In-field hybrid rice nursing seedling, because of the low sowing rates, the existence of insufficient seed strip forming and seed uniformity in trays cause the problem of poor seedling quality and low planting uniformity. To solve the problem, this study presents a furrow opener with press roller for precision drill nursing seedling of hybrid rice. Based on the principle of rolling trenching and analysis of the interaction law between the rollers and subsoil of the seedling tray, ideal seed trenches suitable for hybrid rice precision seedling seeding was constructed. Coupled transplanting experiment was conducted to test the ditching effect. The specific structure and working principle of the furrow opener with press roller are described. The kinetics analysis among the press roller, load adjusting mechanism, seedling tray and subsoil, as well as the soil subsidence calibration test were conducted. The structure and parameters of the key components of furrow opener with press roller are designed and optimized. Discrete element simulation test and the Box-Behnken test method are conducted, the key parameters of the press roller was optimized using multi-objective optimization. Then, verification test is conducted to obtain the key parameters of the press roller with 100 mm of the diameter of the press roller, 45° of the V-shaped tooth angle, and 9mm of the ditch depth. Meanwhile, the verification tests show that the uniform qualified rate was 85.68% and strip forming qualified rate was 87.18%, which is consistent with the optimization results. The field experiments showed that compared with no-ditching drill sowing, the uniform qualified rate and the strip forming qualified rate of the ditching drill sowing increased by 5.78% and 17.17 %, respectively. The highest uniform qualified rate and strip forming qualified rate achieved 86.80% and 87.60%, respectively. According to the coefficient of variation of seedling quality, the consistency of seedlings was significantly improved. In-field rice mechanized transplanting, the missing plant rate of the ditching drill sowing decreased by 1.95%, the rate of 1-3 plants per pot increased by an average of 7.14%, and the root damage rate decreased by an average of 2.27%, compared with the no-ditching drill sowing. When the seeding density is 50, 60, and 70 g/tray, the theoretical yields of the ditching drill sowing were 234.90, 245.55 and 57.15 kg/hm² higher than that of no-ditching drill sowing, which increased by 3.12%, 3.53% and 0.80%, respectively. When the seeding density is 50, 60, and 70 g/tray, the actual yields of the ditching drill sowing were 381.60, 365.85 and 180.15 kg/hm² higher than that of the same sowing density with no-ditching drill sowing, with increased by 5.00%, 5.33% and 2.52%, respectively. The results of the study show that the furrow opener with press roller performs well and can effectively improve the technical level of precision strip seeding and precision coupling transplanting. The results of the study can provide a reference for the research and development of hybrid rice precision strip seeding equipment.

High single-seeding rate and low missed-seeding rate are often required in the process of plug seedling seeding for leafy vegetables. Alternatively, the architecture and substantial parameters of YOLO11 can represent an advanced image detection framework. The remarkable accuracy can be expected to detect the diverse targets in the high-resolution and complex scenes. However, the high precision of the YOLO11 is also confined to the serious computational demands and long inference times, thus restricting its practical deployment in resource-constrained scenarios. In contrast, the lightweight YOLO11n model can be expected to reduce the computational complexity and parameter redundancy for the competitive performance after architectural optimizations. This study aims to propose an improved lightweight model (named Seed-YOLO) using You Only Look Once 11 nano (YOLO11n). The seeding performance was also detected for the three types of leafy vegetable seeds in the plug seedling trays. The model was then deployed on the edge computing device (NVIDIA Jetson Xavier NX). An efficient detection system was developed for the high-performance plug seedling seeding. Four components were utilized to improve the Seed-YOLO model. 1) A Context Anchor Attention (CAA) module was introduced into the backbone network to construct the C2PSA_CAA module. The feature representation of the seed center region was precisely enhanced to capture seed characteristics. The CAA module was a specific network structure to capture the long-range contextual information. Statistical features of the local regions were extracted after average pooling operations, and then strengthened using 1×1 convolutions, thereby enhancing the feature representation of seed central areas. The horizontal (1×11) and vertical (11×1) depth-wise separable strip convolutions were adopted to expand the receptive field for efficient computational complexity, similar to large convolution kernels. An attention weight map was generated via a Sigmoid function. The weight was then applied to the original feature map to realize weighted enhancement of the features. 