@article{fdi:010069981,
  title = {{C}lay content mapping from airborne hyperspectral {V}is-{NIR} data by transferring a laboratory regression model},
  author = {{N}ouri, {M}. and {G}omez, {C}{\'e}cile and {G}orretta, {N}. and {R}oger, {J}. {M}.},
  editor = {},
  language = {{ENG}},
  abstract = {{V}isible {N}ear-{I}nfrared ({V}is-{NIR}, 400-2500 nm) hyperspectral imaging has proven to be a valuable tool for mapping soil properties over bare soils. {T}o date, most predictive models presented in literature, have been built from calibration databases made up of both {V}is-{NIR} imaging spectra (predictor variables) and soil properties (response variables). {N}evertheless, the constitution of such calibration databases is costly both in terms of time and money, as this implies soil sampling over hundreds of pixels of bare soils and physico-chemical soil analysis of these samples. {W}e propose to test the transfer of predictive models built from laboratory database, to {V}is-{NIR} hyperspectral image. {F}our transfer methods were tested (model updating, {R}epfile, {T}ransfer by {O}rthogonal {P}rojection {TOP} and {P}iecewise {D}irect {S}tandardization {PDS}). {T}heir respective performances were evaluated and compared to those obtained using two other predictive models. {T}he first predictive model was obtained by directly applying a partial least square regression ({PLSR}) calibrated in laboratory, onto the hyperspectral image. {T}he second model was obtained by applying a {PLSR} built from {V}is-{NIR} spectra extracted from the hyperspectral image, onto the hyperspectral image. {T}he transferred models are based on the use of standards. {S}tandards were selected in order to (i) take into account or (ii) ignore field soil stratification knowledge. {M}oreover, the impact of the number of standards used in the transferred models was studied. {T}his approach was tested for clay content prediction over the {L}a {P}eyne {V}alley area in {S}outhern {F}rance (24 km(2), in a {M}editerranean context) covered by {H}y{M}ap {V}is-{NIR} airborne data. {T}he results show that 1) applying {PLSR} model from laboratory spectra to hyperspectral image without calibration transfer, provides poor prediction performances ({R}-val(2) = 0.41, {RMSEP} = 103 g/kg), 2) transferring this model from laboratory spectra to hyperspectral image offered good prediction performances ({R}-test(2) median above 0.5 and {RMSEP} median below 70 g/kg), whatever the transfer method used and from only 15 used standards, and 3) taking into account field soil stratification knowledge in the standards selection improves prediction performances ({R}-test(2) median above 0.6 and {RMSEP} median below 60 g/kg) whatever the transfer method and from only 10 used standards. {F}inally, transferring models from the laboratory to a hyperspectral image gave better mapping performances than those obtained using {PLSR} models built from airborne spectra.},
  keywords = {{V}is-{NIR} spectroscopy ; {S}oil mapping ; {C}lay content ; {H}yperspectral airborne data ; {C}alibration transfer ; {P}artial leastsquare regression ; {FRANCE}},
  booktitle = {},
  journal = {{G}eoderma},
  volume = {298},
  numero = {},
  pages = {54--66},
  ISSN = {0016-7061},
  year = {2017},
  DOI = {10.1016/j.geoderma.2017.03.011},
  URL = {https://www.documentation.ird.fr/hor/fdi:010069981},
}