一种实时成像跟踪系统的研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
实时成像跟踪系统克服了雷达成像系统费用昂贵,容易受外界电磁波的干扰,并且难于达到可视化的缺点,在跟踪冷目标方面又较红外优越。它以低价位高清晰度的优势被广泛应用于军事武器、航海导航、工业监控、交通管理等领域。因其具有重大的现实意义和极高的经济价值,应用前景十分广阔,成为各国成像跟踪研究的热点。
     本文以运动的飞行器为目标,分析和研究了近年来国内外成像跟踪的前沿技术,提出并推导了一种高效的、实时的成像跟踪方法,并给出了具体的设计方案和实现方法。文中针对成像跟踪系统中目标识别与跟踪预测两个关键环节,采用并改进了一系列图像分割与跟踪预测算法,并对算法进行了仿真与实验。在目标识别过程中,将遗传算法和最大熵原理相结合应用于图像分割中,充分利用遗传算法简单、快速和稳定性强等特点。经过改进的图像分割算法精度较高,计算量大大减小,增强了实时性。对于跟踪预测方法采用改进的双模跟踪方法,使跟踪的效率更高同时大大提高了跟踪和预测的精度。同时对系统的软、硬件进行了设计,并进行了室内外目标跟踪实验。结果表明该系统有以下特点:对简单背景下的运动目标能自动识别;当目标被短暂遮挡时能预测跟踪;跟踪过程中对目标变化能自适应决策。从实验的结果看,系统基本满足实时性,达到预期的效果。
A real-time imaging tracking system overcomes the shortcomings of the ladar system which needs high cost, is easily disturbed by electromagnetic wave and can't be visualized. Compared with the infrared, it is superior in the aspect of tracking cold-target. The system has been applied in the field of weapons, shipping navigation, industry monitor and traffic control because of its low cost and superior visibility. It is becoming a hot point of imaging tracking in the world because it has very great economy benefit, outstanding practical significance and extensive possibility of
    application.
    In the paper, based on the study of a moving aircraft, some advanced technologies of imaging tracking in recent years are studied. A high efficient and real -time method of imaging tracking is deduced. Further more, the specific design and implementation method is proposed. A series of algorithms about image segmentation, tracking and prediction are adopted and modified in the key parts of the system. These algorithms are simulated and tested. The application of genetic algorithm and maximum entrophy principle to image segmentation shows many attractive advantages such as high accuracy, small calculation amount and excellent real-time property. The dual-mode image tracking and prediction enhances efficiency and accuracy. The hardware and software of the system which are presented in this paper have passed test in lab and outdoors, the experimental results show that the system has characteristics, including automatic recognition of the target; permitting tracking and prediction when the target become shaded;
    the system allows adaptive decision target variation during the process of tracking. The results prove that the system is effective and can be implemented in real time.
引文
[1] 杨宜禾,周维真著.成像跟踪技术导论.西安:西安电子科技大学出版社,1991
    [2] 吴晗平.ATR算法原理分析.现代防御技术,1997(1):54~64
    [3] 田捷.实用图象分析与处理技术.北京:电子工业出版社
    [4] 章毓晋.图像分割.北京:科学出版社,2001
    [5] 江捧岳.图象跟踪算法分析.火力与指挥控制.1998,2(6):47~52
    [6] 马明建,周长城.数据采集与处理技术.西安:西安交通大学出版社,1998
    [7] 宋建中,王延杰.跟踪测量电视系统.光学精密工程.1995,3(5):44~52
    [8] 徐建华.图像处理与分析.北京:科学出版社,1992
    [9] 何斌等.Visual C++数字图象处理,北京:人民邮电出版社,2001
    [10] 李智勇.动态图像分析.北京:国防工业出版社,1999
    [11] 韩思奇,王蕾.图像分割的阈值法综述.系统工程与电子技术.2002,24(6):91~95
    [12] 刘立波.图像分割方法探讨.宁夏农学院学报.2001,22(4):51~56
    [13] 赵荣椿,迟耀斌等.图像分割技术进展.中国体视学与图像分析.1998,3(2):12~128
    [14] Pal N R, Pals K. A Review on image segmentation techniques. Pattern Recognition. 1993,26
    [15] 刘文萍,吴立德. 图像分割中阈值选取方法比较研究.模式识别与人工智能.1997,10(3):271~277
    [16] 赵雪松,陈淑珍.综合全局二值化与边缘检测的图像分割方法.计算机辅助设计与图形学学报.2001,13(2):118~121
    [17] Brown and M A, Blackwell K T, and Khalak H G et al. Multi-Scale Edge Detection and Feature Binding: An Integrates Approach. Pattern recognition. 1998, 31(10): 1479~1490
    [18] Bhabatosh Chanda, Malay k. Kundu and Y. Vani Padmaja. A Multi-Scale Morphologic Edge Detector. Pattern Recognition. 1998, 60:385~401
    [19] Nalwa V and Binford T. On Detecting Edges. IEEE Trans. Pattern Analysis and Machine Intelligence. 1986,8(6): 699~714
    [20] Pal N R Pal S K. Entropy thresholding. Signal Process. 1989, (16): 97~108
    
