基于无人机多光谱影像的小麦封垄前种植行识别方法改进

doi:10.3864/j.issn.0578-1752.2022.20.005

中国农业科学 ›› 2022, Vol. 55 ›› Issue (20): 3926-3938.doi: 10.3864/j.issn.0578-1752.2022.20.005

• 耕作栽培·生理生化·农业信息技术 • 上一篇下一篇

基于无人机多光谱影像的小麦封垄前种植行识别方法改进

马啸^1,²(),陈鹏飞^1,³()

¹中国科学院地理科学与资源研究所/资源与环境信息系统国家重点实验室,北京 100101
²中国科学院大学,北京 100049
³国家科技资源共享服务平台国家地球系统科学数据中心,北京 100101

收稿日期:2021-12-10 接受日期:2022-02-28 出版日期:2022-10-16 发布日期:2022-10-24
通讯作者: 陈鹏飞
作者简介:马啸,E-mail： maxiao20@mails.ucas.ac.cn。
基金资助:
国家自然科学基金(41871344);中国科学院战略先导科技专项(XDA23100100);高分辨率对地观测系统重大专项(21-Y20B01-9001-19/22);国家科技基础条件平台项目(2005DKA32300)

Improvement of Row Detection Method Before Wheat Canopy Closure Using Multispectral Images of UAV Image

MA Xiao^1,²(),CHEN PengFei^1,³()

¹Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences/State Key Laboratory of Resources and Environment Information System, Beijing 100101
²University of Chinese Academy of Science, Beijing 100049
³National Earth System Science Data Center, National Science and Technology Infrastructure of China, Beijing 100101

Received:2021-12-10 Accepted:2022-02-28 Online:2022-10-16 Published:2022-10-24
Contact: PengFei CHEN

摘要/Abstract

摘要：

【目的】为实现小麦精准管理,准确识别其种植行位置具有重要意义。本研究分别针对传统Hough变换法和绿色像元累积法存在的缺陷进行改进,并对改进前、后不同方法在小麦种植行识别上的精度进行对比分析,为小麦种植行精准提取提供技术支撑。【方法】本研究开展了小麦水、氮耦合试验,在小麦拔节前期,基于四旋翼无人机携带RedEdge M传感器获取小麦不同生长条件下多光谱影像。基于上述数据,首先采用超绿超红差分指数和Otsu方法对影像分割、分类,获取植被/土壤二值图;其次,采用3×1线型模板进行形态学开运算,降低边界不规则度并去除噪音;然后,结合无人机影像中小麦种植行排布特点,分别针对传统Hough变换法的峰值检测过程和绿色像素累积法的角度检测过程进行优化,提出改进的小麦种植行识别方法;最后,分别将两种方法改进前、后的识别结果与目视解译种植行位置结果进行对比,基于检出率和作物行识别精度（crop row detection accuracy,CRDA）评价4种方法的优劣。【结果】采用超绿超红差分指数与Otsu方法可以很好对植被/土壤进行分类,分类结果的总体精度达到93.75%,Kappa系数为0.87;形态学运算可以很好地去除图像噪声,减少后期种植行识别误差;改进后Hough变换法通过利用先验知识对峰值检测范围进行约束,有效提升了种植行检测精度,种植行平均检出率从30%提升至67%,CRDA平均值从0.22提升至0.44;改进后绿色像元累积法通过考察整幅影像的绿色像元累积特征,有效提升角度检测精度,种植行平均检出率从14%提升至93%,CRDA平均值从0.12提升至0.69;4种方法的识别精度从高到低依次为改进后绿色像元累积法、改进后Hough变换法、改进前Hough变换法、改进前绿色像元累积法。【结论】本研究较好地改进了传统种植行识别方法,为种植密度大、行间距小的小麦种植行识别提供了技术支撑。

关键词: 小麦, 种植行, Hough变换法, 绿色像元累积法, 无人机影像

Abstract:

【Objective】In order to make precise management of wheat, it is of great significance to accurately identify the location of its rows. In this study, the aim was to improve the traditional Hough transform-based methods and the green pixel accumulation- based methods respectively, and then to analyze the different methods before and after the improvement for wheat row detection, so as to provide the technical support for accurate detection of wheat rows.【Method】A wheat water-nitrogen coupling experiment was established. During the early jointing stage, the multispectral images of wheat under different growth conditions was obtained by the RedEdge M sensors mounted on a four rotor unmanned aerial vehicle (UAV). First, based on these data, these images of the excess green minus excess red (ExGR) index were calculated, and then the Otsu method was performed on it to classify image pixels into vegetation and soil. Second, 3×1 line template was used for morphological opening operation to reduce boundary irregularity and remove noise. Third, according to the characteristics of the wheat rows in the image, the modified wheat row extraction methods were proposed based on the optimization of the peak detection process of the traditional Hough transform-based method and the angel detection process of the traditional green pixel accumulation-based method, respectively. Finally, the detection results before and after improving the two methods were compared with the result of visually interpreted wheat rows, and their performances were evaluated by using the detection rate and crop row detection accuracy (CRDA). 【Result】The combination of the ExGR index and OTSU method could accurately identify the vegetation pixels and soil pixels automatically, and the overall accuracy was 93.75% and the Kappa was 0.87. The morphological opening operation could remove the pattern noise and reduce the error of later crop row detection. The modified Hough transform-based method effectively improved the peak detection accuracy by restricting the peak detection range with prior knowledge; Compare with traditional Hough transform-based methods, the averaged detection rate has increased from 30% to 67%, and the averaged CRDA has increased from 0.22 to 0.44. The modified green pixel accumulation-based method investigated the accumulation characteristics of the green pixels of the entire image, which effectively improved the angel detection accuracy; Compare with traditional green pixel accumulation-based method, the averaged detection rate has increased from 14% to 93%, and the average CRDA has increased from 0.12 to 0.69. The identification accuracy of the four methods from high to low were the modified green pixel accumulation-based method, the modified Hough transform-based method, the traditional Hough transform-based method, and the traditional green pixel accumulation-based method. 【Conclusion】This study improved the traditional crop row detection methods, which provided the technical support for the identification of wheat rows with high planting density and small row spacing.

Key words: wheat, planting rows, Hough transform-based method, green pixel accumulation-based method, UAV image

马啸,陈鹏飞. 基于无人机多光谱影像的小麦封垄前种植行识别方法改进[J]. 中国农业科学, 2022, 55(20): 3926-3938.

MA Xiao,CHEN PengFei. Improvement of Row Detection Method Before Wheat Canopy Closure Using Multispectral Images of UAV Image[J]. Scientia Agricultura Sinica, 2022, 55(20): 3926-3938.

0
/ / 推荐

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: https://www.chinaagrisci.com/CN/10.3864/j.issn.0578-1752.2022.20.005

