Rationale and conceptual framework for classification approaches to assess forest resources and properties
Rationale and conceptual framework for classification approaches to assess forest resources and properties
复制标题
评估森林资源和特性的分类方法的基本原理和概念框架
DOI：
10.1007/978-1-4615-0306-4_10
复制
发表时间：
2003
期刊：
影响因子：
3.7
通讯作者：
C. Woodcock
中科院分区：
文献类型：
--
作者：
J. Franklin;J. Rogan;S. Phinn;C. Woodcock
自早期Landsat任务以来，分类一直是陆地资源应用中数字图像分析的重要工具，当时人们认识到多光谱数字图像由每个像素的多变量测量向量组成。通常构成图像的数十万个这样的向量可以被视为类描述符，而光谱带则是与图像中感兴趣的类别相关的解释变量。这是一种更通用的分类或模式识别方法的应用（Ripley 1996）。本章为选择适当的分类方法来评估森林资源和森林（林冠、林分和景观）属性提供了一个概念框架。它超出了本章的范围，以提供一个全面的审查，最近的文献图像分类。S. E. Franklin（2001）为森林遥感分类提供了一个极好的概述，我们将其作为出发点。Jensen（1996）和Schowengerdt（1997）等教科书对遥感分类的一般问题提供了全面的解释。最近的几篇综述也概述了在森林遥感中使用分类的进展（Wulder 1998年，triietz和Howarth 1999年，Lucas等人出版，Woodcock出版）。
Classification has been an important tool in digital image analysis for land resources applications since early Landsat missions when it was recognized that multispectral digital images are composed of multivariate measurement vectors for each and every pixel. The hundreds of thousands of such vectors typically making up an image could be treated as class descriptors, and the spectral bands as explanatory variables related to categories of interest in the image. This is an application of the more general methodology of classification or pattern recognition (Ripley 1996). This Chapter provides a conceptual framework for selecting appropriate classification approaches to assess forest resources and forest (canopy, stand, and landscape) properties. It is beyond the scope of this Chapter to provide a comprehensive review of recent literature on image classification. S. E. Franklin (2001) provided an excellent overview of classification for the remote sensing of forests, and we use that work as a point of departure. Textbooks such as those by Jensen (1996) and Schowengerdt (1997) provide comprehensive explanations of the general problem of classification in remote sensing. Several recent reviews also outline advances in the use of classification in forest remote sensing Wulder 1998, Trietz and Howarth 1999, Lucas et al. in press, Woodcock in press).
《通过灾害数据的收集和分析研究自然灾害事件——研究成果报告》（1986）
DOI：
--
发表时间：
--
期刊：
影响因子：
--
作者：
通讯作者：
--