Observation-driven generation of texture maps depicting dust accumulation over time

Area of Science:

• Computer Graphics
• Material Appearance Modeling
• Computational Imaging

Background:

• Realism in computer-generated imagery (CGI) is crucial for applications like virtual reality and film.
• Subtle visual cues, such as dust accumulation, significantly impact perceived realism.
• Temporal variations in dust are common in real indoor environments but challenging to simulate.

Purpose of the Study:

• To develop a framework for generating time-dependent dust accumulation textures.
• To enhance the realism of synthetic indoor scenes by simulating dust.
• To provide a physically inspired approach for dust simulation.

Main Methods:

• A framework for generating dust accumulation textures over time.
• Physically inspired algorithms to simulate dust roughness and its effect on surfaces.
• Modulation of dust effects based on illumination and viewing geometries.
• Development guided by empirical insights and observational experiments.

Main Results:

• A method for generating textures representing dust accumulation over time.
• Demonstration of how dust roughness affects substrate appearance.
• Simulation of time-dependent changes in dusty scenes.
• Provision of image sequences illustrating distinct dust accumulation scenarios.

Conclusions:

• The proposed framework effectively generates realistic dust accumulation textures.
• The approach enhances the visual plausibility of synthetic indoor environments.
• The physically inspired method accurately portrays dust effects under varying conditions.