首页 | 本学科首页   官方微博 | 高级检索  
     

一种基于小波变换的照明无关边缘检测和模糊增强方法
引用本文:曹万鹏,陈刚,车仁生,叶东.一种基于小波变换的照明无关边缘检测和模糊增强方法[J].光学精密工程,2007,15(7):1143-1150.
作者姓名:曹万鹏  陈刚  车仁生  叶东
作者单位:哈尔滨工业大学自动化测试与控制系,黑龙江,哈尔滨,150001
摘    要:提出了一种基于小波变换的照明无关边缘检测和模糊增强算法,用于从不均匀的弱照明图像中提取目标边缘。依据照明反射图像形成模板和CCD相机成像公式,推导出图像的小波变换公式。对图像局部区域中边缘与背景像素的小波系数进行比较分析,设计了一种照明无关的小波边缘检测公式。给出一种同时考虑小波模值大小和梯度方向的模糊算子来增强边缘并抑制噪声。最后,采用仿真和真实的图像对该算法进行验证,利用此算法检测阶梯边缘,得到该算法的边缘检测评价标准F系数值为0.984 3,边缘定位精度评价系数Ed值为0.126 5,通过被检测的特征球边缘计算得到的交比值误差为3.72×10-3。实验结果证实,该边缘检测方法能够很好地工作于非均匀的弱照明图像。

关 键 词:边缘检测  小波变换  照明无关  模糊增强
文章编号:1004-924X(2007)07-1143-08
收稿时间:2007-01-14
修稿时间:2007年1月14日

An illumination-independent edge detection and fuzzy enhancement algorithm based on wavelet transform
CAO Wan-peng,CHEN Gang,CHE Ren-sheng,YE Dong.An illumination-independent edge detection and fuzzy enhancement algorithm based on wavelet transform[J].Optics and Precision Engineering,2007,15(7):1143-1150.
Authors:CAO Wan-peng  CHEN Gang  CHE Ren-sheng  YE Dong
Affiliation:Department of Automatic Measurement and Control, Harbin Institute of Technology, Harbin 150001,China
Abstract:An illumination-independent edge detection and fuzzy enhancement algorithm based on wavelet transform is proposed to extract edges out of a non-uniform weak illumination image. By educing the image wavelet transform formula from the image formation model and CCD camera imaging formula, the pixels’ wavelet module coefficients in the edge area of the image are analyzed and compared, and an illumination-independent wavelet edge detection formula insensitive to illumination is designed. Then, the fuzzy enhancement operator is proposed,which increases the difference between the edges,backgroud and noise by taking into account the wavelet module magnitude and gradient direction.Finally,through the synthetic and real images experiments, the performance of proposed edge detection method is evaluated. It is concluded that the edge detection evaluation criterion F of the proposed method for step edge is 0.984 3, the location accuracy Ed is 0.126 5, and the mean error of the cross ratio computed by the detected feature circles edge is 3.72×10-3. The experimental results prove that this edge detection method works well for the non-uniform weak illumination images.
Keywords:edge detection  wavelet transform  illumination-independent  fuzzy enhancement
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《光学精密工程》浏览原始摘要信息
点击此处可从《光学精密工程》下载全文
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