|本期目录/Table of Contents|

[1]朱齐齐,陈西曲.基于改进YOLO v5的轻量级果园苹果检测算法[J].江苏农业科学,2024,52(17):200-208.
 Zhu Qiqi,et al.Lightweight orchard apple detection algorithm based on improved YOLO v5[J].Jiangsu Agricultural Sciences,2024,52(17):200-208.
点击复制

基于改进YOLO v5的轻量级果园苹果检测算法(PDF)
分享到:

《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]

卷:
第52卷
期数:
2024年第17期
页码:
200-208
栏目:
农业工程与信息技术
出版日期:
2024-09-05

文章信息/Info

Title:
Lightweight orchard apple detection algorithm based on improved YOLO v5
作者:
朱齐齐陈西曲
武汉轻工大学电气与电子工程学院,湖北武汉 430023
Author(s):
Zhu Qiqiet al
关键词:
果园苹果YOLO v5s轻量化Fast-C3SIoU嵌入式设备
Keywords:
-
分类号:
TP391.41
DOI:
-
文献标志码:
A
摘要:
为了解决苹果采摘机器人识别算法中涉及到的复杂网络架构及大量参数占用内存巨大、计算需求庞大所导致的检测模型反应缓慢等问题,提出一种改进YOLO v5模型的轻量级果园苹果检测算法。首先,使用带有SE注意力机制的DepthSepConv模块和改进的Fast-C3模块对YOLO v5的Backbone网络部分进行重组,保持较高的精确率的同时减小模型体积;其次,用改进的Fast-C3模块替换整个Neck部分的C3模块,提高模型的准确率;替换颈部网络的普通卷积为Ghostconv,进一步降低模型的参数量与体积;最后,引入SIoU损失函数,使回归精确率和收敛速度得到提高。试验结果表明,该模型对苹果检测mAP为94.0%、模型计算量为8.4G FLOPs、体积仅为7.3 M。对比YOLO v5原模型,在mAP提高0.3百分点的情况下,计算量降低46.84%,模型体积缩减49.31%。于嵌入式平台上进行应用测试,实时检测速率达到了18.76 帧/s,约为原模型检测速率的1.5倍。因此,优化后的YOLO v5轻型模型不仅提升了识别准确性,并明显减少了计算负载量与模型大小,使得它能在嵌入式设备上快速有效地捕捉并分析复杂果园场景下的苹果信息。
Abstract:
-

参考文献/References:

