28 January 2008 Convert a low-cost sensor to a colorimeter using an improved regression method
Author Affiliations +
Closed loop color calibration is a process to maintain consistent color reproduction for color printers. To perform closed loop color calibration, a pre-designed color target should be printed, and automatically measured by a color measuring instrument. A low cost sensor has been embedded to the printer to perform the color measurement. A series of sensor calibration and color conversion methods have been developed. The purpose is to get accurate colorimetric measurement from the data measured by the low cost sensor. In order to get high accuracy colorimetric measurement, we need carefully calibrate the sensor, and minimize all possible errors during the color conversion. After comparing several classical color conversion methods, a regression based color conversion method has been selected. The regression is a powerful method to estimate the color conversion functions. But the main difficulty to use this method is to find an appropriate function to describe the relationship between the input and the output data. In this paper, we propose to use 1D pre-linearization tables to improve the linearity between the input sensor measuring data and the output colorimetric data. Using this method, we can increase the accuracy of the regression method, so as to improve the accuracy of the color conversion.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Yifeng Wu, Yifeng Wu, "Convert a low-cost sensor to a colorimeter using an improved regression method", Proc. SPIE 6807, Color Imaging XIII: Processing, Hardcopy, and Applications, 68070S (28 January 2008); doi: 10.1117/12.764936; https://doi.org/10.1117/12.764936

Back to Top