Maize Leaf Disease Recognition Based on Improved Convolutional Neural Network ShuffleNetV2.

The occurrence of maize diseases is frequent but challenging to manage. Traditional identification methods have low accuracy and complex model structures with numerous parameters, making them difficult to implement on mobile devices. To address these challenges, this paper proposes a corn leaf disease recognition model SNMPF based on convolutional neural network ShuffleNetV2. In the down-sampling module of the ShuffleNet model, the max pooling layer replaces the deep convolutional layer to perform down-sampling. This improvement helps to extract key features from images, reduce the overfitting of the model, and improve the model's generalization ability. In addition, to enhance the model's ability to express features in complex backgrounds, the Sim AM attention mechanism was introduced. This mechanism enables the model to adaptively adjust focus and pay more attention to local discriminative features. The results on a maize disease image dataset demonstrate that the SNMPF model achieves a recognition accuracy of 98.40%, representing a 4.1 percentage point improvement over the original model, while its size is only 1.56 MB. Compared with existing convolutional neural network models such as EfficientNet, MobileViT, EfficientNetV2, RegNet, and DenseNet, this model offers higher accuracy and a more compact size. As a result, it can automatically detect and classify maize leaf diseases under natural field conditions, boasting high-precision recognition capabilities. Its accurate identification results provide scientific guidance for preventing corn leaf disease and promote the development of precision agriculture.

Synthesis of struvite-enriched slow-release fertilizer using magnesium-modified biochar: Desorption and leaching mechanisms.

To improve the retention and slow-release abilities of nitrogen (N) and phosphorus (P), an 82 %-purity struvite fertilizer (MAP-BC) was synthesized using magnesium-modified biochar and a solution with a 2:1 concentration ratio of NH4+ to PO43- at a pH of 8. Batch microscopic characterizations and soil column leaching experiments were conducted to study the retention and slow-release mechanisms and desorption kinetics of MAP-BC. The slow-release mechanism revealed that the dissolution rate of high-purity struvite was the dominant factor of NP slow release. The re-adsorption of NH4+ and PO43- by biochar and unconsumed MgO prolonged slow release. Mg2+ ionized by MgO could react with PO43- released from struvite to form Mg3(PO4)2. The internal biochar exhibited electrostatic attraction and pore restriction towards NH4+, while magnesium modification and nutrient loading formed a physical antioxidant barrier that ensured long-term release. The water diffusion experiment showed a higher cumulative release rate for PO43- compared to NH4+, whereas in soil column leaching, the trend was reversed, suggesting that soil's competitive adsorption facilitated the desorption of NH4+ from MAP-BC. During soil leaching, cumulative release rates of NH4+ and PO43- from chemical fertilizers were 3.55-3.62 times faster than those from MAP-BC. The dynamic test data for NH4+ and PO43- in MAP-BC fitted the Ritger-Peppas model best, predicting release periods of 163 days and 166 days, respectively. The leaching performances showed that MAP-BC reduced leaching solution volume by 5.58 % and significantly increased soil large aggregates content larger than 0.25 mm by 24.25 %. The soil nutrients retention and pH regulation by MAP-BC reduced leaching concentrations of NP. Furthermore, MAP-BC significantly enhanced plant growth, and it is more suitable as a NP source for long-term crops. Therefore, MAP-BC is expected to function as a long-term and slow-release fertilizer with the potential to minimize NP nutrient loss and replace part of quick-acting fertilizer.

Identification of leaf diseases in field crops based on improved ShuffleNetV2.

Rapid and accurate identification and timely protection of crop disease is of great importance for ensuring crop yields. Aiming at the problems of large model parameters of existing crop disease recognition methods and low recognition accuracy in the complex background of the field, we propose a lightweight crop leaf disease recognition model based on improved ShuffleNetV2. First, the repetition number and the number of output channels of the basic module of the ShuffleNetV2 model are redesigned to reduce the model parameters to make the model more lightweight while ensuring the accuracy of the model. Second, the residual structure is introduced in the basic feature extraction module to solve the gradient vanishing problem and enable the model to learn more complex feature representations. Then, parallel paths were added to the mechanism of the efficient channel attention (ECA) module, and the weights of different paths were adaptively updated by learnable parameters, and then the efficient dual channel attention (EDCA) module was proposed, which was embedded into the ShuffleNetV2 to improve the cross-channel interaction capability of the model. Finally, a multi-scale shallow feature extraction module and a multi-scale deep feature extraction module were introduced to improve the model's ability to extract lesions at different scales. Based on the above improvements, a lightweight crop leaf disease recognition model REM-ShuffleNetV2 was proposed. Experiments results show that the accuracy and F1 score of the REM-ShuffleNetV2 model on the self-constructed field crop leaf disease dataset are 96.72% and 96.62%, which are 3.88% and 4.37% higher than that of the ShuffleNetV2 model; and the number of model parameters is 4.40M, which is 9.65% less than that of the original model. Compared with classic networks such as DenseNet121, EfficientNet, and MobileNetV3, the REM-ShuffleNetV2 model not only has higher recognition accuracy but also has fewer model parameters. The REM-ShuffleNetV2 model proposed in this study can achieve accurate identification of crop leaf disease in complex field backgrounds, and the model is small, which is convenient to deploy to the mobile end, and provides a reference for intelligent diagnosis of crop leaf disease.