https://www.chinaagrisci.com/CN/Y2022/V55/I20/3926

图/表 12

图1

图2

图3

图4

图5

图6

图7

图8

表1

图9

图10

图11

参考文献 34

[1]	刘兆辉, 吴小宾, 谭德水, 李彦, 江丽华. 一次性施肥在我国主要粮食作物中的应用与环境效应. 中国农业科学, 2018, 51(20): 3827-3839.
	LIU Z H, WU X B, TAN D S, LI Y, JIANG L H. Application and environmental effects of one-off fertilization technique in major cereal crops in China. Scientia Agricultura Sinica, 2018, 51(20): 3827-3839. (in Chinese)
[2]	霍李龙, 苗芳, 贾丽芳, 王长发. 种植方式对关中灌区冬小麦冠层光合及产量性状的影响. 麦类作物学报, 2017, 37(8): 1098-1104.
	HUO L L, MIAO F, JIA L F, WANG C F. Effect of planting pattern on canopy photosynthesis and yield traits of winter wheat in the irrigation area of central Shaanxi. Journal of Triticeae Crops, 2017, 37(8): 1098-1104. (in Chinese)
[3]	ZHANG X Y, LI X N, ZHANG B H, ZHOU J, TAIN G Z, XIONG Y J, GU B X. Automated robust crop-row detection in maize fields based on position clustering algorithm and shortest path method. Computers and Electronics in Agriculture, 2018, 154: 165-175. doi: 10.1016/j.compag.2018.09.014
[4]	李祥光, 赵伟, 赵雷雷. 缺株玉米行中心线提取算法研究. 农业工程学报, 2021, 37(18): 203-210.
	LI X G, ZHAO W, ZHAO L L. Extraction algorithm of the center line of maize row in case of plants lacking. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(18): 203-210. (in Chinese)
[5]	PLA F, SANCHIZ J M, MARCHANT J A, BRIVOT R. Building perspective models to guide a row crop navigation vehicle. Image and Vision Computing, 1997, 15(6): 465-473. doi: 10.1016/S0262-8856(96)01147-X
[6]	GARCÍA-SANTILLÁN I D, MONTALVO M, GUERRERO J M, PAJARES G. Automatic detection of curved and straight crop rows from images in maize fields. Biosystems Engineering, 2017, 156: 61-79. doi: 10.1016/j.biosystemseng.2017.01.013
[7]	赵学观, 马伟, 高原源, 臧云飞, 何义川, 王秀. 基于逆透视变换的条播作物早期作物行识别. 江苏大学学报(自然科学版), 2019, 40(6): 668-675.
	ZHAO X G, MA W, GAO Y Y, ZANG Y F, HE Y C, WANG X. Identification of early crop row for drillcrops based on reverse perspective transformation. Journal of Jiangsu University(Natural Science Edition), 2019, 40(6): 668-675. (in Chinese)
[8]	孙刚, 黄文江, 陈鹏飞, 高帅, 王秀. 轻小型无人机多光谱遥感技术应用进展. 农业机械学报, 2018, 49(3): 1-17.
	SUN G, HUANG W J, CHEN P F, GAO S, WANG X. Advances in UAV-based multispectral remote sensing applications. Transactions of the Chinese Society for Agricultural Machinery, 2018, 49(3): 1-17. (in Chinese)
[9]	PÉREZ-ORTIZ M, PEÑA J M, GUTIÉRREZ P A, TORRES- SÁNCHEZ J, HERVÁS-MARTÍNEZ C, LÓPEZ-GRANADOS F. A semi-supervised system for weed mapping in sunflower crops using unmanned aerial vehicles and a crop row detection method. Applied Soft Computing, 2015, 37: 533-544. doi: 10.1016/j.asoc.2015.08.027
[10]	郑晓岚, 张显峰, 程俊毅, 任翔. 利用无人机多光谱影像数据构建棉苗株数估算模型. 中国图象图形学报, 2020, 25(3): 520-534.
	ZHENG X L, ZHANG X F, CHENG J Y, REN X. Using the multispectral image data acquired by unmanned aerial vehicle to build an estimation model of the number of seeding stage cotton plants. Journal of Image and Graphics, 2020, 25(3): 520-534. (in Chinese)
[11]	OLIVEIRA H C, GUIZILINI V C, NUNES I P, SOUZA J R. Failure detection in row crops from UAV images using morphological operators. IEEE Geoscience and Remote Sensing Letters, 2018, 15(7): 991-995. doi: 10.1109/LGRS.2018.2819944
[12]	HASSANEIN M, KHEDR M, EL-SHEIMY N. Crop row detection procedure using low-cost UAV imagery system. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2019, XLII-2/W13: 349-356.
[13]	PEÑA J M, TORRES-SÁNCHEZ J, SERRANO-PÉREZ A, DE CASTRO A I, LÓPEZ-GRANADOS F. Quantifying efficacy and limits of unmanned aerial vehicle (UAV) technology for weed seedling detection as affected by sensor resolution. Sensors, 2015, 15(3): 5609-5626. doi: 10.3390/s150305609 pmid: 25756867
[14]	TENHUNEN H, PAHIKKALA T, NEVALAINEN O, TEUHOLA J, MATTILA H, TYYSTJÄRVI E. Automatic detection of cereal rows by means of pattern recognition techniques. Computers and Electronics in Agriculture, 2019, 162: 677-688. doi: 10.1016/j.compag.2019.05.002
[15]	CHEN P F, MA X, WANG F Y, LI J. A new method for crop row detection using unmanned aerial vehicle images. Remote Sensing, 2021, 13(17): 3526. doi: 10.3390/rs13173526
