Rural China's accelerating demographic transition is reshaping rural household migration patterns and socioeconomic development trajectories. Based on data from the 2023 China Rural Homestead Survey, a multinomial logit model is constructed to examine the effect of rural household aging on migration intention and the moderating role of farmers’ property rights cognition. The results show that rural household aging significantly strengthens households’ willingness to remain in their village and reduces the willingness of some household members to migrate, whereas its effect on the intention of all household members to migrate together is not significant. Further analysis reveals that the moderating effect of property rights cognition is not significant in the full sample, but heterogeneity analysis indicates marked conditional features. In villages with relatively abundant policy resources and clearer institutional interpretation, farmers’ cognition of collective property rights can alleviate the inhibitory effect of aging on partial migration intention. In contrast, in villages with insufficient policy support, farmers’ cognition of collective property rights strengthens their willingness to remain in their village. These findings provide empirical evidence from China for understanding the interactions among population aging, rural property rights institutions, and household migration intention in economies undergoing structural transformation. They also offer policy implications for jointly advancing the provision of rural elderly care, the construction of land rights institutions, and the orderly migration of rural households.

Specialty agricultural industries are often constrained by limited labeled data and dependence on domain experts for cultivar identification. Using the high-value Cymbidium goeringii sector as a representative case, this study developed an integrated workflow from data construction to on-site deployment. First, large language model (LLM)-assisted keyword construction and structured retrieval templates were combined with automated image acquisition, quality control, and model-assisted annotation to construct a standardized dataset containing 84,477 images, 110,132 annotated instances, and 20 cultivars. Second, an improved lightweight detection model, YOLO-AgriLite, was developed based on YOLOv5s by incorporating depthwise separable convolutions, ContextAggregation, a bidirectional feature pyramid network, and the SW-MPDIoU loss function. YOLO-AgriLite achieved a precision of 97.1%, recall of 98.7%, mAP50 of 98.6%, and F1-score of 97.9%, with a computational complexity of 10.2 GFLOPs. Finally, the model was optimized using Int8-KL quantization and deployed on an RK3588S-based identification system. In a separate hardware performance test, the three-core NPU configuration achieved an average inference throughput of 80 FPS at a measured power consumption of 9.00 W. An on-site evaluation involving 278 exhibits from 17 cultivars at the 2025 Zhejiang Cymbidium goeringii Festival achieved an identification accuracy of 90.6%. These results demonstrate the feasibility of integrating automated data construction, lightweight vision models, and on-site deployment to support practical cultivar identification in specialty agriculture.

Agricultural specialty industries, exemplified by the high-value Cymbidium goeringii sector, are severely hindered by scarce data and a reliance on rare domain experts for tasks like variety identification. To overcome these dual bottlenecks, this study proposes an end-to-end framework advancing from automatic data collection to intelligent 'brain'-machine realization. First, addressing the data scarcity, a large language model (LLM)-driven pipeline automatically constructed a standardized dataset of 84,477 images across 20 Cymbidium goeringii varieties. Second, to distill expert knowledge into an intelligent "brain", an improved lightweight detection model, YOLO-AgriLite, was developed based on YOLOv5s by incorporating Depthwise Separable Convolutions, ContextAggregation, a Bidirectional Feature Pyramid Network, and the SW-MPDIoU loss function . Finally, for physical "machine" realization, the model was deployed on an RK3588S edge computing device using INT8 quantization. Experimental results demonstrate that YOLO-AgriLite achieved a mean average precision (mAP@0.5) of 98.6%, significantly outperforming newer baselines like YOLOv8s and YOLOv11s. On the edge device, it delivered a high-throughput inference speed of 80 FPS with a full-core power consumption of only 9W. Field validation at a national orchid exhibition confirmed its robustness, achieving a 90.6% real-time accuracy rate .This research provides a highly generalizable 'brain'-machine paradigm that successfully resolves the scarcity of data and experts, enabling low-cost, intelligent decision-making in regional specialty agriculture.