2) Group Shuffle Convolution (GSConv) and GSBottleneck modules were incorporated into the neck network. The C3K2_GS module was then constructed to accelerate the fusion of seed features for the detection accuracy. Among them, the GSConv was a lightweight convolution. After the Shuffle operation, the feature information generated by standard convolution was evenly spread into every part by Depthwise Separable Convolution (DSC) over different channels. The computational complexity and the number of parameters were reduced to maintain the performance. 3) Wise Intersection over Union version 3 (WIoU v3) loss function was adopted to effectively anchor the boxes of average-quality seeds. Thereby, its dynamic non-monotonic focusing mechanism was employed to improve the detection performance. WIoU v3 was often used as a bounding box loss function. The gradient gains of samples were dynamically adjusted with different qualities using a weight factor. While reducing the focus on the high-quality samples, the negative gradients generated by low-quality samples were also mitigated to enhance the overall performance of the model. 4) An XSmall detection head was added to boost the detection accuracy for the small targets of the leafy vegetable seeds. While the original Medium/Large detection heads were removed to reduce the parameter count and size, thus achieving model lightweighting. Experimental results demonstrate that the Seed-YOLO was achieved in a mean average precision at 50% IoU (mAP@0.5) of 96.7% and F1 of 93.79% for the seeding performance of the three leafy vegetable seeds, indicating the improvements of 5.4 and 8.87 percentage points, compared with the YOLO11n’s 91.3% and 84.92%, respectively. Notably, the model’s parameter count was reduced to 1.58 million, which was a 38.7% decrease from YOLO11n’s 2.58 million. The model was deployed on the NVIDIA Jetson platform. A graphical user interface was developed in a real-time detection system for the plug seedling seeding. When operating at a seeding rate of 120 trays per hour, the system achieved the accuracy of 99.19% for the single-seed seeding prediction, 94.79% for the reseeding prediction, and 93.43% for the missed seeding prediction, with an average computation time of 121 milliseconds per tray. This finding can also provide valuable support for the detection systems of the plug seedling seeding in leafy vegetables.

Extra-wide film sowing can be expected to strengthen cotton seedlings during production. Among them, the handover row is reduced to increase the effective lighting surface. This study aims to clarify the hydrothermal coupling characteristics of the seed beds for the mechanized sowing with the extra-wide membranes. The seedbeds after mechanized sowing of cotton were taken as the research objects. A systematic investigation was then made on the influencing on the growth and development of cotton. Two treatments were carried out: 4.4-meter extra-wide film (T1) and 2.05-meter membrane (T2). The seed beds were hydrothermally simulated to monitor the cotton growth. The results showed that the HYDRUS model was performed better to simulate the hydrothermal properties of the seed bed after sowing. The simulated values were in the good agreement with the measured ones. The determination coefficient R² was above 0.97, and the mean absolute error (MAE) and root mean square error (RMSE) were low, respectively. The moisture content and temperature of the T1 increased by 9.7% and 5.3%, respectively, compared with the T2. The deep moisture content and temperature increased by 2.8% and 3.6%, respectively, and the water storage and effective accumulated temperature of tillage layer increased by 9.3% and 18%, respectively. The emergence rates of the T1 and T2 were 91.7% and 89.8%, respectively. The better growth was observed in the T1. The plants height and stem diameter of the T1 increased by 6.7% and 11.5%, respectively, during the whole growth period, compared with the T2. The number of the cotton bolls and the weight of a single boll in the T1 were 6.16 and 6.3 g, respectively. The cotton bolls were distributed in the middle and upper layers. While the number of cotton bolls and the weight of a single boll in the T2 were 6.03 and 5.9 g, respectively. The cotton bolls were distributed in the middle and lower layers. The yield of the seed cotton increased by 9.1%. Comprehensive analysis show that the mechanized sowing of the extra-wide film shared a better hydrothermal environment of the seed bed, due to the great variation in the structure of the seed bed. Much more nutrients were gained to rapidly, uniformly, evenly and strongly promote the emergence of the cotton seedlings. The better growth was achieved, according to the indicators of the cotton growth and development. The overall growth period increased by 3-4 days. The number of cotton bolls increased to optimize the distribution structure of the cotton bolls. The yield of seed cotton was also enhanced under the efficient mechanized planting.

