首页 | 官方网站   微博 | 高级检索  
     

反投影构造二维旋转不变小波用于边缘检测
引用本文:曾理,黄建元,王珏. 反投影构造二维旋转不变小波用于边缘检测[J]. 计算机工程与应用, 2006, 42(10): 31-33,68
作者姓名:曾理  黄建元  王珏
作者单位:重庆大学数理学院ICT中心,重庆,400044
摘    要:通过数值反投影的方法,从一维小波得到二维小波的离散形式。该二维小波具有旋转不变的特点,所以图像在每个方向上的边缘都能得到有效的检测。在非极大值抑制中,同时使用这种二维旋转不变小波与二维Mallat方向小波,阈值的设置更加合理,从而使边缘的定位更加准确。试验表明这种二维旋转不变小波结合Mallat方向小波有更好的检测效果。

关 键 词:小波  反投影  旋转不变  边缘检测
文章编号:1002-8331-(2006)10-0031-03
收稿时间:2005-09-01
修稿时间:2005-09-01

The Numerical Construction of 2D Rotationally Invariant Wavelet by Back-projection and its Application in Image Edge Detection
Zeng Li,Huang Jianyuan,Wang Jue. The Numerical Construction of 2D Rotationally Invariant Wavelet by Back-projection and its Application in Image Edge Detection[J]. Computer Engineering and Applications, 2006, 42(10): 31-33,68
Authors:Zeng Li  Huang Jianyuan  Wang Jue
Affiliation:ICT Center,College of Mathematics and Physics,Chongqing University,Chongqing 400044
Abstract:By back-projection we can gain the discrete form of 2D wavelet from 1D wavelet.The 2D wavelet is rotationally invariant so that the edge in each direction can be detected effectively.The 2D rotationally invariant wavelet and the 2D Mallat directional wavelets are used in Non-Max Suppression,the thresholds are more reasonable and the edge are more exactly.Experiments show that associative algorithm of the 2D rotationally invariant wavelet transform and 2D Mallat wavelet transform is better in edges detections.
Keywords:wavelet   back-project   rotationally invariant   edge-detection
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-23

京公网安备 11010802026262号