An Encoder–Decoder Convolution Network With Fine-Grained Spatial Information for Hyperspectral Images Classification

Abstract

Convolutional Neural Network (CNN) is widely used in Hyperspectral Images (HSIs) classification. However, the fine-grained spatial (FGS) details are discarded during a sequence of convolution and pooling operations for most of CNN-based HSIs classification methods. To address this issue, a unified encoder-decoder framework is proposed to integrate high-level semantics and FGS details for HSIs classification, denoted by FGSCNN. The encoder, including a series of convolution and pooling layers, captures the high-level semantic information with low resolution feature maps. The decoder fuses the high-level low-resolution semantic and the fine-grained high-resolution spatial information, namely, to get the FGS features with high-level semantics. The deconvolution layers and skip connection are used in the decoder to retain the FGS details, while, convolution layers are also used to combine the FGS features with high-level semantics. Based on the encoder-decoder framework, a unified loss function is exploited to integrate the highlevel semantic information and FGS details with an end-to-end manner for HSIs classification. Experiments conducted on the three public datasets, i.e. the Indian Pines, Pavia University and Salinas, demonstrate the effectiveness of the proposed method on HSIs classification. INDEX TERMS Convolutional neural networks (CNNs), encoder-decoder, hyperspectral image (HSI) classification, information fusion.

A Pixel Cluster CNN and Spectral-Spatial Fusion Algorithm for Hyperspectral Image Classification With Small-Size Training Samples

“…In [27], a novel supervised deep feature extraction algorithm combined siamese CNN with linear SVM was introduced. Chen et al [28] [29]- [32]. It also has been verified that the spatial correlation across HSIs can provide complementary information to spectral features and should be taken into account [33]- [39].…”

Tropical cyclone intensity estimation through convolutional neural network transfer learning using two geostationary satellite datasets

“…Transfer learning is a machine learning technique in which a learning model developed for a first learning task is reused as the starting point for another learning model to perform a second task (Taherkhani et al, 2020). Due to the difficulty of achieving high accuracy in computer vision and other domains when using finite training datasets, deep learning models often require vast datasets (Cao et al, 2016;Gorban et al, 2020;Li et al, 2020). Transfer learning offers a viable solution to this issue by transferring knowledge from the source domain to the target domain and enhancing the accuracy of deep learning models (Pan et al, 2011;Yang et al, 2017;Liu et al, 2018;Jiang et al, 2019).…”

Hyperspectral Image Classification Based on Superpixel Pooling Convolutional Neural Network with Transfer Learning

“…Some representative deep learning techniques include convolutional neural network (CNN) [9,10], recurrent neural network (RNN) [11,12], generative adversarial network [13,14] and convolutional auto-encoder [15,16]. Recently, a series of deep learning-based classification frameworks have been widely used in the field of remote sensing [17][18][19][20][21][22]. Combining the deep CNN with multiple feature learning, a joint feature map for HSI classification was generated, which makes the developed method have high classification performance on test datasets [18].…”

“…A feature learning model based on unsupervised segmented denoising auto-encoder was depicted to learn both spectral and spatial features [24]. Even though these deep learning-based HSI classification methods can achieve satisfactory results, a large number of training samples are usually required (e.g., labeling 200 pixels per class) [18][19][20][21]. Therefore, it is necessary to investigate the classification performance of these models in case of finite training samples.…”