基于图像匹配的应用研究
|
摘要
图像匹配是计算机视觉和图像处理中一项很关键的技术,所以研究图像匹配有着很具体的意义。本文对各种图像匹配的方法进行了深入细致的研究。现阶段图像匹配的方法虽然很多,但是研究可行的图像匹配技术一直是人们不断追求的目标。
图像匹配也是近20年来发展起来的一种模式识别的技术,可将人的视觉认知和理解过程用现代信息技术和计算机技术来完成,它以研究根据图像特征进行识别和分类作为主要内容,广泛应用于文字识别、指纹识别、遥感、医学诊断、工业产品检测、卫星航空图片解释等领域。
图像匹配一般分为基于灰度匹配的方法和基于特征匹配的方法两大类。本文详细描述了各种图像匹配算法,对其中一些经典的算法和成熟的算法进行了研究和实验。本文主要对基于灰度的匹配方法,研究了基于不变矩的算法,该方法在图像旋转、平移的时候都有很好的效果。该匹配方法可提高匹配的可靠性和定位精度的快速匹配方法,并与经典的NMI特征方法做了对比,说明了新方法的优势所在。
本文采用图像传感器采集原始的信息,从而避免了与对象的直接接触和人工检查,所以更符合工业生产中在线检测的要求。实验表明,该方法应用于工业生产的流水线上,可以有效地提高生产率和保证产品的质量。
Image matching is an important research topic in the fields of computer vision and image processing. An intensive study and a great deal of work has been done to study image matching in this paper. Nowadays, there are many methods of image mathching has been done, but image mosaic technology is always researched by the people all the time.
During the recent 20 years, the image mathing has developed to be a technique of Pattern Recognition, which achieves human ability of visual cognition and comprehension by modern information processing and computer technology. Image matching principally completes images classification according to feature extraction, diffusely applied to kinds of fields including character recognition, fingerprint identification, remote sensing, medical diagnosis, industrial products test, satellite photos and aerial photographs interpretation and so on. With the large of experiment, this method shows the better performance than the NMI feature.
Image matching methods cover two categories: one is based on the gray, the other is feature. There are kinds of method of image matching being describing in this paper, and do the experiment for some classical and mature methods. Based on the gray method, this paper mainly focuses on the study of Moment Invariants, which can deal with the images existing shift and rotation well.
In this paper, the system gets original information by image sensor, which avoid direct contact with objects being processed and manual inspection. So this system can meet the requirement that industrial parts should be inspected on-line. The experiment result indicates, if applying this system to industrial production line can improve the productivity and guarantee the quality of products.
图像匹配也是近20年来发展起来的一种模式识别的技术,可将人的视觉认知和理解过程用现代信息技术和计算机技术来完成,它以研究根据图像特征进行识别和分类作为主要内容,广泛应用于文字识别、指纹识别、遥感、医学诊断、工业产品检测、卫星航空图片解释等领域。
图像匹配一般分为基于灰度匹配的方法和基于特征匹配的方法两大类。本文详细描述了各种图像匹配算法,对其中一些经典的算法和成熟的算法进行了研究和实验。本文主要对基于灰度的匹配方法,研究了基于不变矩的算法,该方法在图像旋转、平移的时候都有很好的效果。该匹配方法可提高匹配的可靠性和定位精度的快速匹配方法,并与经典的NMI特征方法做了对比,说明了新方法的优势所在。
本文采用图像传感器采集原始的信息,从而避免了与对象的直接接触和人工检查,所以更符合工业生产中在线检测的要求。实验表明,该方法应用于工业生产的流水线上,可以有效地提高生产率和保证产品的质量。
Image matching is an important research topic in the fields of computer vision and image processing. An intensive study and a great deal of work has been done to study image matching in this paper. Nowadays, there are many methods of image mathching has been done, but image mosaic technology is always researched by the people all the time.
During the recent 20 years, the image mathing has developed to be a technique of Pattern Recognition, which achieves human ability of visual cognition and comprehension by modern information processing and computer technology. Image matching principally completes images classification according to feature extraction, diffusely applied to kinds of fields including character recognition, fingerprint identification, remote sensing, medical diagnosis, industrial products test, satellite photos and aerial photographs interpretation and so on. With the large of experiment, this method shows the better performance than the NMI feature.
