期刊文献+

仿射不变性特征提取在目标识别中的应用 被引量:2

Affine Invariability Feature Extraction in Application of Target Recognition
下载PDF
导出
摘要 从不同角度、距离获取的图像中提取不受视点因素影响的仿射不变特征,是图像目标识别、图像几何校正、景象匹配、图像检索等领域的共性问题。从仿射几何的角度出发,在对仿射变换、仿射不变性进行研究的基础上,利用仿射几何的不变性提取仿射不变特征量。针对同底面积比基元进行目标识别存在的问题,引入改进局部不变量,并提出3种基元特征提取算法,以排除锯齿点干扰进行角点提取。Matlab仿真分析过程中,对图像进行基元特征提取,采用改进局部不变量进行计算、比较及识别,能够有效建立图像特征值比较模型。实验结果表明:该局部不变量能较好地识别相同目标,用于不同目标间的分类识别时能取得好的分类效果。 It is common problems in image target recognition, image geometric correction, scene matching, image retrieval and other areas that extracted from viewpoint factors affine invariant feature from different angles and distance of image acquisition. Based on the affine transformation, affine invariance study use of affine geometric invariability extraction of affine invariant feature quantity starting from the point of view of affine geometry. According to the same base area ratio primitives for target recognition problems into improving local invariant, and puts forward three kinds of primitive feature extraction algorithm to eliminate sawtooth point interference with angle point extraction. Do image elementary feature extraction, using improved local invariant calculation, comparison and recognition in Matlab simulation analysis process. It can effectively establish image characteristic value comparison model. The experimental results show that the local invariant can more accurately recognize the same goal, and it can obtain better classification effect when used for different target classification recognition between.
出处 《兵工自动化》 2013年第8期84-87,共4页 Ordnance Industry Automation
关键词 仿射变换 仿射不变性 不变特征提取 affine transform affine invariant feature invariant feature extraction
  • 相关文献

参考文献4

二级参考文献16

  • 1张宜.基于内容的图象检索技术研究综述[J].广西广播电视大学学报,2003,14(3):25-28. 被引量:7
  • 2刘九芬,黄达人,黄继武.图像水印抗几何攻击研究综述[J].电子与信息学报,2004,26(9):1495-1503. 被引量:43
  • 3邵东波.基于内容的图像检索介绍[J].电脑知识与技术,2006,1(7):175-176. 被引量:1
  • 4Pei S C,Lin C N.Image Normalization for Pattern Pecognition[J].Image and Vision Computing,1995,13(10):711-723.
  • 5Wood J.Invariant Pattern Recognition:A Review[J].Pattern Recognition,1996,29(1):1-17.
  • 6Ping Dong,Gaiatsanos N P.Affine transformation resistant water marking based on image normalization[C]// Image Processing.[S.l.]:[s.n.],2002.
  • 7Zheng D,Liu Y,Zhao J.A Survey of RST Invariant ImageWatermarking Algorithms[C]//Proceedings of Canadian Conference on Electrical and Computer Engineering(CCECE).[S.l.]:[s.n.],2006:2086-2089.
  • 8Charles Sutton, Andrew McCallum. An Introduction to Conditional Random Fields[M]. Foundations and Trends in Machine Learning, 2010.
  • 9余军,陈晓鸥.命名实体识别:One-at-a-time or All-at-once? Word-based or Character-based?[C]//萧国政,何炎祥,孙茂松.中国计算技术与语言问题研究:第七届中丈信息处理国际会议论文集.北京:电子工业出版社.2007:81-89.
  • 10CRF++: Yet another crf toolkit, http://crfpp.sourceforge.net/.

共引文献19

同被引文献22

引证文献2

二级引证文献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