一种高分辨率遥感图像单木树冠信息提取方法

发布时间：2018-03-24 17:21

本文选题：高分辨率遥感图像　切入点：面向对象　出处：《遥感信息》2017年03期

【摘要】：单木树冠信息是森林管理和相关科学研究的基础数据。为解决面向对象方法在树冠相互连接时无法有效提取单木树冠信息的问题,提出了一种基于面向对象技术和水文分析的高分辨率遥感图像单木树冠信息提取方法。方法首先对图像进行预处理,然后利用面向对象方法从融合图像获取树冠的分布范围,利用水文分析技术从全色波段图像获取单木树冠的潜在分布范围,最后将提取的数据进行叠加相交处理,完成单木位置信息提取和单木树冠描绘。实验结果表明,所提出的方法能有效提取单木位置和描绘单木树冠;采用修改的遥感分类精度评价指标进行的精度评价结果为,单木探测总体精度为87.63%,单木树冠描绘总体精度为84.89%。
[Abstract]:Single tree crown information is the basic data of forest management and related scientific research. In this paper, a method of extracting single tree crown information from high resolution remote sensing image based on object oriented technology and hydrologic analysis is proposed. Firstly, the image is preprocessed, and then the distribution range of tree crown is obtained from fusion image by object oriented method. The potential distribution range of single tree crown is obtained from panchromatic band images by hydrological analysis technique. Finally, the extracted data are superimposed and intersected to extract the position information of single tree and depict the single tree crown. The experimental results show that, The proposed method can effectively extract the position of single tree and depict single tree crown, and the accuracy evaluation results of the modified remote sensing classification accuracy evaluation index are as follows: the overall accuracy of single wood detection is 87.63, and the total precision of single tree crown description is 84.89.
【作者单位】：深圳市规划国土发展研究中心;深圳市数字城市工程研究中心;南京信息工程大学地理与遥感学院;
【基金】：国土资源部公益性行业专项(201411014-4) 深圳市2012年测绘地籍工程计划项目([2012]0365)
【分类号】：TP751

【参考文献】