Characterisation targets are mainly used to develop characterisation model for the transformation between the device dependent values and device independent primaries for an imaging device. The model's performance is affected by the colours included in the characterisation target. A better characterization target should contain the number of samples us less as possible and provide reasonable accurate prediction. In this paper, a method is proposed for colour selections for training a camera characterization model. Comparison results using the data collected show the new method outperformed the available methods.
展开▼
