[1]林祖香,王英东,马荣,等.基于YOLO-L的自然环境中澳洲坚果果实的检测和识别[J].江苏农业学报,2024,(11):2102-2110.[doi:doi:10.3969/j.issn.1000-4440.2024.11.014]
　LIN Zuxiang,WANG Yingdong,MA Rong,et al.Macadamia (Macadamia integrifolia Maiden & Betche) detection and recognition in natural environments based on YOLO-L[J].,2024,(11):2102-2110.[doi:doi:10.3969/j.issn.1000-4440.2024.11.014]
点击复制

基于YOLO-L的自然环境中澳洲坚果果实的检测和识别()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2024年11期
页码:
2102-2110
栏目:
农业信息工程
出版日期:
2024-11-30

文章信息/Info

Title:
Macadamia (Macadamia integrifolia Maiden & Betche) detection and recognition in natural environments based on YOLO-L
作者:
林祖香1; 王英东1; 马荣1; 韦云松1; 李子文1; 李加强1; 何超2
(1.西南林业大学机械与交通学院,云南昆明650224；2.德宏职业学院,云南德宏678400)
Author(s):
LIN Zuxiang1; WANG Yingdong1; MA Rong1; WEI Yunsong1; LI Ziwen1; LI Jiaqiang1; HE Chao2
(1.College of Mechanical and Transportation Engineering, Southwest Forestry University, Kunming 650224, China;2.Dehong Vocational College, Dehong 678400, China)
关键词:
图像处理; 深度学习; YOLOv9模型; 澳洲坚果
Keywords:
image processing; deep learning; YOLOv9 model; macadamia (Macadamia integrifolia Maiden &;  Betche
分类号:
TP301.6；S664
DOI:
doi:10.3969/j.issn.1000-4440.2024.11.014
文献标志码:
A
摘要:
针对自然环境下果实重叠、相互遮挡和目标小的澳洲坚果果实检测准确率低的问题,提出一种改进YOLOv9模型的识别方法(YOLO-L)。首先,引入BiFormer注意力机制,该机制通过双层路由注意力机制实现了动态、查询感知的稀疏注意力分配,能够很好地捕获特征表征,增强网络对全局特征的关注度；其次,采用VoVGSCSP模块代替YOLOv9中的CBFuse模块,提高了复杂场景下小目标的检测效果；最后,将YOLOv9模型默认的损失函数替换成排斥损失函数,解决了果实排列密集和漏检的问题,进一步提升了澳洲坚果果实检测的平均精度。通过消融试验和对比试验来验证模型的有效性,发现YOLO-L模型的平均精度均值、精确率、召回率和F1值分别达到96.2%、92.3%、88.2%和90.2%。与YOLOv9模型相比,YOLO-L模型的平均精度均值提升了4.9个百分点。总体而言,YOLO-L模型能够在自然环境下准确识别被遮挡、重叠的澳洲坚果果实,且检测精度高。研究结果可为澳洲坚果产业的智能采摘提供有效的技术支持。
Abstract:
Aiming at the issue of low detection accuracy for macadamia nuts in natural environments due to overlapping, mutual occlusion, and small targets, an improved YOLOv9 model recognition method (YOLO-L) was proposed. Firstly, the BiFormer attention mechanism was introduced, which achieved dynamic and query-aware sparse attention allocation through the Bi-level routing attention mechanism. This mechanism was capable of effectively capturing feature representations and enhanced the network’s focus on global features. Secondly, the VoVGSCSP module was used to replace the CBFuse module in YOLOv9, which improved the detection performance for small targets in complex scenes. Lastly, the default loss function of the YOLOv9 model was replaced with an exclusion loss function, which solved the problems of dense fruit arrangement and missed detections, and further enhanced the average accuracy of macadamia nut detection. The effectiveness of the model was validated through ablation and comparative experiments. It was found that the mean average precision, precision, recall, and F1 score of YOLO-L model reached 96.2%, 92.3%, 88.2%, and 90.2%, respectively. Compared with the YOLOv9 model, the mean average precision of the YOLO-L model was improved by 4.9 percentage points. Overall, the YOLO-L model can accurately identify occluded and overlapped macadamia nuts in natural environments with high detection accuracy. The research results can provide effective technical support for the intelligent harvesting in the macadamia industry.