This study aims to ensure the high-quality operation requirements of vegetable seedlings in a plant factory. A series of tests were performed on the precision seeding device of the vegetable pot tray with high accuracy. A device was designed and optimized for online miss-seeding detection and intelligent reseeding, according to the pneumatic roller-type device of precision seeding. Furthermore, a programmable logic controller (PLC) was used as the control core, in order to ensure the stability of the control system in the production line. The real-time miss-seeding detection of suction holes was realized to predict the miss-seeding location of the pots. The intelligent fixed-point was timely completed for the precise reseeding of the missed-seeding location pots. The performance of miss-seeding detection was then optimized in the production process, according to the structural characteristics of the pneumatic roller-type seed-metering device. A specific arrangement of photoelectric sensors was used to realize the real-time detection of miss-seeding suction holes. Subsequently, a dynamic reseeding matrix was constructed corresponding to the suction holes and pots of the pot tray, according to the pot's position and number of the pot tray. The miss-seeding location of the pot tray was also predicted. Furthermore, the intelligent reseeding device was designed and optimized to realize intelligent and accurate reseeding at the miss-seeding holes. The miss-seeding location of the pots was extracted from the data from the dynamic reseeding matrix. Taking vegetable seeds named Zhongshuang No.11 as the test materials, the suction holes miss-seeding detection and pots miss-seeding location prediction were carried out on the reliability of miss-seeding detection. The results that the average accuracies of miss-seeding suction holes detection and pots miss-seeding location prediction were 98.82% and 100%, respectively. Furthermore, the Box-Behnken method was used to evaluate the operational performance of the intelligent reseeding device. The relationship between the performance indexes (single seed qualified index, multiple seeding index, and miss-seeding index) and the influencing factors (negative pressure of suction needle, diameter of suction needle, and position pressure of vibrator) was constructed using multi-objective optimization. The optimal working parameters of the intelligent reseeding device were determined to be the negative suction pressure of 10.19 kPa, the diameter of 0.67 mm, and the vibration pressure of 0.07 MPa after a large number of tests and analyses. In this case, the values of working parameters were rounded to facilitate the test. It was found that the average single-seed qualified index was 94.80%, the multiple seeding index was 2.94%, and the miss-seeding index was 2.26%. The intelligent reseeding device under this condition fully met the reseeding requirements of the miss-seeding detection and reseeding device. The performance test of the miss-seeding detection and reseeding device was carried out to verify the model. Once the productivity of the miss-seeding detection and reseeding device was 100 plates/h, the single-seed qualified index of the miss-seeding detection and reseeding device increased to 98.18%, compared with 93.96% before reseeding. When the productivity of the miss-seeding detection and reseeding device was 300 plates/h, the single-seed qualified index of the miss-seeding detection and reseeding device increased to 97.89%, compared with 93.18% before reseeding. The test results fully met the requirements of high precision seeding in vegetable pot seeding devices in plant factory and field conditions. The practical application value was offered to improve the seeding performance of vegetable pot seeding devices. The finding can provide technical support for the production of high-quality vegetable pot seedlings.

Machine transplanting has been the primary technique of rice seedlings in cultivation in recent years. Among them, factory-scale greenhouses have been commonly used for seedling cultivation in northern China. By contrast, paddy-field seedling can be widely used to cultivate rice transplanting in southern China, according to climate, environment, and production costs. However, manual seedlings are limited to the high labor intensity and the waste of chemical fertilizers and pesticides. It is very necessary to improve the irrigation uniformity in the management of water-fertilizer-pesticide in rice paddy fields. In this study, a variable-rate spraying device was designed for water-fertilizer-pesticide irrigation in rice seedlings. A systematic analysis was performed on the large variable-rate sprinkler and the supporting equipment of fertilizer injection at a high-quality level. The overall structure and working principle were clarified for the variable-rate spraying irrigation device of water-fertilizer-pesticide for seedlings; FLUENT software was used to conduct numerical simulation on the sprinkler irrigation variable-diameter pipeline, in order to determine the optimal structural layout of variable-diameter pipeline; The flow rate of sprinkler and the inlet water of main pipe were then determined after experiments. The curved surface model was established for the pressure and electric ball valve opening; A fertilizer injection and dosing mechanism were designed using proportional integral derivative (PID) self-tuning. The real-time equal-proportion fertilizer or medicine injection was realized after optimization. The online control system of fertilizer injection and variable-rate spraying was constructed with the mitsubishi programmable logic controller (PLC) as the control core. The Box-Behnken experiment was adopted to explore the uniform performance and influencing factors of the device for spraying irrigation. The single objective optimization was applied to determine the key parameters of the device for sprinkler irrigation. The optimal combination of parameters was obtained as follows: The inlet water pressure of the main pipe was 0.20 MPa, the electric ball valve opening was 90°, and the nozzle spraying angle was 80°. Christiansen uniformity of the device was 92.69% in this case; There was a linear model between the mass concentration and electrical conductance (EC) value of the potassium chloride fertilizer solution. The variable-rate irrigation and fertilization experiments were conducted on the device. The EC values of the water-fertilizer mixture were 1.65, 1.66, and 1.68 mS/cm, respectively, in the three spraying irrigation levels. The average values of sprinkler intensity were 900.85, 1092.04, and 1263.67 mm/h, respectively. The fertilization uniformity coefficients were 85.21%, 87.86%, and 91.62%, respectively. There was a small fluctuation of EC values for the water-fertilizer mixtures under the three sprinkler irrigation levels of Ⅰ, Ⅱ, and Ⅲ. The average irrigation intensity of sprinklers varied greatly, where the coefficient of uniformity of fertilizer application was higher than 85%. The field test of rice seedlings was carried out using the variable-rate spraying device of water-fertilizer-pesticide for seedlings. The rice varieties were selected as Huahang51 conventional rice and Guang8You165 hybrid rice. The uniformity of seedling growth was higher than 95% at 13, 21, and 28 d after seeding. There were excellent quality indexes and rooting performance of the seedlings. The various quality indicators of rice seedlings fully met the requirements of machine transplanting. The experiment results showed that the variable-rate spraying device of water-fertilizer-pesticide system shared an accurate, real-time, and proportional fertilizer injection, irrigation, and pesticide spraying in rice paddy fields. The finding can provide a practical application to improve the mechanization level of rice paddy field seedlings for the high-quality of seedlings.

