Remote Sensing Image Change Detection Research Articles

Spatial-Temporal Based Multihead Self-Attention for Remote Sensing Image Change Detection

The neural network-based remote sensing image change detection method faces a large amount of imaging interference and severe class imbalance problems under high-resolution conditions, which bring new challenges to the accuracy of the detection network. In this work, to address the imaging interference caused by different imaging angles and times, the siamese strategy and multi-head self-attention mechanism are used to reduce the imaging differences between the dual-temporal images and fully exploit the inter-temporal information. Secondly, a learnable multi-part feature learning module is used to adaptively exploit features from different scales to obtain more comprehensive features. Finally, a mixed loss function strategy is used to ensure that the network converges effectively and excludes the adverse interference of a large number of negative samples to the network. Extensive experiments show that our method outperforms numerous methods on LEVIR-CD, WHU, and DSIFN datasets.

Superpixel-based multiobjective change detection based on self-adaptive neighborhood-based binary differential evolution

With strong penetrability and high resolution, synthetic aperture radar (SAR) images have been widely used in remote sensing image change detection. With the essence of heuristics, metaheuristics are suitable for the needs of most real-life optimization problems according to the expected solution quality and allowable calculation time. To improve the accuracy of change detection, this paper proposes a novel efficient metaheuristic change detection procedure. More specifically, we develop a superpixel-based multiobjective change detection method based on superpixel-wise feature representation and self-adaptive neighborhood-based binary difference evolution. During the clustering for feature analysis, a multiobjective optimization problem (MOP) is modeled which take the likelihood and Bhattacharyya distance between changed and unchanged classes as two objective functions. To solve the MOP, this paper firstly applies the superpixel binary representation technique into the encoding process of multiobjective evolutionary algorithm so that the dimension of pixel coding space can be substantially reduced. Then a self-adaptive neighborhood-based binary differential evolution strategy is proposed to explore the optimal change detection strategy, where a novel mutation operator integrating neighborhood information is designed for the improvement of change detection performance. Experimental results on three real SAR datasets confirm that the proposed method can make a better performance on change detection and the metaheuristic algorithm has good convergence and stability.

Feature fusion network based on siamese structure for change detection

Remote sensing image change detection is to analyze the change information of two images from the same area at different times. It has wide applications in urban expansion, forest detection, and natural disaster. In this paper, Feature Fusion Network is proposed to solve the problems of slow change detection speed and low accuracy. The MobileNetV3 block is adopted to efficiently extract features and a self-attention module is applied to investigate the relationship between heterogeneous feature maps (image features and concatenated features). The method is tested in data sets SZTAKI and LEVIR-CD. With 98.43 percentage correct classification, it is better than other comparative networks, and its space complexity is reduced by about 50%. The experimental results show that it has better performance and can improve the accuracy or speed of change detection.

Change Detection for High-Resolution Remote Sensing Images Based on a Multi-Scale Attention Siamese Network

To address the problems in remote sensing image change detection such as missed detection of features at different scales and incomplete region detection, this paper proposes a high-resolution remote sensing image change detection model (Multi-scale Attention Siamese Network, MASNet) based on a Siamese network and multi-scale attention mechanism. The MASNet model took the Siamese structure of the ResNet-50 network to extract features of different simultaneous images and then applied the attention module to feature maps of different scales to generate multi-scale feature representations. Meanwhile, an improved contrastive loss function was adopted to enhance the learning of change features and improving the imbalance problem between unchanged and changed samples. Furthermore, to address the current time-consuming and laborious phenomenon of manually annotating datasets, we provided a change detection dataset from Yunnan Province in China (YNCD) that contains 1540 pairs of 256 × 256 bi-temporal images with a spatial resolution of 1 m. Then, model training and change detection applications were studied by expanding a small number of experimental area samples into the existing public datasets. The results showed that the overall accuracy of the MASNet model for change detection in the experimental area is 95.34%, precision rate is 79.78%, recall rate is 81.52%, and F1 score is 80.64%, which are better than those of six comparative models (FC-EF, FC-Siam-Diff, FC-Siam-Conc, PAN, MANet, and STANet). This verifies the effectiveness of the MASNet model as well as the feasibility of change detection by expanding existing public datasets.

