Cover illustration
The cover image vividly captures the essence of smart agriculture and unmanned farming, where a hand holding a tablet acts as the command center, displaying real-time data and images for monitoring and managing the farm. Farmers can observe plant growth and remotely control water and fertilizer systems in vertical farms via the tablet, ensuring optimal resource management. Drones equipped with sensors and cameras fly over fields, collecting data on crops, soil, and weather, w [Detail] ...
Download coverWith the development of smart agriculture, accurately identifying crop diseases through visual recognition techniques instead of by eye has been a significant challenge. This study focused on apple leaf disease, which is closely related to the final yield of apples. A multiscale fusion dense network combined with an efficient multiscale attention (EMA) mechanism called Incept_EMA_DenseNet was developed to better identify eight complex apple leaf disease images. Incept_EMA_DenseNet consists of three crucial parts: the inception module, which substituted the convolution layer with multiscale fusion methods in the shallow feature extraction layer; the EMA mechanism, which is used for obtaining appropriate weights of different dense blocks; and the improved DenseNet based on DenseNet_121. Specifically, to find appropriate multiscale fusion methods, the residual module and inception module were compared to determine the performance of each technique, and Incept_EMA_DenseNet achieved an accuracy of 95.38%. Second, this work used three attention mechanisms, and the efficient multiscale attention mechanism obtained the best performance. Third, the convolution layers and bottlenecks were modified without performance degradation, reducing half of the computational load compared with the original models. Incept_EMA_DenseNet, as proposed in this paper, has an accuracy of 96.76%, being 2.93%, 3.44%, and 4.16% better than Resnet50, DenseNet_121 and GoogLeNet, respectively, proved to be reliable and beneficial, and can effectively and conveniently assist apple growers with leaf disease identification in the field.
In response to the demand of automatic fruit identification and harvesting, this paper presents a human-robot collaborative picking robot based on somatosensory interactive servo control. The robot system mainly consisted of four parts: picking execution mechanism, hand information acquisition system, human-machine interaction interface, and human-robot collaborative picking strategy. A six-degree-of-freedom robotic arm was designed as the picking execution mechanism. The D-H method is employed for both forward and inverse kinematic modeling of the robotic arm. A four-step inverse kinematic optimal solution selection method, including mechanical interference, correctness, rationality, and smoothness of motion, is proposed. The working principle and use of the Leap Motion controller for hand information acquisition were introduced. Data from three types of hand movements were collected and analyzed. Spatial mapping method between the Leap Motion interaction space and operating range of the robotic arm was proposed to achieve a direct correspondence between the cubic interaction box and the cylindrical space of the fan ring of the robotic arm. The test results demonstrated that the average response time of the double-click picking command was 332 ms. The average time consumption for somatosensory control targeting was 6.5 s. The accuracy rate of the picking gesture judgment was 96.7%.
The futures trading market is an important part of the financial markets and soybeans are one of the most strategically important crops in the world. How to predict soybean future price is a challenging topic being studied by many researchers. This paper proposes a novel hybrid soybean future price prediction model which includes two stages of data preprocessing and deep learning prediction. In the data preprocessing stage, futures price series are decomposed into subsequences using the ICEEMDAN (improved complete ensemble empirical mode decomposition with adaptive noise) method. The Lempel-Ziv complexity determination method was then used to identify and reconstruct high-frequency subsequences. Finally, the high frequency component is decomposed secondarily using variational mode decomposition optimized by beluga whale optimization algorithm. In the deep learning prediction stage, a deep extreme learning machine optimized by the sparrow search algorithm was used to obtain the prediction results of all subseries and reconstructs them to obtain the final soybean future price prediction results. Based on the experimental results of soybean future price markets in China, Italy, and the United States, it was found that the hybrid method proposed provides superior performance in terms of prediction accuracy and robustness.
Leaf blast is a significant global problem, severely affecting rice quality and yield, making swift, non-invasive detection crucial for effective field management. This study used hyperspectral remote sensing technology via an unmanned aerial vehicle to gather spectral data from rice crops. ANOVA and the Relief-F algorithm were used to identify spectral bands sensitive to the disease and developed a new vegetation index, the rice blast index (RBI). This RBI was compared with 30 established vegetation indexes, using correlation analysis and visual comparison to further shortlist six superior indexes, including RBI. These were evaluated using the K-nearest neighbor (KNN) and random forests (RF) classification models. RBI demonstrated superior detection accuracy for leaf blast in both the KNN model (95.0% overall accuracy and 93.8% kappa coefficient) and the RF model (95.1% overall accuracy and 92.5% kappa coefficient). This study highlights the significant potential of RBI as an effective tool for precise leaf blast detection, offering a powerful new mechanism and theoretical basis for enhanced disease management in rice cultivation.
