Mapping Tree Species in Coastal Portugal Using Statistically Segmented Principal Component Analysis and Other Methods

Hyperspectral sensors record radiances in a large number of wavelengths of the electromagnetic spectrum and can be used to distinguish different tree species based on their characteristic reflectance signatures. Reflectance spectra were measured from airborne hyperspectral AISA Eagle/Hawk imagery in order to identify different Mediterranean tree species at a coastal test site in Portugal. A spectral range from 400 to 2450 nm was recorded at 2-m spatial resolution. The hyperspectral data are divided into five spectral data ranges. The chosen ranges for segmentation are based on statistical properties as well as on their wavelengths, as radiances of a particular wavelength may overlap with neighboring wavelengths. Principal component analysis (PCA) is applied individually to each spectral range. The first three principal components (PCs) of each range are chosen and are fused into a new data segment of reduced dimensionality. The resulting 15 PCs contain 99.42% of the information content of the original hyperspectral image. These PCs were used for a maximum likelihood classification (MLC). Spectral signatures were also analyzed for the hyperspectral data, and were validated with ground data collected in the field by a handheld spectro-radiometer. Different RGB combinations of PC bands of segmented PC image provide distinct feature identification. A comparison with other classification approaches (spectral angle mapper and MLC of the original hyperspectral imagery) shows that the MLC of the segmented PCA achieves the highest accuracy, due to its ability to reduce the Hughes phenomenon.

[1]  John R. Miller,et al.  Correlating leaf area index of Ponderosa pine with hyperspectral casi data , 1992 .

[2]  Fuan Tsai,et al.  Spectrally segmented principal component analysis of hyperspectral imagery for mapping invasive plant species , 2007 .

[3]  N. Coops,et al.  Assessing the utility of airborne hyperspectral and LiDAR data for species distribution mapping in the coastal Pacific Northwest, Canada , 2010 .

[4]  Paul M. Mather,et al.  An evaluation of Landsat TM spectral data and SAR-derived textural information for lithological discrimination in the Red Sea Hills, Sudan , 1998 .

[5]  Roberta E. Martin,et al.  Invasive species detection in Hawaiian rainforests using airborne imaging spectroscopy and LiDAR. , 2008 .

[6]  K. Jusoff Land Use and Cover Mapping with Airborne Hyperspectral Imager in Setiu, Malaysia , 2009 .

[7]  George Alan Blackburn,et al.  Mapping reedbed habitats using texture-based classification of QuickBird imagery , 2011 .

[8]  Prashant K. Srivastava,et al.  Appraisal of land use/land cover of mangrove forest ecosystem using support vector machine , 2014, Environmental Earth Sciences.

[9]  H. Cetin COMPARISON OF SPACEBORNE AND AIRBORNE HYPERSPECTRAL IMAGING SYSTEMS FOR ENVIRONMENTAL MAPPING , 2004 .

[10]  A. Hastings,et al.  Mapping marshland vegetation of San Francisco Bay, California, using hyperspectral data , 2005 .

[11]  Kalliopi Radoglou,et al.  Forests of the Mediterranean region : gaps in knowledge and research needs , 2000 .

[12]  John A. Richards,et al.  Remote Sensing Digital Image Analysis: An Introduction , 1999 .

[13]  Pavan Kumar,et al.  Geospatial Strategy for Tropical Forest-Wildlife Reserve Biomass Estimation , 2013, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[14]  Fred A. Kruse,et al.  The Spectral Image Processing System (SIPS) - Interactive visualization and analysis of imaging spectrometer data , 1993 .

[15]  Ryuei Nishii,et al.  Hughes phenomenon in the spatial resolution enhancement of low resolution images and derivation of selection rule for high resolution images , 1997, IGARSS'97. 1997 IEEE International Geoscience and Remote Sensing Symposium Proceedings. Remote Sensing - A Scientific Vision for Sustainable Development.

[16]  E. L. Hunter,et al.  An assessment of two classification methods for mapping Thames Estuary intertidal habitats using CASI data , 2002 .

