An Overview on Remote Sensing Image Classification Methods with a Focus on Support Vector Machine

支持向量机 过度拟合 人工智能 计算机科学 上下文图像分类 机器学习 模式识别(心理学) 核(代数) 领域(数学) 分类器(UML) 高光谱成像 图像(数学) 数据挖掘 人工神经网络 数学 组合数学 纯数学
作者
Haodong Li
标识
DOI:10.1109/conf-spml54095.2021.00019
摘要

With the growing demand for better performance of remote sensing (RS) image classification, a variety of methods have been proposed in RS image classification field in recent years. In general, there are two categories of RS image classification methods: pixel-based (PB) approach and object-based (OB) approach. In this paper, RS image classification methods are reviewed from the perspective of PB approach and OB approach and, specifically, the development and characteristics of a promising methodology for RS image classification named support vector machine (SVM) are surveyed. SVM is particularly popular in the RS field since it can deal with small-sized training dataset and provide higher classification accuracy than some traditional methods like maximum likelihood classifier. Besides, SVM has advantages of high memory-efficiency and strong generalization. However, SVM-based approaches also suffer from some problems. For instance, SVM-based methods tend to overfit due to inappropriate choice of kernel functions and it is inefficient for them to determine the optimum kernel function parameters as well as to process hyperspectral images. This paper also proposes the improvement of SVM-based methods aiming to address the limitations and improve the performance of SVM in RS image classification field. Moreover, future directions for SVM in RS image classification field are presented, expecting to help researchers to find possible research focuses in the future.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
1秒前
老赵是真的帅完成签到,获得积分20
1秒前
Paralloria发布了新的文献求助10
2秒前
晨曦完成签到,获得积分10
4秒前
llllhh发布了新的文献求助10
5秒前
zxy发布了新的文献求助10
5秒前
6秒前
6秒前
7秒前
9秒前
9秒前
11秒前
隐形曼青应助i3utter采纳,获得10
11秒前
Paralloria完成签到,获得积分10
12秒前
Transition发布了新的文献求助30
13秒前
13秒前
14秒前
烟花应助momo采纳,获得10
14秒前
congenialboy发布了新的文献求助10
16秒前
Ai完成签到,获得积分10
17秒前
18秒前
19秒前
21秒前
ddddansu发布了新的文献求助10
23秒前
hsuan风向仪发布了新的文献求助100
25秒前
酷波er应助然大宝采纳,获得10
25秒前
ukmy发布了新的文献求助10
25秒前
26秒前
26秒前
LUO完成签到 ,获得积分10
27秒前
Danny完成签到,获得积分10
30秒前
30秒前
31秒前
青黛发布了新的文献求助10
31秒前
明理的天抒完成签到 ,获得积分10
32秒前
哈哈哈应助科研通管家采纳,获得10
33秒前
33秒前
田様应助科研通管家采纳,获得10
33秒前
34秒前
34秒前
高分求助中
A new approach to the extrapolation of accelerated life test data 1000
ACSM’s Guidelines for Exercise Testing and Prescription, 12th edition 500
‘Unruly’ Children: Historical Fieldnotes and Learning Morality in a Taiwan Village (New Departures in Anthropology) 400
Indomethacinのヒトにおける経皮吸収 400
Phylogenetic study of the order Polydesmida (Myriapoda: Diplopoda) 370
基于可调谐半导体激光吸收光谱技术泄漏气体检测系统的研究 350
Robot-supported joining of reinforcement textiles with one-sided sewing heads 320
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3989797
求助须知:如何正确求助?哪些是违规求助? 3531914
关于积分的说明 11255516
捐赠科研通 3270597
什么是DOI,文献DOI怎么找? 1805008
邀请新用户注册赠送积分活动 882181
科研通“疑难数据库(出版商)”最低求助积分说明 809190