摘要
目标识别是图像处理、模式识别和计算机视觉领域的研究热点之一,广泛应用于日常生活、工业应用和军事活动中的各个领域。从人类认知事物的角度来看,目标的特征描述和相似性评价是目标识别中两个关键环节。轮廓作为描述目标特征的一种主要方式,在目标识别中发挥着重要作用。因此,如何有效描述轮廓特征以及合理评价特征之间相似程度对于获得目标识别的最佳结果至关重要。
本论文围绕轮廓特征的提取、描述和评价这一主题展开研究,重点关注了轮廓特征在轮廓平滑、轮廓起始点配准和轮廓识别三个方面的应用。对于目标轮廓,现有轮廓平滑算法很难在保留轮廓特征和抑制噪声之间取得平衡,现有轮廓起始点配准算法无法适用于仿射目标或者具有较高的计算复杂度,现有仿射目标识别算法很难满足实时性的要求,现有部分遮挡目标识别算法没有考虑轮廓局部特征与目标轮廓之间的关系及其对目标识别结果的影响。本论文的主要工作和所取得的研究成果具体如下:
1.提出了一个轮廓特征描述框架。在该框架下,从全局特征、局部特征和结构特征三个层面对现有轮廓特征的描述方法进行总结和分类,并且从可信度、计算复杂度和适用范围三个方面分析了各类特征描述方法的特点,为不同应用环境下选取最佳的特征描述方法以及评价准则提供了借鉴。
2.提出了基于分段轮廓平滑的仿射目标识别算法。为了抑制噪声对目标轮廓的干扰从而准确描述轮廓特征,预先对含噪轮廓进行分段平滑处理。利用曲率将目标轮廓划分为特征区域和非特征区域,对两类区域分别采用不同的高斯平滑滤波器进行差异化处理,在保留轮廓特征和抑制噪声之间取得了较好的平衡。在此基础上,利用仿射不变矩和最小距离分类器对仿射变换下含噪轮廓进行识别。仿真实验结果验证了分段轮廓平滑算法改善识别结果的有效性。
3.提出了基于轮廓起始点配准的快速仿射目标识别算法。针对现有轮廓起始点配准算法存在计算代价较高的不足,提出了基于联合仿射不变弧长的轮廓起始点配准算法,实现了起始点的高效配准。在此基础上,利用离散小波变换的尺度系数构造了满足平移不变性的级联仿射不变函数。与同类基于小波变换的识别算法相比,大大降低了算法的整体复杂度。仿真结果表明,所提出的轮廓起始点配准算法和仿射目标识别算法对噪声具有较好的鲁棒性,并且满足了实时性的要求。
4.提出了基于轮廓特征的部分遮挡目标识别算法。从目标的轮廓特征入手,提出了基于局部曲率分布的轮廓划分方案,建立了完整描述轮廓局部特征的轮廓分段数据库。在此基础上,提出了基于轮廓特征完整描述的部分遮挡目标识别算法,采用形状上下文距离衡量轮廓特征之间的相似度,并结合轮廓分段的可信度评价了部分遮挡目标的匹配程度。仿真实验表明,所提出的识别算法所需训练样本相对较少,能够高效、准确地识别不同遮挡情况下的遮挡目标。更进一步,为了增强在不同遮挡情况下描述局部特征的稳定性,结合目标的结构信息对轮廓分段进行分类,得到了轮廓局部特征的多层描述模型。借鉴哲学中“整体与部分之间关系”的概念,论文深入探讨了整体轮廓与轮廓分段之间的关系,给出了评价轮廓分段的两个参数:重要性和局部性。在此基础上,提出了基于轮廓特征多层描述和评价的部分遮挡目标识别算法。为了准确衡量遮挡目标的匹配程度,将两个评价参数与轮廓特征之间的相似度相结合提出了加权部分相似度。与现有相似性评价准则相比,加权部分相似度合理反映了局部特征相似性与整体目标相似性之间的关系,可以获得遮挡目标的最佳匹配结果。仿真实验表明,所提出的识别算法在不同遮挡情况下均能获得稳定的识别结果。
Object recognition is always one of the hot topics in the field of image processing, pattern recognition and computer vision, and it is widely used in every area of the daily lives, industrial applications and military activities. Contour is one of the apporaches to describe the object and plays an important role in object matching and recognition. From the angle of human being's cognitive things, the description and evaluation of object features are two key parts for object recognition. Therefore, it is very important to obtain the optimal object matching that how to effectively describe contour features and properly evaluate the similarity between different features.