参考文献/References:

[1]MERELES CEUPPENS L G, FERRO E A, ALVARENGA SOSA N L, et al. Chemical composition of Macadamia integrifolia (Maiden and Betche) nuts from Paraguay[J]. International Food Research Journal, 2017, 24(6):2599-2608.
[2]苏继龙,刘远力,吴金东. 香蕉采摘机械的研究现状与发展趋势[J]. 中国农业大学学报,2019,24(7):131-137.
[3]MOREIRA G, MAGALHES S A, PINHO T, et al. Benchmark of deep learning and a proposed HSV colour space models for the detection and classification of greenhouse tomato[J]. Agronomy,2022,12(2):356.
[4]柳军,孔杰,皮杰,等. 果蔬分拣设备研究现状及发展趋势[J]. 中国农机化学报,2024,45(8):120-125.
[5]邢卓冉,丁松爽,张凯,等.计算机视觉与深度学习技术在烟叶生产上的研究进展[J/OL].中国农业科技导报,2024:1-11
[2024-08-01]. https://doi.org/10.13304/j.nykjdb.2023.0379.
[6]廖越,李智敏,刘偲. 基于深度学习的人——物交互关系检测综述[J]. 中国图象图形学报,2022,27(9):2611-2628.
[7]DONG C, ZHANG Z W, YUE J, et al. Automatic recognition of strawberry diseases and pests using convolutional neural network[J]. Smart Agricultural Technology,2021,1:100009.
[8]ZHENG Z H, XIONG J T, LIN H, et al. A method of green Citrus detection in natural environments using a deep convolutional neural network[J]. Frontiers in Plant Science,2021,12:705737.
[9]金梅,李义辉,张立国,等. 基于注意力机制改进的轻量级目标检测算法[J]. 激光与光电子学进展,2023,60(4):385-392.
[10]刘芳,刘玉坤,林森,等. 基于改进型YOLO的复杂环境下番茄果实快速识别方法[J]. 农业机械学报,2020,51(6):229-237.
[11]侯依廷,饶元,宋贺,等. 复杂大田场景下基于改进YOLOv8的小麦幼苗期叶片数快速检测方法[J]. 智慧农业(中英文),2024,6(4):128-137.
[12]吴小燕,郭威,朱轶萍,等. 基于改进YOLOv8s的大田甘蓝移栽状态检测算法[J]. 智慧农业(中英文),2024,6(2):107-117.
[13]赵鹏飞,钱孟波,周凯琪,等. 改进YOLOv7-Tiny农田环境下甜椒果实检测[J]. 计算机工程与应用,2023,59(15):329-340.
[14]AHMAD T, MA Y L, YAHYA M, et al. Object detection through modified YOLO neural network[J]. Scientific Programming,2020,2020:8403262.
[15]SAPKOTA R, QURESHI R, CALERO M F, et al. YOLOv10 to its genesis: a decadal and comprehensive review of the you only look once (YOLO) series[EB/OL]. (2024-07-26)
[2024-08-16]. https://arxiv.org/pdf/2406.19407.
[16]LAWAL M O. Tomato detection based on modified YOLOv3 framework[J]. Scientific Reports,2021,11(1):1447.
[17]CHEN W K, LU S L, LIU B H, et al. Detecting Citrus in orchard environment by using improved YOLOv4[J]. Scientific Programming,2020,2020:8859237.
[18]YAO J,QI J M,ZHANG J,et al. A real-time detection algorithm for kiwifruit defects based on YOLOv5[J]. Electronics,2021,10(14):1711.