EMS-CDNet: an efficient multi-scale-fusion change detection network for very high-resolution remote sensing images

ABSTRACT Remote sensing image change detection (RSICD) is an essential measure for monitoring the earth’s surface changes. In recent years, the explosive growth of very high-resolution (VHR) satellite sensors and the booming innovations in deep learning technology have significantly boosted RSICD development. However, most of the current RSICD models focus on locating accurate change areas while ignoring the efficiency of their method, which limits the practical application of RSICD models, especially for large-scale and emergency RSICD tasks. In this paper, we propose an Efficient Multi-scale-fusion Change Detection Network (EMS-CDNet) for bi-temporal RSICD tasks. Our EMS-CDNet pays more attention to the model’s inference speed and the accuracy-efficiency trade-off rather than only pursuing detection accuracy. We designed a multi-scale fusion module for EMS-CDNet, which adopts multi-scale and multi-branch operations to extract multi-scale features simultaneously and aggregate features at different feature levels. In addition to EMS-CDNet’s ability to achieve sufficient feature extraction, the multi-scale image input within the designed module alleviates the influence of image registration errors in practical applications, thereby strengthening EMS-CDNet’s value for practical RSICD tasks. We also integrated a novel partition unit in EMS-CDNet to lighten the model while maintaining the detection ability of small targets, thus shortening its processing time without a severe accuracy decrease. We conducted experiments on two state-of-the-art (SOTA) public RSICD datasets and our own collected dataset. The public datasets were utilized to comparatively measure the overall accuracy and efficiency measurement of EMS-CDNet, and the dataset of images we collected was used to observe EMS-CDNet’s performance under the influence of image registration errors. Our experimental results show that EMS-CDNet achieved a better accuracy-efficiency trade-off than the SOTA public datasets methods. For example, EMS-CDNet reduced the inference time by about 33% while maintaining identical detection accuracy to CLNet (the optimal method among the comparison methods). Furthermore, EMS-CDNet achieved higher accuracy on our collected dataset, with an F1 of 74% and mIoU of 0.806, demonstrating its robustness to image registration errors and showing its value for practical RSICD applications.

Multi-scale feature progressive fusion network for remote sensing image change detection

Presently, research on deep learning-based change detection (CD) methods has become a hot topic. In particular, feature pyramid networks (FPNs) are widely used in CD tasks to gradually fuse semantic features. However, existing FPN-based CD methods do not correctly detect the complete change region and cannot accurately locate the boundaries of the change region. To solve these problems, a new Multi-Scale Feature Progressive Fusion Network (MFPF-Net) is proposed, which consists of three innovative modules: Layer Feature Fusion Module (LFFM), Multi-Scale Feature Aggregation Module (MSFA), and Multi-Scale Feature Distribution Module (MSFD). Specifically, we first concatenate the features of each layer extracted from the bi-temporal images with their difference maps, and the resulting change maps fuse richer semantic information while effectively representing change regions. Then, the obtained change maps of each layer are directly aggregated, which improves the effective communication and full fusion of feature maps in CD while avoiding the interference of indirect information. Finally, the aggregated feature maps are layered again by pooling and convolution operations, and then a feature fusion strategy with a pyramid structure is used, with layers fused from low to high, to obtain richer contextual information, so that each layer of the layered feature maps has original semantic information and semantic features of other layers. We conducted comprehensive experiments on three publicly available benchmark datasets, CDD, LEVIR-CD, and WHU-CD to verify the effectiveness of the method, and the experimental results show that the method in this paper outperforms other comparative methods.

