基于Faster R-CNN的枸杞开花期与果实成熟期识别技术  

Automatic Identification Technology of Lycium barbarum Flowering Period and Fruit Ripening Period Based on Faster R-CNN

作　　者：朱永宁 周望[2] 杨洋 李剑萍[1,3,4] 李万春[1,3,4] 金红伟 房峰 ZHU Yong-ning;ZHOU Wang;YANG Yang;LI Jian-ping;LI Wan-chun;JIN Hong-wei;FANG Feng(Key Laboratory for Meteorological Disaster Monitoring and Early Warning and Risk Management of Characteristic Agriculture in Arid Regions,CMA,Ningxia Yinchuan 750002,China;Aerospace Newsky Technology Co.Ltd,Wuxi 214000;Key Laboratory of Meteorological Disaster Prevention and Reduction of Ningxia,Yinchuan 750002;Ningxia Meteorological Science Institute,Yinchuan 750002)

机构地区：[1]中国气象局旱区特色农业气象灾害监测预警与风险管理重点实验室,银川750002 [2]航天新气象科技有限公司,无锡214000 [3]宁夏气象防灾减灾重点实验室,银川750002 [4]宁夏气象科学研究所,银川750002

出　　处：《中国农业气象》2020年第10期668-677,共10页Chinese Journal of Agrometeorology

基　　金：中国气象局旱区特色农业气象灾害监测预警与风险管理重点实验室开放研究基金(CAMF-201813);第四批宁夏青年科技人才托举工程项目(TJGC2019058);宁夏回族自治区重点研发计划(2019BEH03008);宁夏回族自治区重点研发项目(2017BY080)。

摘　　要：以宁夏16套枸杞农田实景监测系统2018年和2019年拍摄的图像作为资料,结合枸杞开花期和果实成熟期的植物学特征,利用更快速的基于区域的卷积神经网络(Faster R-CNN)方法对图像进行训练、分类,构建枸杞开花期和果实成熟期的识别算法,以平均精确率(AP)和平均精度均值(mAP)作为模型的评价指标,并将自动识别结果与专家目视判断结果和田间观测记录进行对比。结果表明:当网络结构中重要超参数批尺寸(batch size)和迭代次数(iterations)分别取值64和20000时,mAP值达到0.74,在测试集上对花和果实的识别效果好于其它参数。基于Faster R-CNN判识的枸杞开花期和果实成熟期与专家目视判断的差异在2~5d,这两种方法的判断对象和判断标准一致,可比性强,专家目视判断的结果可以作为自动识别技术的验证标准,用来优化并调整算法。自动识别结果与同期田间观测记录的差异在0~12d,差异的主要原因是这两种方法的判识对象和标准不一致,难以利用田间观测的结果优化自动识别算法。From 2018 to 2019,16 sets of Lycium barbarum farmland monitoring systems had been built in Ningxia.Each system took 10 images every day,and over 30,000 images of the growth of Lycium barbarum trees were taken in two years.To study the recognition technology of the flowering period and fruit ripening period of Lycium barbarum based on these images,three methods were used in this paper to judge the developmental stage of Lycium barbarum.The first one was the field observation method.In this method,two fields where the real-life monitoring system was installed were selected,and the Lycium barbarum trees in the two fields were manually observed once in every two days during the growing season.The Lycium barbarum trees selected by manual observation should be consistent with the ones photographed by the farmland monitoring systems.The second method was expert visual judgment,in which 5 experienced experts were invited to judge all the images.The judgment standard was as follows.If there were 5 features in a certain developmental period in an image,it was considered that this Lycium barbarum tree had reached the universal period of this developmental period.If 5 out of 10 images on a certain day reached the universal period of this developmental period,it was considered that the Lycium barbarum population in the filed had entered this developmental period.Based on the opinions of the experts,the result of the expert visual judgment was given.The third method is the automatic recognition method.In this method,more than 3000 images with characteristics of Lycium barbarum flowering and fruit ripening were screened out from all the images.Removed the images with lens fouling or unsatisfactory field of view,and finally,the number of remaining image samples was 1210.To avoid the phenomenon of underfitting or overfitting due to too few or too many images of a certain category involved in training,rotation,cropping and flipping were used for data enhancement.The data enhanced samples were divided according to the format of the

关 键 词：枸杞 开花期识别 果实成熟期识别 发育期识别 Faster R-CNN 图像识别 

分 类 号：S567.19[农业科学—中草药栽培] S127[农业科学—作物学]