Multi-temporal Images Research Articles

A Heterogeneity-Enhancement and Homogeneity-Restraint Network (HEHRNet) for Change Detection from Very High-Resolution Remote Sensing Imagery

Change detection (CD), a crucial technique for observing ground-level changes over time, is a challenging research area in the remote sensing field. Deep learning methods for CD have made significant progress in remote sensing intelligent interpretation. However, with very high-resolution (VHR) satellite imagery, technical challenges such as insufficient mining of shallow-level features, complex transmission of deep-level features, and difficulties in identifying change information features have led to severe fragmentation and low completeness issues of CD targets. To reduce costs and enhance efficiency in monitoring tasks such as changes in national resources, it is crucial to promote the practical implementation of automatic change detection technology. Therefore, we propose a deep learning approach utilizing heterogeneity enhancement and homogeneity restraint for CD. In addition to comprehensively extracting multilevel features from multitemporal images, we introduce a cosine similarity-based module and a module for progressive fusion enhancement of multilevel features to enhance deep feature extraction and the change information utilization within feature associations. This ensures that the change target completeness and the independence between change targets can be further improved. Comparative experiments with six CD models on two benchmark datasets demonstrate that the proposed approach outperforms conventional CD models in various metrics, including recall (0.6868, 0.6756), precision (0.7050, 0.7570), F1 score (0.6958, 0.7140), and MIoU (0.7013, 0.7000), on the SECOND and the HRSCD datasets, respectively. According to the core principles of change detection, the proposed deep learning network effectively enhances the completeness of target vectors and the separation of individual targets in change detection with VHR remote sensing images, which has significant research and practical value.

Оценка деформационных процессов фотограмметрическим способом в Agisoft Metashape

Active development of digital cameras and computer software has made close-range photogrammetry an extremely popular way to solve various engineering tasks, in particular, to control deformations of constructions and mining objects. Automated measurement of tie points enables reviving traditional techniques of deformation measurement through single shot photogrammetric approach (a zero basis, pseudo parallaxes) at a qualitatively new level. The paper deals with the contour mapping equal displacement method adaptation offered by the first author in the 1970s for processing in Agisoft Metashape. It is proposed to create a virtual basis for two single multi-temporal images by adding a matrix of empty pixels to them. This enables processing those as stereo pair, building virtual point clouds and altitude matrices. In order to test the technique, the authors considered the deflection deformation of various objects under lab conditions and shot their photos with a non-metric camera before and after loading. The obtained images were processed in Metashape, and then the above products were created and contours of equal deformations were plotted. It was found out that the accuracy of displacement value estimation with this method exceeds the spatial resolution of the images more than three times. The technology is applicable for objects deformations monitoring shifting mainly in a single plane, in future it is planned to test it in field at studying slope processes

Histogram matching-based semantic segmentation model for crop classification with Sentinel-2 satellite imagery

ABSTRACT Accurate and near-real-time crop mapping from satellite imagery is crucial for agricultural monitoring. However, the seasonal nature of crops makes it challenging to rely on traditional machine learning methods and previous samples generated within specific domains. In this study, we improved the histogram matching method for color correction of multi-temporal images and tested the performance and prediction classification accuracy of three semantic segmentation models based on weak samples. Classification experiments were conducted for nine categories in two cities in Henan province from 2019 to 2022 using 10 m resolution Sentinel-2 images with different feature selection schemes. We trained the models using classified and recorrected results in four selected sites in 2019 and 2020, and designed experiments to assess the performance of the improved histogram matching method and verify the transferability of semantic segmentation models across regions and years. The experimental results showed that the UNet++ model with feature selection and improved histogram matching methods outperformed other models, such as DeepLab V3+ and UNet, in crop classification transfer cases, with better model performance and higher classification accuracy. The UNet++ model without training samples achieved optimal overall accuracy, Kappa coefficient, and mean F1-score values from 2019 to 2022, exceeding 87%, 82%, and 65%, respectively. Moreover, the representative error of weak samples and prediction classification results were analyzed to improve the model robustness. As an application of transfer-learning in crop mapping, the proposed model effectively addressed the classification problem of multispectral satellite imagery with missing labels.

