The rapid development and increasing availability of high-resolution satellite (HRS) images provides increased opportunities to monitor large scale land cover. However, inefficiency and excessive independence on expert knowledge limits the usage of HRS images on a large scale. As a knowledge organization and representation method, ontology can assist in improving the efficiency of automatic or semi-automatic land cover information extraction, especially for HRS images. This paper presents an ontology-based framework that was used to model the land cover extraction knowledge and interpret HRS remote sensing images at the regional level. The land cover ontology structure is explicitly defined, accounting for the spectral, textural, and shape features, and allowing for the automatic interpretation of the extracted results. With the help of regional prototypes for land cover class stored in Web Ontology Language (OWL) file, automated land cover extraction of the study area is then attempted. Experiments are conducted using ZY-3 (Ziyuan-3) imagery, which were acquired for the Jiangxia District, Wuhan, China, in the summers of 2012 and 2013.The experimental method provided good land cover extraction results as the overall accuracy reached 65.07%. Especially for bare surfaces, highways, ponds, and lakes, whose producer and user accuracies were both higher than 75%. The results highlight the capability of the ontology-based method to automatically extract land cover using ZY-3 HRS images.

Image quality is important for taking full advantage of satellite data. As a common indicator, the National Imagery Interpretability Scale (NIIRS) is widely used for image quality assessment and provides a comprehensive representation of image quality from the perspective of interpretability. The ZY-3 (Ziyuan-3) satellite is the first civil high resolution mapping satellite in China, which was established in 2012. So far, there has been no reports on adopting NIIRS as the common indicator for the quality assessment of that satellite image data. This lack of a common quality indicator results in a gap between satellite data users around the world and those in China regarding the understanding of the quality and usability of ZY-3 data. To overcome the gap, using the general image-quality equation (GIQE), this study evaluates the ZY-3 sensor-corrected (SC) panchromatic nadir (NAD) data in terms of the NIIRS. In order to solve the uncertainty resulting from the exceeding of the ground sample distance (GSD) of ZY-3 data (2.1 m) in GIQE (less than 2.03 m), eight images are used to establish the relationship between the manually obtained NIIRS and the GIQE predicted NIIRS. An adjusted GIQE is based on the relationship and verified by another five images. Our study demonstrates that the method of using adjusted GIQE for calculating NIIRS can be used for the quality assessment of ZY-3 satellite images and reveals that the NIIRS value of ZY-3 SC NAD data is about 2.79.

The need for integrating geospatial information (GI) data from various heterogeneous sources has seen increased importance for geographic information system (GIS) interoperability. Using domain ontologies to clarify and integrate the semantics of data is considered as a crucial step for successful semantic integration in the GI domain. Nevertheless, mechanisms are still needed to facilitate semantic mapping between GI ontologies described in different natural languages. This research establishes a formal ontology model for cross-lingual geospatial information ontology mapping. By first extracting semantic primitives from a free-text definition of categories in two GI classification standards with different natural languages, an ontology-driven approach is used, and a formal ontology model is established to formally represent these semantic primitives into semantic statements, in which the spatial-related properties and relations are considered as crucial statements for the representation and identification of the semantics of the GI categories. Then, an algorithm is proposed to compare these semantic statements in a cross-lingual environment. We further design a similarity calculation algorithm based on the proposed formal ontology model to distance the semantic similarities and identify the mapping relationships between categories. In particular, we work with two GI classification standards for Chinese and American topographic maps. The experimental results demonstrate the feasibility and reliability of the proposed model for cross-lingual geospatial information ontology mapping.

