2026-01-23
Eliseo Roma, Lorenzo Leso, Santo Orlando, Mariangela Vallone, Matteo Barbari
Access to outdoor space for dry cows is strongly influenced by environmental conditions, feed availability, and social dynamics within the barn. Understanding the interplay among these factors can support precision herd management strategies aimed at enhancing cow welfare and health. The aim of this study was to assess how environmental conditions, grazing forage availability and social dynamics influence the behavior of a dairy cow herd, based on their entry and exit from outdoor spaces monitored using the new technologies. The experiment was conducted for 154 days on a commercial dairy farm located in Mantua (Italy) with 35 Holstein cows during the dry period. The availability of outdoor forage was monitored by mowing on several test plots, while meteorological conditions were monitored using a weather station. The cows were equipped with radio frequency identification (RFID) tags to record the free access the pasture. Social interactions were monitored based on the temporal association of cow passages through the gate. To identify a leader-follower relationship, a k-means algorithm was applied to log the frequency of time intervals between successive passages. Correlation analysis results between the number of daily passes and rumination and feeding time revealed strong correlations, with R² values of 0.71*** and 0.65***, respectively. Temperature-humidity index (THI) was the most influent parameter, while RH, solar radiation and rainfall appeared to have less significant impact. However, wind speed during the day and humidity at night has greatest negative influence. The dynamic social network analysis (SNA) showed that the 42% of passages were associated with a leader-follower relationship and three key findings were observed: i) the cow established preferential relationships with specific members of the herd independently of animal's status (heifer or dry), age, and number of calvings; ii) some individuals are more skilled at establishing connections, while others tend to be more solitary; iii) when a significant bond is lost, it is often replaced by another.
2026-01-23
Jiaxing Xie, Yazhong Chen, Shuai Zhao, Jiaxin Wang, Liye Chen, Fan Luo, Zihao Chen, Weixian Sha, Peng Gao, Weixing Wang, Hongshan Liu
This study developed a rapid method for monitoring water stress in tea trees using unmanned aerial vehicle (UAV) multispectral imaging combined with machine learning models. The study analyzed tea trees subjected to five different irrigation treatments. UAV multispectral imaging and ground-based measurements were employed to explore the relationships between canopy-air temperature differences, soil moisture content, spectral reflectance, vegetation indices, and canopy temperature. The ordering points to identify the clustering structure (OPTICS) algorithm was used to cluster soil moisture content and canopy-air temperature difference data. The clustering process determined the radius parameter (ϵ) from an ordered decision graph, establishing the upper and lower limits of the canopy-air temperature difference. These limits were then correlated with the vapor pressure deficit (VPD) lower limit equation for fully irrigated tea trees to validate the effectiveness of the OPTICS algorithm. Several machine learning regression models, including random forest (RF), support vector regression (SVR), K-nearest neighbors (KNR), and backpropagation (BP) neural networks, were used to create an inversion model for predicting canopy temperature based on vegetation indices. A strong correlation (coefficient of 0.74) was observed between the canopy-air temperature difference bounds derived from OPTICS clustering and the crop water stress index (CWSI) values calculated from empirical model data. Spectral reflectance at 450 nm, 560 nm, and 650 nm remained stable, while reflectance at 730 nm and 840 nm increased significantly with rising canopy temperature. The RF model demonstrated robust performance, achieving an R² value greater than 0.8, and effectively predicted canopy temperature using vegetation indices. By combining density-based OPTICS clustering to establish water stress index limits and leveraging vegetation indices for canopy temperature inversion, this study presents a rapid and accurate method for calculating the water stress index in tea trees. The findings provide a foundation for UAV-based remote sensing applications in monitoring tea tree water stress and highlight potential advancements in precision irrigation management.
