|
|||||
|
|
| 一种基于Hop-Along的遥感图像道路提取算法 | |
| 引用本文： | 夏静,胡茂林,卜华龙.一种基于Hop-Along的遥感图像道路提取算法[J].合肥工业大学学报(自然科学版),2009,32(12). |
| 作者姓名： | 夏静 胡茂林 卜华龙 |
| 作者单位： | 1. 巢湖学院,数学系,安徽,巢湖,238000;安徽大学,数学与计算科学学院,安徽,合肥,230039 2. 安徽大学,数学与计算科学学院,安徽,合肥,230039 3. 巢湖学院,数学系,安徽,巢湖,238000 |
| 基金项目： | 国家863高技术研究发展计划资助项目，科技部支持计划资助项目，巢湖学院青年教师资助项目 |
| 摘 要： | 基于数值形态学的道路提取是遥感图像处理的常见问题,通常结构元素的选取复杂多变,降低了算法的自动化程度.文章结合数值形态学和Hop-Along算法,对影像进行预处理、阈值分割,得到包含道路信息的二值影像;使用数值形态学的腐蚀运算及形态重建,得到主要道路网络;使用局部Hop-Along算法进行多线拟合,得到单像素的道路中心线.通过对高分辨率遥感图像的Matlab实验证明,算法大幅降低了选取结构元素的次数,提高了自动化程度,且保持了数值形态学的几何学特性. |
| 关 键 词： | 道路提取 数值形态学 局部Hop-Along算法 |
Algorithm of road extraction from high resolution remote sensing images based on Hop-Along |
|
| XIA Jing,HU Mao-lin,BU Hua-long.Algorithm of road extraction from high resolution remote sensing images based on Hop-Along[J].Journal of Hefei University of Technology(Natural Science),2009,32(12). | |
| Authors: | XIA Jing HU Mao-lin BU Hua-long |
| Abstract: | Road extraction based on mathematic morphology is one of the most important issues in remote sensing image processing,and the selecting of the structure elements needs the human experience or experimental validity,which influences its stability and reduce its automatization.This paper combines mathematic morphology with the local Hop-Along algorithm.First,the image is preprocessed,the threshold is divided,and then the binary image is obtained.Secondly,the erode operator and reconstruction operator are used to obtain the main road network.At last,the local Hop-Along algorithm is adopted as the polyline fitting method to centerline the road network.The experiments in Matlab environment show that the presented algorithm can reduce the number of parameters,improve the efficiency and keep the superiority of mathematic morphology. |
| Keywords: | road extraction mathematic morphology local Hop-Along algorithm |
| 本文献已被 万方数据 等数据库收录！ | |