Sampling optimization for printer characterization by greedy search

Area of Science:

• Color Science
• Digital Printing
• Computational Imaging

Background:

• Accurate printer color characterization is crucial for consistent and acceptable print output.
• The complex, nonlinear relationship between printer inputs (RGB/CMYK) and colorimetry necessitates extensive sampling.
• Existing sampling methods are often inefficient, yielding diminishing returns with increased sample density.

Purpose of the Study:

• To develop a more efficient method for printer color characterization.
• To reduce the number of samples required for achieving a target color accuracy.
• To overcome limitations of previous sampling techniques that either underutilize the color space or make unwarranted assumptions.

Main Methods:

• An optimization-based approach to printer color characterization is presented.
• The method avoids assumptions about the underlying color relationship beyond tessellation and interpolation.
• A computationally intensive initial optimization is performed once during system engineering.

Main Results:

• A significant reduction in the number of samples required for accurate color characterization was achieved.
• The proposed method demonstrates improved efficiency compared to simple sampling techniques.
• The computational cost is confined to the engineering phase, remaining transparent to end-users.

Conclusions:

• The novel approach offers a more efficient and effective solution for printer color characterization.
• This method allows for achieving high color accuracy with fewer samples, optimizing the characterization process.
• The trade-off of initial computational cost is justified by long-term gains in efficiency and accuracy.