Object-Based Change Detection of Very High Resolution Images by Fusing Pixel-Based Change Detection Results Using Weighted Dempster–Shafer Theory

Object-based characterization of vegetation heterogeneity with sentinel images proves efficient in a highly human-influenced National Park of Côte d'Ivoire

Forest monitoring requires more automated systems to analyze high ecosystem heterogeneity. The traditional pixel-based detection method has proven to be less and less effective. A novel change detection method is therefore proposed to detect changes in forest cover using satellite images at very high spatial resolution. This is object-oriented classification, which groups pixels into interpreted objects, based on their spectral values, spatial, and textural properties. Using sentinel and Lansat images, we tested for the first time in the West African rainforest zone the effectiveness of this method for better detection, delineation, and analysis of land use and occupation types. The mean shift algorithm was used in both the segmentation and classification processes. Next, we compared the proposed object-oriented method with a pixel-based image classification detection method by implementing both methods under the same conditions. High detection accuracy (> 90%) and an overall Kappa greater than 0.90 were obtained by the object-oriented method, which is about 20% higher than the pixel-based method. The object-based method was free of salt and pepper effects and was less prone to image misregistration in terms of change detection accuracy and mapping results. This study demonstrates that the object-based classifier is a much better approach than the classical pixel-based classifier. In addition, it shows the problems of detecting heterogeneous landscapes and explains the observed confusions between the types of vegetation formations specific to tropical wetlands. The results obtained are encouraging and the contribution of high-resolution images and the object-based method to better discrimination of tropical wetland vegetation is discussed.

Multi-Scale LBP Texture Feature Learning Network for Remote Sensing Interpretation of Land Desertification

Land desertification is a major challenge to global sustainable development. Therefore, the timely and accurate monitoring of the land desertification status can provide scientific decision support for desertification control. The existing automatic interpretation methods are affected by factors such as “same spectrum different matter”, “different spectrum same object”, staggered distribution of desertification areas, and wide ranges of ground objects. We propose an automatic interpretation method for the remote sensing of land desertification that incorporates multi-scale local binary pattern (MSLBP) and spectral features based on the above issues. First, a multi-scale convolutional LBP feature extraction network is designed to obtain the spatial texture features of remote sensing images and fuse them with spectral features to enhance the feature representation capability of the model. Then, considering the continuity of the distribution of the same kind of ground objects in local space, we designed an adaptive median filtering method to process the probability map of the extreme learning machine (ELM) classifier output to improve the classification accuracy. Four typical datasets were developed using GF-1 multispectral imagery with the Horqin Left Wing Rear Banner as the study area. Experimental results on four datasets show that the proposed method solves the problem of ill classification and omission in classifying the remote sensing images of desertification, effectively suppresses the effects of “homospectrum” and “heterospectrum”, and significantly improves the accuracy of the remote sensing interpretation of land desertification.

Object-Oriented Change Detection Method Based on Spectral–Spatial–Saliency Change Information and Fuzzy Integral Decision Fusion for HR Remote Sensing Images

Spectral features in remote sensing images are extensively utilized to detect land cover changes. However, detection noise appearing in the changing maps due to the abundant spatial details in the high-resolution images makes it difficult to acquire an accurate interpretation result. In this paper, an object-oriented change detection approach is proposed which integrates spectral–spatial–saliency change information and fuzzy integral decision fusion for high-resolution remote sensing images with the purpose of eliminating the impact of detection noise. First, to reduce the influence of feature uncertainty, spectral feature change is generated by three independent methods, and spatial change information is obtained by spatial feature set construction and the optimal feature selection strategy. Secondly, the saliency change map of bi-temporal images is obtained with the co-saliency detection method to complement the insufficiency of image features. Then, the image objects are acquired by multi-scale segmentation based on the staking images. Finally, different pixel-level image change information and the segmentation result are fused using the fuzzy integral decision theory to determine the object change probability. Three high-resolution remote sensing image datasets and three comparative experiments were carried out to evaluate the performance of the proposed algorithm. Spectral–spatial–saliency change information was found to play a major role in the change detection of high-resolution remote sensing images, and the fuzzy integral decision strategy was found to effectively obtain reliable changed objects to improve the accuracy and robustness of change detection.

