Soil elements refer to different types of soil with unique colors, textures, and particle sizes. Their interpretation is essential for agriculture, ecological environment and land permeability assessment. This typically requires experts with dual knowledge in geology and remote sensing. With the increasing volume of remote sensing data, the traditional “visual interpretation" and “field survey" technique is no longer sufficient to meet the demands. Because of the challenges such as the fine structure of soil, complex and variable natural scenes, and strong spatial variability, there remains a considerable gap between the accuracy of deep learning-based methods and expert interpretation. To improve the accuracy of intelligent soil elements interpretation, this study proposes a soil interpretation framework coupling implicit knowledge with multispectral image (SIFCIM). This framework quantifies implicit knowledge, such as interpretation symbol and terrain feature, into matrix data (interpretation symbol distance field and digital elevation model). To align with the SIFCIM, an Implicit-Knowledge-Guided Adaptive Feature Fusion Network (IAFFNet) is constructed, which enhances the utilization efficiency of auxiliary features through an adaptive implicit feature fusion module and a global feature dependence module. Experimental results demonstrate that IAFFNet outperforms interpretation methods with single remote sensing image, achieving approximately 4.34% and 6.62% improvements in overall pixel accuracy and mean intersection over union, respectively. These results validate the effectiveness and robustness of the implicit-knowledge-guided approach in soil elements interpretation. To our knowledge, this work is the first to apply the concept of implicit knowledge to soil elements interpretation, providing a novel insight for related research.

Building extraction refers to the automatic identification and separation of buildings from the background in remote sensing images. It plays a significant role in urban planning, land management, and disaster monitoring. Deep-learning methods have shown advantages in building extraction, but they still face challenges such as variations in building types, object occlusions, and complex backgrounds. To address these issues, SDSNet, a deep convolutional network that incorporates global multi-scale feature extraction and cross-level feature fusion, is proposed. SDSNet consists of three modules: semantic information extraction (SIE), multi-level merge (MLM), and semantic information fusion (SIF). The SIE module extracts contextual information and improves recognition of multi-scale buildings. The MLM module filters irrelevant details guided by high-level semantic information, aiding in the restoration of edge details for buildings. The SIF module combines filtered detail information with extracted semantic information for refined building extraction. A series of experiments conducted on two distinct public datasets for building extraction consistently demonstrate that SDSNet outperforms the state-of-the-art deep-learning models for building extraction tasks. On the WHU building dataset, the overall accuracy (OA) and intersection over union (IoU) achieved impressive scores of 98.86% and 90.17%, respectively. Meanwhile, on the Massachusetts dataset, SDSNet achieved OA and IoU scores of 94.05% and 71.6%, respectively. SDSNet exhibits a unique advantage in recovering fine details along building edges, enabling automated and intelligent building extraction. This capability effectively supports urban planning, resource management, and disaster monitoring.

Water is an important element in the ecological environment, and different types of water (e.g., rivers, lakes, and ponds) have different impacts on the ecology. The extraction and classification of different types of water bodies has significant implications for the water resource management and water environment monitoring. Current research on the water-body types classification is relatively limited compared to water body extraction. Existing methods typically adopt a two-stage architecture, where the first stage extracts water bodies at the pixel level, and the second stage classifies the water bodies into different types using rule-based thresholds classifier and morphological features at object level. However, methods in the second stage suffer from overfitting, lack of robustness, and confusion in object segmentation. Despite these challenges, the deep learning methods could capture the high-level semantic features, which are effective for the classification of different types of water bodies. In this paper, a novel Water-Scene Enhancement Deep Model (WSEDM) was proposed for identifying multiple types of water bodies. The WSEDM consists of a pixel-wise water body extraction using Edge-Otsu and a patch-wise water-body types classification through deep learning model. In order to improve the accuracy of patch-wise water body classification, a novel multimodal feature fusion network(CASANet) was designed for the fusing of optical and SAR images. The water-body types classification was conducted on three international wetland cities in the urban agglomeration in the middle reaches of the Yangtze River. The 10m water-body types map achieved an overall accuracy of 94.6%. The proposed CASANet is also validated through comparison and transferability experiments, which further confirmed the superior performance.

Geological remote sensing interpretation plays a pivotal role in the field of regional geological mapping, encompassing the analysis of rock, soil, and water features. However, these geological elements can be obscured by the surrounding geographical environment and can undergo modifications caused by geological activities. The former hinders the effectiveness of satellite remote sensing data, resulting in the invisibility of element features, while the latter leads to the complex distribution of element features and significant spatial variations of geological elements. Consequently, existing deep learning-based models for interpreting geological elements often exhibit limited accuracy. To address these issues, this study proposes the Contextually Enhanced Multiscale Feature Fusion Network (CEMFFNet) for the efficient interpretation of geological elements. First, the context enhancement module is employed to extract abundant feature information and reinforce contextual features, aiming to capture essential features and strengthen their interconnections. Second, the multiscale feature fusion module incorporates the SimAM attention mechanism to adaptively learn features from different channels, emphasizing the feature information that contributes to interpretation results and maximizing the comprehensive and crucial feature information for each element. Extensive experiments demonstrate the superior performance of both the context enhancement module and the multiscale feature fusion module compared to several representative deep learning networks in terms of overall interpretation accuracy on two datasets. The model demonstrated improvements in oPA and mIoU of 2.4% and 2.8%, respectively, on the Landsat 8 dataset, and 3.5% and 3.2%, respectively, on the Sentinel-2 dataset.

Road extraction is crucial in urban planning, rescue operations, and military applications. Compared to traditional methods, using deep learning for road extraction from remote sensing images has demonstrated unique advantages. However, previous convolutional neural networks (CNN)-based road extraction methods have had limited receptivity and failed to effectively capture long-distance road features. On the other hand, transformer-based methods have good global information-capturing capabilities, but face challenges in extracting road edge information. Additionally, existing excellent road extraction methods lack validation for the Chinese region. To address these issues, this paper proposes a novel road extraction model called the global–local information fusion network (GLNet). In this model, the global information extraction (GIE) module effectively integrates global contextual relationships, the local information extraction (LIE) module accurately captures road edge information, and the information fusion (IF) module combines the output features from both global and local branches to generate the final extraction results. Further, a series of experiments on two different Chinese road datasets with geographic robustness demonstrate that our model outperforms the state-of-the-art deep learning models for road extraction tasks in China. On the CHN6-CUG dataset, the overall accuracy (OA) and intersection over union (IoU) reach 97.49% and 63.27%, respectively, while on the RDCME dataset, OA and IoU reach 98.73% and 84.97%, respectively. These research results hold significant implications for road traffic, humanitarian rescue, and environmental monitoring, particularly in the context of the Chinese region.