To address the dual challenges of excessive energy consumption and operational inefficiency inherent in the reliance of current agricultural machinery on direct supervision, this study developed an enhanced YOLOv8n-SS pedestrian detection algorithm through architectural modifications to the baseline YOLOv8n framework. The proposed method had superior performance in dense agricultural contexts while improving detection capabilities for pedestrian distribution patterns under complex farmland conditions, including variable lighting and mechanical occlusions. The main innovations were: (1) integration of spatial pyramid dilated (SPD) operations with conventional convolution layers to construct SPD-Conv modules, which effectively mitigated feature information loss while enhancing small-target detection accuracy; (2) incorporation of selective kernel attention mechanisms to enable context-aware feature selection and adaptive feature extraction. Experimental validation revealed significant performance improvements over the original YOLOv8n model. This enhanced architecture achieved 7.2% and 9.2% increases in mAP0.5 and mAP0.5:0.95 metrics respectively for dense pedestrian detection, with corresponding improvements of 7.6% and 8.7% observed in actual farmland working environments. The proposed method ultimately provides a computationally efficient and robust intelligent monitoring solution for agricultural mechanization, facilitating the transition from conventional agricultural practices toward sustainable, low-carbon production paradigms through algorithmic optimization.
This research explored a novel multimodal approach for disease management in cauliflower crops. With the rising challenges in sustainable agriculture, the research focused on a patch spraying method to control disease and reduce crop losses and environmental impact. For non-destructive disease assessment, a spectral sensor was used to collect spectral information from diseased and healthy cauliflower parts. The spectral data sets were analyzed using decision tree and support vector machine (SVM) algorithms to identify the most accurate model for distinguishing diseased and healthy plants. The chosen model was integrated with a low-volume sprayer (50‒150 L·ha‒1), equipped with an electronic control unit for targeted spraying based on sensor-detected regions. The decision tree model achieved 89.9% testing accuracy, while the SVM model achieved 96.7% accuracy using hyperparameters: cost of 10.0 and tolerance of 0.001. The research successfully demonstrated the integration of spectral sensors, machine learning, and targeted spraying technology for precise input application. Additionally, the optimized sprayer achieved a 72.5% reduction in chemical usage and a significant time-saving of 21.0% compared to a standard sprayer for black rot-infested crops. These findings highlight the potential efficiency and resource conservation benefits of innovative sprayer technology in precision agriculture and disease management.
Artificial intelligence-based automatic systems can reduce time, human error and post-harvest operations. By using such systems, food items can be successfully classified and graded based on defects. For this context, a machine vision system was developed for fruit grading based on defects. The prototype consisted of defective fruit detection and mechanical sorting systems. Image processing algorithms and deep learning frameworks were used for detection of defective fruit. Different image processing algorithms including pre-processing, thresholding, morphological and bitwise operations combined with a deep leaning algorithm, i.e., convolutional neural network (CNN), were applied to fruit images for the detection of defective fruit. The data set used for training CNN model consisted of fruit images collected from a publicly-available data set and captured fruit images: 1799 and 1017 for mangoes and tomatoes, respectively. Subsequent to defective fruit detection, the information obtained was communicated to microcontroller that further actuated the mechanical sorting system accordingly. In addition, the system was evaluated experimentally in terms of detection accuracy, sorting accuracy and computational time. For the image processing algorithms scheme, the detection accuracy for mango and tomato was 89% and 92%, respectively, and for CNN architecture used, the validation accuracy for mangoes and tomatoes was 95% and 94%, respectively.
Weed management is a crucial aspect of modern agriculture as invasive plants can negatively impact crop yields and profitability. Long-established methods of weed control, such as manual labor and synthetic herbicides, have been widely used but come with their own set of challenges. These methods are often time-consuming, labor-intensive, and pose environmental risks. Herbicides have been the primary method of weed control due to their efficiency and cost-effectiveness. However, over-reliance on herbicides has led to environmental contamination, weed resistance, and potential health hazards. To address these issues, researchers and industry experts are now exploring the integration of machine learning into chemical weed management strategies. As technology advances, there is a growing interest in exploring innovative and sustainable weed management approaches. This review examines the potential of machine learning in chemical weed management. Machine learning offers innovative and sustainable approaches by analyzing large data sets, recognizing patterns, and making accurate predictions. Machine learning models can classify weed species and optimize herbicide usage. Real-time monitoring enables timely intervention, preventing invasive species spread. Integrating machine learning into chemical weed management holds promise for enhancing agricultural practices, reducing herbicide usage and minimizing environmental impact. Validation and refinement of these algorithms are needed for practical application.
