BPYOLO-An Algorithm for Detecting Bagging and Fruiting in Peach Orchards

Aiming at the demand for accurate fruit and bag detection in the automated management of peach orchards, this paper proposes an improved YOLOv8 target detection algorithm, BPYOLO, in order to enhance the detection accuracy and real-time performance in complex environments. Firstly, MANet (Mixed Aggregation Network) is introduced on the basis of YOLOv8 to enhance the feature extraction ability; Secondly, a new detection head p2_SEAMHead is designed by combining the SEAM module with the P2 small target detection layer to improve the recognition effect of occluded and overlapped fruits. The experimental results show that BPYOLO achieves <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$94.6 \%, 94.3 \%$</tex>, and 97.6% in precision, recall, and mAP on the self-built dataset, respectively, which are <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$4.6 \%, 8.8 \%$</tex>, and 5% higher than the original YOLOv8, and the inference speed is 1.1ms, which meets the real-time monitoring requirements. The method provides effective technical support for the intelligent management of peach orchards and has the potential to be promoted to other fruit and vegetable detection tasks.