This dissertation focuses on the extraction, description and evaluation of contour features, and studies the effect of contour feature applied to the following aspects:the smoothing of noisy contour, the alignment of the contour starting-point and the recognition of object contour. As for the object contour, current contour smoothing algorithms cannot achieve a better tradeoff between feature preservation and noise suppression. And, existed starting-point alignment algorithms cannot be applied to the affine-transformed objects or are of high computational complexity. Moreover, state of art of the affine object recognition algorithms cannot meet the requirement of real-time applications. Furthermore, many occluded object recognition algorithms did not consider the influence on object recognition of the relationship between local contour features and the whole object. The main content and researching results of this dissertation are listed as follows:
1. Constructing a kind of framework of describing contour features. Within this framework, current contour feature description methods are summarized and classified into three levels:global features, local features and primitive features. Further, those methods are analyzed in three aspects:reliability, computational complexity and the application range. And it is beneficial for choosing the optimal manner of describing contour feature in the cases of various applications.
2. Proposing the affined object recognition algorithm based on segmented contour-smoothing. To suppress the noise and obtain the accurate feature description, the noisy contour is smoothed in advance. The contour is divided by curvature into two categories:feature zone and non-feature zone. By smoothing contour parts located in those zones by the Gaussian filters with different variance, the tradeoff between noise suppression and feature preservation is well achieved. After the contour-smoothing processing, the affined contour can be recognized by utilizing the affine invariant moment and minimum distance classifier. Simulation results testify the improvement by the segmented contour smoothing on recognition results.
3. Proposing the fast affined object recognition algorithm based on the alignment of contour starting-point. To decrease the high computational complexity of existed starting-point alignment algorithms, the algorithm based on the joint affine invariant arc-length is proposed and aligns the starting-point of contour successfully and efficiently. Then, the cascaded affine invariant function with translation invariance is constructed by using the multi-level approximation coefficients of discrete wavelet transform. Compared with similar recognition algorithms, the proposed algorithm greatly reduces the overall computational complexity. Simulation results show that the proposed alignment algorithm and recognition algorithm are robust to the noise and meet the requirement of real-time application.
4. Proposing the partially occluded object recognition algorithm based on contour features. Starting with the analysis of contour feature, the contour-splitting scheme based on the distribution of local curvature is proposed and builds the contour fragment database, which describe integrally local contour features. Then, the partially occluded object recognition algorithm based on features description integrity is proposed. The shape context distance is introduced to measure the similarity between contour features. And, the matching degree between occluded objects is measured by the combination of shape context distance and the reliability of contour fragments Simulations show that the proposed recognition algorithm needs less training samples and achieves the goal efficiently and accurately. Further, to strengthen the stability of local feature description under various occluded cases, contour fragments obtained by the contour-splitting scheme are classified into different categories with the help of object structure. Then the multi-level description model of local contour features is built. Quoted from the philosophical issue that "the relation between the wholeness and the part", this dissertation explores in depth the relation between the whole contour and contour fragments. And two parameters, importance and partiality, are presented to evaluate contour fragments. On that basis, the partially occluded object recognition algorithm based on multi-level description and evaluation of contour features is proposed. To measure the matching degree, the two evaluation parameters and the similarity between local contour features are combined to introduce the weighted partial similarity. Compared with current similarity measurements, the weighted partial similarity reflects properly the relation of the similarity between the local contour features and that between the corresponding whole contours. So, the partially occluded object recognition algorithm based on multi-level description and evaluation of local contour features is proposed. Simulations show that the proposed recognition algorithm achieves the stable recognition results under various occluded cases.
引文
[1]Gonzalez R C, Woods R E. Digital image processing [M]. London:Prentice Hall,2nd Edition, 2002.
[2]Li W, Piech V, Gilbert C. Contour saliency in primary visual cortex [J]. Neuron,2006,50: 951-962.
[3]Siddiqi K, Kimia B B. Parts of visual form:computational aspects [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1995,17(3):239-251.
[4]Hoffman D D, Singh M. Salience of visual parts [J]. Cognition,1997,63(1):29-78.
[5]Mori S, Nishida H, Yamada H. Optical character recognition [M]. New York:John Wiley & Sons Inc.,1999.