Facility environmental conditions directly affect the growth and quality of crops. The conventional method of laying multiple sensor nodes is costly and difficult to communicate, while the expensive ground robot monitoring has shortcomings such as universality and interference with agricultural activities. A compact autonomous inspection system for greenhouse environmental information sensing and three-dimensional (3D) visualization was developed, composing of a sensing module, a driver module, and a control and communication interface, to monitor the carbon dioxide concentration, relative humidity, temperature and illumination intensity information at different spatial positions of greenhouse autonomously and automatically. Sensing module connected and controlled by the driver module autonomously moved and acquired environmental information. Meanwhile the control and communication interface, based on Qt and Open Graphics Library (OpenGL) technologies, performed functions of remote controlling, data receiving and visual display. ZigBee, 4G wireless components and Ali Cloud platform cooperated to realize the hierarchical command communication and data transmission among the system. Based on Ultra Wideband (UWB) positioning technology, combined with the trilateral positioning algorithm, the fusion of sensing data and positioning data was realized the 3D visualization, further improving the display dimension of facility environmental information. The practical measurement showed that the sensing module, driven by the driver module, can monitor the facility environmental information at different spatial positions in real time. UWB positioning method realized the accurate positioning of the monitoring device, the average maximum errors of the distance information and height were 0.114 m and 0.012 m, which fully meets the positioning requirements of the facility environmental monitoring. The control and communication interface ran normally, operated friendly and displayed visually. This greenhouse 3D environmental sensing autonomous inspection system can effectively reflect the uneven environmental field of the facility agriculture with low cost and good compatibility, providing effective technical support for the accurate control of the facility environment and the high quality growth of the facility crops, with huge potential for agriculture application.

As technology continues to advance and new application scenarios emerge, accurate and rapid on-site detection sensors are essential for monitoring and control the quality and safety of food. The excellent photography and image processing capabilities of smartphone has greatly facilitated the miniaturization, integration, and visual monitoring of sensors, making them essential tools for portable real-time determination. The “on–off-on” fluorescence sensor has garnered significant attention due to its remarkable features, including high sensitivity, excellent selectivity, strong anti-interference ability, and customizable probe structure. In this context, portable sensors with intelligent analysis and sensing systems have been extensively reported according to the “on–off-on” fluorescent signals, and highlight their immense potential and value in various fields. However, to date, a comprehensive review of smartphone-based on-site detection platforms driven by the “on–off-on” fluorescence mechanism remains absent. This review aims to fill this gap by providing an elaborate explanation of the “on–off-on” fluorescence response process, along with a detailed introduction to commonly used fluorescent materials and their underlying design principles. Furthermore, the latest applications of portable sensors based on the “on–off-on” fluorescence approach in detection of food safety are thoroughly reviewed. Finally, an in-depth analysis was conducted on the challenges and future development directions of the smartphone based sensing systems. This review aims to provide a valuable theoretical basis for comprehending and designing portable sensors, ultimately contributing to more effective and accurate real-time detection of pollutants.

Ethylene is one of the important indicators to reflect the ripeness and senescence of the fruit. Ethylene sensor can be applied for fast ethylene detection nondestructively, yet high temperature condition of the common metal oxide semiconductors restrains its agricultural applications. Herein, porous-ZnO/carbon nanofibers (CNFs) three-dimensional structure, derived from Zeolite imidazole framework-8 (ZIF-8) /polyacrylonitrile (PAN) electrospun nanofibers, was proposed and further optimized. The results showed that porous-ZnO/CNFs exhibited increasing resistance to reducing ethylene at room temperature, well explained by the Fermi levels' relative position of the sensing system. Repeatability, selectivity and stability were demonstrated. The addition of porous-ZnO (10 % composting mass ratio) resulted in a 428 % increase in the ethylene-sensing response compared with CNFs, and a 64.2 % improvement compared with ZnO/CNFs. The outstanding sensing performance can be attributed to the synergistic effect of porous and interconnected morphology, the formed p-n heterojunction, as well as oxygen-vacancy effect. This strategy is beneficial for designing efficient gas sensors at room temperature, and exploring the application of the advanced gas sensors in intelligent agriculture field.

Cross-sensitivity among chemical gas sensors leads to inaccurate identification of mixed gas. Pattern recognition algorithms are usually applied to improve the recognition accuracy. However, the abilities of current methods to process sequence property of response data are not strong enough. Especially, they cannot deal with bidirectional cross-sensitive issue, leading to recognition errors. Bidirectional Recurrent Neural Network (BRNN) could well learn bidirectional association features between word sequences in natural language processing field, which is similar to the bidirectional interaction of cross-sensitivity. In this study, an improved deep BRNN model was constructed to solve the cross-sensitivity problem of chemical gas sensor array. A chemical gas sensor array with four units was fabricated and response data was thoroughly obtained. Data preprocessing methods, model structure hyperparameters and optimizers were studied. Finally, an improved deep BRNN model was developed with 3 layers and 100 hidden_size, training with Adamax optimizer. A recognition accuracy of 98.93% was achieved, attributing to the model's excellent learning ability to the bidirectional cross-sensitivity rules among gas sensors. This improved BRNN model provided a novel idea to eliminate cross-sensitivity, exhibiting good potential for recognizing mixed gas analyte accurately.