[16]	ZHAO C J, CHEN P F, HUANG W J, WANG J H, WANG Z J, JIANG A N. Effects of two kinds of variable-rate nitrogen application strategies on the production of winter wheat (Triticum aestivum). New Zealand Journal of Crop and Horticultural, 2009, 37(2): 149-155.
[17]	OLSON D, CHATTERJEE A, FRANZEN D W, DAY S S. Relationship of drone-based vegetation indices with corn and sugarbeet yields. Agronomy Journal, 2019, 111(5): 2545-2557. doi: 10.2134/agronj2019.04.0260
[18]	SØGAARD H T, OLSEN H J. Determination of crop rows by image analysis without segmentation. Computer and Electronics in Agriculture, 2003, 38(2): 141-158. doi: 10.1016/S0168-1699(02)00140-0
[19]	HOUGH P V C. Method and means for recognizing complex patterns: US Patent 3069654. 1962-12-18 [2021-12-07].
[20]	苏伟, 蒋坤萍, 闫安, 刘哲, 张明郑, 王伟. 基于无人机遥感影像的育种玉米垄数统计监测. 农业工程学报, 2018, 34(10): 92-98.
	SU W, JIANG K P, YAN A, LIU Z, ZHANG M Z, WANG W. Monitoring of planted lines for breeding corn using UAV remote sensing image. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(10): 92-98. (in Chinese)
[21]	MEYER G E, NETO J C. Verification of color vegetation indices for automated crop imaging applications. Computers and Electronics in Agriculture, 2008, 63(2): 282-293. doi: 10.1016/j.compag.2008.03.009
[22]	KERKECH M, HAFIANE A, CANALS R. Deep leaning approach with colorimetric spaces and vegetation indices for vine diseases detection in UAV images. Computers and Electronics in Agriculture, 2018, 155: 237-243. doi: 10.1016/j.compag.2018.10.006
[23]	高铭阳, 张锦水, 潘耀忠, 段雅鸣, 张杜娟. 结合植被指数与作物高度反演冬小麦叶面积指数. 中国农业资源与区划, 2020, 41(8): 49-57.
	GAO M Y, ZHANG J S, PAN Y Z, DUAN Y M, ZHANG D J. Retrieval of winter wheat leaf area index based on vegetation index and crop height. Chinese Journal of Agricultural Resources and Regional Planning, 2020, 41(8): 49-57. (in Chinese)
[24]	OTSU N. A threshold selection method from gray-level histograms. IEEE Transactions on Systems, Man, and Cybernetics, 1979, 11: 62-66.
[25]	王宁, 陈方, 于博. 基于形态学开运算的面向对象滑坡提取方法研究. 遥感技术与应用, 2018, 33(3): 520-529.
	WANG N, CHEN F, YU B. A object-oriented landslide extraction method based on morphological opening operation. Remote Sensing Technology and Application, 2018, 33(3): 520-529. (in Chinese)
[26]	姜国权, 柯杏, 杜尚丰, 陈娇. 基于机器视觉和随机方法的作物行提取算法. 农业机械学报, 2008, 39(11): 85-88, 93.
	JIANG G Q, KE X, DU S F, CHEN J. Detection algorithm of crop rows based on machine vision and randomized method. Transactions of the Chinese Society for Agricultural Engineering, 2008, 39(11): 85-88, 93. (in Chinese)
[27]	ZHAO H X, WANG Y X, SUN Z B, XU Q, LIANG D. Failure detection in eucalyptus plantation based on UAV images. Forests, 2021, 12(9): 1250. doi: 10.3390/f12091250
[28]	VIDOVIĆ I, CUPEC R, HOCENSKI Z. Crop row detection by global energy minimization. Pattern Recognition, 2016, 55: 68-86. doi: 10.1016/j.patcog.2016.01.013
[29]	JIANG G Q, WANG Z H, LIU H M. Automatic detection of crop rows based on multi-ROIs. Expert Systems with Applications, 2015, 42(5): 2492-2441.
[30]	PUSDÁ-CHULDE M, ROBAYO A, GIUSTI A D, GARCÍA- SANTILLÁN I. Detection of crop lines and weeds in corn fields based on images obtained from a drone //NAIOUF M, RUCCI E, CHICHIZOLA F, DE GIUSTI L. Cloud Computing, Big Data & Emerging Topics. Cham: Springer, 2021: 31-45.
[31]	BAH M D, HAFIANE A, CANALS R. Crownet: Deep network for crop row detection in UAV image. IEEE Access, 2020, 8: 5189-5200. doi: 10.1109/ACCESS.2019.2960873
[32]	陈子文, 李伟, 张文强, 李云伍, 李明生, 李慧. 基于自动Hough变换累加阈值的蔬菜作物行提取方法研究. 农业工程学报, 2019, 35(22): 314-322.
	CHEN Z W, LI W, ZHANG W Q, LI Y W, LI M S, LI H. Vegetable crop row extraction method based on accumulation threshold of Hough transformation. Transactions of the Chinese Society of Agricultural Engineering, 2019, 35(22): 314-322. (in Chinese)
[33]	COMBA L, GAY P, PRIMICERIO J, AIMONINO D R. Vineyard detection from unmanned aerial systems images. Computers and Electronics in Agriculture, 2015, 114: 78-87. doi: 10.1016/j.compag.2015.03.011
[34]	王侨, 孟志军, 付卫强, 刘卉, 张振国. 基于机器视觉的玉米苗期多条作物行线检测算. 农业机械学报, 2021, 52(4): 208-220.
	WANG Q, MENG Z J, FU W Q, LIU H, ZHANG Z G. Detection algorithm of multiple crop row lines based on machine vision in maize seedling stage. Transactions of the Chinese Society for Agricultural Machinery, 2021, 52(4): 208-220. (in Chinese)

