High-resolution remote sensing images contain rich color and texture information, but due to the inherent limitations of 2-D data, achieving high-quality semantic segmentation remains a challenge. Multimodal data fusion technology has emerged as an effective approach to overcome this issue. To accurately capture the semantic information in remote sensing images, this study designs a multimodal fusion Transformer-based DeepLabv3+ model for remote sensing semantic segmentation, named FTransDeepLab. Specifically, the network learns features from two modalities and is inspired by the DeepLab architecture. We extended the encoder by stacking the multiscale Segformer, encoding the input images into highly representative spatial features. Additionally, we introduced the multimodal feature rectification (MFR) module and the multimodal feature fusion (MFF) module. The MFR, composed of a channel attention module and a spatial attention module, enhances the model’s ability to capture essential features and improves performance by focusing on both global and local contexts. The MFF module utilizes a cross-attention mechanism to optimize the feature fusion process, which enhances representation learning by facilitating the interaction between diverse information and integrates features from different modalities. Finally, in the decoding path, the extracted high-level features are concatenated with low-level features to optimize the feature representation and upsampled to restore the size of input image. Extensive results on two datasets, the International Society for Photogrammetry and Remote Sensing (ISPRS) Vaihingen and Potsdam, have confirmed that the proposed FTransDeepLab can achieve superior performance compared to the state-of-the-art segmentation methods.

Weakly supervised change detection (WSCD) of bi-temporal remote sensing (RS) images has gained attention for its ability to reduce reliance on labor-intensive pixel-level change masks. Recent methods leverage image-level weak supervision to generate change pseudomasks for detecting changed objects, typically using class activation map (CAM) technique combined with DenseCRF or the Segment Anything Model (SAM). However, these methods still face two main challenges: first, CAMs tend to produce weak or false activations for changed objects, and second, DenseCRF and SAM lead to unreliable pseudomask generation, particularly when complex variations occur within objects in bi-temporal images. To address these challenges, a change selection network (CSNet) is proposed to enhance the quality of change activation maps and pseudomasks, improving their ability to accurately extract changed regions in bi-temporal RS images. First, a change activation selection (CAS) module is designed to generate a weight mask that selects and aggregates change-representing features, effectively highlighting missed change activations and strengthening weak activations. Second, a bi-temporal image selection (BIS) strategy is developed, incorporating two selection rules to filter out image pairs with poor-quality mask derived from SAM, while retaining those with high-quality results. Finally, a change pseudomask generation (CPG) module integrated with an atrous-spatial pyramid pooling (ASPP) classifier is developed to predict accurate change pixels for final pseudomask generation. Experimental results demonstrate that the proposed CSNet outperforms existing WSCD methods, achieving 79.32% IoU in change pseudomasks for the WHU-CD dataset, 68.12% for the GZ-CD dataset and 75.44% for the GVLM dataset. This study proposes a novel method that enhances the performance of the weakly supervised paradigm in RS CD.

In recent years, the inherent 2-D characteristics of optical images have led to a plateau in semantic segmentation performance. The complementary nature of light detection and ranging (LiDAR) point clouds and camera images can effectively enhance semantic segmentation capabilities, and thus, research into multimodal joint semantic segmentation is garnering increasing attention. However, the domain gaps between different dimensions present challenges for the fusion of multimodal data. In this article, we introduce a novel asymmetric multimodal interaction augmented network (AMIANet), which directly processes heterogeneous data from images and point clouds. The treatment of the disparities in modal data ensures consistency in the features of both modes. Through the newly developed synergistic multimodal interaction module (SMI Module), AMIANet is capable of combining the complementary characteristics of cross-modal data. This is achieved by interactively fusing and extracting precise and rich structural information from point cloud features to enhance image characteristics. The experimental results on the N3C-California, WHU-RRDSD, and ISPRS Vaihingen datasets demonstrate that AMIANet surpasses benchmark methods and current state-of-the-art (SOTA) approaches. The code will be available athttps://github.com/2012153946/AMIANet.

Weakly supervised building extraction methods, utilizing image-level labels, offer a cost-effective solution by significantly reducing the need for pixel-level annotation in high-resolution (HR) remote sensing (RS) images. These methods often focus on class activation map (CAM) optimization based on features extracted from individual images, missing out on the benefits of associating building features from multiple RS images (i.e., n images) to improve CAMs. This limitation leaves room for improvement in both CAM optimization and pseudo-mask generation. To address this, we propose the building-aware clustering and activation refinement network (BAC-AR-Net), a novel weakly supervised network to enhance weakly supervised building extraction performance. The building-aware clustering (BAC) module aggregates and clusters feature maps from multiple building samples to obtain common features of buildings. The common features are subsequently used to extract regions with similar building semantics, thereby enhancing the accuracy and completeness of building coverage in CAMs. Additionally, the activation refinement module is designed to generate pseudo-masks with clear boundaries and an effective separation of buildings and background. Experiments were conducted on the ISPRS Potsdam and Vaihingen datasets as well as a self-built building dataset to verify the effectiveness of our proposed method. The results show the proposed method outperforms both the weakly supervised semantic segmentation and weakly supervised building extraction methods that use image-level labels, achieving IoU accuracies of 0.8556, 0.8163, and 0.7797 on the respective datasets. This study introduces a novel weakly supervised learning framework to the RS application, with a particular focus on building extraction and semantic segmentation tasks.

In China, the haze is breaking out, and it cannot be solved in a few decade years. In this paper, we used haze simulation laboratory, simulated the particulate concentration through the control experiment, and measured the soil reflectance under the different the particulate concentration. Then we analyzed the effect of different hazes on the different soil moisture reflectance: the reflectance is relatively larger when the concentration increases, But when the concentration of the particles increases to 150 ug/m3, the curve becomes smooth. While the difference of the different soil moisture reflectance is decreasing. The different haze conditions (particulate matter concentration) makes the measured soil spectra curve change, so Some drought monitoring model based on spectral feature space(PDI, MPDI, SPSI, MSPSI) inevitably influence.