2025-12-18
Bowen Sun, Xiangchao Kong, Xingkai Peng, Changxi Chen, Wanchao Zhang, Guangyu Zhao, Yongmin Guo, Kaisi Yang
This study investigates the acoustic characteristics of normal and abnormal calls in white-feathered broilers to propose a method for early detection of non-healthy conditions. Vocalizations were collected from 2-week-old broilers over a 21-day period and analyzed using time-domain and frequency-domain features, including maximum amplitude, effective amplitude, fundamental frequency, and pulse index. Significant differences were identified between normal calls and abnormal calls influenced by laryngeal mucus, with support vector machines and random forest classifiers achieving accuracies of 97.8% and 98.76%, respectively. Unlike previous empirical feature aggregation methods, this research employs statistically validated feature selection aligned with physiological mechanisms, enhancing interpretability and performance. The proposed framework offers a practical, automated solution for on-farm monitoring of broiler vocalizations, contributing to early detection of abnormal signs and improved management in precision poultry farming.
2025-12-15
Lale Ghanizadeh Hesar, Ahmet Ince, Kamil Ekinci, Huseyin Gucum
Liquid manure is a rich source of nutrients for crops, but when applied by traditional methods (broadcast) it causes loss of important nutrients such as nitrogen by evaporation and leaching simultaneously causes environmental and groundwater pollution. In this study, a high-performance liquid manure injection tool (prototype) was designed, developed and evaluated under actual field conditions. The prototype injection system consists of a prototype liquid distributor wheel that injects the liquid subsurface at certain amounts and intervals without cultivating the soil. Since the liquid manure is injected subsurface, so does not remain on the soil surface, it does not need to be mixed, and alternative to other methods in terms of nitrogen loss and availability to the plant. Laboratory and field studies were conducted to explore this system in liquid manure application. The range of slurry application rates was 4000-20000 L ha -1 , at the base of system pressure and forward speed. The trials to determine efficiency of system, image analysis methods were used to quantify the percentage of the surface area covered with manure, and ammonia emission rate were determined by employing a wind tunnel and a dynamic chamber. The results showed that injecting slurry reduced NH 3 emissions most effectively to 70% compared to the traditional surface spreading method. No statistically significant effect of manure application depth on ammonia emission was observed. The manure cover decreased at used by injection system. The machine demonstrated its performance by successfully injecting liquid manure into the soil and preventing nitrogen losses since the fertilizer had minimal contact with the air.
2025-12-15
Shuzhen Yang, Jiahong Du, Dongjian Zhang, sangsang Li
In the intelligent cultivation of mushrooms within greenhouses, monitoring during the blooming period is crucial. This stage involves the formation and differentiation of young fruiting bodies, where timely detection of mushroom pileus is essential for automated environmental control. However, accurately detecting and counting immature caps remains challenging due to their small size, similar morphology, dense clustering, and complex background interference in greenhouse environments. To address these issues, this paper proposes an improved detection system, named SCS-YOLO11, based on the YOLOv11s architecture. To address the challenges of small-target detection in mushroom pileus recognition, we propose a coupled multi-scale attention (CMCA) module that effectively integrates global context and multi-scale spatial features. Additionally, a lightweight SPConv module is introduced to reduce computational cost while maintaining feature expressiveness, and a compact spatial-channel attention module (SCAM) further enhances feature discrimination in the detection head. It jointly models spatial and channel attention to guide the model to focus on key mushroom cap regions across multi-scale feature maps. Compared with the baseline YOLO11s model, SCS-YOLO11s shows remarkable improvements. Its precision increases from 79% to 84%, and mAP rises from 74.6% to 79%, with only 2.13M parameters and 3.6G FLOPs, demonstrating high efficiency. When applied to mushroom datasets, experiments show that its performance surpasses other YOLO-series models. SCS-YOLO11 strikes a balance between detection accuracy and computational efficiency, making it a promising solution for real-time monitoring of small mushroom pileus in the complex and dynamic settings of greenhouse mushroom cultivation.
