[1]李刚,陈琳,卢承方,等.CBSYOLO:复杂环境下的轻量化葡萄果实检测算法[J].江苏农业科学,2025,53(13):241-253.
Li Gang,et al.CBSYOLO:a lightweight grape fruit detection algorithm in complex environments[J].Jiangsu Agricultural Sciences,2025,53(13):241-253.
点击复制
CBSYOLO:复杂环境下的轻量化葡萄果实检测算法(
)
《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]
- 卷:
-
第53卷
- 期数:
-
2025年第13期
- 页码:
-
241-253
- 栏目:
-
农业工程与信息技术
- 出版日期:
-
2025-07-05
文章信息/Info
- Title:
-
CBSYOLO:a lightweight grape fruit detection algorithm in complex environments
- 作者:
-
李刚; 陈琳; 卢承方; 郜潘栓; 龚程杰
-
长江大学计算机科学学院,湖北荆州 434000
- Author(s):
-
Li Gang; et al
-
-
- 关键词:
-
葡萄果实识别; YOLO 11n; 轻量化模型; 目标检测; 深度学习
- Keywords:
-
-
- 分类号:
-
S126;TP391.41
- DOI:
-
-
- 文献标志码:
-
A
- 摘要:
-
针对复杂采摘环境中葡萄果实因叶片和枝蔓遮挡、果实分布密集且簇间边界模糊,以及传统检测模型计算量大、不易部署于移动终端等问题。基于YOLO 11n模型,提出一种改进的轻量化葡萄果实检测模型CBS-YOLO。首先,设计了全新的C3k2_Faster_CAA模块,使用FasterBlock模块替换原主干网络中C3k2的Bottleneck模块,以提升模型的计算效率;在此基础上,引入上下文锚点注意力机制,进一步增强模型的上下文特征提取能力。针对葡萄果实因紧密堆叠和复杂背景引发的目标区分困难问题,在颈部网络中融入BiFPN结构,以增强模型对重叠果实及密集分布目标的检测能力,同时有效减少背景干扰。此外,为应对采摘环境中光照变化及叶片遮挡带来的检测难题,在检测头中引入多头注意力机制,构建Detect_SEAM模块,从而更好地捕捉果实区域的关键特征,提升模型在复杂环境下的鲁棒性和检测性能。结果表明,CBS-YOLO在综合性能上均优于主流检测模型,其精确率、召回率、mAP50的全类别均值分别达92.2%、90.3%、94.6%,较基准模型YOLO 11n分别提升2.8、2.1、1.9百分点。同时,模型参数量和浮点运算量分别减少44.7%、36.5%。该模型在保持高精度的同时实现了轻量化效果,为复杂场景下的葡萄果实高效检测提供了有效方案,并为资源有限设备的实际应用提供了技术支持。
- Abstract:
-
-
参考文献/References:
[1]展晓城. 小葡萄撬动大产业[J]. 农家致富,2024(21):14.
[2]刘柯含,张芮,高彦婷,等. 中国葡萄产业现状分析及其发展对策[J]. 中国果树,2024(7):132-138.
[3]申艺方,聂超超,赵会娟. 浅析自动控制技术在农业机械中的应用[J]. 南方农机,2023,54(22):83-85.
[4]Zhang J,Kang N B,Qu Q J,et al. Automatic fruit picking technology:a comprehensive review of research advances[J]. Artificial Intelligence Review,2024,57(3):54.
[5]Shi M X. Image recognition for fruit-picking robots[J]. Journal of Image Processing Theory and Applications,2023,6(1):132-139.
[6]乔艳军. 基于深度学习的采摘机器人水果识别定位与采后自动分级技术研究[D]. 天津:天津理工大学,2022:1-137.
[7]Fang X M,Zhu H Y,Li S Y,et al. Round fruit picking system based on machine vision[J]. Journal of Physics:Conference Series,2024,2926(1):012015.
[8]郭武元. 基于视觉的苹果采摘机器人目标识别算法研究[D]. 柳州:广西科技大学,2023:1-150.
[9]李天华,孙萌,娄伟,等. 采摘机器人分割与识别算法的研究现状[J]. 山东农业科学,2021,53(10):140-148.
[10]梁敖,代东南,牛思琪,等. 基于改进YOLO v5s的草莓成熟度实时检测算法[J]. 山东农业科学,2024,56(11):156-163.
[11]李国进,黄晓洁,李修华,等. 采用轻量级网络MobileNet v2的酿酒葡萄检测模型[J]. 农业工程学报,2021,37(17):168-176,317.