In the context of food security, drying is a crucial postharvest process for paddy grain because it significantly impacts the quality of both paddy and rice. To conserve energy during the drying process, deep bed dryers are used as convective dryers that use a combination of ambient airflow and heating, thus relying on airflow, temperature, and relative humidity (RH) as the primary drying parameters. Consequently, an aeration system is necessary so that the drying air can penetrate the thick pile of paddy grain and distribute evenly throughout the drying chamber. This analysis aimed to determine the most optimal aeration system by using computational fluid dynamics (CFD) and the AHP-TOPSIS method. The quantitative and visual analysis of the airflow velocity, pressure, temperature, and RH was conducted using CFD on four different dryer aeration systems models, which were then ranked by preference value using the AHP-TOPSIS method. Model 4, with a sloping floor and circular pipe formation, was found to have the most optimal aeration system (preference value of 0.788) for a paddy grain deep bed dryer prototype.
Mulching practices substantially affect soil CO2 emissions from agricultural ecosystems. However, the impacts of mulching practices and their enduring effects on soil CO2 fluxes in humid plantations have not been investigated. To address this research gap, a field experiment was conducted in a Moso bamboo (Phyllostachys edulis) plantation in a humid area of China to investigate the effects of various durations of straw mulching and its enduring effects on soil CO2 fluxes and soil organic carbon (SOC). Straw mulching significantly increased the soil CO2 flux by about 18 times relative to the control, mainly due to the increase in soil temperature during the mulching stage. During the period of enduing effect, straw mulching still significantly increased the soil CO2 flux by 230%–270% relative to the control, primarily due to the enhancement of microbial activity resulting from improved soil nutrient contents, demonstrating that straw mulching had an enduring positive impact on soil CO2 flux. Additionally, straw mulching significantly increased SOC by 27%–72% during the mulching and period of enduing effect. These results indicated that straw mulching in plantations in humid regions could be a potential carbon storage strategy by increasing soil carbon content.
Soil erosion models are effective tools for assessing soil erosion indicators and simulating erosion processes. China has some of the most severe soil erosion in the world. To better apply soil erosion models to address soil erosion issues, it is necessary to understand the development process and current status of soil erosion modeling research in China. In this study, a combination of bibliometric analysis and statistical methods was used to review and organize Chinese soil erosion models (1982–2022) from various perspectives, including keywords, model operations, model classification, model spatiotemporal scales, and model geographical applications. This findings of this analysis indicate that the study of soil erosion models in China mainly focuses on large scales (regional and large river basins) using empirical models including USLE, RUSLE, and CSLE. The research areas are primarily concentrated in southeastern and central China. The research content has gradually shifted from studying soil erosion characteristics to analyzing influencing factors, spatiotemporal evolution of erosion, and erosion process and morphology stages. However, there are several issues in current Chinese soil erosion modeling research. These include a lack of validation of model application results with field measurements, limited application areas for the models, and relatively weak research on erosion process mechanisms. On this basis, it is recommended that future research should increase the observation of soil erosion processes and establish methods for data or mathematical formula conversion based on different geographical environments. Also, there is a need to strengthen research on erosion process mechanisms. The findings of this study should provide a valuable resource for researchers to future understand the development process and current issues of Chinese soil erosion models, providing insights for future research directions.
Chickens are one of the most important domesticated animals, serving as an important protein source. Studying genetic variations in chickens to enhance their production performance is of great potential value. The emergence of next-generation sequencing has enabled precise analysis of single nucleotide polymorphisms and insertions/deletions in chicken, while third-generation sequencing achieves the accurate structural variant identification. However, the high cost of third-generation sequencing technology limits its application in population studies. The graph-based pan-genome strategy can overcome this challenge by enabling the detection of structural variations using cost-effective next-generation sequencing data. This study constructed a graph-based pan-genome for chickens using 12 high-quality genomes. This pan-genome used linear genome GRCg6a as the reference genome, containing variant information from two commercial and nine native chicken breeds. Compared to the linear genome, the pan-genome provided significant improvements in the efficiency of structural variation identification. On the basis of the graph-based pan-genome, high-frequency structural variations related to high egg production in Leghorn chicken were predicted. Additionally, it was discovered that potential structural variations was associated with highland adaptation in Tibetan chickens according to next-generation sequencing and transcriptomics data. Using the pan-genome graph, a new strategy to identify structural variations related to traits of interest in chickens is presented.