With the reform of the separation of the ownership,contractual,and management rights in rural homesteads,how to implement the ownership,contractual,and management rights in the homesteads system has become a hot topic of current research.However,the existing studies mainly analyzed the impact of the implementation of ownership rights on the reform of the homestead system and the problems of the current homestead ownership system,while the study and comparison of the characteristics of exercising subjects under the perspective of the implementation of homestead ownership rights are relatively insufficient.Based on this,this paper analyzed the similarities and differences in the forms,functions and operations of the rural collective economic organizations,villagers'committees and villagers'councils,as well as the problems that exist in the agency forms of the ownership subjects in the homesteads system reform according to the characteristics of the current homesteads ownership subjects.It was found that there are significant differences in the forms,functions and operation mechanisms of the three types of ownership agents in the current homesteads reform,and none of them can fully perform the role of the homesteads ownership agent,so it is necessary to select the homesteads ownership agent according to local conditions,divide and improve the functions of the existing peasant self-government organizations,and consider the extension of the powers scope of homesteads ownership.

Precise and timely classification of land cover types plays an important role in land resources planning and management. In this paper, nine kinds of land cover types in the acquired hyperspectral scene are classified based on the kernel collaborative representation method. To reduce the spectral shift caused by adjacency effect when mining the spatial-spectral features, a correlation coefficient-weighted spatial filtering operation is proposed in this paper. Additionally, by introducing this operation into the kernel collaborative representation method with Tikhonov regularization (KCRT) and discriminative KCRT (DKCRT) method, respectively, the weighted spatial-spectral KCRT (WSSKCRT) and weighted spatial-spectral DKCRT (WSSDKCRT) methods are constructed for land cover classification. Furthermore, aiming at the problem of difficulty of pixel labeling in hyperspectral images, this paper attempts to establish an effective land cover classification model in the case of small-size labeled samples. The proposed WSSKCRT and WSSDKCRT methods are compared with four methods, i.e., KCRT, DKCRT, KCRT with composite kernel (KCRT-CK), and joint DKCRT (JDKCRT). The experimental results show that the proposed WSSKCRT method achieves the best classification performance, and WSSKCRT and WSSDKCRT outperform KCRT-CK and JDKCRT, respectively, obtaining the OA over 94% with only 540 labeled training samples, which indicates that the proposed weighted spatial filtering operation can effectively alleviate the spectral shift caused by adjacency effect, and it can effectively classify land cover types under the situation of small-size labeled samples.

科学评估农户社会资本状况与群体差异,以提升农户社会资本水平,发挥社会资本对乡村治理的正向作用,推动实现乡村治理现代化.基于乡村治理的视角,从社会网络、社会信任、社会声望、社会参与和社会规范5个维度构建指标体系,运用CRITIC法确定指标权重测度农户社会资本状况,并采用独立样本t检验、单因素方差分析法分析了不同群体间的差异.结果发现:各维度的权重大小依次为社会规范、社会信任、社会参与、社会声望、社会网络.农户社会资本指数均值为0.608,5个维度得分从高到低依次为社会信任(0.755)>社会参与(0.631)>社会声望(0.620)>社会规范(0.588)>社会网络(0.417).新老两代农户间社会资本不存在显著差异;不同教育背景农户间的社会资本指数及社会网络、社会参与、社会规范具有显著差异,且均随着家庭成员最高学历的升高呈上升趋势;不同成员身份农户的社会资本指数及社会网络、社会信任、社会声望、社会参与具有显著差异,家庭成员中有村干部、党员、军人(含退伍)的农户的这几个指标均高于普通农户.调研地区农户社会资本状况处于较低水平,加强教育培训、提升政治素养可以提高农户的社会资本水平,并提出了立足根本问题和群体差异,采取差异化手段培育农户社会资本的对策.

为破解因农户分化引发的农村居民集中居住意愿的影响,以长江中下游部分省份农村地区的222份实地调查样本为例,采用社会分化理论、行为决策理论以及认知行为理论构建了农户分化对农村居民集中居住影响的研究框架,并开展了相关的定量分析,研究结果表明:(1)农村居民具有一定的集中居住意愿基础,且集中居住意愿与其分化程度存在明显的相关性.(2)农户的收入结构分化对其集中居住意愿具有显著影响,收入水平分化对其影响不显著.(3)农户认知变量在农户分化对农村居民集中居住意愿的影响中发挥着中介作用.基于此结论,提出应从尊重农民权益、提升政策合理性与透明性、促进农民增收以及增加农民非农收入等角度制定相关政策建议.