[12]肖张娜,罗陆锋,陈明猷,等. 基于改进YOLO-v4的果园环境下葡萄检测[J]. 智能化农业装备学报(中英文),2023,4(2):35-43.
[13]孙俊,吴兆祺,贾忆琳,等. 基于改进YOLO v5s的果园环境葡萄检测[J]. 农业工程学报,2023,39(18):192-200.
[14]张传栋,亓璐,丁华立. 基于改进YOLO v8n模型的多品种葡萄簇检测方法[J]. 中国农机化学报,2024,45(9):220-226.
[15]Chen J L,Chen H,Xu F,et al. Real-time detection of mature table grapes using ESP-YOLO network on embedded platforms[J]. Biosystems Engineering,2024,246:122-134.
[16]陈文骏,饶元,王丰仪,等. 葡萄多模态目标检测和语义分割数据集[J]. 中国科学数据,2025,10(1):89-104.
[17]Wang C Y,Mark Liao H Y,Wu Y H,et al. CSPNet:a new backbone that can enhance learning capability of CNN[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Seattle,WA,USA.IEEE,2020:1571-1580.
[18]Chen J,Kao S,He H,et al. Run,Dont Walk:chasing higher FLOPS for faster neural networks[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver,BC,Canada.IEEE,2023:12021-12031.
[19]Cai X H,Lai Q X,Wang Y W,et al. Poly kernel inception network for remote sensing detection[C]//2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle,WA,USA.IEEE,2024:27706-27716.
[20]Liu S,Qi L,Qin H F,et al. Path aggregation network for instance segmentation[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City,UT,USA.IEEE,2018:8759-8768.
[21]Tan M X,Pang R M,Le Q V. EfficientDet:scalable and efficient object detection[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle,WA,USA.IEEE,2020:10778-10787.
[22]Wang Y D,Zhang J,Kan M N,et al. Self-supervised equivariant attention mechanism for weakly supervised semantic segmentation[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle,WA,USA.IEEE,2020:12272-12281.
[23]Ouyang D L,He S,Zhang G Z,et al. Efficient multi-scale attention module with cross-spatial learning[C]//ICASSP 2023—2023 IEEE International Conference on Acoustics,Speech and Signal Processing. Rhodes Island,Greece.IEEE,2023:1-5.
[24]Hu J,Shen L,Sun G. Squeeze-and-excitation networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City,UT,USA.IEEE,2018:7132-7141.
[25]Yang L X,Zhang R Y,Li L,et al. SimAM:a simple parameter-free attention module for convolutional neural networks[J]. Proceedings of the International Conference on Machine Learning,2021,139:11863-11874.
[26]Li H L,Li J,Wei H B,et al. Slim-neck by GSConv:a better design paradigm of detector architectures for autonomous vehicles[EB/OL]. (2022-06-06)[2025-01-06]. http://export.arxiv.org/abs/2206.02424v2.
[27]Dai X Y,Chen Y P,Xiao B,et al. Dynamic head:unifying object detection heads with attentions[EB/OL]. (2021-06-15)[2025-01-06] .https://arxiv.org/abs/2106.08322v1.
[28]李军,杨飞帆,龚胜,等. 基于视觉的轻量化路面异常检测算法[J/OL]. 吉林大学学报(工学版):1-9(2024-09-23)[2025-01-06]. https://doi.org/10.13229/j.cnki.jdxbgxb.20240802.
[29]Ren S Q,He K M,Girshick R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149.
[30]Liu W,Anguelov D,Erhan D,et al. SSD:single shot multibox detector[J]. European Conference on Computer Vision,2016:21-37.
[31]Selvaraju R R,Cogswell M,Das A,et al. Grad-CAM:visual explanations from deep networks via gradient-based localization[C]//2017 IEEE International Conference on Computer Vision. Venice,Italy.IEEE,2017:618-626.
备注/Memo
- 备注/Memo:
-
收稿日期:2025-01-06
基金项目:国家科技重大专项(编号:2021DJ1006);中国高校产学研创新基金(编号:2020ITA03012、2023IT269)。
作者简介:李刚(1998—),男,河南驻马店人,硕士研究生,主要从事计算机视觉与目标检测研究。E-mail:17633826560@qq.com。
通信作者:陈琳,博士,教授,主要从事深度学习与人工智能研究。E-mail:chenlin@yangtzeu.edu.cn。
更新日期/Last Update:
2025-07-05