首页 | 期刊介绍 | 编委会 | 道德声明 | 投稿指南 | 常用下载 | 过刊浏览 | In English
引用本文:
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  下载PDF阅读器  关闭
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 102次   下载 0  
分享到: 微信 更多
基于点云与影像融合的黄河三角洲互花米草提取方法
曹裕超1, 王健1, 邵为真2, 孙文潇1, 曾静静3
1.山东科技大学;2.山东新汇建设集团有限公司;3.青岛市勘察测绘研究院
摘要:
针对卫星遥感技术在对滨海湿地互花米草监测时受分辨率、气候条件等多种因素限制存在一定局限性且通过单一的影像数据提取互花米草时精度不稳定的问题,提出了基于无人机点云与影像融合的面向对象互花米草提取方法。以黄河三角洲自然保护区为研究对象,获取了该区域的点云和多光谱影像。先将点云进行地面滤波后提取植被点云并转化为二值图像,然后将滤波后的二值图像与多光谱影像进行特征组合优化以提高分类精度,并基于FNEA算法分割融合影像后采用改进的最近邻算法进行面向对象分类,最终得到的互花米草生产者精度和用户精度分别达到了82.53%和86.43%,较未融合点云的提取精度分别提高了22.34%和7.66%,表明本文提出的方法能够有效提高互花米草的提取精度,对互花米草的监测研究具有参考价值。
关键词:  黄河三角洲  互花米草  无人机  数据融合  面向对象分类
DOI:
分类号:
基金项目:山东省重大科技创新工程项目(2019JZZY020103)
Extraction method of Spartina alterniflora in Yellow River Delta based on fusion of point cloud and image
caoyuchao1, wangjian1, shaoweizhen2, sunwenxiao1, zengjingjing3
1.Shandong University of Science and Technology;2.Shandong Xinhui Construction Group Limited Company;3.Qingdao Institute of Geotechnical Investigation and Surveying Research
Abstract:
In view of the limitations of satellite remote sensing technology in monitoring Spartina alterniflora in coastal wetland due to many factors such as resolution and climate conditions, and the accuracy of extracting Spartina alterniflora from a single image data is unstable, this paper proposes an object-oriented extraction method of Spartina alterniflora based on UAV point cloud and image fusion. Taking the Yellow River Delta Nature Reserve as the research object, the point cloud and multispectral images of the region were obtained. Firstly, the point cloud is ground filtered, then the vegetation point cloud is extracted and transformed into a binary image, secondly, features of the filtered binary image and multispectral image are combined and optimized to improve the classification accuracy. After segmenting the fused image based on FNEA algorithm, the improved nearest neighbor algorithm is used for object-oriented classification, the producer accuracy and user accuracy of Spartina alterniflora reached 82.53% and 86.43% respectively, which was 22.34% and 7.66% higher than that without point cloud fusion, the results show that the method proposed in this paper can effectively improve the extraction accuracy of Spartina alterniflora, and has reference value for the monitoring and research of Spartina alterniflora.
Key words:  Yellow River Delta  Spartina alterniflora  UAV  data fusion  object-oriented classification
版权所有 《海洋科学》 Copyright©2008 All Rights Reserved
主管单位:中国科学院 主办单位:中国科学院海洋研究所
地址:青岛市市南区福山路32号  邮编:266071  电话:0532-82898755  E-mail:marinesciences@qdio.ac.cn
技术支持:北京勤云科技发展有限公司