摘要
为快速准确地获取植株冠层氮素含量及空间分布特征,对大尺度的果园进行精准动态的管理,以宽行窄株小冠模式、宽行窄株篱壁模式和传统栽培模式3种栽培模式的120棵柑橘树为研究对象,通过测定冠层氮素含量并提取无人机遥感影像多光谱数据中的纹理指数和植被指数,运用随机森林算法(RF)建立基于植被指数、纹理指数以及融合植被指数和纹理指数的柑橘冠层氮素反演模型,并比较融合植被指数和纹理指数的支持向量机(SVM)、BP神经网络算法(BP)和RF的模型反演精度。结果显示:在随机森林算法中,融合植被指数和纹理指数比单独的植被指数或纹理指数更能准确预测柑橘冠层氮素含量;植被指数训练集
柑橘是世界第一大类水果,也是我国南方栽培面积最广、经济地位最重要的果
基于无人机遥感反演农作物长势参数(形态参数、生理生化参数、胁迫参数、产量参数)的相关研究近年来发展迅速,已成为国内外农业遥感领域的热
高分辨率遥感影像除了具备光谱信息外,还具有丰富的纹理信息,能够有效反映植被内部的结构信息,有效缓和光谱饱和现
研究区位于江西省赣州市信丰县绿萌柑橘基地(24°29′~27°09′N、113°54′~116°38′E),属亚热带丘陵山区湿润季风气候。本研究选择宽行窄株小冠模式、宽行窄株篱壁模式和传统栽培模式的柑橘树作为研究对
于2020年11月16日上午10:00晴朗无云的天气下,利用大疆精灵4多光谱版无人机对柑橘果园进行遥感影像采集,设置飞行高度为100 m,镜头焦距为5.74 mm。遥感影响空间分辨率为5 cm,包含5个波段:蓝波段(blue,B)、绿波段(green,G)、红波段(red,R)、红边波段(rededge band,RE)和近红外波段(near infrared band,NIR)。辐射定标:DN(digital number)值与地面表观反射率之间的回归方程为:y=0.01743x-0.23758(x为地球表观反射率,y为DN值),得到表观反射率影像。
提取17个植被指数,构建基于多光谱的柑橘冠层氮素含量反演模型(
名称 Name | 公式 Formulation | 参考文献 Reference |
---|---|---|
归一化差值植被指数 Normalized difference vegetation index | INDV= |
[ |
差值植被指数 Difference vegetation index | IDV= |
[ |
增强植被指数 Enhanced vegetation index |
[ | |
土壤修正植被指数 Soil adjusted vegetation index |
[ | |
非线性植被指数 None liner index | INL = |
[ |
绿色归一化植被指数 Green normalized difference vegetation index | IGNDV= |
[ |
绿色比植被指数 Green ratio vegetation index | IGRV= |
[ |
归一化叶绿素指数 Normalized pigment chlorophyll index | INPC= |
[ |
作物氮反射指数 Nitrogen reflectance index | INR= |
[ |
土壤调节植被指数 Optimized soil adjusted vegetation index | IOSAV= |
[ |
植物衰老反射率指数 Plant senescing reflectance index | IPSR= |
[ |
比植被指数 Ratio vegetation index | IRV= |
[ |
冠层结构不敏感植被指数 Structure insensitive pigment index | ISIP= |
[ |
三角植被指数 Triangle vegetation index | ITV= |
[ |
可见光大气阻抗植被指数 Visible atmospherically resistant vegetation index | IVAR= |
[ |
宽范围动态植被指数 Wide dynamic vegetation index | IWDRV= |
[ |
转化叶绿素吸收反射指数 Transformation chlorophyll absorption reflection index | ITCAR= |
[ |
注: R为光谱反射率:为近红外反射率、为红光反射率、为绿光反射率、为蓝光反射率。Note:R is spectral reflectance: is near-infrared reflectance, is the red light reflectivity, is the green light reflectivity, is the blue light reflectivity.