分类结果 Classification result	参考数据 Reference data
分类结果 Classification result	植被 Vegetation	土壤 Soil	总计 Summary	用户精度 User’s accuracy (%)
植被 Vegetation	1346	97	1443	93.28%
土壤 Soil	103	1654	1757	94.14%
总计 Summary	1449	1751	3200
生产者精度 Producer‘s accuracy (%)	92.89%	94.46%

基于无人机多光谱影像的小麦封垄前种植行识别方法改进

Improvement of Row Detection Method Before Wheat Canopy Closure Using Multispectral Images of UAV Image

RichHTML

PDF

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 34

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	陈吉浩, 周界光, 曲翔汝, 王素容, 唐华苹, 蒋云, 唐力为, $\boxed{\hbox{兰秀锦}}$, 魏育明, 周景忠, 马建. 四倍体小麦胚大小性状QTL定位与分析[J]. 中国农业科学, 2023, 56(2): 203-216.
[2]	严艳鸽, 张水勤, 李燕婷, 赵秉强, 袁亮. 葡聚糖改性尿素对冬小麦产量和肥料氮去向的影响[J]. 中国农业科学, 2023, 56(2): 287-299.
[3]	徐久凯, 袁亮, 温延臣, 张水勤, 李燕婷, 李海燕, 赵秉强. 畜禽有机肥氮在冬小麦季对化肥氮的相对替代当量[J]. 中国农业科学, 2023, 56(2): 300-313.
[4]	古丽旦,刘洋,李方向,成卫宁. 小麦吸浆虫小热激蛋白基因Hsp21.9的克隆及在滞育过程与温度胁迫下的表达特性[J]. 中国农业科学, 2023, 56(1): 79-89.
[5]	王浩琳,马悦,李永华,李超,赵明琴,苑爱静,邱炜红,何刚,石美,王朝辉. 基于小麦产量与籽粒锰含量的磷肥优化管理[J]. 中国农业科学, 2022, 55(9): 1800-1810.
[6]	唐华苹,陈黄鑫,李聪,苟璐璐,谭翠,牟杨,唐力为,兰秀锦,魏育明,马建. 基于55K SNP芯片的普通小麦穗长非条件和条件QTL分析[J]. 中国农业科学, 2022, 55(8): 1492-1502.
[7]	马小艳,杨瑜,黄冬琳,王朝辉,高亚军,李永刚,吕辉. 小麦化肥减施与不同轮作方式的周年养分平衡及经济效益分析[J]. 中国农业科学, 2022, 55(8): 1589-1603.
[8]	刘硕,张慧,高志源,许吉利,田汇. 437个小麦品种钾收获指数的变异特征[J]. 中国农业科学, 2022, 55(7): 1284-1300.
[9]	王洋洋,刘万代,贺利,任德超,段剑钊,胡新,郭天财,王永华,冯伟. 基于多元统计分析的小麦低温冻害评价及水分效应差异研究[J]. 中国农业科学, 2022, 55(7): 1301-1318.
[10]	职蕾,者理,孙楠楠,杨阳,Dauren Serikbay,贾汉忠,胡银岗,陈亮. 小麦苗期铅耐受性的全基因组关联分析[J]. 中国农业科学, 2022, 55(6): 1064-1081.
[11]	秦羽青,程宏波,柴雨葳,马建涛,李瑞,李亚伟,常磊,柴守玺. 中国北方地区小麦覆盖栽培增产效应的荟萃（Meta）分析[J]. 中国农业科学, 2022, 55(6): 1095-1109.
[12]	蔡苇荻,张羽,刘海燕,郑恒彪,程涛,田永超,朱艳,曹卫星,姚霞. 基于成像高光谱的小麦冠层白粉病早期监测方法[J]. 中国农业科学, 2022, 55(6): 1110-1126.
[13]	宗成, 吴金鑫, 朱九刚, 董志浩, 李君风, 邵涛, 刘秦华. 添加剂对农副产物和小麦秸秆混合青贮发酵品质的影响[J]. 中国农业科学, 2022, 55(5): 1037-1046.
[14]	马鸿翔, 王永刚, 高玉姣, 何漪, 姜朋, 吴磊, 张旭. 小麦抗赤霉病育种回顾与展望[J]. 中国农业科学, 2022, 55(5): 837-855.
[15]	冯子恒,宋莉,张少华,井宇航,段剑钊,贺利,尹飞,冯伟. 基于无人机多光谱和热红外影像信息融合的小麦白粉病监测[J]. 中国农业科学, 2022, 55(5): 890-906.