Monitoring Seasonal Growth of Eucalyptus Plantation under Different Forest Age and Slopes Based on Multi-Temporal UAV Stereo Images

Eucalyptus grows rapidly and can grow up to 10 m per year, making them an important source of wood supply. Guangxi is a typical karst terrain, with hills and plains being the main growth areas for Eucalyptus. The differences in terrain can affect the seasonal growth of Eucalyptus plantations at different forest ages, which in turn affects the stocking of Eucalyptus. Currently, indiscriminate logging and management not only cause resource waste but also bring economic losses to operators. In this study, an unmanned aerial vehicle (UAV) was used to obtain multi-temporal stereo images to investigate the seasonal growth differences in structural parameters of individual Eucalyptus trees under different forest ages and slopes, providing data support for the precise management of Eucalyptus resources. The results showed that the tree height, crown width, diameter at breast height (DBH), and aboveground biomass (AGB) of individual trees, extracted based on UAV stereo images, are highly correlated with the field measured data, with an R2 of 0.99, 0.78, 0.75, and 0.92, and RMSE of 0.21 m, 0.16 m, 1.17 cm, and 3.79 kg/tree, respectively. The growth of Eucalyptus structural parameters varies in different seasons, with faster growth in spring and autumn, accounting for 76.39%, 73.75%, 73.65%, and 73.68% of the total annual growth, respectively. The growth of different structural parameters of individual trees is closely related to forest age, with tree height, crown width, and DBH gradually slowing down with the increase in forest age, while AGB shows a trend of first increasing and then decreasing. The differences in terrain also have a certain impact on the growth of individual Eucalyptus tree structural parameters. For individual 1-year-old and 3-year-old Eucalyptus trees, those located on gentle slopes grow faster in spring and autumn than those located on flat land. However, those located on flat land grow faster in summer and winter than those located on gentle slopes. For tree height, crown width, DBH, and AGB, the maximum annual growth differences between Eucalyptus trees on gentle slopes and flat are 3.17 m, 0.26 m, 1.9 cm, and 9.27 kg/tree, respectively. The results indicate that based on UAV stereo images, the individual tree structural parameters in Eucalyptus plantations under different forest ages and slopes can be extracted, as well as seasonal growth can be monitored, providing data support for logging, planting, and management of Eucalyptus plantations.

Enhancing Crop Mapping Precision through Multi-Temporal Sentinel-2 Image and Spatial-Temporal Neural Networks in Northern Slopes of Tianshan Mountain

Northern Slopes of Tianshan Mountain (NSTM) in Xinjiang hold significance as a principal agricultural hub within the region’s arid zone. Accurate crop mapping across vast agricultural expanses is fundamental for intelligent crop monitoring and devising sustainable agricultural strategies. Previous studies on multi-temporal crop classification have predominantly focused on single-point pixel temporal features, often neglecting spatial data. In large-scale crop classification tasks, by using spatial information around the pixel, the contextual relationships of the crop can be obtained to reduce possible noise interference. This research introduces a multi-scale, multi-temporal classification framework centered on ConvGRU (convolutional gated recurrent unit). By leveraging the attention mechanism of the Strip Pooling Module (SPM), a multi-scale spatial feature extraction module has been designed. This module accentuates vital spatial and spectral features, enhancing the clarity of crop edges and reducing misclassifications. The temporal information fusion module integration features from various periods to bolster classification precision. Using Sentinel-2 imagery spanning May to October 2022, datasets for cotton, corn, and winter wheat of the NSTM were generated for the framework’s training and validation. The results demonstrate an impressive 93.03% accuracy for 10 m resolution crop mapping using 15-day interval, 12-band Sentinel-2 data for the three crops. This method outperforms other mainstream methods like Random Forest (RF), Long Short-Term Memory (LSTM), Transformer, and Temporal Convolutional Neural Network (TempCNN), showcasing a kappa coefficient of 0.9062, 7.52% and 2.42% improvement in Overall Accuracy compared to RF and LSTM, respectively, which demonstrate the potential of our model for large-scale crop classification tasks to enable high-resolution crop mapping on the NSTM.

