B, J L, Li X R 2016 IEEE Trans. Signal Process. 64 3221Google Scholar
[17] 王海梅, 洪敏 2018 火力与指挥控制 43 78
Wang H M, Hong M 2018 Fire Control & Command Control 43 78
[18] Ward M 2011 IEEE Aerospace Conference. Big Sky, Montana, USA, March 05–12, 2011 p1
[19] 侯旺, 于起峰, 雷志辉, 刘晓春 2014 63 074208Google Scholar
Hou W, Yu Q F, Lei Z H, Liu X C 2014 Acta Phys. Sin. 63 074208Google Scholar
[20] Wang F X, Zhou P T, Li K, et al. 2020 IUCrJ 7 793Google Scholar
[21] 李可 2021 博士学位论文 (北京: 中国科学院大学)
Li K 2021 Ph. D. Dissertation (Beijing: University of Chinese Academy of Sciences) (in Chinese)
[22] 鞠晓璐, 李可, 余福成, 许明伟, 邓彪, 李宾, 肖体乔 2022 71 144101Google Scholar
Ju X L, Li K, Yu F C, Xu M W, Deng B, Li B, Xiao T Q 2022 Acta Phys. Sin. 71 144101Google Scholar
[23] Aguiar-Conraria L, Soares M J 2011 The Continuous Wavelet Transform: A Primer (NIPE-Universidade do Minho) No. 16/2011
[24] Sadowsky J 1994 Johns Hopkins APL. Tech. Digest 15 306
[25] Huang N E, Shen Z, Long S R, et al. 1998 Proc. R. Soc. A-Math. Phys. Eng. Sci. 454 903Google Scholar
[26] Guan J, Zhang J, Liu N B, Li B 2009 IEEE Rad Conf Pasadena, CA, USA, May 04–08, 2009 p1
[27] Huang N E, Wu Z H 2008 Rev. Geophys. 46 RG2006Google Scholar
[28] Huang N E, Shen S S P 2014 Hilbert Huang Transform and its Applications (2nd Ed.) (Singapore: World Scientific) p13
[29] 孙玉宇 2007 硕士学位论文 (哈尔滨: 哈尔滨工业大学)
Sun Y Y 2007 M. S. Thesis (Harbin: Harbin Industrial University) (in Chinese)
[30] Bruderer B, Boldt A 2001 Ibis 143 178Google Scholar
[31] 张园, 谢红兰, 杜国浩, 许明伟, 薛艳玲, 肖体乔 2021 核技术 44 060101
Zhang Y, Xie H L, Du G H, Xu M W, Xue Y L, Xiao T Q 2021 Nucl. Tech. 44 060101
-
图 2 低可见度目标的时频分析结果 (a) 原始图像及4个特征点; (b) 图(a)中蓝色虚线位置的灰度轮廓曲线, 黑色曲线代表背景, 红色曲线代表鸟飞过时的灰度分布; (c) 图(a)中标示的4个特征点傅里叶变换频谱; (d)飞鸟、(e)建筑物、(f)黄浦江以及(g)天空的连续小波变换时频图; (h)飞鸟、(i)建筑物、(j)黄浦江以及(k)天空的希尔伯特-黄变换时频图
Fig. 2. Low visibility target image and time-frequency analysis: (a) Raw image and 4 feature points; (b) grayscale line profile at the position of blue dashed line in panel (a), where the black and red curve represent before and after the bird’s flight, respectively; (c) Fourier transform spectrum of 4 feature points marked in panel (a); continuous Wavelet transform time-frequency graphs of (d) flying bird, (e) house, (f) Huangpu River and (g) sky marked in panel (a); Hilbert-Huang transform time-frequency graphs of (h) flying bird, (i) buildings, (j) Huangpu River and (k) sky marked in panel (a).