Image matching methods cover two categories: one is based on the gray, the other is feature. There are kinds of method of image matching being describing in this paper, and do the experiment for some classical and mature methods. Based on the gray method, this paper mainly focuses on the study of Moment Invariants, which can deal with the images existing shift and rotation well.
In this paper, the system gets original information by image sensor, which avoid direct contact with objects being processed and manual inspection. So this system can meet the requirement that industrial parts should be inspected on-line. The experiment result indicates, if applying this system to industrial production line can improve the productivity and guarantee the quality of products.
引文
[1]王军,图像匹配算法的研究进展,大气与环境光学学报,2007, 2(1):30~35
[2]亢院兵,图像检索技术历史、现状与研究进展,内将科技,2008,18(2):20~24
[3]张成,基于内容的图像检索技术研究:[硕士学位论文],延安:延边大学,2005
[4]赵锋,赵荣椿,文理分割及特征提取方法综述,中国体视学与图像分析,1998,3(4):238~245
[5]stepahne mallat,A wavelet Tour of Signal Processing,Second Edition,1999
[6]Wan X,Kuo C J,A new approach to image retrievak with hierarchical color clustering, IEEE Transactions on Circuits and systems for Video Technology,1988,8(5):628~643
[7]Malki J,Boujemaa N,Nastar C,Winter A, Region queries without segmentation for image retrieval by content, Proc Of Visual Information and Information Systems.London,UK:Springer-Verlag, 1999, 114~122
[8]黄诚,王国营,一种基于颜色聚合向量的图像检索方法,计算机工程,2006,,3(2):194~199
[9]戴声扬,章毓晋,图像检索中的两层描述和非对称区域匹配,电子学报,2005,17(8):57~61
[10]张耀宗,图像匹配算法的研究与应用:[硕士学位论文],太原:太原理工大学,2010
[11]冈萨雷斯,数字图像处理,北京:电子工业出版社,2007
[12]饶俊飞,基于灰度的图像匹配方法研究:[硕士学位论文],武汉:武汉理工大学,2005
[13]董安国,图像匹配最大互相关快速算法,浙江万里学院学报,2005,18(4):13~15
[14]Meltzer, J. Soatto, S. Edge descriptors for robust wide-baseline correspond- dence, 2008 IEEE Conference on Computer Vision and Pattern Recognition, 2008
[15]张德丰,图像多尺度边缘检测算法研究,计算机工程与设计,2008, 29(21):5524~5527
[16]蒋建国,孙洪艳,基于mean-shift算法的人脸实时跟踪方法,计算机应用研究,2008, 25(7):23~36
[17]张克军,刘哲,图像理解原理的数学评价,计算机工程与设计,2007, 28(8):54~56
[18]闫宇壮,王易平,基于爬山法的快速图像匹配,科技导报,2008, 26(5):27~30
[19]Rao Naveed Iqbal, M Junaid Khan, Wide Baseline Stereo Based on Back- ground Modeling, International Conference on Machine Vision, ICMV 2007, 12~17
[20]王培容,陈鸿雁,李姣君等,一种快速的具有旋转不变性的模板匹配方法,计算机工程与设计,2006, 27(14):2667~2672
[21]冈萨雷斯,数字图像处理(MATLAB版),北京:电子工业出版社,2007
[22]王钰,孙小惟,基于圆投影及直方图不变特征的图像匹配方法,自动化技术与应用,2007,6(8):80~82
[23]薛骏,程俊,王黎等,基于圆投影与矢量和的旋转不变量检测方案,计算机工程,2009,35(21):181~183
[24]Lang Su, Zheng Gao, Radial Projection Fourier Transform and its Application for Scene Matching with Rotation Invariance, Computer and Information Science, 2009,2(4):31~46
