【摘要】 影像匹配在多级分辨率构成的金字塔影像系统上进行由粗到精的匹配，一个GIS需要能处理多比例尺的要求，一个影像分析系统要能有效地进行多尺度的影像信息的分析。因此人们必然需要研究多分辨率信号处理的理论，其中小波理论就是最近发展的并且是目前公认为有前途的一种理论。本文介绍了小波理论的基本概念和性质，阐述了利用高斯和三次样条作为尺度函数生成小波的方法，并运用它们来检测影像的边缘。通过与Sobel和零交叉提取边缘方法实验结果的比较，说明了该方法的优越性。这一理论对模式识别，计算机视觉，影像分析，偏微分方程的求解和分形分析是有意义的。更多还原

【Abstract】 :The image pyramid with multi-level resolution is needed for image matching, a mod ern GIS should have the multiscale function, and an image analysis system should effe ctively analyse the multiresolution information content of images. Therefore, the devel opment of new theory which concentrates on the muUiresolution signals is necessary. The wavelet theory is nowadays such a greatly potential theory. For this reason, we introduce the basic concepts and properties of the wavelet theory. In this paper, the me thods using Gaussian and Cubic splines functions as the scaling function to generate wavelets are described for image edge detection from digital airphotographys. Comparing with the operators of Zero crossing and Sobel, new approach posses many advantages. This theory is valuable for pattern recognition, computer vision, image analysis, the re solutions of partial differential equations and the analysis of fractals.更多还原