Performance Evaluation of a Cicada-Inspired Subsoiling Tool Using DEM Simulations.

Subsoiling practice is an essential tillage practice in modern agriculture. Tillage forces and energy consumption during subsoiling are extremely high, which reduces the economic benefits of subsoiling technology. In this paper, a cicada-inspired biomimetic subsoiling tool (CIST) was designed to reduce the draught force during subsoiling. A soil-tool interaction model was developed using EDEM and validated using lab soil bin tests with sandy loam soil. The validated model was used to optimize the CIST and evaluate its performance by comparing it with a conventional chisel subsoiling tool (CCST) at various working depths (250-350 mm) and speeds (0.5-2.5 ms-1). Results showed that both simulated draught force and soil disturbance behaviors agreed well with those from lab soil bin tests, as indicated by relative errors of <6.1%. Compared with the CCST, the draught forces of the CIST can be reduced by 17.7% at various working depths and speeds; the design of the CIST obviously outperforms some previous biomimetic designs with largest draught force reduction of 7.29-12.8%. Soil surface flatness after subsoiling using the CIST was smoother at various depths than using the CCST. Soil loosening efficiencies of the CIST can be raised by 17.37% at various working speeds. Results from this study implied that the developed cicada-inspired subsoiling tool outperforms the conventional chisel subsoiling tool on aspects of soil disturbance behaviors, draught forces, and soil loosening efficiencies. This study can have implications for designing high-performance subsoiling tools with reduced draught forces and energy requirements, especially for the subsoiling tools working under sandy loam soil.

Applicability of hybrid bionic optimization models with kernel-based extreme learning machine algorithm for predicting daily reference evapotranspiration: a case study in arid and semiarid regions, China.

The accurate prediction of daily reference crop evapotranspiration (ETO) enables effective management decision-making for agricultural water resources; this is crucial for developing water-efficient agriculture. To improve the accuracy of ETO forecasts in data-deficient areas, this study uses a decision tree algorithm (classification and regression tree [CART]) to obtain the effects of various factors on ETO at typical stations in arid and semiarid regions of China. A combination of factors with considerable influence on the model was selected as the input for constructing a kernel-extreme-learning-machine (KELM) daily reference evapotranspiration prediction model, and three bionic optimization algorithms (i.e., sparrow search optimization algorithm, Harris Hawks optimization algorithm, and lion swarm optimization algorithm) were integrated to optimize KELM prediction model parameters and improve the accuracy of daily reference evapotranspiration prediction. The results indicate that temperature (maximum or minimum temperature) is the primary factor influencing ETO, and the range of importance is 0.399-0.554. RH and Ra are also key factors influencing ETO; the hybrid model optimized using the bionic optimization algorithm provides advantages over the independent KELM model, and the SSA-KELM model has the highest accuracy among hybrid models, with a root-mean-square error of 0.408-1.964, R2 of 0.545-0.982, mean absolute error of 0.273-1.086, and Nash-Sutcliffe efficiency coefficient of 0.658-0.967. The top five factors extracted using the CART algorithm are recommended as inputs for constructing the SSA-KELM model for ETO estimation in arid and semiarid regions of China, and this model can also serve as a reference for ETO forecasting in similar regions.

Optimization of Controlled Water and Nitrogen Fertigation on Greenhouse Culture of Capsicum annuum.

This study investigated the effects of different combinations of irrigation and nitrogen levels on the growth of greenhouse sweet peppers, assessing yield, quality, water use efficiency (WUE), and partial factor productivity from applied N (PFPN). By using controlled drip irrigation, the optimal conditions for efficient, large-scale, high-yield, and high quality production of sweet peppers in Northwest China were determined. Using the local conventional irrigation and nitrogen regime as a control (105% ET0, N: 300 kg·hm-2), three alternative irrigation levels were also tested, at 90%, 75%, and 60% ET0. These were combined with nitrogen levels at 100%, as the control, and 75%, 50%, and 25%, resulting in 16 combination treatments. The results show that different supplies of water and nitrogen nutrition had a significant impact on the growth, yield, WUE, PFPN, and quality of fruit. The treatments of W0.90N0.75, W0.90N0.50, W0.75N0.75, and W0.75N0.50 can better maintain the "source-sink" relationship of peppers. They increased the economic yield, WUE, and PFPN. A principal component analysis was performed to evaluate indicators of fruit quality, revealing that the treatment of W0.75N0.50 resulted in the best fruit quality. For greenhouse sweet peppers produced in Northwest China, the combination of W0.90N0.75 resulted in the highest economic yield of 34.85 kg·hm-2. The combination of W0.75N0.75 had the highest WUE of 16.50 kg·m-3. The W0.75N0.50 combination treatment had the highest fruit quality score. For sustainable ecological development and in view of limited water resources in the area, we recommend the W0.75N0.50 combination treatment, since it could obtain the optimal fruit quality, while its economic yield and WUE were 9% and 4% less than the maximum, respectively. This study provides a theoretical basis for the optimal management of water and nitrogen during production of greenhouse sweet peppers in Northwest China.