[19]ZHANG Z X, LU X Q, CAO G J, et al. ViT-YOLO:transformer-based YOLO for object detection[C]//IEEE. 2021 IEEE/CVF International Conference on Computer Vision Workshops. Montreal,BC,Canada:IEEE,2021:2799-2808.
[20]TERVEN J, CRDOVA-ESPARZA D M, ROMERO-GONZLEZ J A. A comprehensive review of YOLO architectures in computer vision:from YOLOv1 to YOLOv8 and YOLO-NAS[J]. Machine Learning and Knowledge Extraction,2023,5(4):1680-1716.
[21]YAN B, FAN P, LEI X Y, et al. A real-time apple targets detection method for picking robot based on improved YOLOv5[J]. Remote Sensing,2021,13(9):1619.
[22]LI C, LIN J Q, LI Z, et al. An efficient detection method for Litchi fruits in a natural environment based on improved YOLOv7-Litchi[J]. Computers and Electronics in Agriculture,2024,217:108605.
[23]YANG G L, WANG J X, NIE Z L, et al. A lightweight YOLOv8 tomato detection algorithm combining feature enhancement and attention[J]. Agronomy,2023,13(7):1824.
[24]WANG C Y, YEH I H, MARK LIAO H Y. YOLOv9:learning what you want to learn using programmable gradient information[M]. Cham:Springer Nature Switzerland,2024:1-21.
[25]BAKIRCI M, BAYRAKTAR I. YOLOv9-enabled vehicle detection for urban security and forensics applications[C]//IEEE. 2024 12th International Symposium on Digital Forensics and Security,San Antonio,TX,USA:IEEE,2024:1-6.
[26]LIU S H, YANG Y C, JING X J, et al. BiFormer:an end-to-end deep learning approach for enhanced image-based photoplethysmography and heart rate accuracy[M]. Amsterdam: IOS Press, 2024: 205-214.
[27]WANG X L, XIAO T T, JIANG Y N, et al. Repulsion loss:detecting pedestrians in a crowd[C]//IEEE. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City,UT,USA:IEEE,2018:7774-7783.
[28]LI H L, LI J, WEI H B, et al. Slim-neck by GSConv:a lightweight-design for real-time detector architectures[J]. Journal of Real-Time Image Processing,2024,21(3):62.
[29]GOUTTE C, GAUSSIER E. A probabilistic interpretation of precision,recall and F-score,with implication for evaluation[M]. Berlin,Heidelberg:Springer Berlin Heidelberg,2005:345-359.
[30] FLACH P, KULL M. Precision-recall-gain curves: PR analysis done right[J]. Advances in Neural Information Processing Systems, 2015, 28:838-846.

