@article{fdi:010067593,
  title = {{S}emi-blind source separation for the estimation of the clay content over semi-vegetated areas using {VNIR}/{SWIR} hyperspectral airborne data},
  author = {{O}uerghemmi, {W}. and {G}omez, {C}{\'e}cile and {N}aceur, {S}. and {L}agacherie, {P}.},
  editor = {},
  language = {{ENG}},
  abstract = {{V}isible, near-infrared and short wave infrared ({VNIR}/{SWIR}) hyperspectral imagery has proven to be a useful technique for mapping the soil surface properties over bare soils pixels. {M}ultivariate regression models are usually built linking a set of soil surface properties (response {Y}-variables) to a set of imaging reflectance spectra over bare soil pixels (predictor {X}-variables), and then, they are applied to all bare soil pixels to map the soil surface properties. {T}he applicability of {VNIR}/{SWIR} hyperspectral imagery for soil properties mapping decreases when surfaces are partially covered by vegetation. {T}he objective of this research was to develop a "{D}ouble-{E}xtraction" approach for clay content estimation over semi-vegetated surfaces and to evaluate its performance using {VNIR}/{SWIR} {H}y{M}ap airborne data acquired in a {M}editerranean region over an area of 24 km(2). {T}he "{D}ouble-{E}xtraction" approach consists of 1) an extraction of a soil reflectance spectrum, {S}-soil, using a semi-blind source separation ({SBSS}) technique applied to couples of semi-vegetated spectra and 2) an extraction of clay content from the soil reflectance spectrum {S}-soli using a multivariate regression method. {T}he source separation approach is semi blind due to the use of available knowledge about expected soil and vegetation spectra. {T}he multiplicative algorithm of {L}ee & {S}eung, belonging to the family of non-negative matrix factorization ({NMF}) methods, is used to solve the blind source separation ({BSS}) problem. {T}he multivariate regression method used in this study is the partial least squares regression ({PLSR}) method. {T}he "{D}ouble-{E}xtraction" approach was compared to a "{D}irect" approach consisting of the application of the multivariate regression model built from bare soil spectra over the semi-vegetated area. {O}ur results showed poor prediction performances for both approaches when applied to all pixels; however, a slight improvement was observed when correcting the bias prediction that occurs when using the {PLSR} model. {C}onversely, satisfactory prediction performances were obtained by restricting the prediction to the weakly vegetated area ({NDVI} < 0.55) that covered 63% of the study area. {T}he resulting clay map over this restricted vegetated area exhibited patterns of variations that matched the previous expertise acquired on the spatial structures of soils in this area.},
  keywords = {{B}lind source separation ; {N}on-negative matrix factorization ; {P}artial least squares regression ; {C}lay content ; {S}emi-vegetated pixels ; {H}yperspectral imaging ; {S}pectroscopy ; {FRANCE} ; {ZONE} {MEDITERRANEENNE}},
  booktitle = {},
  journal = {{R}emote {S}ensing of {E}nvironment},
  volume = {181},
  numero = {},
  pages = {251--263},
  ISSN = {0034-4257},
  year = {2016},
  DOI = {10.1016/j.rse.2016.04.013},
  URL = {https://www.documentation.ird.fr/hor/fdi:010067593},
}