[6]Hu M K. Visual pattern recognition by moment invariants [J]. IRE Transaction on Information Theory,1962, IT-8:179-187.
[7]Chang S L, Chen L S, Chung Y C Chen S W. Automatic license plate recognition [J]. IEEE Transaction on Intelligent Transportation Systems,2004,5(1):42-53.
[8]马蓓.车辆识别技术在视频监控中的应用[D].西安:西安电子科技大学,硕士论文,2010.
[9]张道德.机械零件图像识别的关键技术研究与实现[D].武汉:华中科技大学,硕士论文,2008.
[10]廖凯.基于形状分析的飞机图像识别[D].南京:南京理工大学,硕士论文,1999.
[11]刘进.不变量特征的构造及在目标识别中的应用[D].武汉:华中科技大学,博士论文,2004.
[12]Tieng Q M, Boles W W. Wavelet-based affine invariant representation:a tool for recognizing planar objects in 3D space [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1997, 19(8):846-857.
[13]Rube I E, Ahmed M, Kamel M. Wavelet approximation-based affine invariant shape representation functions [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,2006,28(2):323-327.
[14]Avrithis Y, Xirouxakis Y, Kollias S. Affine invariant curve normalization for object shape representation classification and retrieval [J]. Machine Vision and Applications,2001,13(2): 80-94.
[15]Yang H S, Lee S U, Lee K M. Recognition of 2D Object Contours Using starting-point-independent wavelet coefficient matching [J]. Journal of Visual Communication and Image Representation,1998,9(2):171-181.
[16]Kith K, Zahzah E.2D shape recognition using discrete wavelet descriptor under similitude transform [C]. Proceedings of International Workshop Combinatorial Image Analysis, Auckland, New Zealand,2004, pp.679-689.
[17]Yadav R B, Nishchal N K, Gupta A K, et.al.. Retrieval and classification of shape-based objects using Fourier, generic Fourier, and wavelet-Fourier descriptors technique:a comparative study [J]. Optics and Lasers in Engineering,2007,45(6):695-708.
[18]Zhang H M, Wang Z, Han L Q, et.al.. Locating the starting point of closed contour based on half-axis-angle [C]. Proceedings of IEEE International Conf. Machine Learning and Cybernetics, Shanghai, China,2004, pp.3899-3903.
[19]Zhang D S, Lu G J. Review of shape representation and description techniques [J]. Pattern Recognition,2004,37(1):1-19.
[20]Guggenheimer H W. Differential Geometry [M]. New York:McGraw-Hill,1963.
[21]Arbter K, Snyder W E, Burkhardt H, et.al.. Application of affine-invariant Fourier descriptors to recognition of 3-D objects [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1990, 12(7):640-647.
[22]Khalil M I, Bayoumi M M. A dyadic wavelet affine invariant function for 2D shape recognition [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,2001,23(10):1152-1164.
[23]Khalil M I, Bayoumi M M. Affine invariant function for object recognition using dyadic wavelet transform [J]. Pattern Recognition Letters,2002,23(12):57-72.
[24]Zhang D S, Lu G J. Evaluation of similarity measurement for image retrieval [C]. Proceedings of IEEE International Conference of Neural Networks & Signal Processing, Nanjing, China,2003, pp. 14-17.
[25]Liu H C, Srinath M D. Partial shape classification using contour matching in distance transformation [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1990, 12(11):1072-1079.
[26]Latecki L J, Lakamper R, Wolter D. Optimal partial shape similarity [J]. Image Vision Computer, 2005,23(2):227-236.
[27]Wu C Q, Bai X, Li Q N, et al.. Contour grouping with partial shape similarity [A]. In Wada T, Huang F, Lin S. Advance in Image and Video Technology, Lecture Notes in Computer Science [C],2009,5414, pp.167-178.
[28]Lu C E, Adluru N, Ling H B, Zhu G X, et al.. Contour based object detection using part bundles [J]. Computer Vision and Image Understanding,2010,114:827-834.
[29]Sebastian T B, Kimia B B. Curves vs. skeletons in object recognition [J]. Signal Processing,2005, 85:247-263.
[30]Bronstein A M, Bronstein M M, Bruckstein M A, et al.. Partial similarity of objects, or how to compare a centaur to ahorse [J]. International Journal of Computer Vision,2008,84(2):163-183.
[31]Bai X, Wang X, Latecki L J, et al.. Active skeleton for non-rigid object detection [C]. Proceedings of International Conference on Computer Vision, Kyoto, Japan,2009, pp.575-582.
