Abstract:A deep neural network is suitable for remote sensing image pixel-wise classification because it effectively extracts features from the raw data. However, remote sensing images with higher spatial resolution exhibit smaller inter-class differences and greater intra-class differences; thus, feature extraction becomes more difficult. The attention mechanism, as a method that simulates the manner in which humans comprehend and perceive images, is useful for the quick and accurate acquisition of key features. In this study, we propose a novel neural network that incorporates two kinds of attention mechanisms in its mask and trunk branches; i.e., control gate (soft) and feedback attention mechanisms, respectively, based on the branches’ primary roles. Thus, a deep neural network can be equipped with an attention mechanism to perform pixel-wise classification for very high-resolution remote sensing (VHRRS) images. The control gate attention mechanism in the mask branch is utilized to build pixel-wise masks for feature maps, to assign different priorities to different locations on different channels for feature extraction recalibration, to apply stress to the effective features, and to weaken the influence of other profitless features. The feedback attention mechanism in the trunk branch allows for the retrieval of high-level semantic features. Hence, additional aids are provided for lower layers to re-weight the focus and to re-update higher-level feature extraction in a target-oriented manner. These two attention mechanisms are fused to form a neural network module. By stacking various modules with different-scale mask branches, the network utilizes different attention-aware features under different local spatial structures. The proposed method is tested on the VHRRS images from the BJ-02, GF-02, Geoeye, and Quickbird satellites, and the influence of the network structure and the rationality of the network design are discussed. Compared with other state-of-the-art methods, our proposed method achieves competitive accuracy, thereby proving its effectiveness.

Pixel-Wise Classification Method For High Resolution Remote Sensing Imagery Using Deep Neural Networks

Very High Resolution Remote Sensing Imagery Classification Using a Fusion of Random Forest and Deep Learning Technique—Subtropical Area for Example

Superpixel-Based Long-Range Dependent Network for High-Resolution Remote-Sensing Image Classification

Aerial Scene Classification Via Multilevel Fusion Based on Deep Convolutional Neural Networks.

Emmcnn: An Etps-Based Multi-Scale And Multi-Feature Method Using Cnn For High Spatial Resolution Image Land-Cover Classification

Optimizing Spatial Relationships in GCN to Improve the Classification Accuracy of Remote Sensing Images

Attention-Mechanism-Containing Neural Networks for High-Resolution Remote Sensing Image Classification

Deep Feature Fusion for High-Resolution Aerial Scene Classification

Remote Sensing Image Semantic Segmentation Method Based on a Deep Convolutional Neural Network and Multiscale Feature Fusion

Fuzzy neighbourhood neural network for high-resolution remote sensing image segmentation

Mapping Impervious Surfaces in Town–Rural Transition Belts Using China’s GF-2 Imagery and Object-Based Deep CNNs

Densely Based Multi-Scale and Multi-Modal Fully Convolutional Networks for High-Resolution Remote-Sensing Image Semantic Segmentation

Object-Based Convolutional Neural Network for High-Resolution Imagery Classification.

A Deep Neural Network Combined With Context Features for Remote Sensing Scene Classification

A Fast Deep Perception Network for Remote Sensing Scene Classification

Scene Classification Of High Resolution Remote Sensing Images Using Convolutional Neural Networks

High-Resolution Remote Sensing Image Classification Method Based on Convolutional Neural Network and Restricted Conditional Random Field

An Attention-Fused Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Imagery

Remote sensing image segmentation based on the fuzzy deep convolutional neural network

Classification of High Resolution Urban Remote Sensing Images Using Deep Networks by Integration of Social Media Photos

DenseNet-Based Depth-Width Double Reinforced Deep Learning Neural Network for High-Resolution Remote Sensing Image Per-Pixel Classification