摘要
为使苹果采摘机器人在复杂果树背景下能快速准确地检测出苹果,提出一种轻量化YOLO(You only look once)卷积神经网络(Light-YOLOv3)模型与苹果检测方法。首先,对传统YOLOv3深度卷积神经网络架构进行改进,设计一种同构残差块串联的特征提取网络结构,简化目标检测的特征图尺度,采用深度可分离卷积替换普通卷积,提出一种融合均方误差损失和交叉熵损失的多目标损失函数;其次,开发爬虫程序,从互联网上获取训练数据并进行标注,采用数据增强技术对训练数据进行扩充,并对数据进行归一化,针对Light-YOLOv3网络训练,提出一种基于随机梯度下降(Stochastic gradient descent,SGD)和自适应矩估计(Adaptive moment estimation,Adam)的多阶段学习优化技术;最后,分别在计算机工作站和嵌入式开发板上进行了复杂果树背景下的苹果检测实验。结果表明,基于轻量化YOLOv3网络的苹果检测方法在检测速度和准确率方面均有显著的提高,在工作站和嵌入式开发板上的检测速度分别为116.96、7.59 f/s,F1值为94.57%,平均精度均值(Mean average precision,mAP)为94.69%。
An apple detection method(Light-YOLOv3)based on lightweight YOLO(You only look once)convolutional neural network was proposed for apple picking robots to detect apples quickly and accurately in the complex background of fruit trees.Firstly,in order to improve the traditional YOLOv3 deep convolutional neural network architecture,a feature extraction network structure containing cascaded homogeneous residual blocks was designed,and the dimensionality of the feature map for object detection was simplified.In this architecture,the conventional convolution was replaced by the depth wise separable convolution,and a multi-objective loss function was defined in terms of the mean square error loss and the cross entropy loss.Secondly,the training data was obtained from the Internet by means of a crawler program,and then labelled.The data augmentation technique was used to expand the training data and normalize it.Thirdly,a multi-stage learning optimization approach based on stochastic gradient descent(SGD)and adaptive moment estimation(Adam)was proposed to train Light-YOLOv3 network.Finally,an apple detection experiment in the complex background of fruit trees was performed on a computer workstation and an embedded processor,respectively.The experimental results showed that the apple detection method based on LightYOLOv3 network improved the detection speed and accuracy significantly,i.e.,the detection speed on the computer workstation and the embedded processor can reach 116.96 f/s,7.59 f/s,F1 value can reach 94.57%,and the mean average precision(mAP)can reach 94.69%.
作者
武星
齐泽宇
王龙军
杨俊杰
夏雪
WU Xing;QI Zeyu;WANG Longjun;YANG Junjie;XIA Xue(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China;Agricultural Information Institute,Chinese Academy of Agricultural Sciences,Beijing 100081,China)
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2020年第8期17-25,共9页
Transactions of the Chinese Society for Agricultural Machinery
基金
国家自然科学基金项目(61973154)
国防基础科研计划项目(JCKY2018605C004)
中央高校基本科研业务费专项资金项目(NS2019033)
南京航空航天大学研究生创新基地(实验室)开放基金项目(kfjj20190516)。
