张杰, 胡海棠, 张丽, 李存军, 周静平, 谢春春. 基于高分辨率影像的蔡家河流域人工林Crop Science单木提取与缺失检测[J]. 西南林业大学学报, 2019, 39(1): 139-145. DOI: 10.11929/j.swfu.201807030
引用本文: 张杰, 胡海棠, 张丽, 李存军, 周静平, 谢春春. 基于高分辨率影像的蔡家河流域人工林Crop Science单木提取与缺失检测[J]. 西南林业大学学报, 2019, 39(1): 139-145. DOI: 10.11929/j.swfu.201807030
Jie Zhang, Haitang Hu, Li Zhang, Cunjun Li, Jingping Zhou, Chunchun Xie. Individual Tree Extraction and Deletion Detection of Plantation in Caijiahe Basin Based on High Resolution Image by Crop Science[J]. Journal of Southwest Forestry University, 2019, 39(1): 139-145. DOI: 10.11929/j.swfu.201807030
Citation: Jie Zhang, Haitang Hu, Li Zhang, Cunjun Li, Jingping Zhou, Chunchun Xie. Individual Tree Extraction and Deletion Detection of Plantation in Caijiahe Basin Based on High Resolution Image by Crop Science[J]. Journal of Southwest Forestry University, 2019, 39(1): 139-145. DOI: 10.11929/j.swfu.201807030

基于高分辨率影像的蔡家河流域人工林Crop Science单木提取与缺失检测

Individual Tree Extraction and Deletion Detection of Plantation in Caijiahe Basin Based on High Resolution Image by Crop Science

  • 摘要: 以北京市延庆区蔡家河流域人工林为研究对象,以蔡家河流域平原造林区的高分辨遥感影像为数据源,利用ENVI 5.4的Crop Science工具包分别对阔叶林和针叶试验林样区进行单木提取、缺失单木检测,并对研究结果进行精度评价和对比分析。结果表明:Crop Science对人工幼林进行单木提取效果较好,总体精度达到94%以上;缺失单木的识别受林木排列规整程度影响较大,排列越规整,提取效果越好。本研究探索了一种基于高分辨率遥感数据进行人工林地单木定位、计数及缺失单木查找的简便可行的方法,有利于林业管护人员快速获取高精度林木监测信息。

     

    Abstract: Taking the plantation of Caijiahe Basin in Yanqing District of Beijing as the research object, the high-resolution remote sensing image of plain afforestation area of Caijiahe Basin was taken as the data source. Using the Crop Science toolkit of ENVI 5.4, extraction and detection of individual tree were performed on the broad-leaved forest and coniferous forest samples, and the results were evaluated and compared. The results show that the Crop Science toolkit has high efficiency and accuracy on extraction of individual tree for young plantation and the overall accuracy is over 94%. The identification of individual tree is greatly affected by the regularity of the trees distribution. The tree locations with structured rows, can get the better extraction result. In this study, we introduce an effective and convenient tool to quickly obtain the location, count at the individual tree scale, based on high resolution remote sensing data. It is beneficial for forestry management and maintenance personnel to quickly obtain high-precision forest monitoring information.

     

/

返回文章
返回