基于改进K-means算法的冬小麦覆盖度提取研究

doi:10.13304/j.nykjdb.2021.0941

摘要/Abstract

摘要：

为快速、精准地提取冬前分蘖期冬小麦覆盖度，提出了一种基于改进K-means算法的冬小麦覆盖度提取方法。首先将冬小麦图像转换到Lab色彩空间，其次利用蜉蝣算法（Mayfly Algorithm， MA）获取K-means最优初始聚类中心，并用马氏距离代替欧氏距离进行算法改进，最后利用分割得到的二值图像计算冬小麦覆盖度。结果显示，该方法的平均分割精度和平均处理时间分别为94.66%和2.03 s，与过绿指数（excess green，EXG）自适应阈值分割和基于粒子群优化算法（particle swarm optimization，PSO）的K-means（PSO-K-means）分割相比，分割精度分别提高了12.04%和4.18%，处理时间分别减少了2.26和2.94 s。该方法分割效果优于EXG和PSO-K-means分割方法，可用于提取冬小麦覆盖度。

关键词: 冬小麦覆盖度, 改进K-means算法, Lab色彩空间, 蜉蝣算法, 马氏距离

Abstract:

In order to rapidly and accurately extract winter wheat coverage at tillering stage before winter， a new method based on the improved K-means Algorithm was proposed. Firstly， the image of winter wheat was transformed to Lab color space. Secondly， the initial clustering center of the K-means Algorithm was obtained by using the Mayfly Algorithm（MA）. At the same time， the Algorithm was improved by using Mahalanobis distance instead of euclidean distance. Finally， the coverage of winter wheat was calculated by using the segmented binary image. The test results of 100 winter wheat images showed that the average segmentation accuracy and the average processing time of the proposed method were 94.66% and 2.03 s， respectively. Compared with excess green （EXG） adaptive threshold segmentation and K-means segmentation based on particle swarm optimization algorithm（PSO-K-means）， the average segmentation accuracy increased by 12.04% and 4.18%， respectively. And the average processing time was reduced by 2.26 and 2.94 s， respectively. The results showed that the segmentation effect of this method was better than the other two segmentation methods， and it could be used to extract the coverage of winter wheat.

Key words: coverage of winter wheat, improved K-means algorithm, lab color space, Mayfly Algorithm, Mahalanobis distance

中图分类号:

S126

赵文昊, 姬江涛, 马淏, 金鑫, 李雪, 马海港. 基于改进K-means算法的冬小麦覆盖度提取研究[J]. 中国农业科技导报, 2023, 25(1): 83-91.

Wenhao ZHAO, Jiangtao JI, Hao MA, Xin JIN, Xue LI, Haigang MA. Extraction of Winter Wheat Coverage Based on Improved K-means Algorithm[J]. Journal of Agricultural Science and Technology, 2023, 25(1): 83-91.

图/表 7

图1 算法流程

Fig. 1 Flow of Algorithm

图2 蜉蝣算法流程

Fig. 2 Flow of the Mayfly Algorithm

图3 不同光照、疏密条件下3种方法分割效果

Fig. 3 Segmentation effect by 3 methods under different illumination and density conditions

图4 不同通道的颜色值

Fig. 4 Color values of different channels

图5 本文方法在不同颜色空间下的分割效果

Fig. 5 Segmentation effect of this method under different color spaces

图6 不同分割方法的分割准确率

Fig. 6 Segmentation accuracy under different segmentation methods

表1 不同方法下冬小麦分割性能

Table 1 Segmentation performance of winter wheat under different methods

方法 Method	准确率 Accuracy/%	平均处理时间 Average processing time/s
PSO-K-means分割 PSO-K-means segmentation	89.85	4.97
EXG自适应阈值分割 EXG adaptive threshold segmentation	82.62	4.29
本文方法 Method of this paper	94.66	2.03

