随机森林遥感图像分类实验教学软件开发研究  

Software Development for Experimental Teaching of Remote Sensing Image Classification Based on Random Forest

作　　者：苏腾飞[1] 罗艳云[1] 翟涌光[1] 张圣微[1] SU Tengfei;LUO Yanyun;ZHAI Yongguang;ZHANG Shengwei(College of Water Conservancy and Civil Engineering,Inner Mongolia Agricultural University,Hohhot 010018,China)

机构地区：[1]内蒙古农业大学水利与土木建筑工程学院,内蒙古呼和浩特010018

出　　处：《测绘与空间地理信息》2021年第11期80-83,86,共5页Geomatics & Spatial Information Technology

基　　金：内蒙古农业大学实验教学仪器设备研制与标本制作项目基金(YZ2019011);国家自然科学基金项目(61701265)资助。

摘　　要：随着教育与科技的发展,我国众多高校的课程体系都开设了“遥感图像解译”或“遥感原理与应用”等课程。遥感图像的监督分类,是遥感类课程不可或缺的教学内容。然而,在目前相关课程的实验教学中,存在教学内容陈旧、更新频率较低等问题。为了充实高校遥感类课程的实验教学,本文介绍了一套基于随机森林的遥感图像分类实验教学软件,包括软件开发与相关实验的设计。鉴于随机森林算法在遥感领域得到了广泛应用,本实验教学软件旨在使学生掌握更新的遥感数据处理技能,提升其在相关行业中的竞争力。With the development of education and technology,the course systems of many universities of China have adopted the courses such as“Remote Sensing Image Interpretation”and“The Principle and Application of Remote Sensing”.Supervised remote sensing image classification is indispensable in the teaching contents of remote sensing-related education.However,existing experimental teaching has some issues,for example,the contents are outdated,and the updating is infrequent.To enrich the experimental teaching of remote sensing-related courses,this article introduces a pilot software for remote sensing image classification experiment based on random forest,including the development of the software and the design of the related experiment.In light of the fact that random forest has been widely applied in the field of remote sensing,the software aims to equip the students with newer skills for remote sensing data processing,and thus raising their competitiveness in the related industries.

关 键 词：随机森林 遥感图像 分类 实验教学 

分 类 号：P237[天文地球—摄影测量与遥感]