Spectral characterization of a color scanner by adaptive estimation

Area of Science:

• Color Science
• Image Processing
• Metrology

Background:

• Color scanners often exhibit non-linear responses, complicating accurate spectral reflectance estimation.
• Traditional spectral characterization methods can be sensitive to inaccuracies in spectral responsivity estimation.

Purpose of the Study:

• To propose a novel adaptive estimation method for spectral characterization of color scanners.
• To address the challenge of estimating high-dimensional reflectance vectors from low-dimensional scanner responses, especially under non-linear conditions.

Main Methods:

• Investigated the spectral linearity of color scanners.
• Developed an adaptive estimation technique utilizing local statistics of training samples.
• Focused on estimating spectral reflectance without relying on mathematically recovered spectral responsivity.

Main Results:

• The proposed adaptive estimation method effectively estimates spectral reflectance vectors.
• The method demonstrated significant advantages compared to other existing spectral characterization techniques.
• Inaccuracy in spectral responsivity did not critically impact the characterization results.

Conclusions:

• Adaptive estimation provides a robust and advantageous approach for spectral characterization of color scanners.
• The method is particularly effective when scanners deviate from linearity.
• This technique offers improved spectral reflectance estimation accuracy for color scanning applications.