采用灰度共生矩阵方法分别提取红光波段、绿光波段、蓝光波段、近红外波段和红边波段的8种纹理指
名称 Name | 公式 Formulation |
---|---|
同质性 Homogeneity(IHOM) | IHOM= |
平均值 Mean(IMEAN) | IMEAN= |
对比度 Contrast(ICON) | ICON= |
熵 Entropy(IEnt) | IEnt= |
非相似度 Dissimilarity(IDIS) | IDIS= |
变化量 Variance(IVAR) | IVAR= |
二阶矩阵 Second moment(ISM) | ISM= |
相关性 Correlation(ICOR) |
ICOR= 1. 2. |
注: i,j为像素灰度、N为灰度级数、Pi,j是在给定空间距离和方向时,灰度以i(行)为起始点,出现灰度级j(列)的概率。Note: i, j is pixel gray,N is gray level, Pi,j is the probability of gray level j (column ) in a given spatial distance and direction, the gray level with i ( row )as the starting point.
由
模式 Pattern | 平均值 Mean | 最大值 Max | 最小值 Min | 中值 Median | 标准差 Standard deviation |
---|---|---|---|---|---|
宽行窄株篱壁 Wide row and narrow plant fence cultivation | 30.10 | 35.68 | 23.76 | 30.40 | 2.18 |
宽行窄株小冠 Wide row and narrow plant cultivation | 31.57 | 35.99 | 25.66 | 32.25 | 2.01 |
传统栽培模式 Traditional cultivation | 27.64 | 36.02 | 18.51 | 27.93 | 3.07 |
利用基于特征递归消除的机器学习算法,对提取的17个植被指数进行重要性排序,递归重要性最低的指数,筛选至最佳的3个指数并以此确定模型。模型的训练集
对提取的40个纹理指数进行重要性排序,递归消减重要性最低的指数,筛选至最佳的7个指数并以此确定模型(

图1 基于纹理指数的RF预测模型建立中重要性排序
Fig.1 Importance ranking in RF prediction model based on texture index
对提取的57个融合指数进行重要性排序,递归消减重要性最低的指数,筛选至最佳的10个指数数量并以此确定模型(

图2 融合植被指数和纹理指数的RF预测模型建立中重要性排序
Fig.2 Importance ranking in RF prediction model based on vegetation index and texture indexes
使用SVM和BP算法以叶片氮素和筛选的10个重要性排序靠前的参数建立模型。结果显示,SVM算法训练集
基于不同参数、不同算法构建的预测模型的预测值和真实值之间的关系如

图3 柑橘叶片氮素含量预测值与实测值的相关性
Fig.3 Correlations of measured values and predicted values of nitrogen content in citrus leaves
A: 植被指数(RF); B: 纹理指数(RF); C: 植被指数+纹理指数(RF); D: 植被指数+纹理指数(SVM); E: 植被指数+纹理指数(BP)。A: Vegetation index(RF); B: Texture index(RF); C:Vegetation index and texture index(RF); D:Vegetation index and texture parameters(SVM); E:Vegetation index and texture index(BP).
根据上述结果,选择精度最高的融合纹理指数和光谱指数的反演模型对不同栽培模式的柑橘植株冠层氮素含量进行估算,得到3种栽培模式下的柑橘园冠层氮素含量分布图(

图4 基于纹理指数和植被指数融合数据的氮素含量反演图
Fig.4 N content inversion map based on mixed data of texture index and vegetation index
A:老果园总体反演图;B:传统栽培模式;C:新果园总体反演图;D:宽行窄株篱壁式; E:宽行窄株式小冠式。A :Overall inversion map of old orchard;B :Traditional cultivation pattern;C: Overall inversion map of new orchard;D :Wide row and narrow plant fence cultivation pattern;E:Wide row and narrow plant cultivation pattern.