SemiSANet: A Semi-Supervised High-Resolution Remote Sensing Image Change Detection Model Using Siamese Networks with Graph Attention

Change detection (CD) is one of the important applications of remote sensing and plays an important role in disaster assessment, land use detection, and urban sprawl tracking. High-accuracy fully supervised methods are the main methods for CD tasks at present. However, these methods require a large amount of labeled data consisting of bi-temporal images and their change maps. Moreover, creating change maps takes a lot of labor and time. To address this limitation, a simple semi-supervised change detection method based on consistency regularization and strong augmentation is proposed in this paper. First, we construct a Siamese nested UNet with graph attention mechanism (SANet) and pre-train it with a small amount of labeled data. Then, we feed the unlabeled data into the pre-trained SANet and confidence threshold filter to obtain pseudo-labels with high confidence. At the same time, we produce distorted images by performing strong augmentation on unlabeled data. The model is trained to make the CD results of the distorted images consistent with the corresponding pseudo-label. Extensive experiments are conducted on two high-resolution remote sensing datasets. The results demonstrate that our method can effectively improve the performance of change detection under insufficient labels. Our methods can increase the IoU by more than 25% compared to the state-of-the-art methods.

A BUILDING CHANGE DETECTION METHOD BASED ON A SINGLE ALS POINT CLOUD AND A HRS IMAGE

Abstract. For common remote sensing image change detection based on different time phases, it is difficult to solve the problem of surface tilt caused by the different shooting angle and time, which makes it difficult to complete the accurate registration. Some scholars utilized three-dimensional data for change detection to avoid registration problem, however, it costs very high for using 3D data to detect changes. Aim at this problem, this paper proposes a method of combining a single phase of ALS (Airborne Laser Scanning) data and HRS (High Resolution Satellite) image. It is composed of the following four steps: (1) Extracting the shadow area of the new-phase optical image, and superimposing it with the classified old-time point cloud data, and then the disappeared building area and the approximate building area are obtained according to the associated relationship between the shadow area and buildings. (2) Determining the unchanged building areas as positive samples, and taking the remaining area and vegetation area after removing the approximate buildings from the image as negative samples, and then GrabCut algorithm is used to segment new-phase HRS images to obtain building areas. (3) Comparing between building areas obtained in the previous step and the registered old-time ALS data to obtain a new building area with noise. (4) Denoising the results of the previous step to obtain the final new building area. Two datasets are used to verify the method. The detection accuracy of the disappeared buildings is over 85%, and the detection accuracy of the newly added buildings is over 70%.

A Network Combining a Transformer and a Convolutional Neural Network for Remote Sensing Image Change Detection

With the development of deep learning techniques in the field of remote sensing change detection, many change detection algorithms based on convolutional neural networks (CNNs) and nonlocal self-attention (NLSA) mechanisms have been widely used and have obtained good detection accuracy. However, these methods mainly extract semantic features on images from different periods without taking into account the temporal dependence between these features. This will lead to more “pseudo-change” in complex scenes. In this paper, we propose a network architecture named UVACD for bitemporal image change detection. The network combines a CNNs extraction backbone for extracting high-level semantic information with a visual transformer. Here, visual transformer constructs change intensity tokens to complete the temporal information interaction and suppress irrelevant information weights to help extract more distinguishable change features. Our network is validated and tested on both the LEVIR-CD and WHU datasets. For the LEVIR-CD dataset, we achieve an intersection over union (IoU) of 0.8398 and an F1 score of 0.9130. For the WHU dataset, we achieve an IoU of 0.8664 and an F1 score of 0.9284. The experimental results show that the proposed method outperforms some previous state of the art change detection methods.

Hybrid-TransCD: A Hybrid Transformer Remote Sensing Image Change Detection Network via Token Aggregation