Combining Gaussian Process Regression with Poisson Blending for Seamless Cloud Removal from Optical Remote Sensing Imagery for Cropland Monitoring

Constructing optical image time series for cropland monitoring requires a cloud removal method that accurately restores cloud regions and eliminates discontinuity around cloud boundaries. This paper describes a two-stage hybrid machine learning-based cloud removal method that combines Gaussian process regression (GPR)-based predictions with image blending for seamless optical image reconstruction. GPR is employed in the first stage to generate initial prediction results by quantifying temporal relationships between multi-temporal images. GPR predictive uncertainty is particularly combined with prediction values to utilize uncertainty-weighted predictions as the input for the next stage. In the second stage, Poisson blending is applied to eliminate discontinuity in GPR-based predictions. The benefits of this method are illustrated through cloud removal experiments using Sentinel-2 images with synthetic cloud masks over two cropland sites. The proposed method was able to maintain the structural features and quality of the underlying reflectance in cloud regions and outperformed two existing hybrid cloud removal methods for all spectral bands. Furthermore, it demonstrated the best performance in predicting several vegetation indices in cloud regions. These experimental results indicate the benefits of the proposed cloud removal method for reconstructing cloud-contaminated optical imagery.

Exchange means change: An unsupervised single-temporal change detection framework based on intra- and inter-image patch exchange

Change detection is a critical task in studying the dynamics of ecosystems and human activities using multi-temporal remote sensing images. While deep learning has shown promising results in change detection tasks, it requires a large number of labeled and paired multi-temporal images to achieve high performance. Pairing and annotating large-scale multi-temporal remote sensing images is both expensive and time-consuming. To make deep learning-based change detection techniques more practical and cost-effective, we propose an unsupervised single-temporal change detection framework based on intra- and inter-image patch exchange (I3PE). The I3PE framework allows for training deep change detectors on unpaired and unlabeled single-temporal remote sensing images that are readily available in real-world applications. The I3PE framework comprises four steps: (1) intra-image patch exchange method is based on an object-based image analysis (OBIA) method and adaptive clustering algorithm, which generates pseudo-bi-temporal image pairs and corresponding change labels from single-temporal images by exchanging patches within the image; (2) inter-image patch exchange method can generate more types of land-cover changes by exchanging patches between images; (3) a simulation pipeline consisting of several image enhancement methods is proposed to simulate the radiometric difference between pre- and post-event images caused by different imaging conditions in real situations; (4) self-supervised learning based on pseudo-labels is applied to further improve the performance of the change detectors in both unsupervised and semi-supervised cases. Extensive experiments on two large-scale datasets covering Hong Kong, Shanghai, Hangzhou, and Chengdu, China, demonstrate that I3PE outperforms representative unsupervised approaches and achieves F1 value improvements of 10.65% and 6.99% to the state-of-the-art method. Moreover, I3PE can improve the performance of the change detector by 4.37% and 2.61% on F1 values in the case of semi-supervised settings. Additional experiments on a dataset covering a study area with 144 km2 in Wuhan, China, confirm the effectiveness of I3PE for practical land-cover change analysis tasks.

Shoreline change pattern analysis of Nagapattinam coastal stretch, Tamilnadu, India using digital shoreline analysis system (DSAS)

We analysed Nagapattinam coastal zones of southeast coast of India, using Multitemporal satellite images for five equal interval of time period (i.e. 2000, 2005, 2010, 2015, 2020 and 2021). The EPR and LLR methods enumerate the maximum accretion at rates of 22.35, 17.02 m/year and maximum erosion rates of -22.82, -10.84 m/year at the study area coastal stretch. This modification is due to several disasters (like Tsunami, cyclone) and anthropogenic activities (like construction of harbours, excavation of beach sand, industrialisation of garbage dump, urbanisation and discharge of domestic sewage). The Kalman filter model forecasting the shoreline by using statistical analysis shows 242 and 236 m in 2031 and 2041 accretion and erosion of -239 and -226 m in same period. The accretion was mainly observed in Vedaranyam and Thirupoondi and erosion was observed in Tharangambadi, Nagapattinam Poompuhar, Karaikal and Thirumullaivasal region. The outcome of this research ensures to create awareness to protect our shoreline and manage our coastal zones properly with several remedial measures for the future.