模式 Pattern | 平均值 Mean | 最大值 Max | 最小值 Min |
---|---|---|---|
宽行窄株篱壁 Wide row and narrow plant fence cultivation | 30.20 | 32.21 | 27.40 |
宽行窄株小冠 Wide row and narrow plant cultivation | 31.33 | 32.89 | 29.50 |
传统栽培模式 Traditional cultivation pattern | 27.82 | 31.40 | 24.66 |
植物冠层光谱特征是监测植被生长的重要指标,植被冠层的遥感影像中的各种特征参数能较准确地反映目标植株的组织分布结构、营养元素、生物量等多种综合信
本研究比较了3种不同栽培模式的柑橘冠层氮素含量,宽行窄株小冠模式的冠层氮素含量平均值最大(31.57 mg/g),宽行窄株篱壁式次之(30.10 mg/g),传统栽培模式最低(27.64 mg/g),这可能与不同栽培模式的不同树形结构相
利用重要性排序分别筛选出3个植被指数、7个纹理指数和10个融合植被指数和纹理指数,经建模得出测试集
柑橘叶片性状与传感器系统的冠层反射率之间的关系会因传感器和果园种类的不同而改
参考文献 References
郭文武,叶俊丽,邓秀新.新中国果树科学研究70年:柑橘[J].果树学报,2019,36(10):1264-1272.GUO W W,YE J L,DENG X X.Fruit scientific research in New China in the past 70 years:citrus[J].Journal of fruit science,2019,36(10):1264-1272 (in Chinese with English abstract). [百度学术]
叶春,李艳大,舒时富,等.基于高光谱的柑橘叶片氮素营养监测模型[J].江苏农业科学,2018,46(7):223-226.YE C,LI Y D,SHU S F,et al. A hyperspectral based nitrogen nutrition monitoring model for citrus leaves[J].Jiangsu agricultural sciences,2018,46(7):223-226 (in Chinese). [百度学术]
李旭.减氮施肥对柑橘树体氮素含量、果实品质产量和氮肥利用的影响[D].武汉:华中农业大学,2020.LI X.Effects of reduced nitrogen fertilization on nitrogen content,fruit yield and nitrogen fertilizer utilization of citrus trees[D].Wuhan:Huazhong Agricultural University,2020 (in Chinese with English abstract). [百度学术]
张冬强,唐子立,杨勇,等.用于监测柑橘叶片冻害的叶绿素含量光谱反射模型研究[J].农业环境科学学报,2012,31(10):1891-1896.ZHANG D Q,TANG Z L,YANG Y,et al.Model for Citrus leaves chlorophyll content based on spectral signature[J].Journal of agro-environment science,2012,31(10):1891-1896 (in Chinese with English abstract). [百度学术]
房贤一,朱西存,王凌,等.基于高光谱的苹果盛果期冠层叶绿素含量监测研究[J].中国农业科学,2013,46(16):3504-3513.FANG X Y,ZHU X C,WANG L,et al.Hyperspectral monitoring of the canopy chlorophyll content at apple tree prosperous fruit stage[J].Scientia agricultura sinica,2013,46(16):3504-3513 (in Chinese with English abstract). [百度学术]
GALVEZ-SOLA L,GARCÍA-SÁNCHEZ F,PÉREZ-PÉREZ J G,et al.Rapid estimation of nutritional elements on citrus leaves by near infrared reflectance spectroscopy[J/OL].Frontiers in plant science,2015,6:571[2022-12-15].https://doi.org/10.3389/fpls.2015.00571. [百度学术]
ZHENG H B,CHENG T,LI D,et al.Combining unmanned aerial vehicle (UAV)-based multispectral imagery and ground-based hyperspectral data for plant nitrogen concentration estimation in rice[J/OL].Frontiers in plant science,2018,9:936[2022-12-15].https://doi.org/10.3389/fpls.2018.00936. [百度学术]
刘忠,万炜,黄晋宇,等.基于无人机遥感的农作物长势关键参数反演研究进展[J].农业工程学报,2018,34(24):60-71.LIU Z,WAN W,HUANG J Y,et al.Progress on key parameters inversion of crop growth based on unmanned aerial vehicle remote sensing[J].Transactions of the CSAE,2018,34(24):60-71 (in Chinese with English abstract). [百度学术]
梁雪映,樊风雷.基于遥感光谱的作物氮含量估算研究进展[J].中国农业科技导报,2019,21(1):51-61.LIANG X Y,FAN F L.Research progress on crop nitrogen content estimation based on remote sensing spectrum[J].Journal of agricultural science and technology,2019,21(1):51-61 (in Chinese with English abstract). [百度学术]
王鑫梅,张劲松,孟平,等.基于无人机遥感影像的核桃冠层氮素含量估算[J].农业机械学报,2021,52(2):178-187.WANG X M,ZHANG J S,MENG P,et al.Estimation of nitrogen content in walnut canopy based on UAV remote sensing image[J].Transactions of the CSAM,2021,52(2):178-187 (in Chinese with English abstract). [百度学术]