For the purpose of evaluating the image quality of ZY-3 (Ziyuan-3) satellite imagery and from the perspective of data users in various professions, this paper proposed a user-oriented quality assessment to assess the panchromatic and multispectral images of ZY-3, which included the assessment of its image statistical characteristics, image visual effect, and interpretability. In order to better understand the quality level of the ZY-3 image product and the scope of its applications, a comparison of image quality factors was conducted between ZY-3 image and SPOT-5 image due to the close similarity between their functionalities. As far as the ZY-3 image was concerned, features of its gray-scale statistics showed the advantage in its wider data range, higher standard deviation, and a better histogram presenting in image. The analysis of texture from the two images also revealed better performance of the ZY-3 image than that of SPOT-5 with respect to homogeneity (HOM), contrast (CON), entropy (ENT), and angular second moment (ASM). Moreover, the result of visual effect assessment also confirmed the good performance of ZY-3 data. An evaluation of the interpretability was conducted by using the National Imagery Interpretability Rating Scale (NIIRS). According to the relationship between NIIRS and mapping scale, the suitable mapping scales of the images from both ZY-3 and SPOT-5 were estimated from the viewpoint of image quality. As our survey shows, ZY-3 image products are marked by good gray scale characteristics and texture effect. According to its moderate NIIRS rating and result of mapping scale estimation, its maximum mapping scale potential is about 1:16 000.

Accurate mapping of winter wheat is essential for the advancement of precision agriculture and food security. However, classical semantic segmentation models frequently encounter difficulties in precise edge extraction, omission, and classification due to the presence of dense distributions and intra-class diversity. This study proposes a novel method for the extraction of winter wheat from remote sensing data using the GF-2 satellite. The method incorporates a multitask learning framework-Vision Transformer-based model (namely MCFormer) that combines semantic segmentation and boundary detection. Furthermore, the normalized difference vegetation index (NDVI) and land surface temperature (LST) derived from Landsat 8 images was included to enhance the representation of winter wheat's spectral characteristics. The method is evaluated in comparison to frequently used U-Net-, SegNet-, SegFormer-, and MANet-based winter wheat extraction methods in northern Anhui Province. The results indicate that the MCFormer-based method achieves the intersection over union (IoU), F1 score, recall, precision and overall accuracy (OA) of 0.9790, 0.9893, 0.9953, 0.9835, and 0.9900, respectively, outperforming the U-Net-, SegNet-, SegFormer-, and MANet-based methods. The incorporation of multitask learning with NDVI and LST data has been demonstrated to enhance several key performance metrics, including improvements in the IoU, F1 score, recall, precision and OA by 5.95%, 3.65%, 3.75%, 2.79%, and 2.24%, respectively. Our proposed approach improves the accuracy of winter wheat extraction from remote sensing images, which has the potential to facilitate precision agriculture and enhance food security.

Accurate identification of bare soil is critical for land management, soil conservation, and disaster risk assessment. Remote sensing imagery plays a pivotal role in bare soil mapping, yet deep convolutional networks typically require large annotated datasets, which are difficult to obtain due to the complexity of bare soil boundaries. Few-shot learning and self-supervised learning have emerged as promising approaches to address the data efficiency challenges in machine learning. However, each of these paradigms presents specific challenges that hinder their direct application to bare soil extraction in remote sensing imagery. To overcome these limitations, we propose a novel method called MAE-Guided Few-Shot (MGFS) for bare soil recognition, which combines the advantages of few-shot learning and self-supervised learning. Unlike traditional few-shot segmentation methods that rely solely on limited labeled datasets for fine-tuning, MGFS utilizes an MAE (Masked AutoEncoder) module to pretrain the encoder on a large corpus of unlabeled bare soil images, thereby enhancing its feature extraction capabilities. In addition, MGFS incorporates an Automatic High-Representative Sample Selection (AHRSS) strategy to ensure that the fine-tuning samples are representative of the broader dataset, effectively minimizing potential biases in sample selection. Experimental results demonstrate that the proposed method significantly outperforms existing few-shot learning approaches for bare soil extraction, achieving an Intersection over Union (IoU) of 69.23$\pm$ 0.45% under the 1-shot condition and 74.88$\pm$ 0.78% under the 5-shot condition. These results provide a new reference for bare soil extraction under few-shot learning conditions.