[1]戴修斌,张辉,舒华忠,等.基于正交矩模糊和仿射混合不变量的图像识别算法[J].东南大学学报(自然科学版),2011,41(1):52-57.[doi:10.3969/j.issn.1001-0505.2011.01.011]
 Dai Xiubin,Zhang Hui,Shu Huazhong,et al.Image recognition algorithm based on combined blur and affine invariants of legendre moment[J].Journal of Southeast University (Natural Science Edition),2011,41(1):52-57.[doi:10.3969/j.issn.1001-0505.2011.01.011]
点击复制

基于正交矩模糊和仿射混合不变量的图像识别算法()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
41
期数:
2011年第1期
页码:
52-57
栏目:
计算机科学与工程
出版日期:
2011-01-20

文章信息/Info

Title:
Image recognition algorithm based on combined blur and affine invariants of legendre moment
作者:
戴修斌张辉舒华忠罗立民
(东南大学计算机科学与工程学院, 南京 210096)
Author(s):
Dai XiubinZhang HuiShu HuazhongLuo Limin
(School of Computer Science and Engineering, Southeast University, Nanjing 210096, China)
关键词:
模糊和仿射形变图像识别图像矩模糊和仿射混合不变量Legendre正交矩
Keywords:
recognition of blurred and affinely transformed image combined blur and affine moment invariant orthogonal Legendre moment
分类号:
TP391
DOI:
10.3969/j.issn.1001-0505.2011.01.011
摘要:
为了识别含有模糊和仿射混合形变的图像,提出了一种新的基于正交矩模糊和仿射混合不变量的图像识别算法.该算法首先使用归一化方法构造了基于Legendre正交矩的仿射不变量,并结合Legendre正交矩的模糊不变量提出了Legendre正交矩的模糊和仿射混合不变量; 然后将该混合不变量作为描述算子,将欧几里德范数作为分类尺度,以最近邻法则作为分类器,对图像进行识别.实验结果表明,与其他基于非正交矩的混合不变量相比,基于Legendre正交矩的模糊和仿射混合不变量在混合形变下能够获得更好的不变性,不会带来信息冗余问题,并且对噪声鲁棒性较好; 此外,该图像识别算法比其他算法具有更高的识别率,特别是在图像含有较大噪声的情况下.
Abstract:
To recognize the images distorted by combined blur and affine transformation, a new image recognition algorithm based on combined blur and affine invariants of orthogonal moment is proposed. In this algorithm, the normalization method is used to construct affine invariants of orthogonal Legendre moment. The combined blur and affine moment invariants of orthogonal Legendre moment is proposed with the help of the blur invariants. Then the new combined invariants, the Euclidean norm and nearest neighbor rules are taken as the descriptors, the measurement of classification and classifier respectively to recognize the image. The experimental results show that compared with the non-orthogonal moment invariants, the combined blur and affine invariants of orthogonal Legendre moment can achieve better invariance, less redundant information and better robustness to noise. Furthermore, this algorithm can obtain higher recognition rates, especially when the images are heavily corrupted by noise.

参考文献/References:

[1] Flusser J,Suk T,Zitov B.Moments and moment invariants in pattern recognition[M].Oxford:John Wiley &Sons,2009:1-5.
[2] Flusser J,Suk T.Pattern recognition by affine moment invariants[J].Pattern Recognition,1993,26(1):167-174.
[3] Reiss T H.The revised fundamental theorem of moment invariants[J].IEEE Trans Pattern Anal Mach Intell,1991,13(8):830-834.
[4] Rothe I,Susse H,Voss K.The method of normalization to determine invariants[J].IEEE Trans Pattern Anal Mach Intell,1996,18(4):366-375.
[5] Flusser J,Suk T.Degraded image analysis:an invariant approach[J].IEEE Trans Pattern Anal Mach Intell,1998,20(6):590-603.
[6] Flusser J,Boldy J,Zitov B.Moment forms invariant to rotation and blur in arbitrary number of dimensions[J].IEEE Trans Pattern Anal Mach Intell,2003,25(2):234-246.
[7] Zhang H,Shu H Z,Luo L M,et al.Blurred image recognition by Legendre moment invariants[J].IEEE Trans Image Process,2010,19(3):596-611.
[8] Zhang Y,Wen C,Zhang Y,et al.Determination of blur and affine combined invariants by normalization[J].Pattern Recognition,2002,35(1):211-221.
[9] Suk T,Flusser J.Combined blur and affine moment invariants and their use in pattern recognition[J].Pattern Recognition,2003,36(12):2895-2907.
[10] Shu H Z,Zhou J,Han G N,et al.Image reconstruction from limited range projections using orthogonal moments[J].Pattern Recognition,2007,40(2):670-680.
[11] Suk T,Flusser J.Affine normalization of symmetric objects[J].Lecture Notes in Computer Science,2005,3708:100-107.

备注/Memo

备注/Memo:
作者简介:戴修斌(1980—),男,博士,hustdxb@gmail.com.
基金项目:国家重点基础研究发展计划(973计划)资助项目(2010CB732503)、中国博士后科学基金资助项目(20100471360).
引文格式: 戴修斌,张辉,舒华忠,等.基于正交矩模糊和仿射混合不变量的图像识别算法[J].东南大学学报:自然科学版,2011,41(1):52-57.[doi:10.3969/j.issn.1001-0505.2011.01.011]
更新日期/Last Update: 2011-01-20