| Point-pattern Matching Using Irreducible Matrix and Relative Invariant |
| |
| 作者姓名： | ZHANG Lihua |
| |
| 作者单位： | ZHANGLihua ( 张立华)，XU Wenli ( 徐文立) Departmentof Automation，Tsinghua University，Beijing 100084 |
| |
| 基金项目： | 教育部科学技术研究项目，9500320， |
| |
| 摘 要： | ＩｎｔｒｏｄｕｃｔｉｏｎＰｏｉｎｔｐａｔｔｅｒｎｍａｔｃｈｉｎｇｉｓａｎｉｍｐｏｒｔａｎｔｔｏｐｉｃｉｎｔｈｅｆｉｅｌｄｓｏｆｃｏｍｐｕｔｅｒｖｉｓｉｏｎａｎｄｐａｔｔｅｒｎｒｅｃｏｇｎｉｔｉｏｎ．Ｉｔｓｍａｉｎｔａｓｋｉｓｔｏｐａｉｒｕｐｔｈｅｐｏｉｎｔｓｉｎｔｗｏｉｍａｇｅｓｏｆｔｈｅｓａｍｅｓｃｅｎｅｗｈｅｎｔｈｅｒｅｉｓａｇｅｏｍｅｔｒｉｃｔｒａｎｓｆｏｒｍａｔｉｏｎｒｅｌａｔｉｎｇｔｈｅｔｗｏｉｍａｇｅｓ．Ｔｈｅｔｒａｎｓｆｏｒｍａｔｉｏｎｍａｙｂｅｒｉｇｉｄｔｒａｎｓｆｏｒｍａｔｉ…
|
| 关 键 词： | point-pattern matching irreducible matrix affine transformation similarity transformation relative invariant |
Point-pattern Matching Using Irreducible Matrix and Relative Invariant |
| |
| ZHANG Lihua.Point-pattern Matching Using Irreducible Matrix and Relative Invariant[J].Tsinghua Science and Technology,1999,4(4):1602-1605. |
| |
| Authors: | ZHANG Lihua XU Wenli |
| |
| Abstract: | Point pattern matchingisanimportantproblem inthefieldsofcomputervision and patternrecognition.In this paper,new algorithms based onirreducible matrix andrelativeinvariantfor matchingtwosets ofpoints withthe same cardinality are proposed.Theirfundamentalideaistransformingthetwo dimensionalpointsets with n points intothe vectorsin n dimensional space. Considering these vectors as one dimensional point patterns,these new algorithms aim atreducingthe point matching problem to thatofsorting vectorsin n dimensionalspace aslong asthe sensornoise does notalterthe order ofthe elementsinthe vectors.Theoreticalanalysis and simulationresults show thatthe new algorithms are effective . |
| |
| Keywords: | point-pattern matching irreducible matrix affine transformation similarity transformation relative invariant |
| 本文献已被 CNKI 等数据库收录！ |
