摘要
文中利用深度学习思想来实现影像光谱维和空间维的特征提取,同时尝试加入稀疏约束的限制条件,并结合Softmax分类器,实现高光谱遥感影像的分类。实验结果表明,加入空间特征的基于堆栈稀疏自编码的分类方法能够得到很好的分类效果。
In this paper,deep learning is used to extract the spectral and spatial features. As the same time,we combine the constraint condition of sparse constraint and the Softmax classifier to classify the hyperspectral remote sensing image. The experimental results show that the classification method based on stack sparse self – coding with spatial features can obtain a good classification.
出处
《矿山测量》
2017年第6期53-58,共6页
Mine Surveying
关键词
影像分类
稀疏表达
深度学习
自动编码机
Image classification
Sparse representation
Deep learning
Auto-encoder