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

IF 8.2 Q1 AGRICULTURE, MULTIDISCIPLINARY
Baoling Ma , Zhixin Hua , Yuchen Wen , Hongxing Deng , Yongjie Zhao , Liuru Pu , Huaibo Song
{"title":"Using an improved lightweight YOLOv8 model for real-time detection of multi-stage apple fruit in complex orchard environments","authors":"Baoling Ma ,&nbsp;Zhixin Hua ,&nbsp;Yuchen Wen ,&nbsp;Hongxing Deng ,&nbsp;Yongjie Zhao ,&nbsp;Liuru Pu ,&nbsp;Huaibo Song","doi":"10.1016/j.aiia.2024.02.001","DOIUrl":null,"url":null,"abstract":"<div><p>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 proposed. The ShuffleNetv2 basic modules and down-sampling modules were alternately connected, replacing the Backbone of YOLOv8n model. The Ghost modules replaced the Conv modules and the C2fGhost modules replaced the C2f modules in the Neck part of the YOLOv8n. ShuffleNetv2 reduced the memory access cost through channel splitting operations. The Ghost module combined linear and non-linear convolutions to reduce the network computation cost. The 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 boxes. The Squeeze-and-Excitation (SE) was embedded in the Backbone and Neck part of YOLOv8n to enhance the representation ability of feature maps. The algorithm ensured high precision while having small model size and fast detection speed, which facilitated model migration and deployment. Using 9652 images validated the effectiveness of the model. 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 fps. The detection speeds on the Jetson Xavier NX development board were 3.17 fps. Comparisons 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 speed. The research can provide reference for the development of smart devices in apple orchards.</p></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"11 ","pages":"Pages 70-82"},"PeriodicalIF":8.2000,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2589721724000023/pdfft?md5=6fc303d1eb23f5151de28ee6f36c2d3d&pid=1-s2.0-S2589721724000023-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Artificial Intelligence in Agriculture","FirstCategoryId":"1087","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2589721724000023","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"AGRICULTURE, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0

Abstract

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 proposed. The ShuffleNetv2 basic modules and down-sampling modules were alternately connected, replacing the Backbone of YOLOv8n model. The Ghost modules replaced the Conv modules and the C2fGhost modules replaced the C2f modules in the Neck part of the YOLOv8n. ShuffleNetv2 reduced the memory access cost through channel splitting operations. The Ghost module combined linear and non-linear convolutions to reduce the network computation cost. The 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 boxes. The Squeeze-and-Excitation (SE) was embedded in the Backbone and Neck part of YOLOv8n to enhance the representation ability of feature maps. The algorithm ensured high precision while having small model size and fast detection speed, which facilitated model migration and deployment. Using 9652 images validated the effectiveness of the model. 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 fps. The detection speeds on the Jetson Xavier NX development board were 3.17 fps. Comparisons 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 speed. The research can provide reference for the development of smart devices in apple orchards.

使用改进的轻量级 YOLOv8 模型实时检测复杂果园环境中的多阶段苹果果实
为了在智能果园中对苹果果实的整个生长期进行有效监控。我们提出了一种名为 YOLOv8n-ShuffleNetv2-Ghost-SE 的轻量级模型。ShuffleNetv2 基本模块和下采样模块交替连接,取代了 YOLOv8n 模型的 Backbone。在 YOLOv8n 的 Neck 部分,Ghost 模块取代了 Conv 模块,C2fGhost 模块取代了 C2f 模块。ShuffleNetv2 通过通道分割操作降低了内存访问成本。Ghost 模块结合了线性和非线性卷积,降低了网络计算成本。Wise-IoU(WIoU)取代了计算边界框回归损失的 CIoU,可动态调整锚框质量阈值和梯度增益分配策略,优化预测边界框的大小和位置。YOLOv8n 的骨干和内核部分嵌入了挤压激励算法(SE),以增强特征图的表示能力。该算法在确保高精度的同时,还具有模型体积小、检测速度快的特点,为模型的迁移和部署提供了便利。使用 9652 幅图像验证了模型的有效性。YOLOv8n-ShuffleNetv2-Ghost-SE 模型的精确度为 94.1%,召回率为 82.6%,平均精确度为 91.4%,模型大小为 2.6 MB,参数为 1.18 M,FLOPs 为 3.9 G,检测速度为 39.37 fps。Jetson Xavier NX 开发板的检测速度为 3.17 fps。与 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 和 ShuflleNetv2-Ghost-CA 等先进模型的比较表明,该方法实现了更小的模型和更快的检测速度。该研究可为苹果园智能设备的开发提供参考。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Artificial Intelligence in Agriculture
Artificial Intelligence in Agriculture Engineering-Engineering (miscellaneous)
CiteScore
21.60
自引率
0.00%
发文量
18
审稿时长
12 weeks
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信