Master Thesis A method to evaluate the photographic color reproducibility was established based on numerical comparison by psychophysical and subjective evaluation. In obtaining the psychophysical evaluation the color charts were photographed on color nega-films, and the colorimetric differences were calculated. They were also subjectively evaluated on their color reproduction qualities. It was found that a reliable correlation exists between the psychophysical and subjective evaluation. Photographic color reproducibility index was induced based on the correlation.
Studies at RICOH Co.
Matrix masking to convert R,G,B densities
to X,Y,Z tristimulus values in color copier
A color patch with 130 color samples ('color chart 129b') was used for the color samples in the experiment. The color patch was scanned by a DFC-a scanner and inputted to an Apollo DN10000 workstation. It was displayed on an Astec display and the R,G,B densities of the 130 color samples were calculated. A convert masking matrix was calculated from the R,G,B densities and the X,Y,Z tristimulus values, which were measured by a colorimeter. Finally, the conversion accuracy using the convert masking matrix was evaluated by color difference formula in the CIE 1976 (L*a*b*) color space. The results are
Conversion of R,G,B densities to X,Y,Z tristimulus values with
consideration of using a non-uniform light source
Samples used in this experiment were a color patch
with 130 color samples ('color chart 129b') , a gray scale ('neutral value scale
33'), a piece of white paper and a piece of black paper. Each sample was photographed
by a color camera and were converted from analog to digital signals using a
Nexus computer. The digitized images were then processed in an Apollo DN10000
workstation.
A masking matrix was calculated to convert R,G,B densities into the X,Y,Z tristimulus
values with consideration of non-uniformity of the light source. The color differences
between X,Y,Z tristimulus values which were measured by the colorimeter
and the converted X,Y,Z tristimulus values were evaluated in the CIE 1976 (L*a*b*)
color space.
Studies at Fuji Xerox Co., Inc.
Analysis of non-uniformity of prints
by IOTs
(Image Output Terminals)
Sine waves, which have different spatial frequencies and amplitudes, were generated corresponding to the non-uniform printout of the IOT. The perceptible region of the non-uniformity in the frequency-amplitude space was then determined by a subjective evaluation based on a rating scale method with five catalogs. Finally, the modulation transfer function of the IOT was obtained.
Studies at LG Electronics Inc.
A study on color correction for color
printer
A color correction algorithm plays a very important role in the reduction
of color difference between the original color sample and the reproduce one.
In this experiment, the correction algorithm was obtained based on the colorimetric
color reproduction principles by a matrix masking method. The average color
difference was calculated to investigate how efficient this color correction
algorithm is.
Auto white balance for color video camera
This study was based on an assumption that the integral of all image signals from CCD becomes white. The field of the CCD was divided into several regions and then the signals of each region were compared with the standard signal level. Those regions whose signal level is within the designated value were considered to be 'approximately white' and were integrated periodically to achieve the white balance.