OSCO L P,MARQUES RAMOS A P,SAITO MORIYA É A,et al.Improvement of leaf nitrogen content inference in Valencia-orange trees applying spectral analysis algorithms in UAV mounted-sensor images[J/OL].International journal of applied earth observation and geoinformation,2019,83:101907[2022-12-15].https://doi.org/10.1016/j.jag.2019.101907. [百度学术]
CEN H Y,WAN L,ZHU J P,et al.Dynamic monitoring of biomass of rice under different nitrogen treatments using a lightweight UAV with dual image-frame snapshot cameras[J/OL].Plant methods,2019,15:32[2022-12-15].https://doi.org/10.1186/s13007-019-0418-8. [百度学术]
贾丹,陈鹏飞.低空无人机影像分辨率对冬小麦氮浓度反演的影响[J].农业机械学报,2020,51(7):164-169.JIA D,CHEN P F.Effect of low-altitude UAV image resolution on inversion of winter wheat nitrogen concentration[J].Transactions of the CSAM,2020,51(7):164-169 (in Chinese with English abstract). [百度学术]
万亮,杜晓月,陈硕博,等.基于无人机多源图谱融合的水稻稻穗表型监测[J].农业工程学报,2022,38(9):162-170.WAN L,DU X Y,CHEN S B,et al.Rice panicle phenotyping using UAV-based multi-source spectral image data fusion[J].Transactions of the CSAE,2022,38(9):162-170 (in Chinese with English abstract). [百度学术]
胡青青,佃袁勇,龚赞,等.不同栽培模式的纽荷尔脐橙果实品质分析[J].华中农业大学学报,2022,41(5):108-115.HU Q Q,DIAN Y Y,GONG Z,et al.Analyzing fruit quality of Newhall navel oranges with different cultivation patterns[J].Journal of Huazhong Agricultural University,2022,41(5):108-115 (in Chinese with English abstract). [百度学术]
栗方亮,孔庆波,张青,等.琯溪蜜柚叶片氮素含量多种高光谱估算模型对比研究[J].果树学报,2022,39(5):882-891.LI F L,KONG Q B,ZHANG Q,et al.Comparative study on several hyperspectral estimation models of nitrogen contents in Guanxi honey pomelo leaves[J].Journal of fruit science,2022,39(5):882-891 (in Chinese with English abstract). [百度学术]
LEMENKOVA P.SAGA GIS for computing multispectral vegetation indices by landsat TM for mapping vegetation greenness[J].Contemporary agriculture,2021,70(1/2):67-75. [百度学术]
FENG W,WU Y P,HE L,et al.An optimized non-linear vegetation index for estimating leaf area index in winter wheat[J].Precision agriculture,2019,20(6):1157-1176. [百度学术]
PEÑUELAS J,GAMON J A,FREDEEN A L,et al.Reflectance indices associated with physiological changes in nitrogen- and water-limited sunflower leaves[J].Remote sensing of environment,1994,48(2):135-146. [百度学术]
YU K,LENZ-WIEDEMANN V,CHEN X P,et al.Estimating leaf chlorophyll of barley at different growth stages using spectral indices to reduce soil background and canopy structure effects[J].ISPRS journal of photogrammetry and remote sensing,2014,97:58-77. [百度学术]
HABOUDANE D,MILLER J R,PATTEY E,et al.Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop canopies:modeling and validation in the context of precision agriculture[J].Remote sensing of environment,2004,90(3):337-352. [百度学术]
HARALICK R M,SHANMUGAM K,DINSTEIN I.Textural features for image classification[J].IEEE transactions on systems,man,and cybernetics,1973,3(6):610-621. [百度学术]
ALI A,IMRAN M.Remotely sensed real-time quantification of biophysical and biochemical traits of citrus (Citrus sinensis L.) fruit orchards:a review[J/OL].Scientia horticulturae,2021,282:110024[2022-12-15].https://doi.org/10.1016/j.scienta.2021.110024. [百度学术]
宋伟,汪军,李华雄,等.川南丘区柑橘整形修剪技术初探[J].四川农业科技,2022(5):25-28.SONG W,WANG J,LI H X,et al.Preliminary study on citrus shaping and pruning techniques in hilly area of southern Sichuan[J].Sichuan agricultural science and technology,2022(5):25-28 (in Chinese). [百度学术]
朱壹,朱二,朱珊.一种篱壁式脐橙种植方法:CN201810798098.8[P].018-07-19.ZHU Y,ZHU E,ZHU S.A hedgerow navel orange planting method.2018.Patent number: CN201810798098.8[P].2018-07-19(in Chinese). [百度学术]