图 3 基于不同频谱分析方法的运动衬度成像结果比对 (a) TSI, (b) FT-MCI, (c) CWT-MCI和(d) HHT-MCI成像结果与原始图像融合, 从红到蓝的颜色变化代表目标位置随时间的演化; (e) 图(a)—(d)白色虚线标记处目标轨迹的归一化灰度分布轮廓曲线
Fig. 3. Comparison of move contrast imaging results based on different spectral analysis methods. The imaging results of (a) TSI, (b) FT-MCI, (c) CWT-MCI and (d) HHT-MCI merged with raw image, and the color evolution from red to blue indicates the change of target position over time. (e) Normalized grayscale line profile of target trajectory marked by white dashed line in panel (a)–(d).
图 4 FT-MCI对时变复杂背景自由运动目标的追迹成像结果 (a)原始图像; (b)时间减影图像; (c)原始图像和时间减影图像标准差随帧数的变化; (d)基于傅里叶变换的运动衬度图像, 黄色箭头标示飞鸟轨迹位置
Fig. 4. FT-MCI imaging results of target trajectory in time-varying complex background: (a) Raw image; (b) time subtraction image; (c) the evolution of standard deviation of raw image and time subtraction image with the number of frames; (d) move contrast image based on Fourier transform, and yellow arrow marked the location of bird trajectory.
图 5 目标特征点的时频分析结果 (a) 原始图像及4个特征点; (b) 图(a)中标记的4个特征点的傅里叶变换频谱; (c) 飞鸟、(d) 建筑物、(e) 天空以及(f) 树林的连续小波变换时频图; (g) 飞鸟、(h) 建筑物、(i) 天空及(j) 树林的希尔伯特-黄变换时频图
Fig. 5. Time-frequency analysis of target feature points: (a) Raw image and 4 feature points; (b) Fourier transform spectrum of 4 feature points marked in panel (a); (c)–(f) continuous wavelet transform time-frequency graphs of (c) flying bird, (d) house, (e) sky and (f) tree marked in panel (a); (g)–(j) Hilbert-Huang transform time-frequency graphs of (g) flying bird, (h) house, (i) sky, and (j) tree marked in panel (a).
表 1 不同方法对低可见度目标追迹成像效果比较
Table 1. Comparison of relative contrast in tracking imaging of low-visibility targets.
Evaluating
indicatorTSI FT-MCI CWT-MCI HHT-MCI CR 0.0046 0.8275 0.8088 0.9873 Improvement — 179.9 175.8 214.6 -
[1] Cheng Y H, Wang J 2014 AMM 490 1283Google Scholar
[2] 徐艳, 王培光, 杨青, 董江涛 2019 68 164203Google Scholar
Xu Y, Wang P G, Yang Q, Dong J T 2019 Acta Phys. Sin. 68 164203Google Scholar
[3] 沈瑜, 王新新 2017 自动化与仪器仪表 4 122Google Scholar
Shen Y, X. W X 2017 Autom. Instrum. 4 122Google Scholar
[4] Husein A M, Calvin, Halim D, Leo R, William 2019 J. Phys. Conf. Ser. 1230 012017Google Scholar
[5] Markandey V, Reid A, Wang S 1996 IEEE Trans Aerosp. Electron. Syst. 32 866Google Scholar
[6] 崔智高, 王华, 李艾华, 王涛, 李辉 2017 66 084203Google Scholar
Cui Z G, Wang H, Li A H, Wang T, Li H 2017 Acta Phys. Sin. 66 084203Google Scholar
[7] 袁国武, 陈志强, 龚健, 徐丹, 廖仁健, 何俊远 2013 小型微型计算机系统 34 668
Yuan G W, Chen Z Q, Gong J, Xu D, Liao R J, He J Y 2013 J. Chin. Comput. Syst. 34 668