An Efficient Lightweight Neural Network for Remote Sensing Image Change Detection

Remote sensing (RS) image change detection (CD) is a critical technique of detecting land surface changes in earth observation. Deep learning (DL)-based approaches have gained popularity and have made remarkable progress in change detection. The recent advances in DL-based methods mainly focus on enhancing the feature representation ability for performance improvement. However, deeper networks incorporated with attention-based or multiscale context-based modules involve a large number of network parameters and require more inference time. In this paper, we first proposed an effective network called 3M-CDNet that requires about 3.12 M parameters for accuracy improvement. Furthermore, a lightweight variant called 1M-CDNet, which only requires about 1.26 M parameters, was proposed for computation efficiency with the limitation of computing power. 3M-CDNet and 1M-CDNet have the same backbone network architecture but different classifiers. Specifically, the application of deformable convolutions (DConv) in the lightweight backbone made the model gain a good geometric transformation modeling capacity for change detection. The two-level feature fusion strategy was applied to improve the feature representation. In addition, the classifier that has a plain design to facilitate the inference speed applied dropout regularization to improve generalization ability. Online data augmentation (DA) was also applied to alleviate overfitting during model training. Extensive experiments have been conducted on several public datasets for performance evaluation. Ablation studies have proved the effectiveness of the core components. Experiment results demonstrate that the proposed networks achieved performance improvements compared with the state-of-the-art methods. Specifically, 3M-CDNet achieved the best F1-score on two datasets, i.e., LEVIR-CD (0.9161) and Season-Varying (0.9749). Compared with existing methods, 1M-CDNet achieved a higher F1-score, i.e., LEVIR-CD (0.9118) and Season-Varying (0.9680). In addition, the runtime of 1M-CDNet is superior to most, which exhibits a better trade-off between accuracy and efficiency.

Mapping and Monitoring of Land Cover/Land Use (LCLU) Changes in the Crozon Peninsula (Brittany, France) from 2007 to 2018 by Machine Learning Algorithms (Support Vector Machine, Random Forest, and Convolutional Neural Network) and by Post-classification Comparison (PCC)

Land cover/land use (LCLU) is currently a very important topic, especially for coastal areas that connect the land and the coast and tend to change frequently. LCLU plays a crucial role in land and territory planning and management tasks. This study aims to complement information on the types and rates of LCLU multiannual changes with the distributions, rates, and consequences of these changes in the Crozon Peninsula, a highly fragmented coastal area. To evaluate the multiannual change detection (CD) capabilities using high-resolution (HR) satellite imagery, we implemented three remote sensing algorithms: a support vector machine (SVM), a random forest (RF) combined with geographic object-based image analysis techniques (GEOBIA), and a convolutional neural network (CNN), with SPOT 5 and Sentinel 2 data from 2007 and 2018. Accurate and timely CD is the most important aspect of this process. Although all algorithms were indicated as efficient in our study, with accuracy indices between 70% and 90%, the CNN had significantly higher accuracy than the SVM and RF, up to 90%. The inclusion of the CNN significantly improved the classification performance (5–10% increase in the overall accuracy) compared with the SVM and RF classifiers applied in our study. The CNN eliminated some of the confusion that characterizes a coastal area. Through the study of CD results by post-classification comparison (PCC), multiple changes in LCLU could be observed between 2007 and 2018: both the cultivated and non-vegetated areas increased, accompanied by high deforestation, which could be explained by the high rate of urbanization in the peninsula.

Remote Sensing Change Detection Based on Multidirectional Adaptive Feature Fusion and Perceptual Similarity

Remote sensing change detection (RSCD) is an important yet challenging task in Earth observation. The booming development of convolutional neural networks (CNNs) in computer vision raises new possibilities for RSCD, and many recent RSCD methods have introduced CNNs to achieve promising improvements in performance. In this paper we propose a novel multidirectional fusion and perception network for change detection in bi-temporal very-high-resolution remote sensing images. First, we propose an elaborate feature fusion module consisting of a multidirectional fusion pathway (MFP) and an adaptive weighted fusion (AWF) strategy for RSCD to boost the way that information propagates in the network. The MFP enhances the flexibility and diversity of information paths by creating extra top-down and shortcut-connection paths. The AWF strategy conducts weight recalibration for every fusion node to highlight salient feature maps and overcome semantic gaps between different features. Second, a novel perceptual similarity module is designed to introduce perceptual loss into the RSCD task, which adds perceptual information, such as structure and semantic information, for high-quality change map generation. Extensive experiments on four challenging benchmark datasets demonstrate the superiority of the proposed network compared with eight state-of-the-art methods in terms of F1, Kappa, and visual qualities.

A Framework for Unsupervised Wildfire Damage Assessment Using VHR Satellite Images with PlanetScope Data

The application of remote sensing techniques for disaster management often requires rapid damage assessment to support decision-making for post-treatment activities. As the on-demand acquisition of pre-event very high-resolution (VHR) images is typically limited, PlanetScope (PS) offers daily images of global coverage, thereby providing favorable opportunities to obtain high-resolution pre-event images. In this study, we propose an unsupervised change detection framework that uses post-fire VHR images with pre-fire PS data to facilitate the assessment of wildfire damage. To minimize the time and cost of human intervention, the entire process was executed in an unsupervised manner from image selection to change detection. First, to select clear pre-fire PS images, a blur kernel was adopted for the blind and automatic evaluation of local image quality. Subsequently, pseudo-training data were automatically generated from contextual features regardless of the statistical distribution of the data, whereas spectral and textural features were employed in the change detection procedure to fully exploit the properties of different features. The proposed method was validated in a case study of the 2019 Gangwon wildfire in South Korea, using post-fire GeoEye-1 (GE-1) and pre-fire PS images. The experimental results verified the effectiveness of the proposed change detection method, achieving an overall accuracy of over 99% with low false alarm rate (FAR), which is comparable to the accuracy level of the supervised approach. The proposed unsupervised framework accomplished efficient wildfire damage assessment without any prior information by utilizing the multiple features from multi-sensor bi-temporal images.