[32]Bai X, Yang X W, Latecki L J. Detection and recognition of contour parts based on shape similarity [J]. Pattern Recognition,2008,41(7):2189-2199.
[33]Bai X, Latecki L J. Path similarity skeleton graph matching [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2008,30(7):1282-1292.
[34]Bai X, Latecki L J, Liu W Y. Skeleton pruning by contour partitioning with discrete curve evolution [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2007,29(3): 449-162.
[35]Pikaz A, Dinstein I. Optimal polygonal approximation of digital curves [J]. Pattern Recognition, 1995,28(3):373-379.
[36]Bruckstein A M, Shapiro G, Shaked C. Evolutions of planer polygons [J]. International Journal of Pattern Recognition and Artificial Intelligence,1995,9:991-1014.
[37]Masood A. Optimized polygonal approximation by dominant point detection [J]. Pattern Recognition,2008,41(1):227-239.
[38]Poyato A C, Cuevas F J M, Carnicer C R M, Salinas R M. Polygonal approximation of digital planar curves through break point suppression [J]. Pattern Recognition,2010,43(1):14-25.
[39]Legault R, Suen C Y. Optimal local weighted averaging methods in contour smoothing [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1997,19(8):801-817.
[40]Li X, Hu W M, Zhang Z F. Corner detection of contour images using spectral clustering [C]. Proceedings of IEEE International Conf. Image Processing, Shanghai, China,2004, pp. 3899-3903.
[41]Xiao Y, Zou J J, Yan H. An adaptive split-and-merge method for binary image contour data compression [J]. Pattern Recognition Letters,2001,22(3/4):299-307.
[42]Yu D G, Yan H. An efficient Algorithm for smoothing binary image contours [C]. Proceedings of International Conference on Pattern Recognition, Vienna, Austria,1996, pp.403-407.
[43]Lee S W, Choi J G, Yun B J. Adaptive contour smoothing based on inter-region contrast [J]. Electronics Letters,2003,39(9):712-714.
[44]Unser M, Aldroubi A, Eden M. B-spline signal processing:Part Ⅰ-theory [J]. IEEE Transaction on Signal Processing,1993,41(2):821-833.
[45]Lee B, Zhuang T. Adopt adaptive B-spline to embellish contours in image segmentation [C]. Proceedings of IEEE International Workshop on Medical Imaging and Augmented Reality, Shatin, Hong Kong,2001, pp.216-221.
[46]Doker T E, Mlsna P A. Efficient region boundary approximation using adaptive smoothing and second-order B-splines [C]. Proceedings of Fifth IEEE Southwest Symposium on Image Analysis and Interpretation, Santa Fe, USA,2002, pp.139-143.
[47]Weickert J.A review of nonlinear diffusion filtering [A]. In Romeny B M H, Florack L, Koenderink J, et.al.. Scale-Space Theory in Computer Vision [C]. Berlin:Springer,1997, pp. 3-28.
[48]Grayson M A. The heat equation shrinks embedded plane curves to round points [J]. Pattern Recognition,1987,26(2):285-314.
[49]Kimia B B, Siddiqi K. Geometric heat equation and nonlinear diffusion of shapes and images [J]. Computer Vision and Image Understanding,1996,64:305-322.
[50]Kimia B B, Tannenbaum A R, Zucker S W. Shapes, shocks, and deformations I:The components of shape and the reaction-diffusion space [J]. International Journal of Computer Vision,1995,15: 189-224.
[51]Perona P, Malik J. Scale space and edge detection using anisotropic diffusion [C]. Proceedings of IEEE Computer Society Workshop on Computer Vision, Miami, USA,1987, pp.16-22.
[52]Lindeberg T. Scale-space theory in computer vision [M]. Boston:Kluwer,1994.
[53]Witkin A P. Scale space filtering [M]. Proceedings of International Joint Conference on Artificial Intelligence, Karlsruhe,1983,1019-1023.
[54]Mokhtarian F, Mackworth A K. A theory of multiscale curvature-based shape representation for planar curves [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1992,14(8): 789-805.
[55]Jang B K, Chin R T. Morphological scale space for 2D shape smoothing [C]. Proceedings of IEEE International Conference on Image Processing, Austin,1994,2:pp.111-115.
