The Utility of AISA Eagle Hyperspectral Data and Random Forest Classifier for Flower Mapping

被引:20
作者
Abdel-Rahman, Elfatih M. [1 ,2 ]
Makori, David M. [1 ,3 ]
Landmann, Tobias [1 ]
Piiroinen, Rami [4 ]
Gasim, Seif [2 ]
Pellikka, Petri [4 ]
Raina, Suresh K. [1 ]
机构
[1] Int Ctr Insect Physiol & Ecol, Nairobi 00100, Kenya
[2] Univ Khartoum, Dept Agron, Fac Agr, Khartoum 13314, Sudan
[3] Univ Helsinki, Dept Geosci & Geog, Helsinki 00560, Finland
[4] Univ KwaZulu Natal, Sch Agr Environm & Earth Sci, Dept Geog, ZA-3209 Pietermaritzburg, South Africa
关键词
AISA Eagle hyperspectral data; random forest classifier; flowering plants; SPECTRAL BAND SELECTION; SPECIES DISCRIMINATION; REGRESSION; ACCURACY; NITROGEN; EO-1;
D O I
10.3390/rs71013298
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Knowledge of the floral cycle and the spatial distribution and abundance of flowering plants is important for bee health studies to understand the relationship between landscape and bee hive productivity and honey flow. The key objective of this study was to show how AISA Eagle hyperspectral data and random forest (RF) can be optimally utilized to produce flowering and spatially explicit land use/land cover (LULC) maps for a study site in Kenya. AISA Eagle imagery was captured at the early flowering period (January 2014) and at the peak flowering season (February 2013). Data on white and yellow flowering trees as well as LULC classes in the study area were collected and used as ground-truth points. We utilized all 64 AISA Eagle bands and also used variable importance in RF to identify the most important bands in both AISA Eagle data sets. The results showed that flowering was most accurately mapped using the AISA Eagle data from the peak flowering period (85.71%-88.15% overall accuracy for the peak flowering season imagery versus 80.82%-83.67% for the early flowering season). The variable optimization (i.e., variable selection) analysis showed that less than half of the AISA bands (n = 26 for the February 2013 data and n = 21 for the January 2014 data) were important to attain relatively reliable classification accuracies. Our study is an important first step towards the development of operational flower mapping routines and for understanding the relationship between flowering and bees' foraging behavior.
引用
收藏
页码:13298 / 13318
页数:21
相关论文
共 64 条
[21]   Collinearity: a review of methods to deal with it and a simulation study evaluating their performance [J].
Dormann, Carsten F. ;
Elith, Jane ;
Bacher, Sven ;
Buchmann, Carsten ;
Carl, Gudrun ;
Carre, Gabriel ;
Garcia Marquez, Jaime R. ;
Gruber, Bernd ;
Lafourcade, Bruno ;
Leitao, Pedro J. ;
Muenkemueller, Tamara ;
McClean, Colin ;
Osborne, Patrick E. ;
Reineking, Bjoern ;
Schroeder, Boris ;
Skidmore, Andrew K. ;
Zurell, Damaris ;
Lautenbach, Sven .
ECOGRAPHY, 2013, 36 (01) :27-46
[22]   Improvements on cross-validation: The .632+ bootstrap method [J].
Efron, B ;
Tibshirani, R .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1997, 92 (438) :548-560
[23]  
Fening Ken O., 2008, International Journal of Biodiversity Science & Management, V4, P104, DOI 10.3843/Biodiv.4.2:4
[24]   Tree Species Discrimination in Tropical Forests Using Airborne Imaging Spectroscopy [J].
Feret, Jean-Baptiste ;
Asner, Gregory P. .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2013, 51 (01) :73-84
[25]   THE RED EDGE POSITION AND SHAPE AS INDICATORS OF PLANT CHLOROPHYLL CONTENT, BIOMASS AND HYDRIC STATUS [J].
FILELLA, I ;
PENUELAS, J .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 1994, 15 (07) :1459-1470
[26]   Rapid changes in flowering time in British plants [J].
Fitter, AH ;
Fitter, RSR .
SCIENCE, 2002, 296 (5573) :1689-1691
[27]   Thematic map comparison: Evaluating the statistical significance of differences in classification accuracy [J].
Foody, GM .
PHOTOGRAMMETRIC ENGINEERING AND REMOTE SENSING, 2004, 70 (05) :627-633
[28]   Atmospheric correction algorithms for hyperspectral remote sensing data of land and ocean [J].
Gao, Bo-Cai ;
Montes, Marcos J. ;
Davis, Curtiss O. ;
Goetz, Alexander F. H. .
REMOTE SENSING OF ENVIRONMENT, 2009, 113 :S17-S24
[29]   Hyperspectral characteristics of canopy components and structure for phenological assessment of an invasive weed [J].
Ge, Shaokui ;
Everitt, James ;
Carruthers, Raymond ;
Gong, Peng ;
Anderson, Gerald .
ENVIRONMENTAL MONITORING AND ASSESSMENT, 2006, 120 (1-3) :109-126
[30]  
GIURFA M, 1995, J COMP PHYSIOL A, V177, P247