Abstract This work uses machine learning approaches to present semantic segmentation for land cover classification in Gambella National Park (GNP). Land cover classification has become more accurate due to developments in remote sensing data. Land cover classification from satellite images has been studied, but the methodologies and satellite data employed so far are not suitable for research regions with the possibility to find heterogeneous land cover classes within small areas. Previous studies found issues with the satellite images coarser spatial resolution, the use of standard statistical methods as classifiers, and the difficulty in optimal patch size selection when patch-based classification is used. To address these issues, we suggested a deep learning-based semantic segmentation model that could be utilized as a pixel-level land cover classification technique. The suggested technique employed high-resolution Sentinel-2 satellite images of our study area (GNP) as a dataset and constructed and assessed pixel-level classification models. As a deep learning-based classification model, we have used the Link-Net architecture and its encoder part was modified further to incorporate the state-of-the-art architecture called ResNet34. The developed models, support vector machine with CNN features (CNN–SVM), random forest with CNN features (CNN-RF), LinkNet model with ResNet-34 as encoder (LinkNet-ResNet34), attain average F1-Score values of 81%,82%, and 87.4% respectively. 

/pdf/semantic-segmentation-model-for-land-cover-classification-2fslyjdn.pdf

Semantic segmentation model for land cover classification from satellite images in Gambella National Park, Ethiopia

This study aims at improving the efficiency of remote sensing scene classification (RSSC) through lightweight neural networks and to provide a possibility for large-scale, intelligent and real-time computation in performing RSSC for common devices. In this study, a lightweight RSSC model is proposed, which is named RSCNet. First, we use the lightweight ShuffleNet v2 network to extract the abstract features from the images, which can guarantee the efficiency of the model. Then, the weights of the backbone are initialized using transfer learning, allowing the model to learn by drawing on the knowledge of ImageNet. Second, to further improve the classification accuracy of the model, we propose to combine ShuffleNet v2 with an efficient channel attention mechanism that allows the features of the input classifier to be weighted. Third, we use a regularization technique during the training process, which utilizes label smoothing regularization to replace the original loss function. The experimental results show that the classification accuracy of RSCNet is 96.75% and 99.05% on the AID and UCMerced_LandUse datasets, respectively. The floating-point operations (FLOPs) of the proposed model are only 153.71 M, and the time spent for a single inference on the CPU is about 2.75 ms. Compared with existing RSSC methods, RSCNet achieves relatively high accuracy at a very small computational cost.

/pdf/rscnet-an-efficient-remote-sensing-scene-classification-3masacml.pdf

RSCNet: An Efficient Remote Sensing Scene Classification Model Based on Lightweight Convolution Neural Networks

Traditional building and water segmentation methods are vulnerable to noise interference, and hence, they could not avoid missed and false detections in the detection process. Excessive deep learning downsampling would lead to significant loss of feature map information, image location information offset, and the overall effect of falling apart. To address these issues, a multiscale location attention network (MSLANet) is proposed. Location-spatial information and channel information are particularly important for edge detail segmentation in building and water cover. The network includes a location channel attention (LCA) unit to focus on tributary details of rivers and segmentation of building edge eaves. Moreover, this article builds a dual-branch multiscale aggregation (DBMSA) unit to obtain deeper multiscale semantic information. Finally, the multiscale fusion unit (MSF) is used to guide the information merging of multiple stages, and the boundary information is improved by splicing the acquired deep multiscale information with the information of the relevant feature extraction layer in the downsampling. The experimental results on several datasets show that the proposed approach outperforms other methodologies in segmentation accuracy.

