一种基于融合技术的遥感图像边缘检测算法  

A remote sensing image edge detection algorithm based on fusion technology

作　　者：刘加伶[1] 杨帆[2] 龚箭[1] 

机构地区：[1]重庆工学院计算机科学与工程学院,重庆400050 [2]重庆邮电大学计算机科学与技术学院,重庆400065

出　　处：《重庆邮电大学学报（自然科学版）》2009年第3期445-448,共4页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：重庆市教委科学技术研究项目(KJ080608)

摘　　要：通过讨论遥感图像的边缘特点,为提取其图像边缘,提出了一种有效的融合多种方法的新的边缘检测算法算法针对遥感图像对象的复杂性、高边缘密度、噪声明显等特性,融合了滑动窗口技术、多阈值技术和模糊增强方法对遥感图像进行了边缘强化,最后利用模糊形态学算子进行了边缘提取实验表明,利用多方法融合,不仅可以弥补单一方法的缺点,还能提高边缘检测的精度。Through discussing the characteristics of remote sensing image, a new algorithm for remote sensing image edge detection based on fusion technology was introduced. In accordance with the complexity, high edge density, and obvious noise in remote sensing image, this algorithm fuses the fuzzy mathematical morphology, the multi-threshold , the method of slipping window and the fuzzy enhancing to strengthen the edges, and it detects the edges of remote sensing image by fuzzy mathematical morphology operators. Simulation result shows that this algorithm solves the problem by using single method, and elevates the edges and keeps the image sharp at the same time.

关 键 词：模糊形态学 模糊增强 多阈值 滑动窗口 边缘检测 融合 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]