摘要
大量的训练样本可有效缓解模型过拟合,从而提高分类效果。在初始标记样本较少的情况下,开展借助不同尺度的同质区快速扩增大量高精度训练样本的实验,并利用初始标记样本和扩增样本训练支持向量机(Support Vector Machine,SVM)分类器,实现对高光谱数据的有效分类。该方法在Pavia Uni-versity、Salinas和Indian Pines三种高光谱数据上均能获得大量高精度的训练样本,分类精度分别达到99%、99%和97%以上。实验结果表明,扩增的大量伪标签样本可以有效训练SVM分类器,提高分类效果。
A large number of training samples can effectively alleviate the overfitting of the model and improve the classification effect.A lot of high-precision training samples are rapidly amplified by using homogenous regions of different scales.The support vector machine classifier is trained with the initial la-beled samples and amplified samples to achieve the effective classification of hyperspectral data.The ma-jority of high-precision training samples based on Pavia University data,Salinas data and Indian Pines da-ta can be obtained by this method,and the accuracy is above 99%,99%and 97%respectively.The experi-ment results show that the large number of pseudo-label samples amplified by the proposed method can ef-fectively train the SVM classifier and improve the classification effect.
作者
刘丽丽
杨春蕾
顾明剑
胡勇
LIU Li-li;YANG Chun-lei;GU Ming-jian;HU Yong(Suzhou Academy,Shanghai Institute of Technical Physics,Chinese Academy of Sciences,Suzhou 215000,China;Shanghai Institute of Technical Physics,Chinese Academy of Sciences,Shanghai 200083,China;Key Laboratory of Infrared Detection and Imaging Technology,Chinese Academy of Sciences,Shanghai 200083,China)
出处
《红外》
CAS
2023年第5期32-45,共14页
Infrared
关键词
高光谱影像
半监督分类
多尺度同质区
训练样本扩增
图像分割
支持向量机
hyperspectral image
semi-supervised classification
multi-scale homogeneous regions
train-ing sample amplification
image segmentation
SVM