Monitoring maize lodging severity based on multi-temporal Sentinel-1 images using Time-weighted Dynamic time Warping

Lodging is one of the most important stresses in maize growth, which affects the yield, quality and mechanical harvesting ability of maize. Monitoring the lodging range and severity of large-scale maize by satellite remote sensing technology can support agricultural insurance claims, planting structure adjustment and field management. The purpose of this study is to investigate the application of Synthetic Aperture Radar (SAR) remote sensing images in monitoring the severity of maize lodging at the county scale. A time series feature dataset was obtained from multi-temporal Sentinel-1 SAR images before and after lodging. The sensitive features of maize lodging severity were screened by Jeffreys-Matusita (J-M) distance. The study compared the performance of three different methods for monitoring maize lodging severity, including Time-weighted Dynamic Time Warping (TwDTW), random forest (RF) and minimum distance (MinD). The results showed that all time phases of SAR images selected in the study had an important contribution to the differentiation of maize lodging severities. Among all features, the VH (Vertical Send and Horizontal Receive) time series features composed of five time-phase had better discrimination for different maize lodging severities. The accuracy of extracting maize lodging severities based on TwDTW was 76% and the Kappa coefficient was 0.70, which was higher than RF and MinD. Thus, by utilizing the multi-time phase features of Sentinel-1, we can effectively monitor the grades of maize lodging at the county scale.

Enhancing RABASAR for Multi-Temporal SAR Image Despeckling through Directional Filtering and Wavelet Transform.

The presence of speckle noise severely hampers the interpretability of synthetic aperture radar (SAR) images. While research on despeckling single-temporal SAR images is well-established, there remains a significant gap in the study of despeckling multi-temporal SAR images. Addressing the limitations in the acquisition of the "superimage" and the generation of ratio images within the RABASAR despeckling framework, this paper proposes an enhanced framework. This enhanced framework proposes a direction-based segmentation approach for multi-temporal SAR non-local means filtering (DSMT-NLM) to obtain the "superimage". The DSMT-NLM incorporates the concept of directional segmentation and extends the application of the non-local means (NLM) algorithm to multi-temporal images. Simultaneously, the enhanced framework employs a weighted averaging method based on wavelet transform (WAMWT) to generate superimposed images, thereby enhancing the generation process of ratio images. Experimental results demonstrate that compared to RABASAR, Frost, and NLM, the proposed method exhibits outstanding performance. It not only effectively removes speckle noise from multi-temporal SAR images and reduces the generation of false details, but also successfully achieves the fusion of multi-temporal information, aligning with experimental expectations.

Interannual comparison of historical floods through flood detection using multi-temporal Sentinel-1 SAR images, Awash River Basin, Ethiopia

Interannual comparison of historical floods through flood detection using multi-temporal Sentinel-1 SAR images, Awash River Basin, Ethiopia

Minimizing vegetation influence on soil salinity mapping with novel bare soil pixels from multi-temporal images

Optical remote sensing satellites provide rapid access to regional topsoil salinization mapping. However, mapping topsoil salinization based on spectral reflectance is always affected by background material like vegetation cover, straw mulching and soil types. In light of these challenges, this study investigates the potential of image fusion, where images of original and bare soil pixels were combined, to minimize the impact of vegetation cover on topsoil salinity mapping. A case study was presented for the typical vegetation cover area using synchronized Sentinel-2 MSI image (named original image) and 255 ground-truth data collected in October 2020, aligning with periods of vegetation cover and salt return. Furthermore, to obtain novel bare soil pixels, multi-temporal Sentinel-2 MSI images were acquired during two distinct intervals: March to May and September to November, spanning the years from 2018 to 2021. The synthetic soil image (SYSI) was obtained by extracting bare soil pixels from multi-temporal images. Two images (original, SYSI) were fused with non-negative matrix factorization (NMF) method, named SYSIfused. Then, the stacking machine algorithm was used for soil salinity mapping under different soil types, with evaluating the impact of SYSIfused on the accuracy of soil salinity prediction. The results showed the SYSIfused outperformed the original image (the R2 of the best models increased by 0.054–0.242, RMSE and MAE decreased by 0.049–0.780 and 0.012–0.546, respectively). Based on the SYSIfused, the order of the effect of soil types was coastal bog solonchaks > alluvial soil > cinnamon soil > coral saline soil > overall samples, and their roles in improving the R2 of the model were 0.141, 0.085, 0.022, 0.012, respectively. Besides, stacking models with the SYSIfused provided the best prediction performances (R2 = 0.742, RMSE = 0.377, MAE = 0.362). This study introduces the concept of merging original images with SYSI, resulting in a significant improvement in the accuracy of soil salinity mapping in areas covered by vegetation.

