全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

基于形状模板匹配的前视红外目标检测方法

Keywords: 尺度空间,分层多阈值,感兴趣区域,模糊集,目标识别

Full-Text   Cite this paper   Add to My Lib

Abstract:

针对前视红外复杂地面固定目标无直接可用基准图、背景干扰严重、目标与背景灰度差异小、不利于目标识别等问题,提出了一种基于形状模板的目标识别方法.首先,在构建高斯多尺度空间的基础上,设计分层多阈值算法,检测感兴趣区域;其次,引入模糊集理论,提取形状特征,分离目标与背景;最后,用改进的Hausdorff距离算法进行精匹配,确定目标.实验结果表明,该算法匹配率与改进的Hausdorff距离算法相比提高了近20%,算法花费时间缩短了2/3;与Nprod算法相比匹配率提高了近30%,时间缩短了1/2,在密度为0.3的椒盐噪声下,匹配率仍能达到70%以上.对于复杂背景下的前视红外固定目标,该方法具有匹配率高、速度快、精度高等优点.

References

[1]  孙君顶,陈素霞,郭启强.基于粗糙集的图像感兴趣区域自动提取算法[J].传感器与微系统,2009,28(3):101-107.SUN Jun-ding,CHEN Su-xia,GUO Qi-qiang.Automaticextraction of regions of interest based on rough sets[J].Transducer and Microsystem Technologies,2009,28(3):101-107.(in Chinese)
[2]  SIMACE Kirmacek B,UNSALAN C.Urban-area andbuilding detection using SIFT keypoints and graph theory[J].IEEE Transactions on Geoscience and RemoteSensing,2009,47(4):1156-1167.
[3]  LAPTEV I.Improving object detection with boostedhistograms[J].Image and Vision Computing,2009,27(5):535-544.
[4]  LINDEBERG T.Scale-space theory:a basic tool foranalyzing structures at different scales[J].Supplement onAdvances in Applied Statistics:Statistics and Images,1994,21(2):224-270.
[5]  刘婧,孙继银,朱俊林.基于模板匹配的前视红外目标识别方法[J].弹箭与制导学报,2010,30(1):17-19.LIU Jing,SUN Ji-yin,ZHU Jun-lin.FLIR targetrecognition algorithm based on template matching[J].Journal of Projectiles,Rockets,Missile and Guidence,2010,30(1):17-19.(in Chinese)
[6]  MIKOLAJCZYK Krystian,SCHMID Cordelia.Indexingbased on scale invariant interest points[C]∥EighthInternational Conference on Computer Vision(ICCV'01).Vancouver:IEEE,2001:525-528.
[7]  DAVID G L.Distinctive image features from scale-invariant keypoints[J].Internation Journal of ComputerVision,2004,60(2):91-110.
[8]  ZADCH L A.Fuzzy sets[J].Information and Control,1965,8(1):78-84.
[9]  HUTTENLOCHER D P,KLANDERMAN G A,RUCHLIDGE W J.Comparing images using theHausdorff distance[J].IEEE Trans Pattern Anal MachIntell,1993,15:850-863.
[10]  SIM D G,KWON O K,PARK R H.Object matchingusing robust Hausdorff distance measure[J].IEEE TransImages Process,1999,8(2):425-429.
[11]  HUTTENLOCHER D P,KLANDERMAN G A,RUCKLIDGE William J.Comparing images using theHausdorff distance[J].IEEE Trans Pattern Analysis andMachine Intelligence,1993,15(9):850-863.
[12]  LING H.Techniques in image retrieval:deformationinsensitivity and automatic thumbnail cropping[D].Maryland:University of Maryland,2006.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133