Multiscale Location Attention Network for Building and Water Segmentation of Remote Sensing Image

Semantic segmentation is an extremely challenging task in high-resolution remote sensing (HRRS) images as objects have complex spatial layouts and enormous variations in appearance. Convolutional neural networks (CNNs) have an excellent ability to extract local features and have been widely applied as a feature extractor for various vision tasks. However, due to the inherent inductive bias of convolution operation, CNNs inevitably have limitations in modeling long-range dependencies. Transformer can capture global representations well but unfortunately ignores the details of local features and has high computational and spatial complexity in processing high-resolution feature maps. In this article, we propose a novel hybrid architecture for HRRS image segmentation, termed Enhancing Multiscale Representations with Transformer (EMRT), to exploit the advantages of convolution operations and Transformer to enhance multiscale representation learning. We incorporate the deformable self-attention mechanism in the Transformer to automatically adjust the receptive field and design an encoder–decoder architecture accordingly to achieve efficient context modeling. Specifically, CNN is constructed to extract feature representations. In the encoder, local features and global representations at different resolutions are extracted by the CNN and Transformer, respectively, and fused in an interactive manner. Moreover, a separate spatial branch is designed to extract multiscale contextual information as queries, and global dependencies between features at different scales are efficiently established by the decoder. Extensive experiments on three public remote sensing datasets demonstrate the superiority of EMRT and indicate that the overall performance of our method outperforms state-of-the-art methods. Code is available at https://github.com/peach-xiao/EMRT.

Enhancing Multiscale Representations With Transformer for Remote Sensing Image Semantic Segmentation

ABSTRACT In the field of land-cover classification with remote sensing images, methods that analyse purely the spectral information of individual pixels generally produce noisy results, due to salt-and-pepper effects. The use of methods that also incorporate spatial contextual information into classification, often defined as contextual or spectral-spatial approaches, is an effective strategy for reducing the occurrence of punctual noises and, consequently, improving accuracy. However, contextual methods still present a critical limitation: an over smoothing performance on certain classes can cause the loss of details on important spatial structures. They may overlook salient punctual and linear objects that can be efficiently classified using purely spectral information, particularly over areas of rapid class transition. This issue is commonly observed with the classification of medium spatial-resolution images that include narrow class structures, such as rivers and roads. To solve this problem, we present a strategy for contextual classification that allows adjusting a trade-off between noise smoothing and the preservation of small spatial details. The proposed strategy comprises a meta-methodology, in the sense that it does not depend on specific pixel-based and contextual classifiers. The meta-methodology for improving contextual classification methods (Meta-CTX) consists in performing a separability analysis, at the pixel level, based on the class membership estimates provided by a pixel-based classifier. The Meta-CTX analyses the distance between class membership estimates in order to identify pixels that are expected to be accurately classified using purely spectral information. The Meta-CTX preserves the per-pixel classification of these pixels. It uses spatial contextual information only to classify pixels that are more susceptible to classification errors. The experimental results indicate that the Meta-CTX can efficiently combine noise smoothing with the preservation of small spatial details in remote sensing image classification.

A meta-methodology for preserving narrow objects when using spatial contextual classifiers for remote sensing data

Enhanced contextual representation with deep neural networks for land cover classification based on remote sensing images

Automatic land cover classification from high-resolution remote sensing (RS) images remains challenging due to the complex composition of classes. Given the potential of a graph to simulate latent class composition, the latest development of graph convolutional network (GCN) has received increasing attention. However, most existing methods only use a single perspective graph structure, which largely limits their ability to capture the complementary features that would better represent the underlying data structure of images. Therefore, this paper proposes a novel multi-view GCN-based representation learning network(MvRLNet) for RS image classification. First, a superpixel-based spectral component decomposes module(SSCDM) is designed to enhance the uniqueness and homogeneity of graph nodes because the mixed superpixels may lead to miscellaneous information on graph aggregations. Second, a multi-view graph learning module(MGLM) is proposed to integrate topology and spectral graph information into a unified network with an effective feature learning strategy. Finally, the effectiveness of the proposed MvRLNet is validated on a variety datasets with different resolutions. The experimental results show that the proposed MvRLNet performs better than state-of-the-art techniques. 

Peng Chang

Papers

Enhanced contextual representation with deep neural networks for land cover classification based on remote sensing images

Multi-view Graph Convolutional Network with Spectral Component Decompose for Remote Sensing Images Classification