参考文献 27

1	何中虎,庄巧生,程顺和,等.中国小麦产业发展与科技进步[J].农学学报, 2018, 8(1):99-106.
	HE Z H, ZHUANG Q S, CHENG S H, et al.. Wheat production and technology improvement in China [J]. J. Agric., 2018, 8(1):99-106.
2	张建朝,赵杨,牛娜,等.新型化学杀雄剂与表面活化剂组合筛选及其诱导小麦雄性不育的效果[J].南方农业学报, 2020, 51(3):531-536.
	ZHANG J C, ZHAO Y, NIU N, et al.. Screening of new chemical hybriding agents with surfactants and the effects on male sterility induction in wheat [J]. J. Southern Agric., 2020, 51(3):531-536.
3	秦伟,朱清科,张学霞,等.植被覆盖度及其测算方法研究进展[J].西北农林科技大学学报(自然科学版), 2006, 34(9):163-170.
	QIN W, ZHU Q K, ZHANG X X, et al.. Review of vegetation covering and its measuring and calculating method [J]. J. Northwest A&F Univ. (Nat. Sci.), 2006, 34(9):163-170.
4	CAI T, MENG X P, LIU T N, et al.. Exogenous hormonal application regulates the occurrence of wheat tillers by changing endogenous hormones [J/OL]. Front. Plant Sci., 2018, 9(21):1886 [2022-11-25]. .
5	ADAMS J E, ARKIN G F. A light interception method for measuring row crop ground cover1 [J]. Soil. Sci. Soc. Am. J., 1977, 41(4): 789-792.
6	卢艳丽,胡昊,白由路,等.植被覆盖度对冬小麦冠层光谱的影响及定量化估产研究[J].麦类作物学报, 2010, 30(1):96-100.
	LU Y L, HU H, BAI Y L, et al.. Effects of vegetation coverage on the canopy spectral and yield quantitative estimation in wheat [J]. J. Triticeae Crops, 2010, 30(1):96-100.
7	SONG W J, MU X H, RUAN G Y, et al.. Estimating fractional vegetation cover and the vegetation index of bare soil and highly dense vegetation with a physically based method [J]. Int. J. Appl. Earth Obs., 2017, 58:168-176.
8	皮新,曾永,贺城.融合多源遥感数据的高分辨率城市植被覆盖度估算[J].遥感学报, 2021, 25(6):1216-1226.
	PI X, ZENG Y, HE C. High-resolution urban vegetation coverage estimation based on multi-source remote sensing data fusion [J]. J. Remote Sens., 2021, 25(6):1216-1226.
9	随金明,宋乃平,薛毅,等.基于照相法的荒漠草原植被覆盖度测定方法[J].中国草地学报, 2019, 41(2):58-64.
	SUI J M, SONG N P, XUE Y, et al.. The method of extracting vegetation coverage in the desert steppe based on photographic [J]. Chin. J. Grassland., 2019, 41(2):58-64.
10	刘哲,乌伟,张善文,等.基于多元岭回归估计小麦种植密度[J].中国农业科技导报, 2020, 22(6):72-80.
	LIU Z, WU W, ZHANG S W, et al.. Estimating wheat planting density based on multiple ridge regression [J]. J. Agric. Sci. Technol., 2020, 22(6):72-80.
11	黄巧义,张木,李苹,等.支持向量机和最大类间方差法结合的水稻冠层图像分割方法[J].中国农业科技导报, 2019, 21(4):52-60.
	HUANG Q Y, ZHANG M, LI P, et al.. Rice canopy image segmentation using support vector machine and otsu’s method [J]. J. Agric. Sci. Technol., 2019, 21(4):52-60.
12	牛亚晓,张立元,韩文霆.基于Lab颜色空间的棉花覆盖度提取方法研究[J].农业机械学报, 2018, 49(10):240-249.
	NIU Y X, ZHANG L Y, HAN W T. Extraction methods of cotton coverage based on lab color space [J]. Trans. Chin. Soc. Agric. Mach., 2018, 49(10):240-249.
13	梁丽秀,叶军立,吴丹,等.水稻表层根系图像分割算法研究[J].中国农业科技导报, 2019, 21(1):70-79.
	LIANG L X, YE J L, WU D, et al.. Study on image segmentation algorithm of rice surface root [J]. J. Agric. Sci. Technol., 2019, 21(1):70-79.
14	吴焕丽,崔可旺,张馨,等.基于改进K-means图像分割算法的细叶作物覆盖度提取[J].农业机械学报, 2019, 50(1):42-50.
