Papers by Keyword: Color Calibration

Abstract: This paper presents the development of a color calibration method based on spectrally reproduced colors produced by multi-primary color narrow-band LED’s. The additive mixture of the primary colors can spectrally reproduce an object color. This approach has been used in our work to overcome the insufficient calibration results obtained by commonly used color targets such as the Macbeth Chart. The method has been tested using a developed prototype of digitally controlled light source calibrated to simulate object color under D65 Daylight standard illumination. The method allows for obtaining of the average color difference for the calibrated XYZ camera of ∆E=0.79 and shows significant improvement in comparison with calibration using standard Macbeth (∆E=1.47). Moreover, we explain the advantages of color calibration with usage of multi-primary light source with regard to the amount of the primary colors.

Abstract: Due to different atmospheric conditions, seasonal changes in vegetation characteristics and other reasons, the remote sensing images captured at different time may be quite different in color, brightness and so on. In this paper, coupled with statistics classification, three common orthogonal space transformations were used to calibrate the color difference respectively. Compared with conventional methods such as the overlapping region correction and histogram matching, the results show that orthogonal transform could achieve better correction effects. The lαβ transform gets the best corrected result among three orthogonal space transform methods.