[56]Weickert J, Ishikawa S, Imiya A. On the history of Gaussian scale-space axiomatics [A]. In Sporring J, Nielsen M, Florack L, et.al.. Gaussian scale-space theory [M]. Dordrecht:Kluwer, 1997.
[57]Latecki L J, Lakamper R. Convexity rule for shape decomposition based on discrete contour evolution [J]. Computer Vision and Image Understanding,1999,73:441-454.
[58]Kpalma K, Ronsin J. Multiscale contour description for pattern recognition [J]. Pattern Recognition Letters,2006,27:1545-1559.
[59]Mokhtarian F. Silhouette-based isolated object recognition through curvature scale space [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1995,17(5):539-544.
[60]Lim K B, Yu W M, Du T H. Bayesian Kernel Inference for 2D Objects Recognition [C]. Proceedings of IEEE 12th International Conference on Multi-Media Modeling, Beijing,2006, pp. 272-279.
[61]Mundy J L, Zisserman A. Geometric invariance in computer vision [M]. Cambridge:MIT Press,
[62]Flusser J, Suk T. Affine moment invariants:a new tool for character recognition [J]. Pattern Recognition Letters,1994,15:433-436.
[63]Heikkila J. Pattern matching with affine moment descriptors [J]. Pattern Recognition,2004,37(9): 1825-1834.
[64]Tzimiropoulos G, Mitianoudis N, Stathaki T. Robust recognition of planar shapes under affine transforms using principal component analysis [J]. IEEE Transaction on Signal Processing Letters, 2007,14(10):723-726.
[65]Ali A, Gilani S A M, Memon N A. Affine invariant contour descriptors using independent component analysis and dyadic wavelet transform [J]. Journal of Computing and Information Technology,2008,16(3):169-181.
[66]Wang Y, Teoh E K.2D affine-invariant contour matching using B-spline model [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,2007,29(10):1853-1858.
[67]Mallat S. A wavelet tour of signal processing [M]. New York:Academic Press,1998.
[68]Tieng Q M, Boles W W. An application of wavelet based affine invariant wavelet representation [J]. Pattern Recognition Letters,1995,16(12):1287-1296.
[69]Bala E, Cetin A E. Computationally efficient wavelet affine invariant functions for shape recognition [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,2004,26(8): 1095-1099.
[70]Kong X D, Luo Q S, Zeng G H, et.al.. A new f based on centroid-radii model and wavelet transform [J]. Optics Communications,2007,273(2):362-366.
[71]http://www.imageprocessingplace.com/root_files_V3/image_databases.htm
[72]Well W M. Statistical approaches to feature-based object recognition [J]. International Journal of Computer Vision.1997,21(1/2):63-98.
[73]Ying Z R, Castanon D. Partially occluded object recognition using statistical models [J]. International Journal of Computer Vision.2002,49(1):57-78.
[74]Boykov Y, Huttenlocher D. A new Bayesian framework for object recognition [C]. Proceedings of IEEE Conference of Computer Vision and Pattern Recognition, Fort Collins, USA,1999, pp. 517-523.
[75]Ying Z. Statistical approaches for partially occluded object recognition [D]. Boston University, 2002.
[76]Mardia K V, Qian W, Shah D, et.al.. Deformable template recognition of multiple occluded objects [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1997,19(9): 1035-1042.
[77]Yung N H C, Lai A H S. Detection of vehicle occlusion using a generalized deformable model [C]. Proceedings of IEEE International Symposium on Circuits and Systems, Monterey, Canada, 1998, pp.154-157.
[78]Michel D, Oikonomidis I, Argyros A. Scale invariant and deformation tolerant partial shape matching [J]. Image and Vision Computing,2011,29(7):459-469.
[79]张桂林,李强,陈益新,等.局部遮挡目标的识别[J].华中理工大学学报,1994,22(5):15-19.
[80]颜佳,吴敏渊.遮挡环境下采用在线Boosting的目标跟踪[J].光学精密工程,2012,20(2):439-446.
[81]Bronstein A M, Bronstein M M, Bruckstein A M, et.al.. Analysis of two-dimensional non-rigid shapes [J]. International Journal of Computer Vision,2008,78:67-88.
[82]Saber E, Xu Y W, Tekalp A M. Partial shape recognition by sub-matrix matching for partial matching guided image labeling [J]. Pattern Recognition,2005,38(10):1560-1573.