MAIMAITIJIANG M,GHULAM A,SIDIKE P,et al.Unmanned aerial system (UAS)-based phenotyping of soybean using multi-sensor data fusion and extreme learning machine[J].ISPRS journal of photogrammetry and remote sensing,2017,134:43-58. [百度学术]
刘俊,毕华兴,朱沛林,等.基于ALOS遥感数据纹理及纹理指数的柞树蓄积量估测[J].农业机械学报,2014,45(7):245-254.LIU J,BI H X,ZHU P L,et al.Estimating stand volume of Xylosma racemosum forest based on texture parameters and derivative texture indices of ALOS imagery[J].Transactions of the CSAM,2014,45(7):245-254 (in Chinese with English abstract). [百度学术]
刘畅,杨贵军,李振海,等.融合无人机光谱信息与纹理信息的冬小麦生物量估测[J].中国农业科学,2018,51(16):3060-3073.LIU C,YANG G J,LI Z H,et al.Biomass estimation in winter wheat by UAV spectral information and texture information fusion[J].Scientia agricultura sinica,2018,51(16):3060-3073 (in Chinese with English abstract). [百度学术]
朱丽娟.基于灰度共生矩阵的红枣纹理特征提取[J].科技风,2022(25):78-80.ZHU L J.Texture feature extraction of red jujube based on gray-level co-occurrence matrix[J].Technology wind,2022(25):78-80 (in Chinese with English abstract). [百度学术]
APOLO-APOLO O E,MARTÍNEZ-GUANTER J,EGEA G,et al.Deep learning techniques for estimation of the yield and size of citrus fruits using a UAV[J/OL].European journal of agronomy,2020,115:126030[2022-12-15].https://doi.org/10.1016/j.eja.2020.126030. [百度学术]
CHLINGARYAN A,SUKKARIEH S,WHELAN B.Machine learning approaches for crop yield prediction and nitrogen status estimation in precision agriculture:a review[J].Computers and electronics in agriculture,2018,151:61-69. [百度学术]
周俊宏,王子芝,廖声熙,等.基于GF-1影像的普达措国家公园森林地上生物量遥感估算[J].农业工程学报,2021,37(4):216-223.ZHOU J H,WANG Z Z,LIAO S X,et al.Remote sensing estimation of forest aboveground biomass in Potatso National Park using GF-1 images[J].Transactions of the CSAE,2021,37(4):216-223 (in Chinese with English abstract).. [百度学术]
李旭青,刘湘南,刘美玲,等.水稻冠层氮素含量光谱反演的随机森林算法及区域应用[J].遥感学报,2014,18(4):923-945.LI X Q,LIU X N,LIU M L,et al.Random forest algorithm and regional applications of spectral inversion model for estimating canopy nitrogen concentration in rice[J].Journal of remote sensing,2014,18(4):923-945 (in Chinese with English abstract). [百度学术]
韩兆迎,朱西存,房贤一,等.基于SVM与RF的苹果树冠LAI高光谱估测[J].光谱学与光谱分析,2016,36(3):800-805.HAN Z Y,ZHU X C,FANG X Y,et al.Hyperspectral estimation of apple tree canopy LAI based on SVM and RF regression[J].Spectroscopy and spectral analysis,2016,36(3):800-805 (in Chinese with English abstract). [百度学术]
CROFT H,CHEN J M,LUO X Z,et al.Leaf chlorophyll content as a proxy for leaf photosynthetic capacity[J].Global change biology,2017,23(9):3513-3524. [百度学术]
SINHA S K,PADALIA H,DASGUPTA A,et al.Estimation of leaf area index using PROSAIL based LUT inversion,MLRA-GPR and empirical models:case study of tropical deciduous forest plantation,North India[J/OL].International journal of applied earth observation and geoinformation,2020,86:102027[2022-12-15].https://doi.org/10.1016/j.jag.2019.102027. [百度学术]
LU J S,EITEL J U H,ENGELS M,et al.Improving unmanned aerial vehicle (UAV) remote sensing of rice plant potassium accumulation by fusing spectral and textural information[J/OL].International journal of applied earth observation and geoinformation,2021,104:102592[2022-12-15].https://doi.org/10.1016/j.jag.2021.102592. [百度学术]
FATEH F S.Prevalence of citrus decline in district sargodha[J].Pakistan journal of agricultural sciences,2017,54(1):9-13. [百度学术]