Using an improved lightweight YOLOv8 model for real-time detection of multi-stage apple fruit in complex orchard environments

作     者：Baoling Ma Zhixin Hua Yuchen Wen Hongxing Deng Yongjie Zhao Liuru Pu Huaibo Song 

作者机构：College of Mechanical and Electronic EngineeringNorthwest A&F UniversityYanglingShaanxi 712100China Key Laboratory of Agricultural Internet of ThingsMinistry of Agricultureand Rural AffairsYanglingShaanxi 712100China Shaanxi Key Laboratory of Agricultural Information Perception and Intelligent ServicesYanglingShaanxi 712100China 

出 版 物：《Artificial Intelligence in Agriculture》 (农业人工智能（英文）)

年 卷 期：2024年第11卷第1期

页      面：70-82页

核心收录：

学科分类：09[农学] 0902[农学-园艺学] 090201[农学-果树学] 

基　　金：supported by the National Key Research and Development Program of China(2019YFD1002401) the National Natural Science Foundation of China(31701326) 

主　　题：Multi-stage apple fruit Deep learning Real-time detection Lightweight model 

摘      要：For the purpose of monitoring apple fruits effectively throughout the entire growth period in smart orchards.A lightweight model named YOLOv8n-ShuffleNetv2-Ghost-SE was *** ShuffleNetv2 basic modules and down-sampling modules were alternately connected,replacing the Backbone of YOLOv8n *** Ghost modules replaced the Conv modules and the C2fGhost modules replaced the C2f modules in the Neck part of the ***2 reduced the memory access cost through channel splitting *** Ghost module combined linear and non-linear convolutions to reduce the network computation *** Wise-IoU(WIoU)replaced the CIoU for calculating the bounding box regression loss,which dynamically adjusted the anchor box quality threshold and gradient gain allocation strategy,optimizing the size and position of predicted bounding *** Squeeze-and-Excitation(SE)was embedded in the Backbone and Neck part of YOLOv8n to enhance the representation ability of feature *** algorithm ensured high precision while having small model size and fast detection speed,which facilitated model migration and *** 9652 images validated the effectiveness of the *** YOLOv8n-ShuffleNetv2-Ghost-SE model achieved Precision of 94.1%,Recall of 82.6%,mean Average Precision of 91.4%,model size of 2.6 MB,parameters of 1.18 M,FLOPs of 3.9 G,and detection speed of 39.37 *** detection speeds on the Jetson Xavier NX development board were 3.17 *** with advanced models including Faster R-CNN,SSD,YOLOv5s,YOLOv7‑tiny,YOLOv8s,YOLOv8n,MobileNetv3_small-Faster,MobileNetv3_small-Ghost,ShuflleNetv2-Faster,ShuflleNetv2-Ghost,ShuflleNetv2-Ghost-CBAM,ShuflleNetv2-Ghost-ECA,and ShuflleNetv2-Ghost-CA demonstrated that the method achieved smaller model and faster detection *** research can provide reference for the development of smart devices in apple orchards.