[83]Fritz G, Paletta L, Bischof. Object recognition using local information content [C]. Proceedings of the 17th International Conference on Pattern Recognition, Cambridge, UK,2004, pp.15-18.
[84]Fang W, Chan K L. Incorporating shape prior into geodesic active contours for detecting partially occluded object [J]. Pattern Recognition,2007,40(8):2162-2172.
[85]Yi X L, Camps O I. Robust occluding contour detection using the Hausdorff Distance [C]. Proceedings of Computer Vision and Pattern Recognition, San Juan, Puerto Rico,1997, pp. 962-968.
[86]Shotton J, Blake A, Cipolla R. Multiscale categorical object recognition using contour fragments [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2008,30(7):1270-1281.
[87]Krolupper F, Flusser J. Polygonal shape description for recognition of partially occluded objects [J]. Pattern Recognition Letter,2007,28(9):1002-1011.
[88]August J, Siddiqi K, Zucker S W. Contour Fragment Grouping and Shared, Simple Occluders [J]. Computer Vision and Image Understanding,1999,76(2):146-162.
[89]Yang X W, Liu H R, Latecki L J. Contour-based object detection as dominant set computation [J]. Pattern Recognition,2012,45(5):1927-1936.
[90]Wu Y N, Si Z Z, Gong H F, et al.. Learning active basis model for object detection and recognition [J]. International Journal of Computer and Vision,2009,38(8):1-38.
[91]Shotton J, Blake A, Cipolla R. Contour-based Learning for object detection [C]. Proceedings of the Tenth IEEE International Conference on Computer Vision, Beijing, China,2005, pp.503-510.
[92]Latecki L J, Lakamper R. Shape similarity measure based on correspondence of visual parts [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2000,22(10):1185-1190.
[93]Sun K B, Super B J. Classification of contour shapes using class segment sets [C]. Proceedings of the Tenth IEEE International Conference on Computer Vision and Pattern Recognition, San Diego, USA,2005, pp.727-733.
[94]Huttenlocher D P, Klanderman G A, Rucklidge W J. Comparing images using the Hausdorff distance [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,1993,15(9): 850-863.
[95]Rucklidge W J. Efficient locating objects using Hausdorff distance [J]. International Journal of Computer Vision,1997,24(3):251-270.
[96]Cyrus D C. Modern Mathematical methods for physicists and engineers [M]. London: Cambridge University Press,2000.
[97]http://www.aiaccess.net/English/Glossaries/GlosMod/e_gm_mahalanobis.htm
[98]Barrow H G, Tenenbaum J M, Bolles R C, Wolf H C. Parametric correspondence and chamfer matching:two new techniques for image matching [C]. Proceedings of Fifth International Joint Conference of Artificial Intelligence,1977, pp.659-663.
[99]Gavrila D M. Multi-feature hierarchical template matching using distance transforms [C]. Proceedings of Fourteenth International Conference on Pattern Recognition, Brisbane, Australia, 1998,1:pp.439-444.
[100]David C L. Linear algebra and its applications [M]. Addison Wesley,4th edition,2011.
[101]Duda R O, Hart P E, Stork D G. Pattern classification [M]. Wiley-Interscience,2nd Edition, 2000.
[102]Belongie S, Malik J, Puzicha J. Shape matching and object recognition using shape contexts [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2002,24(4):509-522.
[103]章毓晋.图像工程[M].北京:清华出版社,第2版,2007.
[104]Marques de Sa J P. Pattern recognition:concepts, methods, and applications [M]. Berlin: Springer-Verlag,2001.
[105]贾沛.特征选择技术研究[D].武汉,华中科技大学,硕士论文,2003.
[106]Zahn C T, Roskies R Z. descriptors for plane closed curves [J]. IEEE Transaction on Computers, 1972,21(3):269-281.
[107]Zhang D S, Lu G. A comparative study of Fourier descriptors for shape representation and retrieval [C]. Proceedings of the Fifth Asian Conference on Computer Vision, Melbourne, Australia,2002, pp.646-651.
[108]Chuang G, Kuo C. Wavelet descriptor of planar curves:theory and applications [J]. IEEE Transaction on Image Processing,2006,5(1):56-70.
[109]Kokkinos I, Yuille A. Inference and learning with hierarchical shape models [J]. International Journal of Computer Vision,2010,93:201-255.
[110]Zhu L, Chen Y, Yuille A. Learning a hierarchical deformable template for rapid deformable object parsing [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2010, 32(6):1029-1042.