[1]冯娟,刘刚,司永胜,等. 苹果采摘机器人激光视觉系统的构建[J]. 农业工程学报,2013,29(增刊1):32-37.
[2]Lehnert C,Sa I,McCool C,et al. Sweet pepper pose detection and grasping for automated crop harvesting[C]//2016 IEEE International Conference on Robotics and Automation. Stockholm,Sweden.IEEE,2016:2428-2434.
[3]王丹丹,宋怀波,何东健. 苹果采摘机器人视觉系统研究进展[J]. 农业工程学报,2017,33(10):59-69.
[4]王卓,王健,王枭雄,等. 基于改进YOLO v4的自然环境苹果轻量级检测方法[J]. 农业机械学报,2022,53(8):294-302.
[5]景亮,王瑞,刘慧,等. 基于双目相机与改进YOLO v3算法的果园行人检测与定位[J]. 农业机械学报,2020,51(9):34-39,25.
[6]何进荣,石延新,刘斌,等. 基于DXNet模型的富士苹果外部品质分级方法研究[J]. 农业机械学报,2021,52(7):379-385.
[7]薛勇,王立扬,张瑜,等. 基于GoogLeNet深度迁移学习的苹果缺陷检测方法[J]. 农业机械学报,2020,51(7):30-35.
[8]Turan M,Almalioglu Y,Araujo H,et al. Deep EndoVO:a recurrent convolutional neural network (RCNN) based visual odometry approach for endoscopic capsule robots[J]. Neurocomputing,2018,275:1861-1870.
[9]Ren Y,Zhu C R,Xiao S P. Object detection based on fast/faster RCNN employing fully convolutional architectures[J]. Mathematical Problems in Engineering,2018,2018:3598316.
[10]Sun X D,Wu P C,Hoi S C H. Face detection using deep learning:an improved faster RCNN approach[J]. Neurocomputing,2018,299:42-50.
[11]Gao F F,Fu L S,Zhang X,et al. Multi-class fruit-on-plant detection for apple in SNAP system using Faster R-CNN[J]. Computers and Electronics in Agriculture,2020,176:105634.
[12]Yang J,He W Y,Zhang T L,et al. Research on subway pedestrian detection algorithms based on SSD model[J]. IET Intelligent Transport Systems,2020,14(11):1491-1496.
[13]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. Las Vegas:IEEE,2016:779-788.
[14]罗志聪,李鹏博,宋飞宇,等. 嵌入式设备的轻量化百香果检测模型[J]. 农业机械学报,2022,53(11):262-269,322.
[15]张恩宇,成云玲,胡广锐,等. 基于SSD算法的自然条件下青苹果识别[J]. 中国科技论文,2020,15(3):274-281.
[16]汪颖,王峰,李玮,等. 用于复杂环境下果蔬检测的改进YOLO v5算法研究[J]. 中国农机化学报,2023,44(1):185-191.
[17]熊俊涛,韩咏林,王潇,等. 基于YOLO v5-Lite的自然环境木瓜成熟度检测方法[J]. 农业机械学报,2023,54(6):243-252.
[18]董丽君,曾志高,易胜秋,等. 基于YOLO v5的遥感图像目标检测[J]. 湖南工业大学学报,2022,36(3):44-50.
[19]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:IEEE,2018:8759-8768.
[20]Lin T Y,Dollár P,Girshick R,et al. Feature pyramid networks for object detection[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu:IEEE,2017:936-944.
[21]Yao T,Zhang Q,Wu X Y,et al. Image recognition method of defective button battery base on improved MobileNetV1[C]//Wang Y,Li X,Peng Y.Chinese Conference on Image and Graphics Technologies.Singapore:Springer,2020:313-324.
[22]Hu J,Shen L,Sun G. Squeeze-and-excitation networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Salt Lake City:IEEE,2018:7132-7141.
[23]Chen J R,Kao S H,He H,et al. Run,dont walk:chasing higher FLOPS for faster neural networks[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver:IEEE,2023:12021-12031.
[24]Han K,Wang Y H,Tian Q,et al. GhostNet:more features from cheap operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle:IEEE,2020:1580-1589.
[25]朱瑞鑫,杨福兴. 运动场景下改进YOLO v5小目标检测算法[J]. 计算机工程与应用,2023,59(10):196-203.

相似文献/References:

[1]龙阳,肖小玲.基于多注意力机制的苹果叶部病害检测方法[J].江苏农业科学,2023,51(23):178.
 Long Yang,et al.Apple leaf disease recognition method based on multi-attention mechanism[J].Jiangsu Agricultural Sciences,2023,51(17):178.
[2]赵方,左官芳,顾思睿,等.基于改进YOLO v5s的温室番茄检测模型轻量化研究[J].江苏农业科学,2024,52(8):200.
 Zhao Fang,et al.Lightweight research of greenhouse tomato detection model based on improved YOLO v5s[J].Jiangsu Agricultural Sciences,2024,52(17):200.
[3]高泉,刘笠溶,张洁,等.基于ActNN-YOLO v5s-RepFPN的番茄病害识别及系统设计[J].江苏农业科学,2024,52(20):220.
 Gao Quan,et al.Tomato disease identification and system design based on ActNN-YOLO v5s-RepFPN[J].Jiangsu Agricultural Sciences,2024,52(17):220.
[4]贺洪江,刘毅祥,王双友.基于改进YOLO v5s的叶菜病虫害检测算法研究[J].江苏农业科学,2025,53(5):244.
 He Hongjiang,et al.Study on foliage vegetable disease and pest detection algorithm based on improved YOLO v5s[J].Jiangsu Agricultural Sciences,2025,53(17):244.
[5]史鹏涛,田政伟,李晓泽,等.基于改进YOLO v5s算法的红枣缺陷检测与分拣方法[J].江苏农业科学,2025,53(5):83.
 Shi Pengtao,et al.Defect detection and sorting method of jujube based on improved YOLO v5s algorithm[J].Jiangsu Agricultural Sciences,2025,53(17):83.

备注/Memo

备注/Memo:
收稿日期:2024-06-26
基金项目:湖北省自然科学基金(编号:2019CFB813)。
作者简介:朱齐齐(1996—),男,安徽阜阳人,硕士研究生,主要研究方向为嵌入式开发。E-mail:zqq13155642201@163.com。
通信作者:陈西曲,博士,教授,主要研究方向为红外成像技术、电子信息处理技术、嵌入式技术、图像处理技术。E-mail:cxqdhl@whpu.edu.cn。
更新日期/Last Update: 2024-09-05