[1]郑文轩,杨瑛.基于改进轻量化YOLO v7的成熟期香梨目标检测方法[J].江苏农业科学,2024,52(20):121-128.
 Zheng Wenxuan,et al.Target detection method for fragrant pears at mature stage based on improved lightweight YOLO v7[J].Jiangsu Agricultural Sciences,2024,52(20):121-128.
点击复制

基于改进轻量化YOLO v7的成熟期香梨目标检测方法()

《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]

卷:
第52卷
期数:
2024年第20期
页码:
121-128
栏目:
果实智能检测
出版日期:
2024-10-20

文章信息/Info

Title:
Target detection method for fragrant pears at mature stage based on improved lightweight YOLO v7
作者:
郑文轩 杨瑛
江苏第二师范学院物理与信息工程学院,江苏南京 211200
Author(s):
Zheng Wenxuanet al
关键词:
目标检测香梨YOLO v7轻量化注意力机制
Keywords:
-
分类号:
S126;TP391.41
DOI:
-
文献标志码:
A
摘要:
为在自然环境下准确地识别和检测香梨果实,以YOLO v7为基础网络模型,针对果园中香梨果实、果叶、枝干之间相互遮挡问题,提出一种改进的轻量化YOLO v7梨果目标检测方法。该方法将MobileNet v3引入YOLO v7 模型中作为其骨干特征提取网络,从而减少网络的参数量,使其更容易部署在移动端和生产实际,在特征融合层引入协同注意力机制CA(coordinate attention)模块,以提高网络的特征表达能力,将原YOLO v7中的损失函数CIoU替换为SIoU,从而提高模型的检测速度和性能。最后利用Grad-CAM 方法产生目标检测热力图,进行特征可视化。结果表明,改进的轻量化YOLO v7模型的平均精度均值(mAP)、精确率、召回率指标分别为96.33%、94.36%、89.28%,检测速度为87.71(帧/s),模型内存占用量与原YOLO v7相比减少21.45 MB;其检测平均精度均值(mAP) 与 Faster R-CNN、YOLO v3、MobileNet v3-YOLO v4、YOLO v5s、YOLO v7模型相比分别提高28.37、9.66、1314、4.58、3.20百分点。研究表明,改进的轻量化YOLO v7模型对成熟期香梨具有很好的目标检测效果和鲁棒性,可为香梨自动化采摘提供有效的技术支持。
Abstract:
-

参考文献/References:

[1]邵园园,王永贤,玄冠涛,等. 基于高光谱成像的肥城桃品质可视化分析与成熟度检测[J]. 农业机械学报,2020,51(8):344-350.
[2]宋怀波,尚钰莹,何东健. 果实目标深度学习识别技术研究进展[J]. 农业机械学报,2023,54(1):1-19.
[3]林景栋,吴欣怡,柴毅,等. 卷积神经网络结构优化综述 [J]. 自动化学报,2020,46(1):24-37.
[4]He K M,Zhang X Y,Ren S Q,et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:770-778.
[5]傅隆生,冯亚利,Tola E,et al. 基于卷积神经网络的田间多簇猕猴桃图像识别方法[J]. 农业工程学报,2018,34(2):205-211.
[6]崔潇,马原东,倪照风,等. 基于深度学习的桃子检测研究[J]. 贵州科学,2019,37(2):82-86.
[7]He K M,Zhang X Y,Ren S Q,et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(9):1904-1916.
[8]Ren S Q,He K M,Girshick R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[EB/OL]. (2015-06-04)[2023-11-20]. https://arxiv.org/abs/1506.01497v3.
[9]Redmon J,Farhadi A. YOLO v3:an incremental improvement[EB/OL]. (2018-04-08)[2023-11-20]. https://arxiv.org/abs/1804.02767v1.
[10]Tian Y N,Yang G D,Wang Z,et al. Apple detection during different growth stages in orchards using the improved YOLO v3 model[J]. Computers and Electronics in Agriculture,2019,157:417-426.
[11]王立舒,秦铭霞,雷洁雅,等. 基于改进YOLO v4-Tiny的蓝莓成熟度识别方法[J]. 农业工程学报,2021,37(18):170-178.
[12]周桂红,马帅,梁芳芳. 基于改进YOLO v4模型的全景图像苹果识别[J]. 农业工程学报,2022,38(21):159-168.
[13]Lyu S L,Li R Y,Zhao Y W,et al. Green Citrus detection and counting in orchards based on YOLO v5-CS and AI edge system[J]. Sensors,2022,22(2):576.
[14]Singh B,Najibi M,Davis S. Efficient multi-scale training[EB/OL]. [2023-11-20]. https://arxiv.org/abs/1805.09300.
[15]Redmon J,Divvala S,Girshick R,et al. You only look once:unified,real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas,NV,USA:IEEE,2016:779-788.
[16]Redmon J,Farhadi A. YOLO 9000:better,faster,stronger[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu,HI:IEEE,2017:7263-7271.
[17]Jiang B R,Luo R X,Mao J Y,et al. Acquisition of localization confidence for accurate object detection[M]//Lecture Notes in Computer Science.Cham:Springer International Publishing,2018:816-832.
[18]彭心睿,潘晴,田妮莉. WCF-MobileNet v3:轻量型新冠肺炎 CXR 图像识别网络[J]. 计算机工程与应用,2023,59(14):224-231.
[19]刘雪,沈长盈,吕学泽,等. 基于改进MobileNet v3-Large的鸡蛋新鲜度识别模型[J]. 农业工程学报,2022,38(17):196-204.
[20]林森,刘美怡,陶志勇. 采用注意力机制与改进YOLO v5的水下珍品检测[J]. 农业工程学报,2021,37(18):307-314.
[21]He J,Sarah E,Ma X,etal. Alpha-IoU:a family of power intersection over union losses for boundingbox regression[C]//Proceedings of the Conference and Workshop on Neural Information Processing Systems.Long Beach,2021:1-10.
[22]Zheng Z H,Wang P,Liu W,et al. Distance-IoU loss:faster and better learning for bounding box regression[J]. Proceedings of the AAAI Conference on Artificial Intelligence,2020,34(7):12993-13000.
[23]Gevorgyan Z. SIoU loss:more powerful learning for bounding box regression[EB/OL]. (2022-05-25)[2023-11-20]. https://arxiv.org/abs/2205.12740.
[24]周绍发,肖小玲,刘忠意,等. 改进的基于YOLO v5s苹果树叶病害检测[J]. 江苏农业科学,2023,51(13):212-220.
[25]Danell J,Khan F. Adaptive color attributes for real-time visual tracking[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition. Columbus,OH,USA:IEEE,2014:1090-1097.
[26]龙燕,杨智优,何梦菲. 基于改进 YOLO v7 的疏果期苹果目标检测方法 [J]. 农业工程学报2023,39(14):191-199.

