期刊文献+

基于类似形原理从单幅透视图识别平面立体 被引量:1

Recognizing Polyhedrons from a Perspective Drawing Based on Principle of Homograph
下载PDF
导出
摘要 根据透视不变性提出了5个描述多边形形状的拓扑特征和几何特征,以这些特征作为约束条件给出透视投影下类似形的定义。根据类似形的定义提出了识别多边形是类似形的新方法。据此提出了基于类似形原理从单幅透视投影线图识别平面立体的新方法。新定义的类似形能将平面立体中的表面分类。匹配过程限制在类似形内进行,大大减少了搜索范围。用拓扑结构对类似形进行定性分析,对噪音不敏感,用几何特征进行定量分析,则能反应平面立体形状的细微差别。该识别方法不仅能识别不同形状的平面立体目标,而且还可以区分形状相同但大小、比例不同的平面立体目标。实验结果表明该方法是有效的。 In this paper, two problems are researched. First, we propose five topology characteristics and geometry characteristics that are invariant under perspective transformation to describe polygon. Then the definition of homograph is given in terms of these characteristics. On the basis of this definition, we give a procedure for recognizing homographs. Second, a new algorithm for recognizing polyhedrons from a perspective line drgwing based on principle of homograph is presented. Homograph as defined by us can be employed to distinguish between surfaces with different characteristics and to classify the surfaces of polyhedron. As homographs' categories are few in number and as it is only necessary to put the polygon to be recognized in one of these few categories, efficiency of recognition is greatly raised. We introduced the topological structure to analyze homograph qualitatively, therefore the algorithm is insensitive to noises; and employ geometrical structure to analyze homograph quantitatively, thus the algorithms can show fine discrimination between polyhedrons. Thus, it can not only recognize polyhedrons of different shape but also distinguish between polyhedrons that are of the same shape but their sizes and proportions are different. Comparing with other recognition algorithms, our algorithm has obtained obvious improvement. The effectiveness of the algorithm is demonstrated preliminarily by computer simulation.
出处 《中国图象图形学报》 CSCD 北大核心 2005年第12期1511-1516,共6页 Journal of Image and Graphics
基金 陕西省自然科学基金项目(2002E224) 江西省自然科学基金项目(0311019)
关键词 类似形 透视图 线图 识别 homograph, perspective drawing, line drawing, recognition
  • 相关文献

参考文献9

  • 1Chin R T, Dayer C R. Model-based recognition in robot vision[J].ACM Computer Surveys, 1986, 18(1) : 67-108.
  • 2Bunke H, Glauser T. Viewpoint independent representation and recognition of polygonal faces in 3D space[J]. IEEE Journal Robotics and Automation, 1993, 9(2) : 457 -463.
  • 3Bamieh B, Figueiredo R J. A general moment-invariants/attributed-Graph method for three-object recognition from a single image[J].IEEE Journal Robotics and Automation, 1986, 2(5) : 31 - 41.
  • 4Lowe D G. Three-dimensional object recognition from single two-dimensional images[J]. Artificial Intelligence, 1987, 31 ( 7 ) :355 -395.
  • 5Gu W K, Yang J Y, Huang T S. Matching perspective views of a polyhedron using circuits[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1987, 9(3): 435 -462.
  • 6Arbter K, Synder W E. Application of affine invariant fourier descriptors to recognition of 3D objects[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990,12 (7) : 640 - 646.
  • 7张桂梅,高满屯,曾接贤,沈允文.基于类似形原理识别平面多边形[J].计算机应用,2004,24(9):7-9. 被引量:2
  • 8Liu J, Lee Y T. A graph-based method for face identification from a single 2D drawing[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001,23(10):1106 - 1119.
  • 9Milan Sonka, Vaclav Hlavac, Roger Boyle. Image Processing and Machine Vision ( Second Edition)[M]. London: An International Thomson Publishing Company, 1998.

二级参考文献5

  • 1Bunke H, Glauser T. Viewpoint Independent Representation and Recognition of Polygonal Faces in 3-D Space[J]. IEEE Journal Robotics and Automation, 1993,9(2):457-463.
  • 2Bamieh B, Figueiredo RJ. A general moment-invariants/attributed-graph method for three-object recognition from a single image[J]. IEEE Journal Robotics and Automation, 1986, 2(5):31-41 .
  • 3Lowe DG. Three-dimensional object recognition from single two-dimensional images[J]. Artificial Intelligence, 1987,31(7):355-395.
  • 4Gu WK, Yang JY, Huang TS. Matching perspective views of a polyhedron using circuits[J]. IEEE Transaction on Pattern Analysis and Machine Intelligence, 1987,9(3):435-462.
  • 5Arbter K, Synder WE. Application of affine invariant fourier descriptors to recognition of 3D objects[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990,12(7):640-646.

共引文献1

同被引文献6

  • 1张桂梅,郭航,高满屯,沈允文.基于类似形原理从单幅轴测投影图识别平面立体[J].西北工业大学学报,2004,22(5):586-590. 被引量:1
  • 2Cheng Y,Wong K C,Kittler J.The recognition of triangle-pairs and quadrilaterals from a single perspective view[C]//The 4th Int Conf Image Processing and its Applications,Maastricht Holland, April 1,1992:425-428.
  • 3Bunke H,Glauser T.Viewpoint independent representation and recognition of polygonal faces in 3-D space [J].IEEE Journal Robotics and Automation, 1993,9(2) : 457-463.
  • 4Lowe D G.Three-dimensional object recognition from single two- dimensional images[J].Artificial Intelligence, 1987,31 (7) : 355-395.
  • 5Gu W K,Yang J Y,Huang T S.Matching perspective views of a polyhedron using circuits[J].IEEE Trans Pattern Analysis and Machine Intelligence, 1987,9(3) :435-462.
  • 6Arbter K,Synder W E.Application of affine invariant fourier descriptors to recognition of 3D objects[J].IEEE Trans on Pattern Analysis and Machine Intelligence, 1990,12(7) :640-647.

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部