李开敬,许燕,周建平,樊湘鹏,魏禹同

• 1:新疆大学机械工程学院

摘要(Abstract)：

为解决棉花幼苗与多种类杂草交叉生长的识别率低、鲁棒性差等问题,以棉花幼苗和田间的七类常见杂草为研究对象,提出了一种基于Faster R-CNN和数据增强的棉田苗期杂草识别方法.采集不同背景下受光照影响的杂草图像4 694张,包括晴天、阴天和雨天.通过对样本图像的数据增强和特征提取网络ResNet-101的参数优化,训练出了一种可识别棉花幼苗与多种类杂草交叉生长的Faster R-CNN网络模型.在相同样本和特征网络下,将该模型与YOLO模型进行对比.结果表明:Faster R-CNN模型在棉田苗期的多种不同杂草识别中具有明显的优势,可实现各种交叉生长的杂草目标识别,平均识别率为92.01%,平均识别时间为0.261 s.

关键词(KeyWords)： 棉花苗期;杂草识别;数据增强;Faster R-CNN

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金地区科学基金项目(51765063)

作者(Author): 李开敬,许燕,周建平,樊湘鹏,魏禹同

DOI: 10.13568/j.cnki.651094.651316.2020.06.03.0001

参考文献(References)：

• [1]WANG A,ZHANG W,WEI X,et al.A review on weed detection using ground-based machine vision and image processing techniques[J].Computers and Electronics in Agriculture,2019,158:226-240.
• [2]陈桂芬,李静,陈航,等.大数据时代人工智能技术在农业领域的研究进展[J].吉林农业大学学报,2018,40(4):502-510.CHEN G F,LI J,CHEN H,et al.Research progress of artificial intelligence technology in agriculture in the era of big data[J].Journal of Jilin Agricultural University,2018,40(4):502-510.(in Chinese)
• [3]袁亮,涂雪滢,巨刚,等.基于机器视觉的番茄实时分级系统设计[J].新疆大学学报(自然科学版),2017,34(1):11-16.YUAN L,TU X Y,JU G,et al.Design of tomato real time grading system based on machine vision[J].Journal of Xinjiang University(Natural Science Edition),2017,34(1):11-16.(in Chinese)
• [4]陈亚军,赵博,李树君,等.基于多特征的杂草逆向定位方法与试验[J].农业机械学报,2015,46(6):257-262.CHEN Y J,ZHAO B,LI S J,et al.Method and experiment of weed reverse location based on multi feature[J].Journal of Agricultural Machinery,2015,46(6):257-262.(in Chinese)
• [5]BRIVOT R,MARCHANT J.Segmentation of plants and weeds for a precision crop protection robot using infrared images[J].IEE Proceedings-Vision,Image and Signal Processing,1996,143(2):118.
• [6]李慧,祁力钧,张建华,等.基于PCA-SVM的棉花出苗期杂草类型识别[J].农业机械学报,2012,43(9):184-189+196.LI H,QI L J,ZHANG J H,et al.Identification of weed types in cotton seedling stage based on PCA-SVM[J].Journal of Aagricultural Machinery,2012,43(9):184-189+196.(in Chinese)
• [7]PIRON A,LEEMANS V,LEBEAU F,et al.Improving in-row weed detection in multispectral stereoscopic images[J].Computers and Electronics in Agriculture,2009,69(1):73-79.
• [8]HERRMANN I,SHAPIRA U,KINAST S,et al.Ground-level hyper spectral imagery for detecting weeds in wheat fields[J].Precision Agriculture,2013,14(6):637-659.
• [9]JAFARI A,MOHTASEBI SS,JAHROMI HE,et al.Weed detection in sugar beet fields using machine vision[J].International Journal of Agriculture&Biology,2006(8):602-605.
• [10]REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:Unified,real-time object detection[C].Las Vegas:Proceedings of the IEEE conference on computer vision and pattern recognition,2016.
• [11]孙哲,张春龙,葛鲁镇,等.基于Faster R-CNN的田间西兰花幼苗图像检测方法[J].农业机械学报.2019,50(7):216-221.SUN Z,ZHANG C L,GE L Z,et al.Field broccoli seedling image detection method based on fast rcnn[J].Journal of Agricultural Machinery,2019,50(7):216-221.(in Chinese)
• [12]周云成,许童羽,郑伟,等.基于深度卷积神经网络的番茄主要器官分类识别方法[J].农业工程学报.2017,33(15):219-226.ZHOU Y C,XU C Y,ZHENG W,et al.Classification and recognition of tomato main organs based on deep convo-lution neural network[J].Journal of Agricultural Engineering.2017,33(15):219-226.(in Chinese)
• [13]张庆英,葛动元,岳卫宏,等.摄像机标定精度研究中的图像及数据处理问题[J].新疆大学学报(自然科学版),2004,21(1):73-76.ZHANG Y Y,GE D Y,YUN W H,et al.Image and data processing in the study of camera calibration accuracy[J].Journal of Xinjiang University(Natural Science Edition),2004,21(1):73-76.(in Chinese)
• [14]余鹰,王乐为,张应龙.基于特征提取偏好与背景色相关性的数据增强算法[J].计算机应用,2019,39(11):3172-3177.YU Y,WANG L W,ZHANG Y L,Data enhancement algorithm based on feature extraction preference and backgrou-nd color correlation[J].Computer application,2019,39(11):3172-3177.(in Chinese)
• [15]李哲.基于GAN的不平衡数据增强算法及其应用研究[D].北京：北京交通大学,2019.LI Z.Research on GAN based unbalanced data enhancement algorithm and its application[D].Beijing:Beijing Jiaotong University,2019.(in Chinese)
• [16]RUSSAKOVSKY O,DENG J,SU H,et al.Image net large scale visual recognition challeng[J].International Journal of Computer Vision,2015,115(3):211-252.