相似文献/References:

[1]刘晨,哈斯亚提·托逊江,热沙来提汗·买买提,等.香梨与牧草套种对产草量及土壤性状的影响[J].江苏农业科学,2013,41(08):209.
 Liu Chen,et al.Effect of pear and grass intercropping on grass yield and soil properties[J].Jiangsu Agricultural Sciences,2013,41(20):209.
[2]罗巍,陈曙东,王福涛,等.基于深度学习的大型食草动物种群监测方法[J].江苏农业科学,2020,48(20):247.
 Luo Wei,et al.Monitoring method of large herbivore population based on deep learning[J].Jiangsu Agricultural Sciences,2020,48(20):247.
[3]陈恩会,褚姝频,王炜,等.基于RetinaNet模型的梨小食心虫智能识别计数方法[J].江苏农业科学,2021,49(24):205.
 Chen Enhui,et al.Intelligent recognition and counting method of Grapholitha molesta based on RetinaNet model[J].Jiangsu Agricultural Sciences,2021,49(20):205.
[4]陶雪阳,施振旦,郭彬彬,等.基于RFID与目标检测的种鹅个体产蛋信息监测方法[J].江苏农业科学,2023,51(5):200.
 Tao Xueyang,et al.Monitoring method of individual egg-laying information of breeding geese based on RFID and object detection[J].Jiangsu Agricultural Sciences,2023,51(20):200.
[5]严陈慧子,田芳明,谭峰,等.基于改进YOLOv4的水稻病害快速检测方法[J].江苏农业科学,2023,51(6):187.
 Yanchen Huizi,et al.Rapid detection method of rice diseases based on improved YOLOv4[J].Jiangsu Agricultural Sciences,2023,51(20):187.
[6]张驰,穆凯代斯罕·伊萨克,吴俊杰,等.基于不同去袋期分析红皮梨花青苷合成酶及相关因子[J].江苏农业科学,2023,51(9):147.
 Zhang Chi,et al.Analysis of anthocyanin synthase and related factors in red pear based on different bag removal period[J].Jiangsu Agricultural Sciences,2023,51(20):147.
[7]周绍发,肖小玲,刘忠意,等.改进的基于YOLOv5s苹果树叶病害检测[J].江苏农业科学,2023,51(13):212.
 Zhou Shaofa,et al.Improved apple leaf disease detection based on YOLOv5s[J].Jiangsu Agricultural Sciences,2023,51(20):212.
[8]姜国权,杨正元,霍占强,等.基于改进YOLOv5网络的疏果前苹果检测方法[J].江苏农业科学,2023,51(14):205.
 Jiang Guoquan,et al.Apple detection method before thinning fruit based on improved YOLOv5 model[J].Jiangsu Agricultural Sciences,2023,51(20):205.
[9]王圆圆,林建,王姗.基于YOLOv4-tiny模型的水稻早期病害识别方法[J].江苏农业科学,2023,51(16):147.
 Wang Yuanyuan,et al.An early rice disease recognition method based on YOLOv4-tiny model[J].Jiangsu Agricultural Sciences,2023,51(20):147.
[10]倪智涛,胡伟健,李宝山,等.一种基于图像分类与目标检测协同的番茄细粒度病害识别方法[J].江苏农业科学,2023,51(22):221.
 Ni Zhitao,et al.A novel method for tomato fine-grained disease recognition based on image classification and target detection[J].Jiangsu Agricultural Sciences,2023,51(20):221.

备注/Memo

备注/Memo:
收稿日期:2023-11-23
基金项目:新疆生产建设兵团科技创新项目(编号:2021CB021)。
作者简介:郑文轩(1980—),男,河南南阳人,博士,教授,从事图像分析、计算机视觉研究。E-mail:wenxuanzhengdx@163.com。
通信作者:杨瑛,博士,教授,从事农业信息技术研究。E-mail:yangyingtlmdx@163.com。
更新日期/Last Update: 2024-10-20