研究目的
To perform virtual relighting of stained glass artwork by acquiring the actual light transport properties of a given artifact using novel methods built upon image based priors.
研究成果
The paper presents a practical workflow for digitizing stained glass, enabling interactive rendering capabilities under various lighting conditions. The method allows for the preservation of volume impurities and heterogeneity in light transport, providing a faithful representation of the visual experience of stained glass artifacts.
研究不足
The technical and application constraints include the difficulty in analyzing stained glass windows due to their size and position, and the challenge of acquiring the full bidirectional scattering distribution function (BSDF) due to its impracticality.
1:Experimental Design and Method Selection:
The methodology involves an acquisition stage with a multiple illumination setup followed by a modelling stage where the acquired data is used to obtain re?ectance models by exploiting prior knowledge on re?ectance properties and in the ?nal rendering stage, the object is rendered in virtual environments.
2:Sample Selection and Data Sources:
The samples or datasets used in the experiment include stained glass artifacts, with selection criteria based on their size and position.
3:List of Experimental Equipment and Materials:
Instruments and devices include a projector-camera setup for controlled illumination and natural illumination for uncontrolled scenarios.
4:Experimental Procedures and Operational Workflow:
The process involves acquiring image pairs of incident illumination and stained glass appearance, modelling stained glass using light transport matrices obtained by solving a linear inverse problem, and exploiting the known approximate geometry of stained glass for a compact representation.
5:Data Analysis Methods:
The approach for analyzing experimental data includes solving for the light transport matrix under controlled and uncontrolled illuminations using sparse regularization and compressive sensing techniques.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容