Deep Convolutional Neural Networks for Hyperspectral Image Classification

Глубокие сверточные нейронные сети для классификации гиперспектральных изображений
Fan Zhang, Heng-Chao Li, Wei Hu, Yangyu Huang, Li Wei
2015-01-01

deep convolutional neural networksfive-layer CNN architecturehyperspectral image classificationspectral-domain CNNsupport vector machines comparison
Recently, convolutional neural networks have demonstrated excellent performance on various visual tasks, including the classification of common two-dimensional images. In this paper, deep convolutional neural networks are employed to classify hyperspectral images directly in spectral domain. More specifically, the architecture of the proposed classifier contains five layers with weights which are the input layer, the convolutional layer, the max pooling layer, the full connection layer, and the output layer. These five layers are implemented on each spectral signature to discriminate against others. Experimental results based on several hyperspectral image data sets demonstrate that the proposed method can achieve better classification performance than some traditional methods, such as support vector machines and the conventional deep learning-based methods.
1
A deep convolutional neural network (five-weight-layer architecture) is applied directly to hyperspectral spectral signatures for classification.
2
Experimental results on several hyperspectral datasets show the proposed CNN outperforms traditional methods such as support vector machines.
3
The CNN architecture consists of input, convolutional, max-pooling, fully connected, and output layers implemented per spectral signature.
4
The proposed CNN also achieves better classification performance than conventional deep learning–based methods on the tested hyperspectral datasets.

Hyperspectral images (spectral signatures) being classified

Classification performance of deep convolutional neural networks applied directly in the spectral domain (five-layer CNN architecture: input, convolution, max-pooling, fully-connected, output) to discriminate spectral signatures

Publication Details
Publication Date
2015-01-01
Journal
Publisher
ISSN
Cited by
1969
Access Type
Author Information
Authors
Fan Zhang
Heng-Chao Li
Wei Hu
Yangyu Huang
Li Wei
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%