[8] Han X W, Gao Y, Zheng L, Zhang Z M, Niu D 2015 Fifth International Conference on Instrumentation & Measurement, Computer, Communication, and Control (IMCCC) Qinhuangdao, China, September 18–20, 2015 p579
[9] 连可, 严明, 李丹, 王厚军 2011 电讯技术 51 49
Lian K, Yan M, Li D, Wang H J 2011 Telecommun. Eng. 51 49
[10] Dong X B, Huang X S, Zheng Y B, Bai S J, Xu W Y 2014 Infrared Phys. Technol. 65 36Google Scholar
[11] 张弘, 赵保军, 毛二可, 朱梦宇 2001 红外与激光工程 30 96
Zhang H, Zhao B J, Mao E K, Zhu M Y 2001 Infrared Laser Eng. 30 96
[12] Lü P Y, Sun S L, Lin C Q, Liu G R 2018 Infrared Phys. Technol. 91 107Google Scholar
[13] Wu Y, Yang Z, Niu W L, Zheng W 2018 IGARSS Valencia, Spain, July 22−27, 2018, p7066
[14] 郑晓枫 2015 硕士学位论文 (杭州: 杭州电子科技大学)
Zheng X F 2015 M. S. Thesis (Hangzhou: Hangzhou Dianzi University) (in Chinese)
[15] 吕腾蛟, 袁子乔, 雷刚, 任泽宇 2021 火控雷达技术 50 78Google Scholar
Lyu T J, Yuan Z Q, Lei G, Ren Z Y 2021 Fire Contrl Radar Technol. 50 78Google Scholar
[16] Liu B, J L, Li X R 2016 IEEE Trans. Signal Process. 64 3221Google Scholar
[17] 王海梅, 洪敏 2018 火力与指挥控制 43 78
Wang H M, Hong M 2018 Fire Control & Command Control 43 78
[18] Ward M 2011 IEEE Aerospace Conference. Big Sky, Montana, USA, March 05–12, 2011 p1
[19] 侯旺, 于起峰, 雷志辉, 刘晓春 2014 63 074208Google Scholar
Hou W, Yu Q F, Lei Z H, Liu X C 2014 Acta Phys. Sin. 63 074208Google Scholar
[20] Wang F X, Zhou P T, Li K, et al. 2020 IUCrJ 7 793Google Scholar
[21] 李可 2021 博士学位论文 (北京: 中国科学院大学)
Li K 2021 Ph. D. Dissertation (Beijing: University of Chinese Academy of Sciences) (in Chinese)
[22] 鞠晓璐, 李可, 余福成, 许明伟, 邓彪, 李宾, 肖体乔 2022 71 144101Google Scholar
Ju X L, Li K, Yu F C, Xu M W, Deng B, Li B, Xiao T Q 2022 Acta Phys. Sin. 71 144101Google Scholar
[23] Aguiar-Conraria L, Soares M J 2011 The Continuous Wavelet Transform: A Primer (NIPE-Universidade do Minho) No. 16/2011
[24] Sadowsky J 1994 Johns Hopkins APL. Tech. Digest 15 306
[25] Huang N E, Shen Z, Long S R, et al. 1998 Proc. R. Soc. A-Math. Phys. Eng. Sci. 454 903Google Scholar
[26] Guan J, Zhang J, Liu N B, Li B 2009 IEEE Rad Conf Pasadena, CA, USA, May 04–08, 2009 p1
[27] Huang N E, Wu Z H 2008 Rev. Geophys. 46 RG2006Google Scholar
[28] Huang N E, Shen S S P 2014 Hilbert Huang Transform and its Applications (2nd Ed.) (Singapore: World Scientific) p13
[29] 孙玉宇 2007 硕士学位论文 (哈尔滨: 哈尔滨工业大学)
Sun Y Y 2007 M. S. Thesis (Harbin: Harbin Industrial University) (in Chinese)
[30] Bruderer B, Boldt A 2001 Ibis 143 178Google Scholar
[31] 张园, 谢红兰, 杜国浩, 许明伟, 薛艳玲, 肖体乔 2021 核技术 44 060101
Zhang Y, Xie H L, Du G H, Xu M W, Xue Y L, Xiao T Q 2021 Nucl. Tech. 44 060101
计量
- 文章访问数: 3201
- PDF下载量: 94
- 被引次数: 0