[111]Schindler K, Suter D. Object detection by global contour shape [J]. Pattern Recognition,2008, 41(12):3736-3748.
[112]Xu D, Xu W L. Description and recognition of object contours using arc length and tangent orientation [J]. Pattern Recognition Letter,2005,26(7):855-864.
[113]Gavrila D M. A Bayesian, exemplar-based approach to hierarchical shape matching [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,2007,29(8):1408-1421.
[114]Safaee-Rad R, Benhabib B, Smith K C, et.al.. Position, rotation and scale-invariant recognition of 2 dimensional objects using a gradient coding scheme [C]. Proceedings of IEEE Pacific RIM Conference on Communications, Computers and Signal Processing, Vicotria, Canada,1989, pp.306-311.
[115]Sonka M, Hlavac V, Boyle R. Image processing, analysis and machine vision [M]. London: Chapman & Hall,1993.
[116]Prokop J R, Reeves R P. A survey of moment-based techniques for unoccluded object representation and recognition [J]. Graphical Models and Image Processing,1992,54(5): 438-460.
[117]Asada H, Brady M. The curvature primal sketch [J], IEEE Transaction on Pattern Analysis Machine Intelligence,1986,8(1):2-14.
[118]Freeman H. On the digital computer classification of geometric line patterns [C]. Proceedings of National Electronics Conference,1962,18:pp.312-324.
[119]Librescu L, Na S. Bending vibration control of cantilevers via boundary moment and combined feedback control laws [J]. Jounal of Vibration and Control,1998,4:733-746.
[120]Pavlidis T. Structural pattern recognition [M]. Berlin:Springer Verlag,1977.
[121]Lindenbaum M, Bruckstein A. On recursive on partitioning of a digitized curve into digital straight segments [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1993,15(9): 949-953.
[122]Groskey W I, Mehrotra R. Index-based object recognition in pictorial data management [J]. Computer Vision Graphics Image Processing,1990,52(3):416-436.
[123]Ferrari V, Fevrier L, Jurie F, Schmid C. Groups of adjacent contour segments for object detection [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2008,30(1) 36-51.
[124]Ling H, Jacobs D W. Shape classification using the inner-distance [J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,2007,29(2):286-299.
[125]Liu M, Tuzel O, Veeraraghavan A, Chellappa R. Fast directional chamfer matching [C]. Proceedings of Computer Vision and Pattern Recognition, San Francisco, USA,2010, pp. 1696-1703.
[126]Marc P S, Chen J S, Medioni G. Adaptive smoothing:a general tool for early vision [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1991,13(6):514-529.
[127]Kasvand T, Otsu N. Regularization of digitized plane curves for shape analysis and recognition [C]. Proceedings of SPIE Conference Architecture and Algorithms for Digital Image Processing, San Diego, USA,1983, pp.44-52.
[128]Shaharay B, Anderson D J. Optimal estimation of contour properties by cross-validated regularization [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1989,11(6): 600-610.
[129]Li X, Chen T. Optimal L1 approximation of the Gaussian kernel with application to scale-space construction [J]. IEEE Transaction on Pattern Analysis Machine Intelligence, 1995,17(10):1015-1019.
[130]Bronstein M M, Bronstein A M. Shape Recognition with Spectral Distances [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,2011,33(5):1065-1071.
[131]Andalo F A, Miranda P A V, Torres R D S, et.al.. Shape feature extraction and description based on tensor scale [J]. Pattern Recognition,2010,43(1):26-36.
[132]Latecki L J, Lakamper R. Application of planar shape comparison to object retrieval in image databases [J]. Pattern Recognition,2002,35(1):15-29.
[133]Bandera A, Marfil R, Antunez E. Affine-invariant contours recognition using an incremental hybrid learning approach [J]. Pattern Recognition Letters,2009,30(14):1310-1320.
[134]Siddiqi K, Shokoufandeh A, Dickinson S, et al.. Shock graphs and shape matching [J]. International Journal of Computer Vision,1999,35(1):13-32.
[135]Canny J A. Computational approach to edge detection [J]. IEEE Transaction on Pattern Analysis Machine Intelligence,1986,8(6):679-698.
[136]曹健.基于局部特征的图像目标识别技术研究[D].北京:北京理工大学,博士论文,2010.
[137]张永亮.基于射影不变量的平面目标识别方法研究[D].长沙:国防科技大学,硕士论文,2004.