Computational Caustic Design for Surface Light Source.

Area of Science:

• Optics and Photonics
• Computer Graphics
• Optical Engineering

Background:

• Designing freeform surfaces for light control is difficult due to simplified source assumptions in current methods.
• Existing caustic lens designs often fail with complex, real-world illumination patterns.

Purpose of the Study:

• To develop a novel method for designing freeform lenses that accurately control light from arbitrary surface light sources.
• To improve the fidelity of light shaping by better representing real-world illumination.

Main Methods:

• Representing surface light sources using an optimized set of point sources.
• Utilizing a differentiable rendering framework to fit point source parameters to real illumination.
• Employing a physically-based rendering approach simulating light transmission using flux.
• Applying a contraction mapping for efficient parameter space exploration during optimization.
• Designing freeform lens shapes based on flux consistency and normal integrability.

Main Results:

• The proposed method accurately represents real surface light sources, outperforming point-source approximations.
• The designed caustic lenses produce light distributions closely matching the target illumination patterns.
• Simulations and physical experiments validate the effectiveness of the novel approach.

Conclusions:

• The developed technique offers a more accurate and versatile approach to freeform lens design for complex lighting scenarios.
• This method advances the capability to precisely control light based on real-world illumination, with applications in various optical systems.