TY - GEN
T1 - Colour matching function learning
AU - Romero-Ortega, Luis
AU - Robles-Kelly, Antonio
PY - 2012
Y1 - 2012
N2 - In this paper, we aim at learning the colour matching functions making use of hyperspectral and trichromatic imagery. The method presented here is quite general in nature, being data driven and devoid of constrained setups. Here, we adopt a probabilistic formulation so as to recover the colour matching functions directly from trichromatic and hyperspectral pixel pairs. To do this, we derive a log-likelihood function which is governed by both, the spectra-to-colour equivalence and a generative model for the colour matching functions. Cast into a probabilistic setting, we employ the EM algorithm for purposes of maximum a posteriori inference, where the M-step is effected making use of Levenberg-Marquardt optimisation. We present results on real-world data and provide a quantitative analysis based upon a colour calibration chart.
AB - In this paper, we aim at learning the colour matching functions making use of hyperspectral and trichromatic imagery. The method presented here is quite general in nature, being data driven and devoid of constrained setups. Here, we adopt a probabilistic formulation so as to recover the colour matching functions directly from trichromatic and hyperspectral pixel pairs. To do this, we derive a log-likelihood function which is governed by both, the spectra-to-colour equivalence and a generative model for the colour matching functions. Cast into a probabilistic setting, we employ the EM algorithm for purposes of maximum a posteriori inference, where the M-step is effected making use of Levenberg-Marquardt optimisation. We present results on real-world data and provide a quantitative analysis based upon a colour calibration chart.
UR - http://www.scopus.com/inward/record.url?scp=84868154570&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-34166-3_35
DO - 10.1007/978-3-642-34166-3_35
M3 - Conference contribution
SN - 9783642341656
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 318
EP - 326
BT - Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, SSPR and SPR 2012, Proceedings
T2 - Joint IAPR International Workshops on Structural and Syntactic PatternRecognition, SSPR 2012 and Statistical Techniques in Pattern Recognition,SPR 2012
Y2 - 7 November 2012 through 9 November 2012
ER -