Heilongjiang Province Key Laboratory of Forest Sustainable Management and Environmental Microorganism Engineering (Northeast Forestry University),Harbin 150040,China
Show less
文章历史+
Received
Accepted
Published
2025-03-10
Issue Date
2025-07-16
PDF (11536K)
摘要
传统松科球果采摘面临效率低、风险高和成本不可控等挑战,针对自动化松科球果采摘对果实的实时识别与定位问题,提出改进的YOLOv5s-7.0 (You Only Look Once)目标检测模型,基于此模型,构建基于双目深度相机的松科球果检测与定位网络。为提高目标检测精度及效率,对YOLOv5s模型进行改进,将部分卷积PConv嵌入到模型的颈部网络neck多分枝堆叠结构中,面对松科球果的复杂场景增强对稀疏特征的处理能力,提升鲁棒性,减轻特征信息的冗余。在骨干网络backbone的深层及backbone与neck的连接处嵌入简单注意力机制SimAM,在不引入过多参数的基础上优化模型复杂背景下特征提取能力和信息传递的有效性。为满足高效率检测定位,基于双目深度相机测距原理和改进的YOLOv5s模型搭建目标检测及实时定位代码,通过深度匹配,构建松科球果检测与定位系统。根据构建的大兴安岭樟子松球果与小兴安岭红松球果数据集,改进后YOLOv5s模型目标检测精确率达96.8%,召回率和平均精度分别达94%、96.3%,松科球果检测与定位系统在x轴、y轴、z轴的平均绝对误差分别为0.644、0.620、0.740 cm,顺、侧、逆光照下定位试验成功率93.3%,暗光下定位成功率83.3%,视场角等其他性能符合松科球果采摘需求。研究提出的松科球果检测与定位系统为机械化采摘的实时目标检测与定位问题提供可靠的解决方案。
Abstract
Traditional methods for harvesting pinecone species face challenges such as low efficiency, high risks, and uncontrollable costs. To address real-time recognition and localization in automated pinecone harvesting, we proposed an improved YOLOv5s-7.0 (you only look once) object detection model and construct a binocular depth camera-based detection and localization network. To improve the accuracy and efficiency of object detection, the YOLOv5s model was improved by embedding partial convolutions (PConv) into the neck module's multi-branch stacked structure to enhance sparse feature processing capability, improve robustness, and reduce feature redundancy in complex scenarios of pinecones. Additionally, the simple attention mechanism (SimAM) was integrated at deep backbone layers and backbone-neck connections to optimize the model’s feature extraction ability and information transmission efficiency in complex backgrounds without significant parameter increases. To meet the requirements of efficient detection and localization, a target detection and real-time localization code was developed using binocular vision principles and the improved YOLOv5s model, and a pinecone detection and localization system was constructed through depth matching. Based on the constructed dataset of Pinus sylvestris var. mongolica cones from the Greater Khingan Mountains and Pinus koraiensis cones from the Lesser Khingan Mountains, the improved YOLOv5s model achieved a precision of 96.8%, a recall of 94.0%, and an average precision (AP) of 96.3% in target detection tasks. The proposed pinecone detection and localization system demonstrated mean absolute errors of 0.644 cm, 0.620 cm, and 0.740 cm along the x-, y-, and z-axes, respectively. Under front, side, and backlighting conditions, the localization success rate reached 93.3%, while in low-light environments, it maintained a success rate of 83.3%. Other performance indicators, including field of view, meet the operational requirements for pinecone harvesting. The proposed pinecone detection and localization system provides a reliable solution for real-time target detection and localization problems in mechanized pinecone harvesting.
为全面评价松科球果检测算法性能,使用查准率(precision,P)、查全率(recall,R)、平均精度(mean average precision,mAP,式中记为mAP)作为试验评价指标。由于算法以联合定位算法的最终应用为目的,为保证检测效率,对速度指标帧率(frames per second,FPS)进行评价,保障算法在实时场景中的性能表现。试验评价指标计算公式为
WANGK Q, ZHANGW H, LUOZ,et al.Design and experiment of hitting pine cone picking robot[J].Transactions of the Chinese Society for Agricultural Machinery,2020,51(8):26-33.
CHENQ, YINC K, GUOZ L,et al.Current status and future development of the key technologies for apple picking robots[J].Transactions of the Chinese Society of Agricultural Engineering,2023,39(4):1-15.
[7]
TANGY C, CHENM Y, WANGC L,et al.Recognition and localization methods for vision-based fruit picking robots:A review[J].Frontiers in Plant Science,2020,11:510.
[8]
GIRSHICKR, DONAHUEJ, DARRELLT,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of 2014 IEEE Conference on Computer Vision and Pattern Recognition.June 23-28,2014,Columbus,OH,USA:IEEE,2014:580-587.
[9]
GIRSHICKR.Fast R-CNN[C]//Proceedings of 2015 IEEE International Conference on Computer Vision.December 07-13,2016,Santiago,Chile:IEEE,2015:1440-1448.
[10]
RENS, HEK, GIRSHICKR,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.
CHENGJ Y, CHENM J, LIT,et al.Detection of peach trees in unmanned aerial vehicle (UAV) images based on improved Faster-RCNN network[J].Acta Agriculturae Zhejiangensis,2024,36(8):1909-1919.
[13]
LIUW, ANGUELOVD, ERHAND,et al.SSD:Single shot multibox detector[C]//Proceedings of the European Conference on Computer Vision.The Netherlands:Springer Cham,2016:21-37.
[14]
REDMONJ, DIVVALAS, GIRSHICKR,et al.You only look once:Unified,real-time object detection[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition.June 27-30,2016,Las Vegas,NV,USA:IEEE,2016:779-788.
SHAOY H, ZHANGD, CHUH Y,et al.A review of YOLO object detection based on deep learning[J].Journal of Electronics & Information Technology,2022,44(10):3697-3708.
ZHAOH, QIAOY J, WANGH J,et al.Apple fruit recognition in complex orchard environment based on improved YOLOv3[J].Transactions of the Chinese Society of Agricultural Engineering,2021,37(16):127-135.
LIUJ, LIY, XIAOL M,et al.Recognition and location method of orange based on improved YOLOv4 model[J].Transactions of the Chinese Society of Agricultural Engineering,2022,38(12):173-182.
[26]
LIT F, FANGWT, ZHAOG N,et al.An improved binocular localization method for apple based on fruit detection using deep learning[J].Information Processing in Agriculture,2023,10(2):276-287.
[27]
ZHUL L, GENGX, LIZ,et al.Improving YOLOv5 with attention mechanism for detecting boulders from planetary images[J].Remote Sensing,2021,13(18):3776.
[28]
CHENJ R, KAOS H, HEH,et al.Run,don't walk:Chasing higher FLOPS for faster neural networks[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition.June 17-24,2023,Vancouver,BC,Canada:IEEE,2023:12021-12031.
[29]
YANGL X, ZHANGR Y, LIL D,et al.SimAM:A simple,parameter-free attention module for convolutional neural networks[C]//International Conference on Machine Learning.July 18-24,2021,Vienna,Austria,2021:11863-11874.
LIL, LIANGJ Y, ZHANGY F,et al.Accurate detection and localization method of citrus targets in complex environments based on improved YOLO v5[J].Transactions of the Chinese Society for Agricultural Machinery,2024,55(8):280-290.
GUOH, CHENH Y, GAOG M,et al.Safflower corolla object detection and spatial positioning methods based on YOLO v5m[J].Transactions of the Chinese Society for Agricultural Machinery,2023,54(7):272-281.