摘要
在分析不同分辨率图像空间特性的基础上,提出了一种基于多方向二进小波变换的图像特征提取方法.该方法首先对图像进行小波变换,然后由二进小波变换的模的局部极大值检测信号突变点位置及奇异性大小,实现图像的边缘特征提取.通过对遥感图像样本的仿真实验表明,基于多方向二进小波变换的图像边缘特征提取方法可以取得较好的边缘特征提取效果.
By analyzing the spatial property of different resolution image, an approach to image detection based on multi-direction two-scale wavelet transform is proposed in this paper. Through this algorithm, a digital image is firstly decomposed with two-scale wavelet transform, and according to the part max module of wavelet transform, the pixel position of grayness changed sharply and the value of singularity can be detected, then the process of edge extraction is fulfilled. Satisfactory and effective results of edge detection have been obtained by analyzing the stylebook of a normal remote-sensing image.
出处
《小型微型计算机系统》
CSCD
北大核心
2005年第6期1054-1056,共3页
Journal of Chinese Computer Systems
基金
国家自然科学基金(60072034)资助
南京气象学院科研基金(Y318)资助
关键词
多方向二进小波
多分辨率分析
局部极大值
特征提取
multi-direction two-scale wavelet
multi-resolution analysis
local max
edge extraction
