摘要
激光雷达在保障空间安全中扮演重要角色,主要用于对空间运动目标的搜索、捕获与跟踪。光学相控阵作为一种激光束控制装置,能够实时、无惯性、高精度地控制多束激光束,故使用光学相控阵的激光雷达与常规激光雷达相比,能够以更灵活的方式、更高效率、更高精度地跟踪运动目标,本文研究基于光学相控阵的运动目标捕获与跟踪方法。
光学相控阵的优势之一是可以灵活、敏捷地改变光束的形状和能量分布,并与图像实时采集系统组成主动照明闭环跟踪系统。系统通过光学相控阵发射的激光网搜索目标,一旦发现运动目标,光学相控阵即自适应改变光束形状,将光能量集中于目标,在该系统内,运动目标捕获与跟踪软件系统首先要快速从激光网图像中发现并分离出目标,其次要能预测目标运动轨迹,向光学相控阵给出控制光束方位的指令,使激光束精确跟踪目标,实现对运动目标的主动照明快速、稳定跟踪。本文的主要研究内容如下:
(1)研究在激光网照明条件下研究捕获运动目标的理论方法。理论上,运动目标的捕获方法由运动目标检测、图像阈值分割和图像滤波三部分构成。在运动目标图像灰度特性的基础上,分别研究了运动目标检测、图像阈值分割和图像滤波的不同方法,然后就其所组成的多种捕获方法进行了仿真对比,通过对比结果确定快速、稳定的捕获方法。该捕获方法首先采用背景差分法检测到运动目标,然后对运动目标图像进行中值滤波以减弱噪声影响,接着采用最大类间方差法分割图像,最后进行形态学滤波消除目标内的孔洞。
(2)在捕获到运动目标的基础上研究跟踪运动目标的方法。运动目标的跟踪方法由运动目标图像跟踪和运动目标空间位置变换两部分构成。其中,运动目标图像跟踪采用波门形心跟踪法,该方法不但可计算当前帧图像目标的形心位置,还可预测下一帧图像目标的形心位置。通过运动目标空间位置变换将图像目标的形心位置变换为目标的实际空间位置,以控制激光网指向,从而主动照亮运动目标,以达到跟踪目标的目的。
(3)通过跟踪实验验证运动目标捕获与跟踪方法的可行性。设计并实现了基于光学相控阵的运动目标捕获与跟踪实验平台,完成了软件编写工作,最终实现了基于光学相控阵运动目标的捕获与跟踪实验。实验结果表明上述捕获与跟踪方法在主动照明条件下,可快速、稳定地跟踪运动目标。
Laser radar plays an important role in the protection of space security, mainly used for spatial moving object searching, acquiring and tracking. As a multiple-laser-beams control in laser radar, optical phased array has such advantage as real time, no inertia and high precision. Compared to conventional laser rader, the laser rader used optical phased array can track moving object with more flexible way, high efficiency, high precision, therefore, method of acquiring and tracking moving object based on optical phased array is researched.
One of the advantages of optical phased array is changing the shape and energy distribution of beam flexibly and agilely, which can form active lighting closed-loop tracking system with real-time image acquisition system. Optical phased array can generate variable laser net to search moving object, once found, Optical phased array changes the shape of the light adaptively, to make energy focus on the moving object. In this system, acquiring and tracking moving object software system should find and segment object from laser net image firstly, next predict the target trajectory and control the direction of laser net to make it cover the moving object all the time, in order to use active lighting to track moving object high efficiently, fast and stably. The main contents and achievements are described as following.
(1) Theoretical method of acquiring moving object is studied in the condition of laser net lighting. In theory, method of acquiring moving object consists of moving object detection, image threshold segmentation and image filtering. On the basis of image gray characteristics, many different methods of moving object detection, image threshold segmentation and image filtering are studied separately, and methods of acquiring moving object are compared with simulation in order to determine fast and stable acquiring method. In the determined method, firstly moving object is detected by background differencing, secondly the noise of moving object image is attenuated by median filter, thirdly the image is segmented by OTSU method, at last the hole of object image is eliminated by morphological filter.
(2) On the basis of acquired moving object, method of tracking moving object is studied, and consists of moving object image tracking and spatial transformation. Moving object image tracking does not only calculate the centroid position of object in current frame, but also predict the centroid position of object in next frame through gate centroid tracking method. And moving object spatial transformation transforms centroid position of imaged object to the real position of object for controlling the direction of laser net, in order to track object with the active lighting.