[17]  J. M. Moore,et al.  Debris provenance mapping in braided drainage using remote sensing , 1993, Geological Society, London, Special Publications.

[18]  Hao Zhang,et al.  Estimation of Above-Ground Biomass Using HJ-1 Hyperspectral Images in Hangzhou Bay, China , 2009, 2009 International Conference on Information Engineering and Computer Science.

[19]  B. N. Baatuuwie,et al.  Evaluation of three classifiers in mapping forest stand types using medium resolution imagery: a case study in the Offinso Forest District, Ghana , 2011 .

[20]  S. Adler-Golden,et al.  Atmospheric Correction for Short-wave Spectral Imagery Based on MODTRAN 4 , 2000 .

[21]  Gail P. Anderson,et al.  Atmospheric correction for shortwave spectral imagery based on MODTRAN4 , 1999, Optics & Photonics.

[22]  Helmi Zulhaidi Mohd Shafri,et al.  The Performance of Maximum Likelihood, Spectral Angle Mapper, Neural Network and Decision Tree Classifiers in Hyperspectral Image Analysis , 2007 .

[23]  David W. Scott,et al.  The Curse of Dimensionality and Dimension Reduction , 2008 .

[24]  Yunseop Kim,et al.  Hyperspectral Image Analysis for Plant Stress Detection , 2010 .

[25]  P. Switzer,et al.  A transformation for ordering multispectral data in terms of image quality with implications for noise removal , 1988 .

[26]  V. Karathanassi,et al.  A comparison study on fusion methods using evaluation indicators , 2007 .

[27]  Jiang Li,et al.  Dimensionality reduction of hyperspectral data using discrete wavelet transform feature extraction , 2002, IEEE Trans. Geosci. Remote. Sens..

[28]  D. Roberts,et al.  Hyperspectral discrimination of tropical rain forest tree species at leaf to crown scales , 2005 .

[29]  Zhihong Pan,et al.  Vegetation management of utility corridors using high-resolution hyperspectral imaging and LiDAR , 2010, 2010 2nd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing.

[30]  Shunlin Liang,et al.  Earth system science related imaging spectroscopy — an assessment , 2009 .

[31]  Richard Bellman,et al.  Adaptive Control Processes: A Guided Tour , 1961, The Mathematical Gazette.

[32]  M. Cochrane Using vegetation reflectance variability for species level classification of hyperspectral data , 2000 .

[33]  Fred Ortenberg,et al.  Hyperspectral Sensor Characteristics: Airborne, Spaceborne, Hand-Held, and Truck-Mounted; Integration of Hyperspectral Data with LIDAR , 2011 .

[34]  J. R. Morrison,et al.  Macroalgae and Eelgrass Mapping in Great Bay Estuary Using AISA Hyperspectral Imagery. , 2009 .

[35]  David A. Landgrebe,et al.  Signal Theory Methods in Multispectral Remote Sensing , 2003 .

[36]  José A. Malpica,et al.  CONSEQUENCES OF THE HUGHES PHENOMENON ON SOME CLASSIFICATION TECHNIQUES , 2011 .

[37]  G. F. Hughes,et al.  On the mean accuracy of statistical pattern recognizers , 1968, IEEE Trans. Inf. Theory.

[38]  Antonio J. Plaza,et al.  Dimensionality reduction and classification of hyperspectral image data using sequences of extended morphological transformations , 2005, IEEE Transactions on Geoscience and Remote Sensing.

[39]  George Alan Blackburn,et al.  Optimising the use of hyperspectral and LiDAR data for mapping reedbed habitats , 2011 .

[40]  George P. Petropoulos,et al.  Hyperion hyperspectral imagery analysis combined with machine learning classifiers for land use/cover mapping , 2012, Expert Syst. Appl..

[41]  Jon Atli Benediktsson,et al.  Neural Network Approaches Versus Statistical Methods in Classification of Multisource Remote Sensing Data , 1989, 12th Canadian Symposium on Remote Sensing Geoscience and Remote Sensing Symposium,.