Soil moisture not only is an important parameter in precision agriculture, but is the main parameter in crop condition monitoring. We developed the method of soil moisture monitoring and evaluation with the new remote sensing platform, modified the existing indexes of PDI and MPDI, established inversion model and invert the soil moisture.

Soil line is important foundation of several drought monitoring models based on spectra feature space. So the accuracy of soil line obtained of study area effect the accuracy of this type monitoring model. More research in factors of soil line, however, the study has yet been rarely involved, that the accuracy of the soil line affecting drought monitoring accuracy. This paper analyzed the accuracy of the soil line impacting the accuracy of drought monitoring, such as PDI, MPDI, SPSI and NPSI.

The surface land albedo is an important parameter of the energy balance. It is very significant for the urban micro-climate and environment to research the variation and analysis of land surface albedo. In this paper, the variation of land surface albedo in Beijing in recent ten years, the correlation of the albedo and underlying surface type, the correlation of the albedo and LST and NDVI were studied based on ETM+ data in 2010 and 1999. The conclusions of the paper are as followed: (1) The average albedo of second, third, fourth and fifth ring in Beijing were gradually increased. The albedo of Beijing area in 2010 reduced 0.0075 compared with 1999; the albedo of main city area within fifth ring road reduced 0.0108; the albedo of the area between third ring road and fourth ring was the biggest drop range area, and it reduced 0.0135; the albedo of the area without fifth ring was the smallest range area, it reduced 0.0048. (2) The average albedo of land cover types in order from high to low: new construction areas, farmland, bare land, woodland, villages, construction zone, paddy fields, water. The change of land surface type led to the albedo difference in Beijing between 2010 and 1999.

In recent years, the Pearl River Delta (PRD) of China experienced serious saline water intrusion in the dry season, leading to water supply crisis. A challenging issue in water supply management is that there is not enough saline water intrusion information for decision making due to limited site-monitoring. Therefore, the objectives of this study is to investigate the spectral characteristics of seawater with different chlorosity, the most fundamental step in remote sensing application. Experiments in laboratory were performed to obtain spectral characteristics of saline water by spectroradiometer. The primary results show that, (1) reflectance of seawater was small between 400 and 850 nm, ranging from 0.01 to 0.14; (2) reflectance spectra of seawater was unique because spectral curves changed slightly in the ranges of 400-700 nm and 850-930 nm, but fluctuated observably in 700-850 nm, i.e., suddenly declined from 700 to 750 nm and then formed a narrow peak (750-770 nm) and a wide peak (770-850 nm); and (3) spectral reflectance increased with the increasing of chlorosity.

Drought index plays an important role in the assessment of drought severity as one sensitive indicator of land drought status. A simple and accurate method of expressing ground drought from remote sensing data is in urgent. Firstly, a brand-new frame has been developed by coordinate system transformation, with (x', y') expressing (soil water content, vegetation coverage), which made the meaning more clearly. Secondly, a new drought index - the modified Shortwave Infrared Perpendicular Water Stress Index (MSPSI) has been proposed, with additive of vegetation fraction f, (expressed as a distance from LAI contour to soil baseline) as a vegetation coverage adjusting factor, which makes ground drought status with vegetation coverage variance be considered as belonging to the new coordinate system. To evaluate the new index, the correlation analysis between the in-situ observation data from the meteorological sites and simulated values from MOD09A1 has been performed on a regional scale. We found them to be in good agreement. To further confirm that the drought index is feasible with higher precision than original drought indices (DIs), comparison experiment has been done.

Leaf Area Index (LAI) is a very important parameter in the area of vegetation quantitative remote sensing. Large range of LAI can reflect the change of eco-system. This article has discussed whether the crop type is a factor to impact the leaf area index retrieval. We choose four types of crops in our research and Hyperspectral Data and leaf area index of these crops were measured. Then the LAI retrieval models were established, which demonstrate the relationships between SVI and LAI. Finally the conclusion can be made that the type of crop is a factor impacting the LAI retrieval. For different crops, the best models are not the same. But the little difference of R2can be omitted. The SR is the best spectral vegetation index for LAI retrieval.

Forest regulating temperature ecosystem services was mainly that forest had the cooling effect in summer and warming effect in winter. Land surface temperature (LST) and normalized difference vegetation index (NDVI) were important parameter of forest regulating the temperature. This paper studied forest temperature changes with time using the time-series of LST and NDVI. Five typical samples were respectively selected from urban areas, farm, coniferous forest, broadleaf forest. The conclusions were followed: farm regulating temperature was less than forest; the lowest value of NDVI of the vegetation occurred at Feb., which lagged the time of the minimum LST occurred at January; the LST of conifer sample point had obvious low ebb, and the low ebb of broad-leaved sample point was not obvious; LST and NDVI were the negative correlation in the day, this was, the better of the vegetation cover, the lower of LST.

The estimation of soil wetness variations is of considerable importance to improve the reliability of flood warning. In this paper, Polarization Ratio Variation Index (PRVI) is presented which, on the basis of long-term multi-temporal Special Sensor Microwave/Image (SSM/I) data, can monitor soil wetness variations at a large scale. Nearly 18-years long time series PRVI are calculated based on the SSM/I data in Huaihe River Basin of China for the 1988-2005 period. Preliminary results achieved for the several important flooding events in Huaihe River Basin from 1988 to 2005 years are described, which confirms the reliability of proposed method.