Contrastive-Learning-Based Time-Series Feature Representation for Parcel-Based Crop Mapping Using Incomplete Sentinel-2 Image Sequences

Parcel-based crop classification using multi-temporal satellite optical images plays a vital role in precision agriculture. However, optical image sequences may be incomplete due to the occlusion of clouds and shadows. Thus, exploring inherent time-series features to identify crop types from incomplete optical image sequences is a significant challenge. This study developed a contrastive-learning-based framework for time-series feature representation to improve crop classification using incomplete Sentinel-2 image sequences. Central to this method was the combined use of inherent time-series feature representation and machine-learning-based classifications. First, preprocessed multi-temporal Sentinel-2 satellite images were overlaid onto precise farmland parcel maps to generate raw time-series spectral features (with missing values) for each parcel. Second, an enhanced contrastive learning model was established to map the raw time-series spectral features to their inherent feature representation (without missing values). Thirdly, eXtreme Gradient-Boosting-based and Long Short-Term Memory-based classifiers were applied to feature representation to produce crop classification maps. The proposed method is further discussed and validated through parcel-based time-series crop classifications in two study areas (one in Dijon of France and the other in Zhaosu of China) with multi-temporal Sentinel-2 images in comparison to the existing methods. The classification results, demonstrating significant improvements greater than 3% in overall accuracy and 0.04 in F1 scores over comparison methods, indicate the effectiveness of the proposed contrastive-learning-based time-series feature representation for parcel-based crop classification utilizing incomplete Sentinel-2 image sequences.

A cross‐cross‐correlation based method for joint coregistration of rotated multitemporal synthetic aperture radar images

AbstractCoregistration is among the most important and challenging tasks when dealing with multiple synthetic aperture radar (SAR) images, especially when they are acquired at different time instants and characterised by low signal to noise power ratio (SNR) that contributes to their coherence reduction. However, even if some technological expedients could be implemented to maintain the same trajectory and to compensate for these inaccuracies during the acquisition campaign, multitemporal SAR images always need additional registration refinements after compression. Usually, to coregister a series of multitemporal SAR images, one of them is selected as the master, and the remainders are separately registered to it. Differently, in this study, a new strategy is developed to jointly coregister a stack of multitemporal SAR images. It is based on the exploitation of the cross‐correlations in turn computed from each couple of cross‐correlations (a.k.a. cross‐cross‐correlations) of the extracted patches. By doing so, the method is capable of exploiting also the respective misregistration information between the slaves during the estimation process. In this respect, this methodology is applied to enhance the registration capabilities of the constrained Least Squares (CLS) optimisation method, which instead does not account for the reciprocal information related to the slaves. Several tests are performed on multitemporal airborne‐measured SAR data. Obtained results show the effectiveness of the proposed algorithm in terms of achieved root mean square error for images affected by respective rotations also in comparison with the CLS counterpart.

Local-scale mapping of tree species in a lower mountain area using Sentinel-1 and -2 multitemporal images, vegetation indices, and topographic information

