In recent years, the integration of deep learning technology based on convolutional neural networks with historical maps has made it possible to automatically extract roads from these maps, which is highly important for studying the evolution of transportation networks. However, the similarity between roads and other features (such as contours, water systems, and administrative boundaries) poses a significant challenge to the feature extraction capabilities of convolutional neural networks (CNN). Additionally, CNN require a large quantity of labelled data for training, which can be a complex issue for historical maps. To address these limitations, we propose a method for extracting roads from historical maps based on an attention generative adversarial network. This approach leverages the unique architecture and training methodology of the generative adversarial network to augment datasets by generating data that closely resembles real samples. Meanwhile, we introduce an attention mechanism to enhance UNet3 + and achieve accurate historical map road segmentation images. We validate our method using the Third Military Mapping Survey of Austria-Hungary and compare it with a typical U-shaped network. The experimental results show that our proposed method outperforms the direct use of the U-shaped network, achieving at least an 18.26% increase in F1 and a 7.62% increase in the MIoU, demonstrating its strong ability to extract roads from historical maps and provide a valuable reference for road extraction from other types of historical maps.

Farmland abandonment significantly influences food and ecological security. To provide a comprehensive understanding of the current research landscape and evolving developments in the field of farmland abandonment, we have summarized the advancements and future trends in this research domain. This review employed CiteSpace software and incorporated geographic visualization techniques to generate knowledge maps and visually analyze literature on farmland abandonment sourced from the Web of Science (WOS) and China National Knowledge Infrastructure (CNKI) core databases, spanning the period from 1993 to 2023. The findings indicated a consistent annual increase in the number of publications on abandoned farmland research. China has emerged as a significant contributor to research in this field, exhibiting a relatively large number of related research publications. The investigation on farmland abandonment spans across multiple disciplines, indicating intersections among various fields. Chinese publications predominantly focus on abandoned farmland studies within the agricultural discipline, while English publications exhibit greater interest in abandoned farmland research within the context of ecological and environmental sciences. Presently, the research hotspots in this field include the alterations in the soil properties of abandoned farmland, factors impacting farmland abandonment, the ecological impact of such abandonment, and prevention and control strategies. It is expected that future research on farmland abandonment will aim to strike a balance between ensuring food security and preserving ecological value to optimize decision-making in governance.

Mangroves, integral to ecological balance and socioeconomic well-being, are facing a concerning decline worldwide. Remote sensing is essential for monitoring their evolution, yet its effectiveness is hindered in developing countries by economic and technical constraints. In addressing this issue, this paper introduces MANGLEE (Mangrove Mapping and Monitoring Tool in Google Earth Engine), an accessible, adaptable, and multipurpose tool designed to address the challenges associated with sustainable mangrove management. Leveraging remote sensing data, machine learning techniques (Random Forest), and change detection methods, MANGLEE consists of three independent modules. The first module acquires, processes, and calculates indices of optical and Synthetic Aperture Radar (SAR) data, enhancing tracking capabilities in the presence of atmospheric interferences. The second module employs Random Forest to classify mangrove and non-mangrove areas, providing accurate binary maps. The third module identifies changes between two-time mangrove maps, categorizing alterations as losses or gains. To validate MANGLEE’s effectiveness, we conducted a case study in the mangroves of Guayas, Ecuador, a region historically threatened by shrimp farming. Utilizing data from 2018 to 2022, our findings reveal a significant loss of over 2900 hectares, with 46% occurring in legally protected areas. This loss corresponds to the rapid expansion of Ecuador’s shrimp industry, confirming the tool’s efficacy in monitoring mangroves despite cloud cover challenges. MANGLEE demonstrates its potential as a valuable tool for mangrove monitoring, offering insights essential for conservation, management plans, and decision-making processes. Remarkably, it facilitates equal access and the optimal utilization of resources, contributing significantly to the preservation of coastal ecosystems.

In this study, we investigated the accuracy of surface models and orthophoto mosaics generated from images acquired using different data acquisition methods at different processing levels in two urban study areas with different characteristics. Experimental investigations employed single- and double-grid flight directions with nadir and tilted (60°) camera angles, alongside the Perimeter 3D method. Three processing levels (low, medium, and high) were applied using SfM software, resulting in 42 models. Ground truth data from RTK GNSS points and aerial LiDAR surveys were used to assess horizontal and vertical accuracies. For the horizontal accuracy test, neither the oblique camera angle nor the double grid resulted in an improvement in accuracy. In contrast, when examining the vertical accuracy, it was concluded that for several processing levels, the tilted camera angle yielded better results, and in these cases, the double grid also improved accuracy. Feature importance analysis revealed that, among the four variables, the data acquisition method was the most important factor affecting accuracy in two out of three cases.