一种小波滤波器的构造与多尺度边缘检测-论文.pdfVIP

_____________________________________________________________________________ 一种小波滤波器的构造与多尺度边缘检测 林 怡 陈 鹰 （同济大学测量与国土信息工程系，上海，200092） 摘要：Marr 曾经指出，人的视觉对于影像的描述具有多尺度的特性.本文基于这一思想，从二维平滑函数出发， 并根据尺度呈级数变化的特点，导出了一组多尺度小波滤波器。该滤波器低通响应关于原点对称，高通响应关于 原点反对称，而且截断误差很小，具有近似的紧支撑性和平滑性。同时给出了9 组滤波器响应系数，可以实现由 粗到精的特征提取，为影像的多尺度边缘检测提供了有用的工具。另一方面，影像经过该滤波器处理后，利用小 波变换系数模的局部极大值来提取图像的边缘特征不仅能够有效的抑制噪声，而且能够以子象素精度确定边缘的 位置。最后，利用该滤波器对建筑物遥感影像进行了边缘检测试验，获得了良好的结果，为遥感影像建筑物边缘 的自动提取打下了基础。 关键词：小波滤波器；多尺度分析；对称与反对称；边缘检测。 The Construction of a Wavelet Filter and Multiscale Edge Detection LIN Yi CHEN Ying (Department of Surveying and Geo-Informatics, Tongji University, Shanghai, 200092) Abstract: Marr has point out that describing images in human vision has multiresolution characteristics. In this paper, according to the peculiarity that the scale exhibit progressional variety, we deduced a group of multiscale wavelet filter from two dimensions smooth function. The lowpass impulse response of the filter is symmetric at origin, the hignpass response of the filter is anti-symmetric at origin, and the truncation error is very small. They have the characteristics of approximation compact support and smoothness. Meanwhile nine group of filter response coefficient are given to realize the feature extraction from coarse to fine, and providing a useful tools for multiscale edge detection. On the other hand, Detecting the modulus maximum of the wavelet transformation coefficient to acquire the image edge feature after image processing by using the filter. The method not only effectively reduces noise, but also gets the edge position at sub-pixel precision. Finally, the effectiveness of our method is tested, which provides a fundamental basis for automatic extracting the building on r