Going Deeper with Contextual CNN for Hyperspectral Image Classification

Abstract

In this paper, we describe a novel deep convolutional neural network (CNN) that is deeper and wider than other existing deep networks for hyperspectral image classification. Unlike current state-of-the-art approaches in CNN-based hyperspectral image classification, the proposed network, called contextual deep CNN, can optimally explore local contextual interactions by jointly exploiting local spatio-spectral relationships of neighboring individual pixel vectors. The joint exploitation of the spatio-spectral information is achieved by a multi-scale convolutional filter bank used as an initial component of the proposed CNN pipeline. The initial spatial and spectral feature maps obtained from the multi-scale filter bank are then combined together to form a joint spatio-spectral feature map. The joint feature map representing rich spectral and spatial properties of the hyperspectral image is then fed through a fully convolutional network that eventually predicts the corresponding label of each pixel vector. The proposed approach is tested on three benchmark datasets: the Indian Pines dataset, the Salinas dataset and the University of Pavia dataset. Performance comparison shows enhanced classification performance of the proposed approach over the current state-of-the-art on the three datasets.

14 pages

References in corpus (1)

• Shapelet-based Sparse Representation for Landcover Classification of Hyperspectral Images

Cited by in corpus (28)

• Deep Learning for Hyperspectral Image Classification: An Overview
• Deep Learning for Classification of Hyperspectral Data: A Comparative Review
• A Survey: Deep Learning for Hyperspectral Image Classification with Few Labeled Samples
• Few-Shot Hyperspectral Image Classification With Unknown Classes Using Multitask Deep Learning
• A Spectral-Spatial-Dependent Global Learning Framework for Insufficient and Imbalanced Hyperspectral Image Classification
• Validating Hyperspectral Image Segmentation
• Efficient Deep Learning of Non-local Features for Hyperspectral Image Classification
• DCN-T: Dual Context Network with Transformer for Hyperspectral Image Classification
• MCUa: Multi-level Context and Uncertainty aware Dynamic Deep Ensemble for Breast Cancer Histology Image Classification
• Unsupervised Segmentation of Hyperspectral Images Using 3D Convolutional Autoencoders
• A lightweight deep learning based cloud detection method for Sentinel-2A imagery fusing multi-scale spectral and spatial features
• Dual-Branch Subpixel-Guided Network for Hyperspectral Image Classification
• Multitask Deep Learning with Spectral Knowledge for Hyperspectral Image Classification
• Anchor Cascade for Efficient Face Detection
• Spectral-spatial classification of hyperspectral images: three tricks and a new supervised learning setting
• Single-stream CNN with Learnable Architecture for Multi-source Remote Sensing Data
• Active Ensemble Deep Learning for Polarimetric Synthetic Aperture Radar Image Classification
• Exploring Cross-Domain Pretrained Model for Hyperspectral Image Classification
• Effective training of deep convolutional neural networks for hyperspectral image classification through artificial labeling
• A machine learning based approach to gravitational lens identification with the International LOFAR Telescope
• Denoising of Three-Dimensional Fast Spin Echo Magnetic Resonance Images of Knee Joints using Spatial-Variant Noise-Relevant Residual Learning of Convolution Neural Network
• Shared Manifold Learning Using a Triplet Network for Multiple Sensor Translation and Fusion with Missing Data
• A CNN with Noise Inclined Module and Denoise Framework for Hyperspectral Image Classification
• SLCRF: Subspace Learning with Conditional Random Field for Hyperspectral Image Classification
• Shallow Network Based on Depthwise Over-Parameterized Convolution for Hyperspectral Image Classification
• Advances in Deep Learning for Hyperspectral Image Analysis--Addressing Challenges Arising in Practical Imaging Scenarios
• Generating Hard Examples for Pixel-wise Classification
• ScaleNet: An Unsupervised Representation Learning Method for Limited Information