Construction of a strawberry ripeness detection model based on the improved YOLOv11n algorithm
ZHENG Zaifei
LYU Changxin
XU Xiuyi
Abstract：Accurate and efficient assessment of strawberry ripeness is crucial for improving fruit quality and standardizing harvesting in modern agriculture.However,traditional manual inspection methods suffer from strong subjectivity and low efficiency.This study proposes an enhanced lightweight object detection model based on YOLOv11n,specifically designed for real-time strawberry ripeness classification in complex greenhouse environments.The model integrates the LightGhostConv module,LightCA attention mechanism,and an optimized C2f module to improve small object feature extraction,reduce computational complexity,and enhance robustness to occlusion and illumination changes.A custom dataset containing different ripeness grades of"Benihoppe"strawberries was collected in Jinzhou,China,under various lighting and occlusion conditions.Experimental results show that the improved model achieves an average precision of 92.6%,with an inference speed of 49 FPS and only 2.24 M parameters,outperforming the original YOLOv11n and other mainstream models such as YOLOv5n,YOLOv8n,and YOLOv10n.Based on this model,a web-based detection system was further developed,which supports image upload,classification,and visualization operations,providing practical support for intelligent harvesting applications.This work offers a feasible solution for lightweight deployment on mobile terminals and has broad application prospects in the fields of smart agriculture and precise fruit harvesting.
Keywords：strawberry ripenesssmall target detectionlightweight modelYOLOv11ncoordinate attentionfacility agriculture
Publication Date：2025-09-15
Online Publishing Date：2025-12-22(First online date of this platform, not the publication date of the document)
Pages：10( 170-179 )