Existing optical remote sensing image change detection (CD) methods aim to learn an appropriate discriminate decision by analyzing the feature information of bitemporal images obtained at the same place. However, the complex scenes in high-resolution (HR) remote images cause unsatisfied results, especially for some irregular and occluded objects. Although recent self-attention-driven change detection models with CNN achieve promising effects, the computational and consumed parameters costs emerge as an impassable gap for HR images. In this paper, we utilize a transformer structure replacing self-attention to learn stronger feature representations per image. In addition, concurrent vision transformer models only consider tokenizing single-dimensional image tokens, thus failing to build multi-scale long-range interactions among features. Here, we propose a hybrid multi-scale transformer module for HR remote images change detection, which fully models representation attentions at hybrid scales of each image via a fine-grained self-attention mechanism. The key idea of the hybrid transformer structure is to establish heterogeneous semantic tokens containing multiple receptive fields, thus simultaneously preserving large object and fine-grained features. For building relationships between features without embedding with token sequences from the Siamese tokenizer, we also introduced a hybrid difference transformer decoder (HDTD) layer to further strengthen multi-scale global dependencies of high-level features. Compared to capturing single-stream tokens, our HDTD layer directly focuses representing differential features without increasing exponential computational cost. Finally, we propose a cascade feature decoder (CFD) for aggregating different-dimensional upsampling features by establishing difference skip-connections. To evaluate the effectiveness of the proposed method, experiments on two HR remote sensing CD datasets are conducted. Compared to state-of-the-art methods, our Hybrid-TransCD achieved superior performance on both datasets (i.e., LEVIR-CD, SYSU-CD) with improvements of 0.75% and 1.98%, respectively.

High resolution representation‐based Siamese network for remote sensing image change detection

Change detection of high-resolution remote sensing images can help to accurately understand the changes in the earth's surface. Advanced methods based on deep features have some limitations, including limited accuracy, poor detection effect, and poor robustness. The main reason is that these frameworks have poor feature extraction capabilities, insufficient context aggregation, and inadequate discrimination capabilities. In order to solve these problems, SiHDNet, a Siamese segmentation network based on deep, high-resolution differential feature interaction, is proposed. Specifically, after the high-resolution features of the dual-temporal image are extracted, the difference map is generated through a special fusion module, which contains sufficient and effective change information. Finally, the final binary change map is obtained through the improved spatial pyramid pooling module. Experiments are conducted on the newly released building change detection data set LEVIR-CD and the challenging remote sensing image change detection data set Google Data Set. Five benchmark methods are chosen. The results of quantitative analysis and qualitative comparison show that SiHDNet is superior to the five benchmark methods. The results of the ablation experiment also verify the effectiveness of this method.

Multiscale Geometric Analysis Fusion-Based Unsupervised Change Detection in Remote Sensing Images via FLICM Model.

Remote sensing image change detection is widely used in land use and natural disaster detection. In order to improve the accuracy of change detection, a robust change detection method based on nonsubsampled contourlet transform (NSCT) fusion and fuzzy local information C-means clustering (FLICM) model is introduced in this paper. Firstly, the log-ratio and mean-ratio operators are used to generate the difference image (DI), respectively; then, the NSCT fusion model is utilized to fuse the two difference images, and one new DI is obtained. The fused DI can not only reflect the real change trend but also suppress the background. The FLICM is performed on the new DI to obtain the final change detection map. Four groups of homogeneous remote sensing images are selected for simulation experiments, and the experimental results demonstrate that the proposed homogeneous change detection method has a superior performance than other state-of-the-art algorithms.

MAFF-Net: Multi-Attention Guided Feature Fusion Network for Change Detection in Remote Sensing Images.

One of the most important tasks in remote sensing image analysis is remote sensing image Change Detection (CD), and CD is the key to helping people obtain more accurate information about changes on the Earth’s surface. A Multi-Attention Guided Feature Fusion Network (MAFF-Net) for CD tasks has been designed. The network enhances feature extraction and feature fusion by building different blocks. First, a Feature Enhancement Module (FEM) is proposed. The FEM introduces Coordinate Attention (CA). The CA block embeds the position information into the channel attention to obtain the accurate position information and channel relationships of the remote sensing images. An updated feature map is obtained by using an element-wise summation of the input of the FEM and the output of the CA. The FEM enhances the feature representation in the network. Then, an attention-based Feature Fusion Module (FFM) is designed. It changes the previous idea of layer-by-layer fusion and chooses cross-layer aggregation. The FFM is to compensate for some semantic information missing as the number of layers increases. FFM plays an important role in the communication of feature maps at different scales. To further refine the feature representation, a Refinement Residual Block (RRB) is proposed. The RRB changes the number of channels of the aggregated features and uses convolutional blocks to further refine the feature representation. Compared with all compared methods, MAFF-Net improves the F1-Score scores by 4.9%, 3.2%, and 1.7% on three publicly available benchmark datasets, the CDD, LEVIR-CD, and WHU-CD datasets, respectively. The experimental results show that MAFF-Net achieves state-of-the-art (SOTA) CD performance on these three challenging datasets.