IntroductionMapping tree species is an important activity that provides the information necessary for sustainable forest management. Remote sensing is a effective tool that offers data at different spatial and spectral resolutions over large areas. Free and open acces Sentinel satellite imagery and Google Earth Engine, which is a powerful cloud computing platform, can be used together to map tree species.MethodsIn this study we mapped tree species at a local scale using recent Sentinel-1 (S-1) and Sentinel-2 (S-2) time-series imagery, various vegetation indices (Normalized Difference Vegetation Index - NDVI, Enhanced Vegetation Index - EVI, Green Leaf Index - GLI, and Green Normalized Difference Vegetation Index - GNDVI) and topographic features (elevation, aspect and slope). Five sets of data were used, in different combinations, together with the Random Forest classifier in order to determine seven tree species (spruce, beech, larch, fir, pine, mixed, and other broadleaves [BLs]) in the studied area.Results and discussionDataset 1 was a combination of S-2 images (bands 2, 3, 4, 5, 6, 7, 8, 8a, 11 and 12), for which an overall accuracy of 76.74% was obtained. Dataset 2 comprised S-2 images and vegetation indices, leading to an overall accuracy of 78.24%. Dataset 3 included S-2 images and topographic features, which lead to an overall accuracy of 89.51%. Dataset 4 included S-2 images, vegetation indices, and topographic features, that have determined an overall accuracy of 89.36%. Dataset 5 was composed of S-2 images, S-1 images (VV and VH polarization), vegetation indices, and topographic features that lead to an overall accuracy of 89.68%. Among the five sets of data, Dataset 3 produced the most significant increase in accuracy, of 12.77%, compared to Dataset 1. Including the vegetation indices with the S-2 images (Dataset 2) gave an accuracy increase of only 1.50%. By combining the S-1 and S-2 images, vegetation indices and topographic features (Dataset 5) there was an accuracy increase of only 0.17%, compared with the S-2 images plus topographic features combination (Dataset 3). However, the input brought by the S-1 images was apparent in the increase in classification accuracy for the mixed and other BL species that were mostly found in hilly locations. Our findings confirm the potential of S-2 images, used together with other variables, for classifying tree species at the local scale.

Automatic zone management definition in Mediterranean environment for Triticum durum

AbstractPrecision agriculture (PA) is an agronomic management that allows the parsimonious use of agronomic inputs according to the crop's actual spatiotemporal needs. The most important agronomic practice of PA is the site‐specific management (SSM) of agronomic inputs. To apply SSM, the number and location of homogeneous zones must be defined. Nowadays, many approaches are used to define the number of zones in an arbitrary and constant way, such as high‐, medium‐, and low‐input zones, without a priori statistical analysis. Two combined field experiments were carried out on durum wheat (Triticum turgidum subsp. durum Desf.) in central Italy to agronomically validate an automatic approach aimed at defining the number and position of the homogeneous zones using multispectral images. For both sites, at stem elongation and anthesis, the dry biomass per square meter and Soil Plant Analysis Development (SPAD) readings were measured, while at crop maturity the yield was measured. This approach uses multispectral images as source data and uses the gap statistic index to determine the correct number of zones. The management zones created were well‐fitted with the experimental design of both sites, and the validation was made by the statistical analysis performed on yield data, SPAD, dry biomass, and normalized difference red edge index with an average difference of 40.2%, 28.5%, 57.5%, and 37.4%, respectively, between the management zones. The approach could be scaled by using the multitemporal multispectral images provided by satellite constellation, which enables farms and societies to take advantage of all the economic and environmental benefits of PA.

Mass movement evaluation in deformed clastic rock with InSAR technique

AbstractThis study uses multitemporal InSAR images to assess different types of mass movements occurring in the Seydikemer (Muğla, SW Türkiye) district. A major highway connecting the Fethiye district and Antalya province and numerous small settlements of various sizes are found in the study area. In terms of summer tourism, they are Türkiye's most well‐known locations. Since the 1960s, at least according to official records, mass movements in this area have been known to negatively impact human life. Some areas have even been declared disaster zones for this reason. However, the movement's size, type, and susceptibility to landslides have all been studied in the last 10–15 years. With satellite images and field studies, distinctive lobate‐shaped movements were classified as inactive and active. During fieldwork, characteristic features such as rough topography and bending trees were easily observed and indicated extremely slow and large‐scale movement. In this study, it was found that movement less than 7–8 mm/year could be classified as ‘extremely slow’ mass movement like a ‘creep’, and moving up to 43 mm/year was evaluated as ‘very slow’ movement and ‘earth slide’ on a rigid surface. The examined moving masses with lobate geometry are activated by natural factors such as valley slope direction, stream bed flow direction, erosion of the toe of the river valleys, earthquakes, excessive precipitation, or human activities such as road opening.

