文章基本信息

标题：Land-Use Classification Using Convolutional Neural Network with Bagging and Reduced Categories
本地全文：下载
作者：Noritaka Shigei ; Kazuki Mandai ; Satoshi Sugimoto 等
期刊名称：Lecture Notes in Engineering and Computer Science
印刷版ISSN：2078-0958
电子版ISSN：2078-0966
出版年度：2019
卷号：2239
页码：7-11
出版社：Newswood and International Association of Engineers
摘要：In this paper, we tackle with the land-use classification from aerial photographs in the hilly and mountainous areas and apply two kinds of approaches to the problem. The one is the ensemble learning approach to improve the classification accuracy for overall classes, and the other is the optimization of the number of classes to improve the classification accuracy for coniferous forest. Our ensemble learning approach adopts Bagging and uses a Convolutional Neural Network (CNN) classifier as a weak learner. The optimization of the number of classes utilizes the spectral clustering algorithm and the confusion matrix of the classification result obtained by a CNN classifier. The effectiveness of the proposed approaches is demonstrated by numerical simulations.
关键词：convolutional neural network; land;use classification; Bagging; aerial photograph