A lightweight pineapple detection network based on YOLOv7-tiny for agricultural robot system

Jiehao Li, Chenglin Li, Shan Zeng, Xiwen Luo, C. L. Philip Chen, Chenguang Yang

Abstract: Highlights•A high-precision lightweight detection network Pineapple-YOLO is proposed for pineapple detection on agricultural robots, improving detection performance by 6.28% with only a 2.21% increase in parameters, maintaining real-time detection.•Extensive experiments conducted on mixed pineapple datasets (varying directions, lighting, and backgrounds) achieve 89.7% average precision, outperforming other detection networks.•The Pineapple-YOLO is deployed on the agricultural robot, and the successful picking rate is maintained at 92% under the average time of 12s, indicating superior performance in engineering applications.