随机森林遥感信息提取研究进展及应用展望

Random Forest Classifier in Remote Sensing Information Extraction:A Review of Applications and Future Development

针对国内外随机森林集成分类方法的相关成果及发展趋势尚未有研究进行梳理与展望这一问题,首先,系统介绍随机森林分类方法的基本原理及应用优势、重要参数及其具体设定;其次,综述该方法在多光谱影像、高光谱数据、雷达及激光测距仪等多源遥感数据信息提取领域以及分类参量遴选中的研究应用;最后,在分类精度检验、可移植性以及算法改进等方面对其发展及应用趋势进行了展望。该研究可为随机森林分类方法初学者提供参考,有助于随机森林分类方法在遥感信息提取领域的推广及应用。

The random forest classifier (RFC) is an ensemble method that produces multiple decision trees,using a randomly selected subset of training samples and feature variables.The classifier has become popular in remote sensing studies due to its classification accuracy,while no literature review has been done to cover its application in remote sensing.The objective of this study is to review the utilization of RFC in remote sensing,and the application of RFC in the classification of multi-sensor images and relevant data selection.Further investigations are recommended into less commonly exploited use of this classifier,such as outliers detecting in training samples and novel approaches improving.