[25]冯世栋,人脸检测及跟踪算法的研究:[硕士学位论文],长春:长春理工大学,2002
[26]王红梅,张科,李言俊,图像匹配研究进展,计算机工程与应用,2004,19(1):42~44
[27]殷伶,图像匹配技术的研究:[硕士学位论文],西安:西安电子科技大学,2010
[28]王星,目标图像匹配算法研究:[硕士学位论文],太原:中北大学,2009
[29]彭皓,局部不变特征提取算法的研究及其在图像识别中的应用:[硕士学位论文],上海:东华大学,2009
[30]门蓬涛,张秀彬,张峰,图像特征识别方法研究,微计算机信息,2004, 20(5):103~105
[31]徐柯,朱煜,基于机器视觉的手机键盘质量检测系统,计算机系统应用,2009,4:7~10
[2]亢院兵,图像检索技术历史、现状与研究进展,内将科技,2008,18(2):20~24
[3]张成,基于内容的图像检索技术研究:[硕士学位论文],延安:延边大学,2005
[4]赵锋,赵荣椿,文理分割及特征提取方法综述,中国体视学与图像分析,1998,3(4):238~245
[5]stepahne mallat,A wavelet Tour of Signal Processing,Second Edition,1999
[6]Wan X,Kuo C J,A new approach to image retrievak with hierarchical color clustering, IEEE Transactions on Circuits and systems for Video Technology,1988,8(5):628~643
[7]Malki J,Boujemaa N,Nastar C,Winter A, Region queries without segmentation for image retrieval by content, Proc Of Visual Information and Information Systems.London,UK:Springer-Verlag, 1999, 114~122
[8]黄诚,王国营,一种基于颜色聚合向量的图像检索方法,计算机工程,2006,,3(2):194~199
[9]戴声扬,章毓晋,图像检索中的两层描述和非对称区域匹配,电子学报,2005,17(8):57~61
[10]张耀宗,图像匹配算法的研究与应用:[硕士学位论文],太原:太原理工大学,2010
[11]冈萨雷斯,数字图像处理,北京:电子工业出版社,2007
[12]饶俊飞,基于灰度的图像匹配方法研究:[硕士学位论文],武汉:武汉理工大学,2005
[13]董安国,图像匹配最大互相关快速算法,浙江万里学院学报,2005,18(4):13~15
[14]Meltzer, J. Soatto, S. Edge descriptors for robust wide-baseline correspond- dence, 2008 IEEE Conference on Computer Vision and Pattern Recognition, 2008
[15]张德丰,图像多尺度边缘检测算法研究,计算机工程与设计,2008, 29(21):5524~5527
[16]蒋建国,孙洪艳,基于mean-shift算法的人脸实时跟踪方法,计算机应用研究,2008, 25(7):23~36
[17]张克军,刘哲,图像理解原理的数学评价,计算机工程与设计,2007, 28(8):54~56
[18]闫宇壮,王易平,基于爬山法的快速图像匹配,科技导报,2008, 26(5):27~30
[19]Rao Naveed Iqbal, M Junaid Khan, Wide Baseline Stereo Based on Back- ground Modeling, International Conference on Machine Vision, ICMV 2007, 12~17
[20]王培容,陈鸿雁,李姣君等,一种快速的具有旋转不变性的模板匹配方法,计算机工程与设计,2006, 27(14):2667~2672
[21]冈萨雷斯,数字图像处理(MATLAB版),北京:电子工业出版社,2007
[22]王钰,孙小惟,基于圆投影及直方图不变特征的图像匹配方法,自动化技术与应用,2007,6(8):80~82
[23]薛骏,程俊,王黎等,基于圆投影与矢量和的旋转不变量检测方案,计算机工程,2009,35(21):181~183
[24]Lang Su, Zheng Gao, Radial Projection Fourier Transform and its Application for Scene Matching with Rotation Invariance, Computer and Information Science, 2009,2(4):31~46
[25]冯世栋,人脸检测及跟踪算法的研究:[硕士学位论文],长春:长春理工大学,2002
[26]王红梅,张科,李言俊,图像匹配研究进展,计算机工程与应用,2004,19(1):42~44
[27]殷伶,图像匹配技术的研究:[硕士学位论文],西安:西安电子科技大学,2010
[28]王星,目标图像匹配算法研究:[硕士学位论文],太原:中北大学,2009
[29]彭皓,局部不变特征提取算法的研究及其在图像识别中的应用:[硕士学位论文],上海:东华大学,2009
[30]门蓬涛,张秀彬,张峰,图像特征识别方法研究,微计算机信息,2004, 20(5):103~105
[31]徐柯,朱煜,基于机器视觉的手机键盘质量检测系统,计算机系统应用,2009,4:7~10