摘要
针对高分辨遥感图像样本量小,以及传统优化支持向量机(SVM)算法易陷入局部最优解、寻优速度慢等问题,提出一种基于深度迁移学习与狮群优化SVM(LSO-SVM)算法对遥感图像场景进行分类.首先,通过自适应对比度增强图像后利用颜色聚合向量提取图像颜色特征;其次,利用3种预训练网络分别提取图像的迁移学习深度特征;最后,将手工提取的图像特征与用3种预训练网络获取的特征使用系列特征融合方法进行融合,并将其输入LSO-SVM进行图像场景分类.结果表明,该算法解决了小样本情况下深度学习较难训练及传统优化SVM算法易陷入局部最优解、寻优速度慢的问题.在80%的训练条件下,数据集UCM Land-Use和RSSCN7的分类精度分别达到99.52%和98.57%.
Aiming at the problem of the small sample size of high-resolution remote sensing images and traditional optimized support vector machine(SVM)algorithms easily falling into local optima and slow optimization speed,we proposed an algorithm based on deep transfer learning and lion swarm optimization SVM(LSO-SVM)to classify remote sensing image scene.Firstly,after enhancing the image through adaptive contrast,color aggregati on vectors were used to extract image color features.Secondly,three kinds of pretrained networks were used to extract the transfer learning depth features of images.Finally,the manually extracted image features and the features obtained using three pretrained networks were fused by using a series of feature fusion met hods,and inputted them into LSO-SVM for image scene classification.The results show that the algorithm solves the problems of difficulty in deep learning training in small sample situations and the t endency of traditional optimized SVM algorithms to fall into local optima and slow search speed.Under 80%training conditions,the classification accuracy of UC M Land-Use and RSSCN7 datasets reaches 99.52%and 98.57%,respectively.
作者
王李祺
侯宇超
高翔
谭秀辉
程蓉
王鹏
白艳萍
WANG Liqi;HOU Yuchao;GAO Xiang;TAN Xiuhui;CHENG Rong;WANG Peng;BAI Yanping(School of Mathematics,North University of China,Taiyuan 030051,China;School of Information and Communication Engineering,North University of China,Taiyuan 030051,China)
出处
《吉林大学学报（理学版）》
CAS
北大核心
2023年第4期863-874,共12页
Journal of Jilin University:Science Edition
基金
国家自然科学基金(批准号:61774137)
山西省基础研究计划项目(批准号:202103021224195,202103021224212,202103021223189,20210302123019)
山西省回国留学人员科研项目(批准号:2020-104,2021-108).
关键词
遥感图像
图像分类
迁移学习
狮群优化算法
颜色聚合向量
remote sensing image
image classification
transfer learning
lion swarm optimization algorithm
color coherence vector
