1. This can be attributed to the phenomenon of repeated detections (i.e., the detection of the same target multiple times), as shown in Fig. 13a. For YOLOv3n-Pose, YOLOv5n-Pose, and YOLOv8n-Pose, the AP 50 and AP 50-95 values perform similarly, ranging from 93.5% to 93.8% and from 92.3% to 93.0%, respectively. The AP 50-95 value of YOLOv8n-CK is the highest, at 6.5%, 6.2%, and 6.9% higher than those of YOLOv3n-Pose, YOLOv5n-Pose, and YOLOv8n-Pose, respectively. This performance can be attributed to the introduction of the CBAM in YOLOv8n-CK, which significantly enhanced the feature extraction capability of the cabbage head detection under complex occlusion conditions;Additionally
2. Development of technology and means for machine harvesting of head cabbage in the non-damaging mode;S Alatyrev;IOP conference series. Earth and environmental science,2020
3. Technology and parameters of cabbage machine harvesting by careful stacking of heads in containers;S S Alatyrev;IOP conference series. Earth and environmental science,2020
4. Image-based size estimation of broccoli heads under varying degrees of occlusion;P M Blok;Biosyst. Eng,2021