The Impact and Prediction of Shoreline Dynamics in Pekalongan, Indonesia

Shoreline in Pekalongan has changed and impacted land use significantly. This research investigates historical changes in shoreline dynamics and their impact on land use from 2003–2021. This research also examines the trend of shoreline prediction for 2031–2041. Multitemporal SPOT5 and SPOT6 satellite images were used as primer data to detect shoreline and land use changes. The DSAS (Digital Shoreline Analysis System) analyzes the shoreline change and generates shoreline prediction. This research shows that the coastal areas of Pekalongan have been impacted by erosion, with 528.08 hectares of area and 56.33 hectares of accretion. The North Pekalongan sub-district is the worst area hit by erosion. The shoreline has retreated 2405.8 meters with a 353.3 m/year erosion rate. Several areas of land use were heavily affected by erosion trends, such as ponds, rice fields, and open land. Shoreline predictions generated by DSAS were heavily reliant on historical rates of shoreline change dynamics. The predicted area of the shoreline is estimated to retreat 49.9–466.2 meters from the existing position in the next 10 and 20 years. This research is expected to provide information to local governments for evaluating the maintenance and development planning of the coastal area

Temporal-agnostic change region proposal for semantic change detection

Remote sensing imagery allows temporal and large-scale observation of the Earth, and advanced techniques such as deep learning have been developed to deal with the massive data. As a result, by combining remote sensing with deep learning, it is possible to realize change detection, which can reveal the change in land cover and land use. Recent work has focused on semantic change detection (SCD), which simultaneously locates where the changes took place while identifying what objects changed between the different periods. Deep learning based SCD methods are typically based on a multi-task structure, which combines the classification and binary change detection (BCD). In this structure, the change region proposals are derived from the transformation (e.g., concatenation, differencing) of multi-temporal images, but after being processed by a convolutional neural network, there can be an incomplete response in the spatio-temporal features. Because of the high similarity of the color, texture, and geometric information of the ground objects in temporal high-resolution remote sensing imagery, this reduces the distinguishability of features and makes it difficult to identify changed areas. Meanwhile, there is less interaction between the BCD and classification, resulting in semantic inconsistency in changed areas. In this paper, the temporal-agnostic change region proposal network (TCRPN) is proposed, which is based on exploring the saliency of the changed regions in single-temporal images to highlight the changed areas in the spatio-temporal features and improve the response to change. As the changed areas can be the salient regions in the single-temporal images, a single-image salient area extractor (SSAE) is proposed to extract the salient areas from the different temporal images, obtaining salient change region candidates. As the change-related salient regions are temporally agnostic, temporal saliency fusion is used to integrate the candidates extracted from the single-temporal images to obtain a more reliable salient area map, where the salient area map is used to enhance the spatio-temporal features. Meanwhile, a shortcut is proposed to further make the change representation more distinguishable and suppress incorrectly highlighted regions. Directly utilizing multi-temporal images for change region proposal can reduce the change information loss and increase the interaction between the classification and BCD tasks to obtain better spatio-temporal features. The TCRPN is simple and flexible, and can be plugged into any multi-task SCD network. We selected four advanced multi-task SCD methods as the baseline, and added the TCRPN in their encoder or decoder part to conduct comprehensive experiments on three large-scale SCD datasets. The experimental results confirmed the effectiveness of the proposed method.

Spatiotemporal landuse dynamics driven by socioeconomic context in the southwestern coastal districts of Bangladesh

Socio-economic aspects and the increasing saltwater intrusion toward the southwestern coastal districts of Bangladesh are the main driving force for the alteration of landuse landcover (LULC) pattern. The present study employed a supervised classification approach using the maximum likelihood algorithm method in the geospatial environment to quantify the spatiotemporal LULC changes in the southwestern coastal districts of Bangladesh. The study used multi-temporal Landsat satellite images and identified the influencing socio-economic factors to find the relation with LULC changes. The study has seen a significant relationship between socioeconomic factors and landuse dynamics. Nearly 1500 km2 area has been converted to shrimp farming in almost all districts previously used for rice cultivation or uncultivated. It is actually due to the profitable shrimp production and saline water ingress toward the inland area that induced people to alter the landuse pattern. From 2010 to 2019, shrimp production increased significantly in the study area. The output of the study could be beneficial to policymakers for implementing the landuse policy and sustainable landuse management focusing on the aquaculture industry in this area.