向量空间的彩色图像边缘检测  

Edge detection of color images in vector space

作　　者：夏泽邑[1,2] 刘冲[1,2] 王跃宗[1,2] 宋近才 

机构地区：[1]大连理工大学微系统研究中心,辽宁大连116024 [2]大连理工大学精密与特种加工教育部重点实验室,辽宁大连116024

出　　处：《光电工程》2005年第1期63-66,共4页Opto-Electronic Engineering

基　　金：国家自然科学基金(50275023);国家自然科学基金重点项目(50135040)

摘　　要：边缘检测是图像处理的重要组成部分之一。拓展小波变换为向量空间形式,并提出了基于向量小波变换的彩色图像边缘检测方法。根据小波变换的特性,检测模值局部极大点作为边缘点。引入大小自适应的卷积模板,小波变换尺度为 2j时,大小为(4×2j-1)×(4×2j-1),使计算量得到有效控制。实验结果证明,该方法继承了小波变换的性能,并充分利用了彩色信息,能完全地检测出彩色图像边缘。One of the fundamental tasks in image processing is edge detection. The wavelet transform was extended into vector form and a color edge detection based on the wavelet transform in vector space was developed. The correlation among the color channels was taken into account. According to the characteristic of wavelet transform, the points where the modulus of the wavelet transform has a local maximum in the direction were detected as edge points. Furthermore, a convolution template with adaptive size was used. The template size is [4(2j)-l]×[4(2j)-l], when the wavelet scale is 2j. The amount of calculation is controlled efficiently. The experiment results of a simulated color image and two real color images demonstrate that the edge detector inherits properties of wavelet transform, which can fully use the color information of color images, and detect entire edges of color images.

关 键 词：边缘检测 彩色图像 小波变换 向量空间 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]