Abstract
A new method is proposed to build up a color reproduction model based on neural networks and visual matching. Training data were obtained from a visual matching experiment and the commonly used CIECAM02 model to set up a link between the device parameters and human color perception. The process of training neural networks is presented. An experiment simulating an office environment was conducted, together with a user study, to verify the performance of the proposed method in identifying and reproducing color.
Original language | English |
---|---|
Pages (from-to) | 11027-11033 |
Number of pages | 7 |
Journal | Applied Optics |
Volume | 59 |
Issue number | 35 |
DOIs | |
Publication status | Published - 10 Dec 2020 |