(3) In order to validate the feasibility of the proposed method of acquiring and tracking moving object, tracking experiments have been performed. The experimental platform with optical phased array is designed to acquire and track moving object, and software is developed to achieve acquire and track moving object based on optical phased array. Experimental results demonstrate that the proposed method above can fast and stably track moving object in the condition of active lighting.
引文
[1]尤政,赵岳生.国外太空态势感知系统发展与展望[J].中国航天,2009(9):40-44.
[2]李颖,张占月,方秀花.空间目标监视系统发展现状及展望[J].国际太空,2004 (6):28-32.
[3] Stokes G H, Kostishack D F. Technology for the Detection of Near-earth Objects[C]. International Conference on Near-Earth Objects, NEW YORK, APR 24-26, 1995. ANNALS OF THE NEWYORK ACADEMY OF SCIENCES, 1997, 822: 544.
[4] Hoots F R, Schurnacher P W, DeVere T. 50 Years of Satellite Tracking and Cataloging in the US[C]. AIAA/AAS Astrodynamics Specialist Conference, AUG 19-23, 2007 Mackinac Isl, ADVANCES IN THE ASTRONAUTICAL SCIENCES, 2008,129: 1251-1277.
[5]吴勤.美国目前太空对抗发展态势分析[J].国际太空,2008(7):22-28.
[6]李少敏,牛威,马鑫.空间目标探测技术研究[J].国防科技,2009,30(3):6-13.
[7]王国强,陈涛,王建立.地基空间目标探测与识别技术[J].长春理工大学学报,2009, 32(2):197-199.
[8] Sharma J, Stokes G H, von Braun C, et al. Toward Operational Space-Based Space Surveillance[J]. Lincoln Laboratory Journal, 2002, 13(2): 309-334.
[9]李怀锋,陈新龙,李志.天基监测空间碎片有效载荷效能评估方法研究.航天电子对抗[J],2009,26(2):27-30.
[10]华灯鑫,宋小全.先进激光雷达探测技术研究进展[J].红外与激光工程,2008, 37(增刊):21-27.
[11] Buzzard G. Advanced laser proximity fuzing [C]. First Annual International Missiles&Rockets Symphosium and Exhibition. Feb 22-24, California, 2000.
[12] Fenn A J, Temme D H, Delaney W P, et al. The Development of Phased-Array Radar Technology[J]. Lincoln Laboratory Journal, 2000, 12(2): 321-340.
[13] Meyer R A. Optical Beam Steering Using a Multichannel Lithium Tantalate Crysta[J]. Applied Optics, 1972, 11(3): 613-616.
[14] Wight D R, Heaton J M, Hughes B T, et al. Novel Phased Array Optical Scanning Device Implemented Using GaAs/AlGaAs technology[J]. Applied Physics Letters, 1991, 59(8): 899-901.
[15] James A T, Yeshaiahu F. Programmable Diffractive Optical Element Using aMultichannel Lanthanum-Modified Lead Zirconate Titanate Phase Modulator[J]. Opt Lett, 1995, 20(13): 1510-1512.
[16] Paul F M, David L C, Larry J F, et al. Optical Phased Array Technology[J]. Proceedings of the IEEE, 1996, 84(2): 268-298.
[17] Fairfax N. Steered Agile Beam [N/OL]. Commerce Business Daily, (1999-05-10): http://www.fbodaily.com/cbd/archive/1999/05(May)/10-May-1999/Asol007.htm.
[18]王莉,石顺祥,李家立.基于GaAs/AlGaAs光波导扫描技术的研究[J].激光与光电子学进展,2005,42(1):37-39.
[19]梁华伟,石顺祥,李家立.新型电光扫描器中光波导阵列特征研究[J].光子学报,2006,35(11):1654-1658.
[20]石顺祥,李家立,王广生.一种新型的光波导阵列电光快速扫描器[J].光学学报,2002,22(11):1318-1322.
[21]刘立人.卫星激光通信Ⅰ链路和终端技术[J].中国激光,2007,34(1):3-20.
[22]倪树新.新体制成像激光雷达发展评述[J].激光与红外,2006,32(增刊):732-736.
[23]倪树新,李一飞.军用激光雷达的发展趋势[J].红外与激光工程,2003,32(2):111-114.
[24] Surendra G., Osama M, Robert F K, et al. Detection and Classification of Vehicles[J]. IEEE Trans.On intelligent transportation systems,2002,3(1):37~47.
[25] Stauffer C, Grimson W. Adaptive background mixture models for real-time tracking[C]. Proceediongs of International Conference on Computer Vision and Patterm Recognition, 1999, 2: 246-252.