相似文献/References:

[1]车金庆,王帆,王艺洁,等.基于视觉注意机制的黄绿色苹果图像分割[J].江苏农业学报,2018,(06):1347.[doi:doi:10.3969/j.issn.1000-4440.2018.06.021]
　CHE Jin-qing,WANG Fan,WANG Yi-jie,et al.A segmentation method of yellow and green apple images based on visual attention mechanism[J].,2018,(11):1347.[doi:doi:10.3969/j.issn.1000-4440.2018.06.021]
[2]车金庆,王帆,吕继东,等.重叠苹果果实的分离识别方法[J].江苏农业学报,2019,(02):469.[doi:doi:10.3969/j.issn.1000-4440.2019.02.030]
　CHE Jin-qing,WANG Fan,LYU Ji-dong,et al.Separation and recognition method for overlapped apple fruits[J].,2019,(11):469.[doi:doi:10.3969/j.issn.1000-4440.2019.02.030]
[3]任胜男,孙钰,张海燕,等.基于one-shot学习的小样本植物病害识别[J].江苏农业学报,2019,(05):1061.[doi:doi:10.3969/j.issn.1000-4440.2019.05.009]
　REN Sheng-nan,SUN Yu,ZHANG Hai-yan,et al.Plant disease identification for small sample based on one-shot learning[J].,2019,(11):1061.[doi:doi:10.3969/j.issn.1000-4440.2019.05.009]
[4]宋晓倩,张学艺,张春梅,等.基于深度迁移学习的酿酒葡萄种植信息提取[J].江苏农业学报,2020,(03):689.[doi:doi:10.3969/j.issn.1000-4440.2020.03.022]
　SONG Xiao-qian,ZHANG Xue-yi,ZHANG Chun-mei,et al.Extraction of wine grape planting information based on deep transfer learning[J].,2020,(11):689.[doi:doi:10.3969/j.issn.1000-4440.2020.03.022]
[5]金沙沙,贾良权,龙伟,等.基于特征选择与骨架提取的种子萌发的芽长、根长检测[J].江苏农业学报,2021,(03):597.[doi:doi:10.3969/j.issn.1000-4440.2021.03.007]
　JIN Sha-sha,JIA Liang-quan,LONG Wei,et al.Detection of seed bud length and root length based on feature selection and skeleton extraction[J].,2021,(11):597.[doi:doi:10.3969/j.issn.1000-4440.2021.03.007]
[6]金寿祥,周宏平,姜洪喆,等.采摘机器人视觉系统研究进展[J].江苏农业学报,2023,(02):582.[doi:doi:10.3969/j.issn.1000-4440.2023.02.033]
　JIN Shou-xiang,ZHOU Hong-ping,JIANG Hong-zhe,et al.Research progress on visual system of picking robot[J].,2023,(11):582.[doi:doi:10.3969/j.issn.1000-4440.2023.02.033]
[7]李颀,郭梦媛.基于深度学习的休眠期苹果树点云语义分割[J].江苏农业学报,2023,(05):1189.[doi:doi:10.3969/j.issn.1000-4440.2023.05.011]
　LI Qi,GUO Meng-yuan.Semantic segmentation of apple tree point cloud in dormant period based on deep learning[J].,2023,(11):1189.[doi:doi:10.3969/j.issn.1000-4440.2023.05.011]
[8]翟先一,魏鸿磊,韩美奇,等.基于改进YOLO卷积神经网络的水下海参检测[J].江苏农业学报,2023,(07):1543.[doi:doi:10.3969/j.issn.1000-4440.2023.07.011]
　ZHAI Xian-yi,WEI Hong-lei,HAN Mei-qi,et al.Underwater sea cucumber identification based on improved YOLO convolutional neural network[J].,2023,(11):1543.[doi:doi:10.3969/j.issn.1000-4440.2023.07.011]
[9]李仁杰,宋涛,高婕,等.基于改进YOLOv5的自然环境下番茄患病叶片检测模型[J].江苏农业学报,2024,(06):1028.[doi:doi:10.3969/j.issn.1000-4440.2024.06.009]
　LI Renjie,SONG Tao,GAO Jie,et al.Tomato diseased leaf detection model based on improved YOLOv5 in natural environment[J].,2024,(11):1028.[doi:doi:10.3969/j.issn.1000-4440.2024.06.009]
[10]许鑫,耿庆,郑凯,等.基于纹理特征与深度学习的小麦图像中的穗粒分割与计数[J].江苏农业学报,2024,(04):661.[doi:doi:10.3969/j.issn.1000-4440.2024.04.010]
　XU Xin,GENG Qing,ZHENG Kai,et al.Segmentation and counting of wheat spikes and grains based on texture features and deep learning[J].,2024,(11):661.[doi:doi:10.3969/j.issn.1000-4440.2024.04.010]

备注/Memo

备注/Memo:
收稿日期:2024-08-23基金项目:云南省高层次人才支持项目(YNWR-QNBJ-2018-066)；云南省科技厅项目(202301BD070001-077)作者简介:林祖香(1997-),女,云南保山人，硕士研究生,主要从事图像识别、目标检测方面的研究。(E-mail)2817702741@qq.com通讯作者:马荣,(E-mail)mr@swfu.edu.cn
更新日期/Last Update: 2025-01-20