Projector radiometric compensation using a 2D spectroradiometer

Area of Science:

• Computer Vision
• Optics
• Human-Computer Interaction

Background:

• Projection mapping (PM) offers immersive augmented reality (AR) without display devices.
• Surface textures on real-world objects cause color distortion in projected imagery.
• Existing compensation methods using RGB cameras lack sufficient accuracy.

Purpose of the Study:

• To develop an accurate projector radiometric compensation method for AR.
• To minimize color differences between target images and projected results.
• To improve alignment with human color perception.

Main Methods:

• Utilized a 2D spectroradiometer (2DSR) for precise color measurement.
• Modeled projector-to-2DSR color transformation differentiably.
• Implemented two optimization strategies: spectral error minimization and differentiable color space error minimization.

Main Results:

• The proposed method achieved significantly more accurate projector radiometric compensation.
• Experimental results demonstrated superior performance compared to RGB camera-based methods.
• The compensation method showed better alignment with human color perception.

Conclusions:

• The developed method effectively addresses color distortion in projection mapping.
• Accurate radiometric compensation enhances the realism of AR experiences.
• The approach offers a viable solution for high-fidelity projection mapping applications.