首页 | 本学科首页   官方微博 | 高级检索  
     

基于IIR指数平滑滤波与非线性Laplacian算子的边缘检测方法
引用本文:游素亚,柳健.基于IIR指数平滑滤波与非线性Laplacian算子的边缘检测方法[J].计算机学报,1997,20(7):667-672.
作者姓名:游素亚  柳健
作者单位:清华大学计算机科学与技术系
基金项目:国家自然科学基金,测绘遥感信息国家重点实验室基金
摘    要:本文提出了一种基于IIR指数平滑滤波与非线性Laplacian算子的边缘检测方法,具有以下特点;采用最佳指数平滑滤波器增强抗噪声能力,且可实现无限大窗口的快速递归运算;利用非线性Laplacian算子只沿边缘增强边缘,充分利用了边缘的局部方向信息;引入边缘强度信息,在二阶零交叉处理根据边缘强度定位边缘;运算量独立于滤波器窗口尺寸,具有较高的计算效率,通过与LOG算子的对比实验和在立体视觉中的应用表

关 键 词:边缘检测  IIR指数滤波器  非线性  Laplacian算子

AN EDGE DETECTION APPROACH VIA IIR EXPONENTIAL SMOOTHING FILTERING AND NONLINEAR LAPLACIAN OPERATOR
YOU Suya, LIU Jian, XU Guangyou.AN EDGE DETECTION APPROACH VIA IIR EXPONENTIAL SMOOTHING FILTERING AND NONLINEAR LAPLACIAN OPERATOR[J].Chinese Journal of Computers,1997,20(7):667-672.
Authors:YOU Suya  LIU Jian  XU Guangyou
Abstract:An edge detection approach via IIR exponential smoothing filtering and nonlinear Laplacian operator is proposed, which has the following advantages. At first, utilizing the exponential kind filter can greatly enhance the ability of smoothing noise, and can implement fast structure recurred filter. Secondly, the nonlinear Laplacian operator can adapt the edge local information and enhance edges only along the edge directions. Thirdly, the edges are located by combination edge strong information with the zero crossings. Finally, the operation which is independent of the size of the filter is efficient. The experimental results of comparing the proposed approach with the LOG operator and the application in stereo vision show that the proposed approach has greatly practical value.
Keywords:Edge detection  IIR exponential filter  nonlinear Laplacian operator    
本文献已被 CNKI 维普 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