基于2010-2018年中国家庭追踪调查(CFPS)数据,研究发现农村住房总体上具有积极的社会效应,农村住房面积能够显著提高农村居民的社会阶层认同,考虑房屋类型、室内环境等条件后的结果依然稳健.进一步分析表明,收入不平等成为调节农村住房能否影响农村居民社会阶层认同的关键变量,截止到2018年,面积更大且更新颖的农村住房更有利于提升高收入群体的社会阶层认同.政府须借此机会完善农村住房建设管理制度,探索将常住人口作为农村住房建设的重要参考指标,宣传节约集约的农村住房消费观念.

农村住房质量安全是保障农村居民安居乐业的基本要求,全国多地为保障农村住房质量安全推出了加强农村建房主体管理的举措,而农村建房主体在保障农村住房质量安全的作用仍尚未形成定论.基于2022年中国31个省(市、自治区)4568个农户的微观调查数据,实证分析农村建房主体对农村住房质量安全的影响.研究结果表明,农村建房主体的专业化程度越高,农户对农村住房质量安全的评价越高,其住房满意度也越高.此外,提高住房安全质量对于农村建房主体在建造建筑结构相对简单房屋时的作用更大.因此,应当加强农村建房主体的培训与管理,积极探索农村建筑工匠资质管理制度,引导农村居民选择规范、有资质的农村建房主体,并注重施工全程的监管,以减少住房质量安全隐患、保障农村居民住房质量安全.

The continuous changes in Land Use and Land Cover (LULC) produce a significant impact on environmental factors. Highly accurate monitoring and updating of land cover information is essential for environmental protection, sustainable development, and land resource planning and management. Recently, Collaborative Representation (CR)-based methods have been widely used in land cover classification from Hyperspectral Images (HSIs). However, most CR methods consider the spatial information of HSI by taking the average or weighted average of spatial neighboring pixels of each pixel to improve the land cover classification performance, but do not take the spatial structure information for pixels into account. To address this problem, a novel Weighted Spatial–Spectral Joint CR Classification (WSSJCRC) method is proposed in this paper. WSSJCRC only performs spatial filtering on HSI through a weighted spatial filtering operator to alleviate the spectral shift caused by adjacency effect, but also utilizes the labeled training pixels to simultaneously represent each test pixel and its spatial neighborhood pixels to consider the spatial structure information of each test pixel to assist the classification of the test pixel. On this basis, the kernel version of WSSJCRC (i.e., WSSJKCRC) is also proposed, which projects the hyperspectral data into the kernel-induced high-dimensional feature space to enhance the separability of nonlinear samples. The experimental results on three real hyperspectral scenes show that the proposed WSSJKCRC method achieves the best land cover classification performance among all the compared methods. Specifically, the Overall Accuracy (OA), Average Accuracy (AA), and Kappa statistic (Kappa) of WSSJKCRC reach 96.21%, 96.20%, and 0.9555 for the Indian Pines scene, 97.02%, 96.64%, and 0.9605 for the Pavia University scene, and 95.55%, 97.97%, and 0.9504 for the Salinas scene, respectively. Moreover, the proposed WSSJKCRC method obtains the promising accuracy with OA over 95% on the three hyperspectral scenes under the situation of small-scale labeled samples, thus effectively reducing the labeling cost for HSI.

德国乡村内生型发展战略对优化乡村地区土地资源利用、提升居民生活质量、促进社会经济发展有着积极作用,为全球乡村发展提供了借鉴经验.本文通过文献综述、历史研究等方式对德国乡村内生型发展战略的背景与意义、流程管控、实施成效进行系统分析,挖掘该战略对于乡村整体发展与建设的推动作用.研究发现,德国在推进乡村内生型发展战略过程中始终以可持续发展为核心理念,建立了完善的法律体系和协调统一的合作模式,创建了优势互补、激发主体参与的条件与环境,并利用现代化技术手段实现精准和高效的规划管理.未来中国可以从四方面入手不断优化乡村规划建设与管理:一是强化可持续发展理念的认识;二是建立完整的法律法规与规则制度;三是推动跨部门、跨区域合作,动员乡村居民参与;四是采用数字化、智能化工具.

