Analyzing the relationship between rural settlements and rural population change under different policy scenarios is key in the sustainable development of China’s urban and rural areas. We proposed a framework that comprised the mixed land use structure simulation (MCCA) model and the human–land coupling development model to assess the spatiotemporal dynamic changes in rural settlements and its’ coupling relationship with the rural population in the economically developed region of Deqing, Zhejiang Province. The results showed that rural settlements and urban land increased by 14.36 and 29.07 km², respectively, over the last 20 years. The expansion of some rural settlements and urban land occurred at the cost of cropland occupation. Rural settlements showed an expansion trend from 2000 to 2020, increasing from 42.69 km² in 2000 to 57.05 km² in 2020. In 2035, under the natural development scenario, the cropland protection scenario, and the rural development scenario, rural settlements are projected to show an expansion trend and Wukang and Leidian are the key regions with rural settlement expansion. The distance to Hangzhou, nighttime light data, distance to rivers, and precipitation are important factors influencing the expansion of rural settlements. The coupling relationship between rural settlements and the rural population developed in a coordinated manner from 2000 to 2020. For 2035, under different scenarios, the coupling relationship between rural settlements and the rural population showed different trends. In the rural development scenario, the highest number of towns with coordinated development between rural settlements and the rural population is in Deqing, predominantly with Type I coupling. Overall, an important recommendation from this study is that the sustainable development of regional land use can be promoted by controlling the occupation of cropland for urban and rural construction, balancing rural settlement expansion and rural population growth, and formulating land use policies that are more suitable for rural development.

Deep convolution neural network (DCNN) technology has achieved great success in extracting buildings from aerial images. However, the current mainstream algorithms are not satisfactory in feature extraction and classification of homesteads, especially in complex rural scenarios. This study proposes a deep convolutional neural network for rural homestead extraction consisting of a detail branch, a semantic branch, and a boundary branch, namely Multi-Branch Network (MBNet). Meanwhile, a multi-task joint loss function is designed to constrain the consistency of bounds and masks with their respective labels. Specifically, MBNet guarantees the details of prediction through serial 4× down-sampled high-resolution feature maps and adds a mixed-scale spatial attention module at the tail of the semantic branch to obtain multi-scale affinity features. At the same time, the low-resolution semantic feature maps and interaction between high-resolution detail feature maps are maintained. Finally, the result of semantic segmentation is refined by the point-to-point module (PTPM) through the generated boundary. Experiments on UAV high-resolution imagery in rural areas show that our method achieves better performance than other state-of-the-art models, which helps to refine the extraction of rural homesteads. This study demonstrates that MBNet is a potential candidate for building an automatic rural homestead management system.

The current research on the change of homesteads has rarely combined spatial information, and no expansion research has been found with a scale as detailed as the level of the homestead plots. Understanding the changing rules of homesteads expansion and its influencing factors are important to homestead reform, optimization of village layout, and cultivated land protection. In order to explore the direction, level, and spatial pattern of homesteads expansion from 2010 to 2020 and the influencing factors, by taking six villages in Deqing County as an example, this research proposed a tract division scheme based on multidistance spatial clustering and intersecting area index and used homesteads expansion index and geographical detector. The results showed that (1) 83.3% of the villages had a decreasing trend of homesteads expansion, and half of the villages’ homesteads remained unchanged in their distribution direction. (2) The average order of factors influencing the expansion of homesteads was: distance from road > altitude > slope > policy restriction > distance from water area > slope aspect. This study helps to understand the spatio-temporal features of homesteads expansion in small-scale areas, provides an empirical reference for deepening homesteads reform and policy formulation, and also provides new methods for quantitative research on homesteads expansion.

The accurate and timely monitoring of land cover types is of great significance for the scientific planning, rational utilization, effective protection and management of land resources. In recent years, land cover classification based on hyperspectral images and the collaborative representation (CR) model has become a hot topic in the field of remote sensing. However, most of the existing CR models do not consider the problem of sample imbalance, which affects the classification performance of CR models. In addition, the Tikhonov regularization term can improve the classification performance of CR models, but greatly increases the computational complexity of CR models. To address the above problems, a local nearest neighbor (LNN) method is proposed in this paper to select the same number of nearest neighbor samples from each nearest class of the test sample to construct a dictionary. This is then introduced into the original collaborative representation classification (CRC) method and CRC with Tikhonov regularization (CRT) for land cover classification, denoted as LNNCRC and LNNCRT, respectively. To verify the effectiveness of the proposed LNNCRC and LNNCRT methods, the classification performance and running time of the proposed methods are compared with those of six popular CR models on a hyperspectral scene with nine land cover types. The experimental results show that the proposed LNNCRT method achieves the best land cover classification performance, and the proposed LNNCRC and LNNCRT methods not only further exclude the interference of irrelevant training samples and classes, but also effectively eliminate the influence of imbalanced training samples, so as to improve the classification performance of CR models and effectively reduce the computational complexity of CR models.

Regional crop layout has changed significantly due to climate, policy, and other factors, which has impacted farmland nutrient balance. Here, we evaluated the impact of crop layout changes on N and P nutrient balance in the West Liaohe River Basin from 2000–2015. The study area has long been in a N and P surplus state. The unit N surplus exhibited a downward trend and the unit P surplus showed an increasing trend. Significant correlations existed between planting areas and nutrient surplus. The N and P surplus layout was mainly concentrated in the West Liaohe River lower reaches basin. The planting area of wheat must be reduced and the areas of maize and soybean must be controlled to adjust the N and P balance and reduce the environmental pollution risk. Chemical fertilizer and seed inputs are the main sources of N input. Furthermore, combining farming and pastoral farming is conducive to improving N and P use efficiency. Manure can be absorbed by farmland, the ratio of organic and chemical fertilizers can be reasonably set, and chemical fertilizer application can be reduced.

It is very significant for rural planning to accurately count the number and area of rural homesteads by means of automation. The development of deep learning makes it possible to achieve this goal. At present, many effective works have been conducted to extract building objects from VHR images using semantic segmentation technology, but they do not extract instance objects and do not work for densely distributed and overlapping rural homesteads. Most of the existing mainstream instance segmentation frameworks are based on the top-down structure. The model is complex and requires a large number of manually set thresholds. In order to solve the above difficult problems, we designed a simple query-based instance segmentation framework, QueryFormer, which includes an encoder and a decoder. A multi-scale deformable attention mechanism is incorporated into the encoder, resulting in significant computational savings, while also achieving effective results. In the decoder, we designed multiple groups, and used a Many-to-One label assignment method to make the image feature region be queried faster. Experiments show that our method achieves better performance (52.8AP) than the other most advanced models (+0.8AP) in the task of extracting rural homesteads in dense regions. This study shows that query-based instance segmentation framework has strong application potential in remote sensing images.