Seedling transplanting with paper chain pots has been one of the most efficient and sustainable crop cultivation, in order to reduce environmental impact for the simple transplantation, and seedling survival rate. Simultaneously, the seedling root systems can promote multiple crop cultivations, the land replanting index, and crop yield in high-yield and efficient agriculture. However, existing continuous automatic transplantation has been limited to the seedling transplanting with the paper chain pots, only suitable for transplanting dense-planted crops with the same plant spacing. In this study, a seedling picking-throwing device was designed with a baffle/conveyor belt and chain chain-breaking mechanism for the chain-pot transplanter, particularly for a series of transplanting operations, such as the seedling picking, transporting, chain breaking, seedling throwing, and planting. Various mechanisms were composed of the baffle-type conveyor belt, chain breaking, driving system, posture correcting, and planting device. The procedure of the chain pot transplanting was as follows. Firstly, the chain pots were placed on the seedling storage tray in the optimal planting period before transplanting. The initial end of the chain-pot seedling tray was pulled and then placed on the seedling picking and throwing timing belt. The chain pots remained engaged with the seedling transplanting baffles. Secondly, the chain-pot transplanting device was started along a predetermined trajectory using the seedling picking-throwing timing belt. The chain pots were broken by the chain-breaking device, in order to form the independent seedling pot. The planting device and the seedling picking-throwing timing belt were also combined to throw the seedling pots into the seedling cups. Finally, the seedling pots were transported to the seedbed and then planted vertically into the soil. A single-factor bench test was conducted to clarify the influence of different parameters on the success rates of the chain breaking and seedling throwing. The chain-pot seedlings were taken as the experiment object in the suitable planting period. Specifically, the optimal ranges were achieved in the seedling age and the inclination angles of the timing belt. The transplanting frequencies were determined as 28-36 d, 0°-6°, and 50-60 plants/min, respectively. A three-factor, three-level orthogonal experiment was conducted to determine the effects of different factors on the success rate of chain breaking and seedling throwing. An optimal configuration was achieved in the seedling age of 29 d, the timing belt inclination angles of 4°, and the transplanting frequencies of 55 plants/min. Furthermore, the success rate of chain breaking and seedling throwing was 92.44% and 94.60%, respectively, under the optimal configuration. A field experiment was carried out under an optimal combination of working parameters, in order to verify the operation performance of the whole machine. It was found that the success rates of chain breaking and seedling throwing were 90.62% and 93.10%, respectively, which fully met the performance requirements of the seedling transplanter. The finding can provide some insights into the operating parameters of fully automatic transplanters for unmanned transplanting.

Early growth of rapeseed can be used to assess the performance of seeder, grain yield, crop management, and fertilizer application. Rapid and accurate detection is critical to rapeseed production and yield. However, the manual seedling survey cannot fully meet the operational demands for the extensive or frequent high-precision seedling, due to the low efficiency and subjectivity. In this study, efficient detection and counting models were proposed for the video stream of rapeseed seedlings using unmanned aerial vehicle (UAV) imagery and machine learning. Multi-head self-attention was added to the YOLO series models. The attention was then reduced to irrelevant semantic information. The model was improved to focus on the target object. The basic receptive field block (BasicRFB) module was selected to replace the original spatial pyramid pooling-fast (SPPF) module. One-dimensional convolution was added to the Neck part. The downsampling was then changed to achieve the target of rapeseed seedlings in the image. The efficient fusion of features was also promoted to focus on the key features among other interference factors. The deep simple online and real-time tracking (DeepSORT) was further combined with the cross-line counting to achieve the continuous tracking and target number counting. In addition, the counting model was deployed on edge computing devices. A real-time target counting was designed using a multi-rotor UAV platform. The edge computing device was used to realize the real-time detection and counting of rape seedlings. The targets of rape seedlings were processed in the video stream that was captured by the camera in real time. The experimental results show that: 1) The improved model with multi-head self-attention was significantly focused on the rape seedling area in the image. A better performance was achieved in extracting the target features than before. 2) The detection accuracy of rapeseed seedlings was improved using BasicRFB and the operator of the Neck part. The detection misjudgment of targets was reduced to effectively alleviate the negative impact of invalid targets in the image background. 3) The improved YOLOv5s was achieved in the AP50 and AP95 scores of 93.1% and 67.5%, respectively. Among them, AP50 was significantly higher by 14.82, 26.37, and 3.3 percentage points, respectively, while AP95 was higher by 25.7, 33.9, and 6.7 percentage points, respectively, compared with the classical target detection, such as Faster R-CNN, SSD, and YOLOX. The counting trial of rapeseed seedlings demonstrated that the counting model achieved the maximum precision of 96.34% with an average of 93.75%. Furthermore, the rapeseed counting efficiency exceeded the well-trained operator with an average increase of 9.52 times. In the case of online real-time counting of rapeseed seedlings, the maximum difference in counting precision was 1.87% on the UAV counting platform under different weather conditions. The excellent generalization, counting precision, and efficiency fully met the real-time requirements for rapeseed seedlings. The finding can provide a powerful reference to assess the quality of rapeseed seeding and field management.

Seedling transplanting with paper chain pots has been one of the most efficient and sustainable crop cultivations, in order to reduce the environmental impact, simple transplantation, and high seedling survival rate. The stronger seedling root system can be realized to promote multiple crop cultivation, land replanting index, and crop yield in high-yield and efficient agriculture, compared with traditional cultivation. However, it is still lacking in the equipment of paper chain pot, in order to fully meet the large-scale production in recent years. In this study, a booklet-making machine was developed to automatically fabricate the paper chain pots. Key processes were also integrated, such as gluing, indentation, paper feeding, and folding, in order to meet the agronomic needs of transplanting seedlings in various crops. A cost-effective solution was then provided for the production. Firstly, a prototype of a paper chain pot nursery booklet-making machine was designed, according to the hexagonal cross-section bowl chain seedling booklet. The motors and pneumatics were combined to drive the coordinated action of various working components. The formation and assembly of chain pots into booklets were realized to significantly reduce manual labor and production costs. Structural parameters were determined to optimize the key components. Among them, the glue coating device was used for the precise adhesive application; The indentation device was used to create the necessary indentation for folding and formation; The paper feeding device was to drive the movement of the paper through the machine; The folding paper collection device was assembled the chain pots into a booklet format. In addition, an operation control system was designed with various functions, such as one-touch start/stop, commissioning, cleaning, and operating instructions. An STM32 microcontroller was selected as the control core, in order to coordinate the movement of all components. As such, each device was optimized using the correct program during production. The working performance tests were carried out on the prototype of the machine using water-soluble adhesive, water-resistant adhesive, and nursery paper. The complete working test of making booklet results show that the average pass rate was 91.55% when the working efficiency was 60 copies/h and the chain bowl length was between 12.16 and 21.28 m. By contrast, the accuracy of paper feeding remained at a low level at different candidate job speeds, with a slip rate below 0.23%. The seedling experiment with the oilseed rape and tomato showed that the germination rate of seedlings with the made paper chain pot nursery booklet was 96.21%, indicating better seedlings growth. The paper chain pot nursery booklet can be easily dragged and unfolded into chains before transplanting, fully meeting the requirements of seedling cultivation and lightweight automatic transplantation. The performance indicators also met the requirements of the paper chain pot nursery booklet making, in order to effectively save manual labor and production costs for high efficiency. This research can provide new ways to realize mechanization and automation during transplanting.