A fast forecast method of chromatic aberration based on color luminosity

2017 
In CIELAB color space, color chromatic aberration[1-3] is the result of a combination of luminosity difference, chroma difference and hue difference, this three differences played individual weight value[4-5] roles on different color valuation objects[6-8]. A fast forecast method of chromatic aberration which is based on color luminosity[9-10] is presented in this paper. Taken ten groups of dyeing fabrics with different concentration and ratio as study object, and CS-650 colorimeter was adopt to measure data which include the value of luminosity difference, chroma difference, hue difference. The chromatic aberration value was calculated by CIEDE CMC(l:c)[11] formula which is well suit for textile and dyeing industry[12]. In this experiment, one hundred evenly spaced sampling points which were chosen from each group were needed to measure and record. The final data of each group were verified by simulation with MATLAB and displayed by 2D diagram one by one. In these diagrams, the data of luminosity difference, chroma difference, hue difference were taken as horizontal axis and chromatic aberration value as the vertical axis. From the ten groups of 2D data chart, it is obvious that the function of luminosity difference and chromatic aberration value is linear, then taken that function into linear fitting by least-squares[13] method. Compared to the actual chromatic aberration value, the fitting function showed that the average error is no more than 0.05 and the maximum error is no more than 0.2. Therefore, within the range of chromatic aberration precision, it is possible to make color-difference formula simplified, and the method of forecasting chromatic aberration based on color luminosity is practicable and will provide new ideas and ways for evenness detection of color chromatic aberration as well.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    19
    References
    0
    Citations
    NaN
    KQI
    []