Nitrogen is a key nutrient for crop growth. Excessive or insufficient nitrogen affects crop growth, yield, and quality. Additionally, excessive nitrogen fertilizer can cause soil and water pollution. Applying panicle fertilizer during the late jointing stage can promote rice panicle growth. Therefore, accurately and timely monitoring of nitrogen status in rice fields during the late jointing stage and timely optimizing fertilization strategies is crucial for ensuring rice yield and environmental protection. This paper integrates multimodal data from unmanned aerial vehicle (UAV) remote sensing and ground observations to construct inversion models for leaf nitrogen content (LNC) and plant nitrogen content (PNC) of rice at the late jointing stage. The research was conducted at the Shapu Experimental Base of the Agricultural Science Research Institute in Zhaoqing City, Guangdong Province, with two field experiments carried out during the late rice seasons of 2021 and 2022. Each of Experiment 1 (2021) and Experiment 2 (2022) included 30 experimental plots, designed with 5 nitrogen fertilizer gradients, 2 planting densities, and 3 replications. Phosphorus and potassium fertilizers were applied uniformly across all plots. UAVs equipped with multispectral and RGB cameras were used to acquire remote sensing images of rice canopies during the late jointing stage. Vegetation indices (VIs) and texture feature values (TFVs) were extracted from the multispectral images, with TFVs derived using the gray level co-occurrence matrix (GLCM) method. Texture indices (TIs) were then constructed by combining TFVs. RGB images were used to generate digital surface models (DSM) for bare ground (pre-transplant) and rice fields (late jointing stage). These DSMs, combined with ground reference methods, were used to construct crop surface models (CSM) to derive estimated canopy heights (ECH) for each plot. Manually collected data included measured canopy height (MCH) and field nitrogen management data (FN) used as ground observations. For each experimental plot, three representative rice plants were selected as samples. After removing the roots, the leaves and stems were separated and dried at 85 ℃ to a constant weight, which was recorded as the aboveground biomass of the leaves and stems. The true values of leaf nitrogen content and stem nitrogen content were obtained using the Kjeldahl method. Combining these values with the dry weight data, the true values of plant nitrogen content were calculated. The maximal information coefficient (MIC) was used as an evaluation metric for feature assessment and selection. Random forest regression algorithms were employed to construct inversion models for rice LNC and PNC, respectively, using the coefficient of determination (R²), root mean squared error (RMSE), and mean absolute error (MAE) as model evaluation metrics. The analysis and experimental results indicate: TIs constructed using combinations of TFVs significantly enhanced the correlation between texture information and LNC and PNC. When the UAV flight height was 100 m, the Ratio Texture Index constructed using a 9×9 sliding window size in the GLCM method showed the best performance, improving the MIC value by 11.48% compared to the best TFV. For conventional machine-transplanted rice planting density, the correlation between TIs and LNC and PNC was best when the GLCM sliding window size was set to 9×9 or 11×11 at a UAV flight height of 100 m. The ECH derived from the CSM showed a high correlation with the manual MCH in the field. Including canopy height (MCH or ECH) as an input feature in the random forest regression model significantly improved the inversion accuracy of rice nitrogen content. The ECH extracted from the CSM showed high estimation accuracy (R² = 0.77, RMSE = 3.4 cm, MAE = 2.8 cm). The inclusion of canopy height (MCH or ECH) in the model construction improved the inversion accuracy for PNC more significantly compared to LNC. Integrating UAV remote sensing and ground observation multimodal data, the random forest regression algorithm significantly improved the inversion accuracy of rice LNC and PNC at the late jointing stage. Considering both inversion accuracy and operational convenience, it is recommended to use a feature combination of VI+TI+ECH+FN in field production. The results demonstrate that constructing random forest regression models by integrating UAV remote sensing and ground observation multimodal data can accurately detect rice LNC and PNC, providing a scientific basis for rice field management and fertilization decision-making.