YOLOv8-Based Recognition of Wolfberry Pistil Assisted Pollination

In this paper, a pistil-assisted pollination recognition method based on improved YOLOv8 is proposed, aiming to improve the detection accuracy and efficiency in the pistil pollination process. Firstly, a weighted bidirectional feature pyramid network (BiFPN) is introduced to replace the PAN-FPN module in YOLOv8 to enhance the multi-scale feature fusion capability for the complex background and small target features of pistil images. Secondly, the Coordinate Attention (CA) mechanism is combined to further enhance the extraction ability of the model for stamen features. In addition, Ghost convolution is used to replace the traditional convolution, which effectively reduces the computational complexity and storage requirements of the model. The experimental results show that the improved YOLOv8 model achieves a mean average precision (mAP) of 90.5% on the self-constructed stamen dataset. The method provides an efficient and accurate solution for pistil-assisted pollination identification, which is suitable for real-time detection and low-match device deployment.