Research on Lightweight Kiwifruit Detection Algorithm Based on YOLOv8-BFW

Aiming at the problems of low recognition efficiency and accuracy and high computing resources of YOLOv8 model in kiwifruit detection, this paper proposes a lightweight kiwifruit detection algorithm YOLOv8-BFW based on YOLOv8. Firstly, the weighted bidirectional feature pyramid network BiFPN module is introduced into the Neck part of YOLOv8 to realize multi-scale feature fusion, so as to improve the efficiency and accuracy of its detection; Secondly, all C2f modules are integrated into the FasterNet network to form a C2f-Faster module to reduce computing resources; Finally, the WIOU boundary loss function is used to replace the original CIOU to improve the regression performance of the network bounding box and accelerate the convergence speed of the model, which can more accurately measure the overlapping area of the box and further improve the efficiency and accuracy of its detection. The experimental results show that the average detection accuracy mAP of the YOLOv8-BFW algorithm on the self-made kiwifruit dataset collected and labeled in this paper reaches 96.9%. Compared with the original YOLOv8 algorithm, the model size is reduced by 32%, the calculation amount is reduced by 22.2%, the parameter quantity is reduced by 29.6%, and the detection speed is increased by 26.7%.