D-TNet: Category-Awareness Based Difference-Threshold Alternative Learning Network for Remote Sensing Image Change Detection

Deep learning-based change detection methods have achieved remarkable success through the feature learning capability of deep convolutions. However, the network structures of existing methods are simply modified from the semantic segmentation models, ignoring the essential characteristics of change detection, thereby limiting their applications. In this work, we propose a category-awareness based difference-threshold alternative learning network (D-TNet) for remote sensing image change detection. Our motivation is to characterize the different change magnitudes for different land cover changes, and represent the semantic content differences of various objects. Thus, our D-TNet consists of a difference map learning path and a threshold map learning path, realizing self-adapting thresholds selection by assigning each pixel a unique threshold. The two paths are alternatively optimized to make the difference map more discriminative, as well as making the threshold map more adaptive. In addition, a category-awareness attention mechanism is introduced in D-TNet, which learns a pixel-to-category relationship to benefit in representing the heterogeneity of land covers. Finally, experimental results on three change detection datasets verify the effectiveness of our D-TNet in both visual and quantitative analysis. Code will be available at: https://www.researchgate.net/profile/Ling-Wan-4.

ISNet: Towards Improving Separability for Remote Sensing Image Change Detection

Deep learning has substantially pushed forward remote sensing image change detection through extracting discriminative hierarchical features. However, as the increasingly high resolution remote sensing images have abundant spatial details but limited spectral information, the use of conventional backbone networks would give rise to blurry boundaries between different semantics among hierarchical features. This explains why most false alarms in the final predictions distribute around change boundaries. To alleviate the problem, we pay attention to feature refinement and propose deep learning networks that deliver improved separability (ISNet). Our ISNet reaps the advantages from two strategies applied to refining bi-temporal feature hierarchies: (i) margin maximization that clarifies the gap between changed and unchanged semantics, and (ii) targeted arrangement of attention mechanisms that directs the use of channel attention and spatial attention for highlighting semantic and positional information, respectively. Specifically, we insert channel attention modules into share-weighted backbone networks to facilitate semantic-specific feature extraction. The semantic boundaries in the extracted bi-temporal hierarchical features are then clarified by margin maximization modules, followed by spatial attention modules to enhance positional change responses. A top-down fusion pathway makes the final refined features cover multi-scale representations and have strong separability for remote sensing image change detection. Extensive experimental evaluations demonstrate that our ISNet achieves state-of-the-art performance on the LEVIR-CD, SYSU-CD, and Season-Varying datasets, in terms of Overall Accuracy (OA), Intersection-of-Union (IoU), and F1 score. Code is available at https://github.com/xingronaldo/ISNet.

Semisupervised Adaptive Ladder Network for Remote Sensing Image Change Detection

Nowadays, due to the difficult acquisition of true labels, a semisupervised neural network has shown great potential for change detection (CD) in remote sensing images. However, most of the traditional semisupervised neural network detection frameworks are complex to train and require additional structural analysis, along with a fixed structure, lacking universality. In this article, a semisupervised adaptive ladder network (SSALN) for remote sensing image CD is proposed, which enables dual-input label-incremental architecture searching with a concise and variable structure. First, SSALN is suitable for CD from two remote sensing images of any type with the characteristic of minimal label dependency and automatic network structure adjustment. The network can generate more reliable pseudolabels through continuous iterations to help limited real labels exploit implicit information, identify the most effective network, and form the ascending network structure optimization. Second, the acquisition of pseudolabels is the fusion of semisupervised and unsupervised CD approaches, which ensures the multiperspective information supplement. Multiple CD maps are fused to generate labels for the next iteration, making the predicting more reliable. Finally, both homogenous images and heterogenous images are tested with experiments. Even if the detection object is switched, it can be well adaptive and compatible without manual modification of the network. Experimental results demonstrate that the proposed method can promote the flow of label information through structure searching and self-circulation in the ascending network optimization; thus, it has outstanding performance on tasks of remote sensing image CD.

