| 脉冲耦合神经网络的遥感影像城市道路提取 |
| |
| 引用本文： | 孙震辉,孟庆岩,孙云晓,张佳晖,张琳琳.脉冲耦合神经网络的遥感影像城市道路提取[J].测绘科学,2018(1):145-152. |
| |
| 作者姓名： | 孙震辉 孟庆岩 孙云晓 张佳晖 张琳琳 |
| |
| 作者单位： | 中国科学院遥感与数字地球研究所,北京100101;中国科学院大学,北京100049;三亚遥感研究所,海南三亚572000
中国科学院遥感与数字地球研究所,北京100101;三亚遥感研究所,海南三亚572000 |
| |
| 基金项目： | 国土资源部城市土地资源监测与仿真重点实验室开放基金资助课题，高分辨率对地观测系统重大专项应用共性技术，海南省科技合作专项资金项目，三亚市专项科研试制项目 |
| |
| 摘 要： | 针对高分辨率遥感影像中阴影对道路提取产生较大干扰的问题,提出了一种基于脉冲耦合神经网络(PCNN)的城市道路提取方法。该方法首先在近红外波段检测并消除阴影和水体的影响,并使用PCNN对消除阴影后的灰度图像进行分割处理;然后使用形态学建筑物指数(MBI)和归一化差分植被指数(NDVI)分别提取出建筑物和植被信息,消除建筑物和植被的影响;最后提取受行道树影响较大的道路,并对处理后的图像作数学形态学法的处理。该文以深圳市SPOT-7高分辨率影像进行实验。实验表明,该方法能保留原始的道路边缘细节信息,并对阴影具有很好的抗干扰作用,提取的道路信息具有很好的连续性和完整性。
|
| 关 键 词： | 城市道路 脉冲耦合神经网络 阴影干扰 高分辨率遥感影像 urban roads PCNN shadow interference high resolution satellite image |
Remote sensing image extraction of urban road based on pulse coupled neural network |
| |
| SUN Zhenhui,MENG Qingyan,SUN Yunxiao,ZHANG Jiahui,ZHANG Linlin.Remote sensing image extraction of urban road based on pulse coupled neural network[J].Science of Surveying and Mapping,2018(1):145-152. |
| |
| Authors: | SUN Zhenhui MENG Qingyan SUN Yunxiao ZHANG Jiahui ZHANG Linlin |
| |
| Abstract: | According to the problem that road extraction is interfered by shadow in high-resolution remote sensing images,a method of urban road extraction based on pulse coupled neural network(PCNN)is proposed.Firstly,the method detects and eliminates the influence of shadow and water in near infrared band,and uses PCNN to segment the gray image.Then the morphological building index(MBI)and the normalized difference vegetation index(NDVI)are used to extract the building and vegetation information respectively to eliminate the influence of buildings and vegetation.Finally,the road which is affected by the street trees is extracted,and the processed images are processed by mathematical morphology.This paper carries on the experiment with SPOT-7 high resolution image in Shenzhen.The experimental results show that the method can retain the original road edge details and have a good anti-jamming effect on the shadow,and the extracted road information has good continuity and integrity. |
| |
| Keywords: | |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