Object-Based Building Change Detection by Fusing Pixel-Level Change Detection Results Generated from Morphological Building Index

Change detection (CD) is an important tool in remote sensing. CD can be categorized into pixel-based change detection (PBCD) and object-based change detection (OBCD). PBCD is traditionally used because of its simple and straightforward algorithms. However, with increasing interest in very-high-resolution (VHR) imagery and determining changes in small and complex objects such as buildings or roads, traditional methods showed limitations, for example, the large number of false alarms or noise in the results. Thus, researchers have focused on extending PBCD to OBCD. In this study, we proposed a method for detecting the newly built-up areas by extending PBCD results into an OBCD result through the Dempster–Shafer (D–S) theory. To this end, the morphological building index (MBI) was used to extract built-up areas in multitemporal VHR imagery. Then, three PBCD algorithms, change vector analysis, principal component analysis, and iteratively reweighted multivariate alteration detection, were applied to the MBI images. For the final CD result, the three binary change images were fused with the segmented image using the D–S theory. The results obtained from the proposed method were compared with those of PBCD, OBCD, and OBCD results generated by fusing the three binary change images using the major voting technique. Based on the accuracy assessment, the proposed method produced the highest F1-score and kappa values compared with other CD results. The proposed method can be used for detecting new buildings in built-up areas as well as changes related to demolished buildings with a low rate of false alarms and missed detections compared with other existing CD methods.

Uncertainty Analysis for Object-Based Change Detection in Very High-Resolution Satellite Images Using Deep Learning Network

Object-based image analysis (OBIA) is better than pixel-based image analysis for change detection (CD) in very high-resolution (VHR) remote sensing images. Although the effectiveness of deep learning approaches has recently been proved, few studies have investigated OBIA and deep learning for CD. Previously proposed methods use the object information obtained from the preprocessing and postprocessing phase of deep learning. In general, they use the dominant or most frequently used label information with respect to all the pixels inside an object without considering any quantitative criteria to integrate the deep learning network and object information. In this study, we developed an object-based CD method for VHR satellite images using a deep learning network to denote the uncertainty associated with an object and effectively detect the changes in an area without the ground truth data. The proposed method defines the uncertainty associated with an object and mainly includes two phases. Initially, CD objects were generated by unsupervised CD methods, and the objects were used to train the CD network comprising three-dimensional convolutional layers and convolutional long short-term memory layers. The CD objects were updated according to the uncertainty level after the learning process was completed. Further, the updated CD objects were considered as the training data for the CD network. This process was repeated until the entire area was classified into two classes, i.e., change and no-change, with respect to the object units or defined epoch. The experiments conducted using two different VHR satellite images confirmed that the proposed method achieved the best performance when compared with the performances obtained using the traditional CD approaches. The method was less affected by salt and pepper noise and could effectively extract the region of change in object units without ground truth data. Furthermore, the proposed method can offer advantages associated with unsupervised CD methods and a CD network subjected to postprocessing by effectively utilizing the deep learning technique and object information.

A Coarse-to-Fine Deep Learning Based Land Use Change Detection Method for High-Resolution Remote Sensing Images

In recent decades, high-resolution (HR) remote sensing images have shown considerable potential for providing detailed information for change detection. The traditional change detection methods based on HR remote sensing images mostly only detect a single land type or only the change range, and cannot simultaneously detect the change of all object types and pixel-level range changes in the area. To overcome this difficulty, we propose a new coarse-to-fine deep learning-based land-use change detection method. We independently created a new scene classification dataset called NS-55, and innovatively considered the adaptation relationship between the convolutional neural network (CNN) and the scene complexity by selecting the CNN that best fit the scene complexity. The CNN trained by NS-55 was used to detect the category of the scene, define the final category of the scene according to the majority voting method, and obtain the changed scene by comparison to obtain the so-called coarse change result. Then, we created a multi-scale threshold (MST) method, which is a new method for obtaining high-quality training samples. We used the high-quality samples selected by MST to train the deep belief network to obtain the pixel-level range change detection results. By mapping coarse scene changes to range changes, we could obtain fine multi-type land-use change detection results. Experiments were conducted on the Multi-temporal Scene Wuhan dataset and aerial images of a particular area of Dapeng New District, Shenzhen, where promising results were achieved by the proposed method. This demonstrates that the proposed method is practical, easy-to-implement, and the NS-55 dataset is physically justified. The proposed method has the potential to be applied in the large scale land use fine change detection problem and qualitative and quantitative research on land use/cover change based on HR remote sensing data.