	WU H L, CUI K W, ZHANG X, et al.. Improving accuracy of fine leaf crop coverage by improved K-means algorithm [J]. Trans. Chin. Soc. Agric. Mach., 2019, 50(1):42-50
15	ZHENG L, ZHANG J, WANG Q. Mean-shif-based color segmentation of images containing green vegetation [J]. Comput. Electron. Agric., 2009, 65(1): 93-98.
16	孙金英,黄云,曹宏鑫,等.油菜地上生物量与冠层光谱植被指数之间的相关分析[J].南方农业学报, 2009, 40(6):716-723.
	SUN J Y, HUANG Y, CAO H X, et al.. Correlation analysis between aboveground biomass of rape and canopy spectral vegetation index [J]. J. Southern Agric., 2009, 40(6):716-723.
17	陈刘凤,林开平,胡宝清,等.基于Landsat8_OLI数据的甘蔗种植面积监测[J].南方农业学报, 2015, 46(11):2068-2072.
	CHEN L F, LIN K P, HU B Q, et al.. Monitoring of sugarcane planting area based on Landsat8 OLI data [J]. J. Southern Agric., 2015, 46(11):2068-2072.
18	LAURENT K, PIERRE H, CECILE D, et al.. Dry-season vegetation mass and cover fraction from SWIR1.6 and SWIR2.1 band ratio: ground-radiometer and MODIS data in the sahel [J]. Int. J. Appl. Earth. Obs., 2015, 39:56-64.
19	刘昕,杨光,林群,等.基于连续小波变换的蒺藜植被覆盖度预测[J].中国农业科技导报,2021,23(10):107-116.
	LIU X, YANG G, LIN Q, et al.. Prediction on the vegetation coverage of tribulus terrestris l. based on the continuous wavelet transform [J]. J. Agric. Sci. Technol., 2021,23(10):107-116.
20	冯海英,冯仲科,冯海霞.一种基于无人机高光谱数据的植被盖度估算新方法[J].光谱学与光谱分析, 2017, 37(11):3573-3578.
	FENG H Y, FENG Z K, FENG H X. A new method for estimating the fractional vegetation cover based on UVA hyperspectrum [J]. Spectrosc. Spect. Anal., 2017, 37(11):3573-3578.
21	李冰,刘镕源,刘素红,等.基于低空无人机遥感的冬小麦覆盖度变化监测[J].农业工程学报, 2012, 28(13):160-165.
	LI B, LIU R Y, LIU S H, et al..Monitoring vegetation coverage variation of winter wheat by low-altitude UAV remote sensing system [J]. Trans. Chin. Soc. Agric. Eng., 2012, 28(13):160-165.
22	BAI X D, CAO Z G, WANG Y, et al.. Crop segmentation from images by morphology modeling in the CIE Lab* color space[J]. Comput. Electron. Agric., 2013, 99:21-34.
23	刘哲,黄文准,王利平.基于改进K-means聚类算法的大田麦穗自动计数[J].农业工程学报, 2019, 35(3):174-181.
	LIU Z, HUANG W Z, WANG L P. Field wheat ear counting automatically based on improved K-means clustering algorithm [J]. Trans. Chin. Soc. Agric. Eng., 2019, 35(3):174-181.
24	ZERVOUDAKIS K, TSAFARAKIS S. A mayfly optimization algorithm [J/OL]. Com. Ind. Eng., 2020, 145:106559 [2021-11-26]. .
25	李华,王菊香,邢志娜,等.改进的K/S算法对近红外光谱模型传递影响的研究[J].光谱学与光谱分析, 2011, 31(2):362-365.
	LI H, WANG J X, XING Z N, et al.. Influence of improved kennard/stone algorithm on the calibration transfer in near-infrared spectroscopy [J]. Spectrosc. Spect. Anal., 2011, 31(2):362-365.
26	祁力钧,程一帆,程浈浈,等.基于M-K聚类法的果树上下冠层体积比测算[J].农业机械学报, 2018, 49(5):57-64.
	QI L J, CHENG Y F, CHENG Z Z, et al.. Estimation of upper and lower canopy volume ratio of fruit trees based on M-K clustering [J]. Trans. Chin. Soc. Agric. Mach., 2018, 49(5):57-64.
27	陈梦涛,余粟.基于Lab颜色空间的自适应K均值彩色图像分割方法[J].软件导刊, 2021, 20(6):230-234.
	CHEN M T, YU S. Adaptive K-means color image seqmentation method based on Lab color space [J] Software Guide, 2021, 20(6):230-234.