A Combined Loss-Based Multiscale Fully Convolutional Network for High-Resolution Remote Sensing Image Change Detection

In the task of change detection (CD), high-resolution remote sensing images (HRSIs) can provide rich ground object information. However, the interference from noise and complex background information can also bring some challenges to CD. In recent years, deep learning methods represented by convolutional neural networks (CNNs) have achieved good CD results. However, the existing methods have difficulty in detecting the detailed change information of the ground objects effectively. The imbalance of positive and negative samples can also seriously affect the CD results. In this letter, to solve the above problems, we propose a method based on a multiscale fully convolutional neural network (MFCN), which uses multiscale convolution kernels to extract the detailed features of the ground object features. A loss function combining weighted binary cross-entropy (WBCE) loss and dice coefficient loss is also proposed, so that the model can be trained from unbalanced samples. The proposed method was compared with six state-of-the-art CD methods on the DigitalGlobe dataset. The experiments showed that the proposed method can achieve a higher <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">F1</i> -score, and the detection effect of the detailed changes was better than that of the other methods.

A Siamese Network Based U-Net for Change Detection in High Resolution Remote Sensing Images

Remote sensing image change detection (RSICD) is a technique that explores the change of surface coverage in a certain time series by studying the difference between multiple remote sensing images (RSIs) collected over the same area. Traditional RSICD algorithms exhibit poor performance on complex change detection (CD) tasks. In recent years, deep learning (DL) techniques have achieved outstanding results in the fields of RSI segmentation and target recognition. In CD research, most of the methods treat multitemporal remote sensing data as one input and directly apply DL-based image segmentation theory on it while ignoring the spatio-temporal information in these images. In this article, a new siamese neural network is designed by combing an attention mechanism (Siamese_AUNet) with UNet to solve the problems of RSICD algorithms. SiameseNet encodes the feature extraction of RSIs by two branches in the siamese network, respectively. The weights are shared between these two branches in siamese networks. Subsequently, an attention mechanism is added to the model in order to improve its detection ability for changed objects. The models are then compared with conventional neural networks using three benchmark datasets. The results show that the Siamese_AUNet newly proposed in this article exhibits better performance than other standard methods when solving problems related to weak CD and noise suppression.

High-Resolution Remote Sensing Image Change Detection Based on Improved DeepLabv3+

High-Resolution Remote Sensing Image Change Detection Based on Improved DeepLabv3+

SwinSUNet: Pure Transformer Network for Remote Sensing Image Change Detection

Convolutional neural network (CNN) can extract effective semantic features, so it was widely used for remote sensing image change detection (CD) in the latest years. CNN has acquired great achievements in the field of CD, but due to the intrinsic locality of convolution operation, it could not capture global information in space-time. The transformer was proposed in recent years and it can effectively extract global information, so it was used to solve computer vision (CV) tasks and achieved amazing success. In this article, we design a pure transformer network with Siamese U-shaped structure to solve CD problems and name it SwinSUNet. SwinSUNet contains encoder, fusion, and decoder, and all of them use Swin transformer blocks as basic units. Encoder has a Siamese structure based on hierarchical Swin transformer, so encoder can process bitemporal images in parallel and extract their multiscale features. Fusion is mainly responsible for the merge operation of the bitemporal features generated by the encoder. Like encoder, the decoder is also based on hierarchical Swin transformer. Different from the encoder, the decoder uses upsampling and merging (UM) block and Swin transformer blocks to recover the details of the change information. The encoder uses patch merging and Swin transformer blocks to generate effective semantic features. After the sequential process of these three modules, SwinSUNet will output the change maps. We did expensive experiments on four CD datasets, and in these experiments, SwinSUNet achieved better results than other related methods.