Method and system for imaging device characterization including look-up table construction via tensor decomposition

摘要

A method and apparatus for characterizing the performance of a printing device comprising printing a target set of patches with the device and measuring the printing device response with the printed target set; compiling a LUT from the printed target set and measured response; and representing the LUT as a tensor. According to one exemplary embodiment, tensor decomposition/parallel factor analysis is employed for compacting the tensor representation of the LUT.

主权项

1. A method of generating a multidimensional look-up-table (LUT) including a plurality of nodes, the LUT a characterization of a color device and the multidimensional LUT mapping a plurality of device dependent color space values to a plurality of respective device independent color space values associated with the plurality of nodes, the method comprising:
defining the multidimensional LUT as a tensor; rendering a training set of patches with the color device, the training set of patches associated with a subset of the plurality of nodes, the training set of patches rendered for device dependent color space values associated with the subset of the plurality of nodes; measuring one or more colors associated with the rendered training set of patches to generate device independent color space values associated with respective device dependent color space values associated with the subset of the plurality of nodes; calculating one or more tensor terms associated with the tensor using the device dependent color space values and device independent color space values associated with the rendered training set of patches; and calculating the plurality of device dependent color space values to the plurality of respective device independent color space values based on the calculated one or more calculated tensor terms to generate the multidimensional LUT, wherein the process of defining the multidimensional LUT as a tensor includes:
a) obtaining an initial estimate for a one term approximated to the tensor:b) improving the one term approximation to the tensor using alternating least squares;c) obtaining an initial estimate for a next term associated with a multi-term approximation to the tensor; andd) improving the multi-term approximation using alternating least squares.