[26] Kamann K, Brandt A. Moving Object Recognition Using an Adaptive Background Memory[C]. In: Cappellini V, ed. Time-Varying Image Processing and Moviung Object Recgnition. Amsterdam: Elservier Science, 1990, 289-296.
[27] Seki, T. Wada, H. Fujiwara, K. Sumi. Background subtraction based on cooccurrence of image variation[C], IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2003, (2): 64-70.
[28] Mckenna SJ, Jabri S, Duric Z, et al. Tracking groups of people[J]. Computer Vision and Image Undestanding, 2000, 80(1): 42-56.
[29] Neri A, Colonnese S, Russb G., et al. Automatic Moving Object and Background Separation[J]. Signal Processing, 1998,66(2):219~232.
[30] Lipton A, Fujiyoshi H, Patil R. Moving target classification and tracking from real-video[C]. In: Proc IEEE Workshop on Applications of Computer Vision, Princeton, NJ, 1998, 8-14.
[31] Anderson C, Bert P, Vander Wal G. Change detection and tracking using pyramidstransformation techniques[C]. In: Pro SPIE Conference on Intelligent Robots and Computer Vision, Cambridge, MA, 1985, 579: 72-78.
[32] C. Stauffer, W. Eric L. Grimson. Learning patterns of activity using real-time tracking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(8): 747-756.
[33] Barron J, Fleet D. Performance of optical flow techniques[J]. Internatronal Journal of Computer Vision,1994,12(l):42~77.
[34] Meyer D, Denzler J, Niemann H. Model based extraction of artificalated objects in image sequences for gait analysis[C]. In: Pro IEEE International Conference on Image Processing, Santa Barbara, California, 1997, 78-81.
[35] K. Jung, Y. S. Ho. Traffic parameter extraction using video-based vehicle tracking[C]. Proceedings of IEEE/IEEJ/JSAI International Conference on Intelligent Transportion Systems, 1999, 764-768.
[36]王建华,梁伟,王春平,等.基于改进模糊熵的图像阈值分割方法研究[J].光电子·激光,2008, 19(6):828-831.
[37] Pal S K, King R A, Hashim A A. Automatic gray-level thresholding through index of fuzziness and entropy [J]. Pattern Recognition Letter,1983,(1):141~146.
[38]刘平,陈斌,阮波.基于边缘信息的图像阈值化分割方法[J].计算机应用,2004, 24(9):28-36.
[39]陈方昕.基于区域生长法的图像分割技术[J].科技信息,2008(15):58-59.
[40] Veenman C, Reinders M, Backer E. A Cellular Coevolutionary Algorithm for Image Segmentation [J]. IEEE Transaction on Image Processing, 2003, 12(3):304-316.
[41]吴晗平.动态多目标自动识别及自适应多波门跟踪[J].光学精密工程,1996, 4(2):53-60.
[42]梁锡宁,杨刚,余学才,等.一种动态模板匹配的卡尔曼滤波跟踪方法[J].光电工程,2010, 37(10):29-33.
[43]谭熊,兰远鸽.基于光流方法的运动估计在OpenCV中的实现[J].测绘与空间地理信息,2010, 33(5): 127-129.
[44]张宏志,张金换,岳卉,等.基于CamShift的目标跟踪算法[J].计算机工程与设计,2006, 27(11):2012-2014.
[45]张健,徐林,吴丽莹,等.液晶光学相控阵可编程光束偏转研究[J].光子学报,2008,37(8):1497-1501.
[46]维可多·索菲尔,衍射光学元件的计算机设计方法[M].金国蕃,谭峭峰,译校.天津科学技术出版社,2007:103-116.
[47] Gonzalez R C, Woods R E. Digital Image Processing[M].阮秋琦,阮宇智,等译.电子工业出版社,2008:479-494.
[48]吴一全,朱兆达.图像处理中阈值选取方法30年(1962—1992)的进展(一)[M].数据采集与处理,1993,8(3):193~201.
[49]刘平,陈斌,阮波.基于边缘信息的图像阈值化分割方法[J].计算机应用,2004, 24(9):28-36.
[50]于起峰,陆宏伟,刘肖琳.基于图像的精密测量与运动测量[M].科学出版社,2007:92-99.
[51]刘瑞祯,于仕琪. OpenCV教程—基础篇[M].北京航空航天大学出版社,2009: 1-5.
[52] http://www.intel.com/technology/computing/opencv/index.htm.