TY - JOUR

T1 - Multispectral colormapping using penalized least square regression

A1 - Dissing,Bjørn Skovlund

A1 - Carstensen,Jens Michael

A1 - Larsen,Rasmus

AU - Dissing,Bjørn Skovlund

AU - Carstensen,Jens Michael

AU - Larsen,Rasmus

PB - Society for Imaging Science and Technology

PY - 2010

Y1 - 2010

N2 - The authors propose a novel method to map a multispectral image into the device independent color space CIE-XYZ. This method provides a way to visualize multispectral images by predicting colorvalues from spectral values while maintaining interpretability and is tested on a light emitting diode based multispectral system with a total of 11 channels in the visible area. To obtain interpretable models, the method estimates the projection coefficients with regard to their neighbors as well as the target. This results in relatively smooth coefficient curves which are correlated with the CIE-XYZ color matching functions. The target of the regression is a well known color chart, and the models are validated using leave one out cross validation in order to maintain best possible generalization ability. The authors compare the method with a direct linear regression and see that the interpretability improves significantly but comes at the cost of slightly worse predictability.

AB - The authors propose a novel method to map a multispectral image into the device independent color space CIE-XYZ. This method provides a way to visualize multispectral images by predicting colorvalues from spectral values while maintaining interpretability and is tested on a light emitting diode based multispectral system with a total of 11 channels in the visible area. To obtain interpretable models, the method estimates the projection coefficients with regard to their neighbors as well as the target. This results in relatively smooth coefficient curves which are correlated with the CIE-XYZ color matching functions. The target of the regression is a well known color chart, and the models are validated using leave one out cross validation in order to maintain best possible generalization ability. The authors compare the method with a direct linear regression and see that the interpretability improves significantly but comes at the cost of slightly worse predictability.

U2 - 10.2352/J.ImagingSci.Technol.2010.54.3.030401

DO - 10.2352/J.ImagingSci.Technol.2010.54.3.030401

JO - Journal of Imaging Science and Technology

JF - Journal of Imaging Science and Technology

SN - 1062-3701

IS - 3

VL - 54

SP - 304011

EP - 304016

ER -