2025-12-01
Qinghua Ren, Yanlin Wu, Qingshuai Zeng, Ning Yang
Agricultural remote sensing image segmentation, which involves classifying each pixel of an image into a specific category, has recently been driven by deep learning methods due to their powerful feature extraction capabilities. This paper presents a systematic review of deep learning-based image segmentation techniques for agricultural remote sensing, along with an overview of current challenges and emerging research trends. First, it outlines the characteristics of agricultural remote sensing tasks and the requirements for remote sensing image acquisition and processing, providing an in-depth analysis of the nature of agricultural remote sensing data. Next, it systematically reviews the evolution of deep learning-based methods, with a focus on summarizing segmentation network architectures, including convolution-based models, transformer-based models, hybrid architectures, lightweight models, and vision-language models. Moreover, it discusses several deep learning paradigms designed for annotation-efficient scenarios, including semi-supervised, weakly supervised, self-supervised, and transfer learning. Then, it offers an in-depth analysis of key challenges, such as data annotation, computational cost, and model generalization. Finally, it summarizes the latest advances in deep learning for agricultural remote sensing image segmentation and outlines potential future research directions, aiming to provide technical references that promote the practical application and successful deployment of deep learning in this critical domain.
2025-11-04
Antonio Garcia-Chica, Angel Mariano Rodriguez-Perez, Rosa Maria Chica, Julio Jose Caparros Mancera, Cesar Antonio Rodriguez Gonzalez
This study explores the design and adaptation of a shipping container into a portable irrigation control station for agricultural operations. The project leverages the structural durability and mobility of containers to offer a versatile and sustainable solution for irrigation management. By integrating irrigation equipment, control systems, and energy storage, this unit provides an efficient and cost-effective alternative to traditional irrigation stations. A key advantage of this innovation is its mobility, allowing the container to be easily relocated between farms using a crane truck. This feature optimizes its use in seasonal crop rotations and in agricultural operations spread across different locations. The system operates autonomously, harnessing photovoltaic solar energy stored in batteries, thereby eliminating reliance on fossil fuels and significantly reducing the environmental impact of agricultural irrigation. The system was designed to irrigate 4 hectares, with a pump flow rate of 26 L/s, a total power load of 3.47 kW, and the capacity to supply a crop area of up to 4 ha under typical operating conditions. Beyond its operational efficiency, the study emphasizes the environmental benefits of repurposing shipping containers, contributing to waste reduction and mitigating ecological degradation. This approach aligns with sustainability principles in agriculture, promoting the responsible and efficient use of water and energy resources in decentralized irrigation systems.
2025-10-20
Yerim Jo, Sangik Lee, Byung-hun Seo, Jong-hyuk Lee, Dongsu Kim, Yejin Seo, Dongwoo Kim, Jimin Shim, Won Choi
Single-layer perforated air ducts made of plastic films are widely used in greenhouses to control the root-zone environment of crops. However, conventional ducts often exhibit non-uniform airflow and thermal distributions along the duct length, making it difficult to maintain consistent environmental conditions in the greenhouse. To address this issue, a double-layer perforated air duct has been developed and implemented in greenhouses. However, it is necessary to quantitatively evaluate its effectiveness in improving environmental uniformity. In this study, computational fluid dynamics (CFD) simulations were conducted to compare the internal airflow and jet flow characteristics between the conventional single-layer duct and the proposed double-layer duct. In addition, three double-layer duct designs with different hole arrangements, sizes, and spacings were analyzed. The double-layer duct significantly improved the uniformity of the jet flow temperature and mass flow rate compared with the single-layer duct. The space between the inner and outer tubes in the double-layer duct acted as both a thermal insulation layer and a pressure chamber, maintaining a high, uniform internal static pressure and a low, consistent air velocity. The maximum improvement in temperature uniformity was 75%, and that in mass flow rate was 42%. The proposed double-layer perforated air duct can contribute to enhanced environmental uniformity in greenhouses by supplying jet flows through its holes at a more consistent temperature and mass flow rate along the duct length.