首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于高分辨率遥感影像的耕地地块提取方法研究
引用本文:庞新华,朱文泉,潘耀忠,贾斌.基于高分辨率遥感影像的耕地地块提取方法研究[J].测绘科学,2009,34(1).
作者姓名:庞新华  朱文泉  潘耀忠  贾斌
作者单位:北京师范大学资源学院,地表过程与资源生态国家重点实验室,北京,100875
基金项目:国家自然科学基金,北京师范大学青年基金 
摘    要:针对耕地地块提取问题,提出了一种基于图像分割的耕地地块提取新方法。该方法以高分辨率遥感影像为基础,借助于边缘提取和数学形态学的方法,通过边缘检测、边缘闭合、区域标号和后处理四个步骤,提取耕地地块。该方法在IDL6.3平台下编程实现。将此方法应用于北京地区QuickBird多光谱遥感影像,结果表明此方法有较好的定位精度,又在一定程度上去除了噪声,具有较好的实用性。

关 键 词:地块提取  图像分割  高分辨率遥感影像  边缘检测  数学形态学

Research on cultivated land parcel extraction based on high-resolution remote sensing image
PANG Xin-hua,ZHU Wen-quan,PAN Yao-zhong,JIA Bin.Research on cultivated land parcel extraction based on high-resolution remote sensing image[J].Science of Surveying and Mapping,2009,34(1).
Authors:PANG Xin-hua  ZHU Wen-quan  PAN Yao-zhong  JIA Bin
Abstract:Cultivated land parcel extraction is critical for national operational monitoring system of crop planting area on large scale,especially for the national scale like China.According to edge detection and mathematical morphology theories,an effective algorithm has been put forward to extract parcels based on high-resolution remote sensing image.The algorithm consists of edge detection,edge closing,region labeling and hysteresis.Experiment with Quickbird images indicates that this algorithm has better performance on not only location accuracy but also denoise.This algorithm is run on IDL 6.3 software.
Keywords:parcel extraction  image segmentation  high-resolution remote sensing image  edge detection  mathematical morphology
本文献已被 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号