开展农民生计研究对解决农民贫困、实现乡村融合发展具有重要意义.在阐述可持续生计理论形成和发展过程的基础上,介绍了英国国际发展署(DFID)可持续生计框架的结构和内容,从贫困与生计脆弱性、资源保护政策与可持续生计、生计活动引入与可持续生计、土地政策与可持续生计等4个方面总结了国内可持续生计研究的主要内容,分析了在理论框架适用性、生计目标有效性、政策条件稳定性等方面存在的问题.最后,提出了未来研究的重点方向,即分析框架的扩展、结合宏观背景的可持续生计分析、宅基地制度改革与可持续生计等.

提取UN Comtrade数据库HS二位编码的中国南非农产品贸易数据,在分析数据基本特征的基础上,采用显性比较优势指数、贸易互补性指数和贸易强度指数,从竞争性、互补性和增长潜力3个方面对2005—2019年中国与南非的农产品贸易发展进行了分析.结果显示:中国与南非贸易往来的主要农产品大部分属于本国的出口优势类农产品,并且各自的优势农产品都具有鲜明的本国特色,竞争性明显;中国与南非双方农产品贸易具有较强的互补性,并且总体上双方的互补性都随着时间的增加而逐渐增强;中国与南非在较多类别的农产品上存在紧密的贸易联系,两国之间的农产品贸易存在较大发展潜力.基于此,提出了促进中国和南非农产品贸易发展的对策建议.

The monitoring of ethylene is of great importance to fruit and vegetable quality, yet routine techniques rely on manual and complex operation. Herein, a chemiresistive ethylene sensor based on reduced graphene oxide (rGO)/tungsten diselenide (WSe2)/Pd heterojunctions was designed for room-temperature (RT) ethylene detection. The sensor exhibited high sensitivity and quick p-type response/recovery (33/13 s) to 10–100 ppm ethylene at RT, and full reversibility and excellent selectivity to ethylene were also achieved. Such excellent ethylene sensing behaviors could be attributed to the synergistic effects of ethylene adsorption abilities derived from the negative adsorption energy and the promoted electron transfer across the WSe2/Pd and rGO/WSe2 interfaces through band energy alignment. Furthermore, its application feasibility to banana ripeness detection was verified by comparison with routine technique through simulation experiments. This work provides a feasible methodology toward designing and fabricating RT ethylene sensors, and may greatly push forward the development of modernized intelligent agriculture.

[目的]将社会资本理论引人农户宅基地退出意愿问题研究,探究社会资本及各构成维度对农户宅基地退出意愿的影响和作用机制,以期为提升农户意愿提供参考.[方法]基于苏北地区沛县和丰县的共411份农户调研数据,构建农户社会资本测度指标体系,并运用Logit模型、中介效应模型实证研究了社会资本对农户宅基地退出意愿的影响及抗险能力的中介效应.[结果](1)社会资本对农户的宅基地退出意愿具有显著的正向影响,社会资本越丰富的农户宅基地退出意愿越强;(2)各构成维度均显著影响农户宅基地退出意愿,且影响方向均为正,效用强度从大到小依次为社会规范(0.116)、社会网络(0.065)、社会参与(0.064)、社会信任(0.032)、社会声望(0.031);(3)抗险能力在社会资本和各维度对农户宅基地退出意愿的影响过程中均具有中介效应,社会资本和各维度不仅可以直接影响农户的宅基地退出意愿,还可以通过抗险能力产生间接影响,但影响以直接效应为主.[结论]社会资本和抗险能力是影响农户宅基地退出意愿的重要因素,政府部门在宅基地退出工作推进中,应立足农户的社会属性,加大农户社会资本培育力度,完善风险分担机制,提高农户收入水平.

The Chinese government has implemented a homestead withdrawal policy to improve the efficiency of rural construction land use. The compensation for rural homestead withdrawal (CRHW) is crucial to the reconstruction and sustainable development of farmers’ livelihoods. This paper analyzed the response mechanisms of farmers’ livelihoods to the CRHW with the combined application of the logistic regression, the mediation effect model, and the moderating effect model. The results indicated that CRHW had a significant positive impact on the sustainable livelihoods of rural households, mainly by improving the physical capital and social capital. In addition, adaptability and livelihood diversity played intermediary and regulatory roles in the positive impacts of the CRHW on sustainable livelihoods, respectively. The conclusions may provide insight into the demand for more reasonable compensation policies to ensure the sustainability of farmers’ livelihoods.