首页 | 本学科首页   官方微博 | 高级检索  
     

图像识别的RSTC不变矩
引用本文:王连亮,陈怀新.图像识别的RSTC不变矩[J].数据采集与处理,2006,21(2):225-229.
作者姓名:王连亮  陈怀新
作者单位:1. 西南电子技术研究所,成都,610036
2. 西南电子技术研究所,成都,610036;四川大学电子信息学院,成都,610064
摘    要:分析了对比度变化对Hu矩的影响,构造了一种新的不变矩形式,该矩不但具有通常的旋转、尺度缩放以及平移不变性,同时还具有对比度变化不变性。以三类具有旋转、尺度缩放、平移以及对比度变化(RSTC)的目标图像为例进行了识别仿真实验,结果表明新的不变矩形式消除了对比度变化带来的影响,增强了三类图像的类内内聚性和类间可分性,实验结果证明能够对具有RSTC变化的目标图像进行有效识别。

关 键 词:模式识别  畸变不变识别  不变矩  对比度
文章编号:1004-9037(2006)02-0225-05
收稿时间:2004-04-15
修稿时间:2005-06-10

Moment Invariants of RSTC for Image Recognition
Wang Lianliang,Chen Huaixin.Moment Invariants of RSTC for Image Recognition[J].Journal of Data Acquisition & Processing,2006,21(2):225-229.
Authors:Wang Lianliang  Chen Huaixin
Affiliation:1. Southwest China Institute of Electronic Technology, Chengdu, 610036, China; 2. School of Electronics and Information Engineering, Sichuan University, Chengdu, 610064, China
Abstract:The influence of contrast change on Hu moments and is analyzed and a new kind of invariant moments is proposed.The moments have the accustomed invariance of rotation,scaling and translation,and contrast change.A simulating experiment is carried out using three kinds of target images with rotation,scaling,translation and contrast change(RSTC).Results show that the new invariant moments have eliminated the influences of contrast change and strengthened the centrality within a class and the separability between classes.They can also be used to recognize the target images with RSTC change effectively.
Keywords:pattern recognition  distortion invariant recognition  moment invariant  contrast
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号