Securing Color Fidelity in 3D Architectural Heritage Scenarios

Area of Science:

• Digital Heritage
• Photogrammetry
• Computer Vision

Background:

• Maintaining color fidelity in 3D heritage modeling is a significant challenge.
• Image color accuracy is crucial for photogrammetric processing and analysis.

Purpose of the Study:

• To develop an automated solution for improving radiometric quality in image datasets for 3D heritage modeling.
• To enhance the performance of camera orientation and dense image matching in photogrammetry.
• To achieve robust automatic color balance, exposure equalization, and faithful color appearance of digitized artifacts.

Main Methods:

• Development of an automated solution for radiometric quality improvement.
• Implementation of ACR scripts in MATLAB, including weighted polynomial regression.
• Focus on automatic color balance, exposure equalization, and RGB-to-gray conversion stability.

Main Results:

• Demonstrated efficiency of the developed methodology through a series of tests.
• Evaluation of color accuracy and characterization of digitized heritage artifacts.
• Improved performances in camera orientation and dense image matching stages of the photogrammetric pipeline.

Conclusions:

• The proposed automated solution effectively improves radiometric quality and color accuracy in 3D heritage modeling.
• The methodology ensures stable color conversion and faithful digital representation of artifacts.
• This work contributes to advancing the field of image-based 3D modeling for cultural heritage preservation.