首页 | 本学科首页   官方微博 | 高级检索  
     

针对高分辨率遥感影像分割的改进连通域标记方法
引用本文:王晶,张艳宁,骆剑承,明冬萍. 针对高分辨率遥感影像分割的改进连通域标记方法[J]. 计算机工程与应用, 2005, 41(10): 37-39,112
作者姓名:王晶  张艳宁  骆剑承  明冬萍
作者单位:西北工业大学计算机学院,西安,710072;中国科学院地理科学与资源研究所,北京,100101
基金项目:国家863高技术研究发展计划项目(编号:2002AA135230)
摘    要:针对高分辨率遥感影像信息量大的特点,提出一种新的连通域标记方法-包围判断式标记,快速完成对影像分割结果的连通域标记。尤其对分割结果中含有大面积连通分割子区域的情况,此方法能大大减少像素4-连通比较的次数,并且能在标记步骤有效地去除部分无意义的分割小区域。复杂度分析和实验结果证明,包围判断式标记方法比传统的顺序式标记方法具有更高的运行效率,更能适应高分辨率遥感影像处理的要求。

关 键 词:高分辨率遥感影像  连通区域标记  包围判断式标记  顺序式标记
文章编号:1002-8331-(2005)10-0037-03

Improved Connected Component Labeling on High-resolution Remote Sensing Image
Wang Jing,Zhang Yanning,Luo Jiancheng,Ming Dongping. Improved Connected Component Labeling on High-resolution Remote Sensing Image[J]. Computer Engineering and Applications, 2005, 41(10): 37-39,112
Authors:Wang Jing  Zhang Yanning  Luo Jiancheng  Ming Dongping
Affiliation:Wang Jing1 Zhang Yanning1 Luo Jiancheng2 Ming Dongping21
Abstract:Circling and Judging Labeling,a new connected component labeling strategy is proposed to implement real-time component labeling for segmented result of high-resolution remote sensing image.In the new strategy,computing times of 4-connected neighbor inspection are greatly reduced especially for segmented result containing connected components of large area.Besides,some meaningless segmented regions are removed during the labeling process.Experimental results and complexity analysis show that the new strategy is more efficient than traditional sequential labeling and more adaptive to high-resolution remote sensing image.
Keywords:high-resolution remote sensing image  connected component labeling  circling and judging labeling  sequential labeling
本文献已被 CNKI 维普 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