Xinjiang, one of China’s most water-scarce provinces, produces 25% of the world’s cotton. However, changes in water consumption of cotton production in Xinjiang under two climate change scenarios is unclear. This study considered three irrigation techniques (i.e., furrow, micro (drip) and sprinkler irrigation) and simulated the blue and green water footprints of cotton production in Xinjiang at a 5-arcmin grid level in response to climate change scenarios in the 2050s and 2090s. Taking the period 2000–2018 as the baseline, results showed that this footprint of cotton in Xinjiang for the baseline period was 4264 m3·t–1, with blue water accounting for 83%. Under climate change scenarios, Xinjiang was predicted to have an increasing drought trend and intensifying pressure on water resources. Owing to increased CO2 concentrations, the water footprint of cotton tended to decrease by 19.3% and 35.7% under two Shared Socioeconomic Pathway scenarios—SSP2-4.5, representing a moderate socioeconomic development path with lower emissions, and SSP5-8.5, indicating a scenario of high growth with higher emissions—respectively, for the 2090s. The blue water footprint was predicted to have an overall decrease. However, its proportion of the total would increase slightly, with the highest increase being 3.4%. The green water footprint was also predicted to have decreasing trend, with reductions of 33.7% (SSP2-4.5) and 47.2% (SSP5-8.5), respectively. Of the three irrigation techniques, sprinkler irrigation was predicted to have the greatest water conservation potential, with a reduction of up to 40.1%.
Analyzing the changes in agricultural carbon emissions (ACE) and their influencing factors can provide a sound basis for accurately estimating the carbon balance of agroecosystems. Such analyses can serve as a reference for developing policies to mitigate global climate change and promote sustainable agricultural development. Using the carbon emission calculation framework of the Intergovernmental Panel on Climate Change, this study examined the spatiotemporal characteristics of ACE, including total amount, intensity, structure and their influencing factors, in Fujian Province from 2002 to 2022. The logarithmic mean scale index model and Tapio decoupling model were used, with the GM (1,1) model to forecast carbon emissions from 2023 to 2040. The results indicate that both the total emissions and intensity of ACE had fluctuating downward trends and agricultural material inputs were the largest contributors to ACE. Additionally, total ACE was found to have a spatial pattern higher in the west and lower in the east and agricultural production efficiency was the primary factor in reducing ACE. ACE was clearly decoupled from economic development and is projected to continually decline after 2023.
The macropores of biochar provide a suitable habitat for microbial growth, and its high carbon content serves as an energy source for beneficial microbes. This study evaluated the potential of biochar as a carrier for Trichoderma in managing Sclerotinia sclerotiorum in chickpeas. Biochar application reduced plant disease severity by 36.5% and increased plant root mass by 23.3%. For this, three types of biochar, wheat straw, organic kitchen waste, and hardwood were tested with Trichoderma, analyzing such as organic C, total N, P, K, Mg, and Ca; pH, and ash content. Trichoderma populations were monitored with biochar carrier of different mesh sizes (250, 150, 75, and 45 µm) for up to 6 weeks after inoculation. Hardwood biochar at 150 µm supported the highest Trichoderma population, reaching 33.5 × 105 CFU·g−1 after 6 weeks. Hardwood biochar also achieved the maximum disease suppression compared to other biochar types. This research highlights the dual role of biochar in enhancing plant growth and controlling disease, contributing to the standardization of biochar use in agricultural practices.
Crop rotation is a beneficial and sustainable agricultural practice that facilitates increased opportunities for smallholders. This study investigated the impact of eight commonly used crop rotations in China on soil properties and microbial communities. The faba bean (Vicia faba) rotation increased soil water content, total carbon, total nitrogen, total phosphorus and organic carbon content by 29.1%, 40.9%, 55.9%, 18.9%, and 61.6%, respectively, compared to other rotations. The faba bean rotation also exhibited increased soil microbial biomass and soil respiration rates. The effect sizes of the faba bean rotation on soil properties were larger than those of other rotations. The richness and diversity of the microbial community were significantly higher in the faba bean rotation than in other rotations. Desulfobacterota and Planctomycetota had a positive correlation with soil multifunctionality. The faba bean rotation was potentially beneficial to soil fertility and water-use efficiency, creating a favorable niche for microbial growth. With increased microbial activity and potential for nutrient mineralization, legume-microbe interactions had been improved through crop rotation. This resulted in enhancing nutrient cycling efficiency in the faba bean rotation, potentially improving soil properties.
