Review of precision rice hill-drop drilling technology and machine for paddy

Design of an active seed throwing and cleaning unit for pneumatic rice seed metering device

Design and field evaluation of hill-drop pneumatic central cylinder direct-seeding machine for hybrid rice

The double-hole rice vacuum seed meter is critical equipment for the planting precision of rice direct seeding. The effects of shaped holes and seed disturbance on the precision of rice bunch planting were investigated to improve the precision of bunch planting with the double-hole rice vacuum seed meter. A test bench with the rice vacuum seed meter was set up to analyze the trends in the quality of feed index, miss index, and multiple index of seed meters with different shaped holes at different speeds and vacuum pressures. Based on the optimal hole structure, different seed disturbance structures were designed to investigate the influence of the seed disturbance structure on the precision of bunch planting. A multiple linear regression model was established for the relationship between the disturbance structure, vacuum pressure, rotational speed, and the precision of bunch planting. Discrete element numerical simulation experiments were carried out to analyze the effect of disturbance structures on seeds. The planting precision of the seed meter with the shaped hole was significantly higher than that of the seed meter without the shaped hole while the shaped hole B was the optimum structure. Disturbance structure affects the quality of feed index, multiple index rate, and miss index. The planting precision of the seed disturbance structure II was better than the other structures. At a speed of 60 rpm and vacuum pressures of 2.0 kPa, 2.4 kPa, and 2.8 kPa, the qualities of feed index of seed disturbance structure II were 90%, 91.11%, and 89.17%, respectively, and the miss indexes were 2.96%, 1.94%, and 1.57%, respectively. At high rotational speeds, the precision of rice bunch planting with the seed disturbance structure is better than that without the seed disturbance structure. In the simulation test, the seed velocity and total force magnitude of the meter without disturbance structures were less than those with the disturbed structure. Simulation experiments showed that the seed disturbance structure breaks up the stacked state of seeds. Research has shown that the shaped hole holds the seed in a stable suction posture, which helps to increase the seed-filling rate. Seed disturbance improves seed mobility, thereby enhancing the precision of bunch planting.

Weed detection is a crucial step in achieving intelligent weeding for vegetables. Currently, research on vegetable weed detection technology is relatively limited, and existing detection methods still face challenges due to complex natural conditions, resulting in low detection accuracy and efficiency. This paper proposes the YOLOv8-EGC-Fusion (YEF) model, an enhancement based on the YOLOv8 model, to address these challenges. This model introduces plug-and-play modules: (1) The Efficient Group Convolution (EGC) module leverages convolution kernels of various sizes combined with group convolution techniques to significantly reduce computational cost. Integrating this EGC module with the C2f module creates the C2f-EGC module, strengthening the model’s capacity to grasp local contextual information. (2) The Group Context Anchor Attention (GCAA) module strengthens the model’s capacity to capture long-range contextual information, contributing to improved feature comprehension. (3) The GCAA-Fusion module effectively merges multi-scale features, addressing shallow feature loss and preserving critical information. Leveraging GCAA-Fusion and PAFPN, we developed an Adaptive Feature Fusion (AFF) feature pyramid structure that amplifies the model’s feature extraction capabilities. To ensure effective evaluation, we collected a diverse dataset of weed images from various vegetable fields. A series of comparative experiments was conducted to verify the detection effectiveness of the YEF model. The results show that the YEF model outperforms the original YOLOv8 model, Faster R-CNN, RetinaNet, TOOD, RTMDet, and YOLOv5 in detection performance. The detection metrics achieved by the YEF model are as follows: precision of 0.904, recall of 0.88, F1 score of 0.891, and mAP0.5 of 0.929. In conclusion, the YEF model demonstrates high detection accuracy for vegetable and weed identification, meeting the requirements for precise detection.