    
    [21] Thomos P. Entropy:a new definition and its application. IEEE Trans Syst Man Cybern. 1991(21): 1~260
    [22] N R Pal, S K Pal. Object-background segmentation using new definitions of entropy[J].IEEE Proceedings. Part E. 1989,136(4):284~295
    [23] 陈果,左洪福.图像分割的二维最大熵遗传算法.计算机辅助设计与图形学学报.2002,14(6):530~534
    [24] 张二虎,刘世昌等.一种基于熵的图象分割方法.微电子学与计算机.1998(4):43~45
    [25] 袁占亭,洪毅,王青.二维熵图像阈值分割的遗传算法.甘肃工业大学学报.1999,25(1):68~72
    [26] 刘京南,陈从顺等.一种快速二维熵阈值分割算法.计算机应用研究.2002(1):67~70
    [27] 陈国良等.遗传算法及其应用.北京:人民邮电出版社,1996
    [28] 周明,孙树栋.遗传算法原理及应用.北京:国防工业出版社,1999
    [29] 郑宏,潘励.基于遗传算法的图像阈值的自动选取.中国图象图形学报.1999,4(4):327~330
    [30] 金聪,彭嘉雄.利用遗传算法实现数字图像分割.小型微型计算机系统.2002,23(7):875~877
    [31] 向健勇,徐军.一种实用的红外图像分割算法研究.西安电子科技大学学报.1997,24(3):416~420
    [32] Goldberg. D. Genetic algorithms in search, optimization and Machine Learning[M]. USA:Addison-Wesley Publising Company. 1988, 7(10): 59~308
    [33] 孙即祥等.模式识别中的特征提取与计算机视觉不变量.北京:国防工业出版社,2001
    [34] 王积分,张新荣,计算机图象识别.北京:中国铁道出版社,1988
    [35] 张元,王广宇等.数字图像的不变特性与特征提取.郑州大学学报.2000,32(4):57~59
    [36] 郭飚,陈曾平.一种高分辨率雷达二维图象的特征提取与目标识别方法.信号处理.1999,15(3):281~284
    [37] 郭旭平,李在铭.图像目标的可识别与智能跟踪.系统工程与电子技术.1998(8),8~10
    [38] Tom V, Peli T, Leung M and Bondaryk J. Morphology-based algorithm for
    
    point-target detection in infrared backgrounds. SPIE. 1993, 1954: 2~11
    [39] 贺利洁,江晓安.运动点目标的检测识别方法.佳木斯工学院学报.1997,15(3):196~201
    [40] 刘桂雄,申柏华等.基于改进的Hough变换图像分割方法.光学精密工程.2002,10(3):257~260
    [41] 刘上乾,向健勇等.图像制导系统算法设计.西安电子科技大学学报.1997,24(2):241~246
    [42] Rosenfeld A. Image analysis and computer vision:1994. CVIU. 1995, 62(1): 90~143
    [43] Hassan M H. Real-time signal and image thresholding using advanced low-pass filtering. SPIE. 1988, 977: 30~38
    [44] 王虎元,雷万云.多模实时图像处理跟踪算法研究.弹箭与制导学报.1996(4):5~11
    [45] 王春平,朱元昌等.电视跟踪系统中图像目标的预测跟踪算法研究.测控技术.1999,18(3):45~47
    [46] 张思芳,李正明.一种实时成像跟踪系统的研究.光学精密工程.2002,10(6):573~577
    [47] 王鸿钰.步进电机控制入门.上海:同济大学出版社
    [48] 李圣怡等.Windows环境下软硬件接口技术.长沙:国防科技大学出版社,2001
    [49] William Buchanan.PC接口通信与Windows编程.北京:中国电力出版社,2001
    [50] 曲仕茹,史忠科.基于飞行器图像的目标跟踪方法研究.飞行力学.2001,19(4):81~84

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700