[1]	张静, 郭思梦, 韩迎春, 雷亚平, 邢芳芳, 杜文丽, 李亚兵, 冯璐. 基于无人机RGB图像的棉花产量估算[J]. 中国农业科技导报, 2022, 24(11): 112-120.
[2]	黄昊, 谢圣桥, 陈度, 王恒. 深度学习在苹果产业链中的应用与研究进展[J]. 中国农业科技导报, 2022, 24(10): 79-89.
[3]	卢闯, 胡海棠, 覃苑, 淮贺举, 李存军. 基于无人机多光谱影像的春玉米田管理分区研究[J]. 中国农业科技导报, 2022, 24(9): 106-115.
[4]	赵越, 卫勇, 单慧勇, 穆志民, 张健欣, 吴海云, 赵辉, 胡建龙. 基于深度学习的高分辨率麦穗图像检测方法[J]. 中国农业科技导报, 2022, 24(9): 96-105.
[5]	张文麒, 吴升, 郭新宇, 温维亮, 卢宪菊, 赵春江. 植株自旋转多视角重建技术在小麦植株三维表型获取中的应用评估[J]. 中国农业科技导报, 2022, 24(8): 87-98.
[6]	刘海涛, 韩鑫, 兰玉彬, 伊丽丽, 王宝聚, 崔立华. 基于YOLOv4网络的棉花顶芽精准识别方法[J]. 中国农业科技导报, 2022, 24(8): 99-108.
[7]	赵瑞雪, 孙坦, 杨晓蓉, 王剑. 数据密集型环境下农业科研信息化发展路径研究[J]. 中国农业科技导报, 2022, 24(7): 1-7.
[8]	方逵, 李成, 何潇, 陈益能. 基于三维重建的多角度葡萄叶病害识别方法研究[J]. 中国农业科技导报, 2022, 24(7): 86-96.
[9]	郑建华, 赵瑞雪, 赵华, 李思经, 王国刚. 智慧农业信息服务发展的国际经验与启示[J]. 中国农业科技导报, 2022, 24(6): 9-18.
[10]	李炜, 朱德利, 王青, 曾绍华. 监测生长状态和环境响应的作物数字孪生系统研究综述[J]. 中国农业科技导报, 2022, 24(6): 90-105.
[11]	吴建伟, 黄杰, 熊晓菲, 高晗, 秦向阳. 基于AI的桃树病害智能识别方法研究与应用[J]. 中国农业科技导报, 2022, 24(5): 111-118.
[12]	赵管乐, 刘勤, 彭培好. 基于Sentinel⁃2影像与MaxEnt模型识别云南华坪县芒果种植区[J]. 中国农业科技导报, 2022, 24(3): 111-119.
[13]	许鑫, 马兆务, 熊淑萍, 马新明, 程涛, 李海洋, 赵锦鹏. 基于气候年型的河南省冬小麦产量预测[J]. 中国农业科技导报, 2022, 24(2): 136-144.
[14]	顾生浩, 卢宪菊, 王勇健, 郭新宇, . 数字孪生系统在农业生产中的应用探讨[J]. 中国农业科技导报, 2021, 23(10): 82-89.
[15]	王传旭, 王康, 陈林, 李学, 张红伟. 基于Kriging插值和BP神经网络结合的粮仓温度场预测模型研究及实现[J]. 中国农业科技导报, 2021, 23(9): 96-102.