Improving tone consistency and reducing calibration frequency for color Electrophotography
Abstract
This research aims to improve tone consistency and to reduce calibration frequency for color Electrophotography (EP). EP processes are sensitive to changes in operating conditions. Calibrations are performed to stabilize primary color reproduction. During a calibration, regression models are used to predict the primary color tone values on output media. A principal component regression (PCR) approach is employed for the model coefficient estimation. It is shown that the PCR models can effectively reduce the prediction error by 24.7% compared to existing models. With given the PCR models, a model-based calibration procedure is proposed to stabilize the EP processes. Experimental results show that the proposed model-based calibration procedure provides a 63.3% reduction in tone error over the existing approach. In addition, transient tone fluctuation is investigated. A first-order dynamic model is proposed to characterize the transient tone fluctuation that lasts for less than eleven pages in the beginning of print jobs. The proposed model can accurately predict transient tone fluctuation over pages based on cross-validation and can be used to identify the necessary adjustments to compensate the tone fluctuation. With all the above efforts aiming at improving color consistency, potential in calibration frequency conservation is also investigated. A decision-tree-based approach is proposed to determinate appropriate timing for performing calibration as a trade-off between color consistency and consumable economy. Simulation with historic data shows that the proposed method can effectively reduce calibration frequency by 48.4% while the color consistency is guaranteed.
Degree
Ph.D.
Advisors
Chiu, Purdue University.
Subject Area
Computer Engineering|Electrical engineering|Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.