Share Email Print

Proceedings Paper

Noise tolerant illumination optimization applied to display devices
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Display devices have historically been designed through an iterative process using numerous hardware prototypes. This process is effective but the number of iterations is limited by the time and cost to make the prototypes. In recent years, virtual prototyping using illumination software modeling tools has replaced many of the hardware prototypes. Typically, the designer specifies the design parameters, builds the software model, predicts the performance using a Monte Carlo simulation, and uses the performance results to repeat this process until an acceptable design is obtained. What is highly desired, and now possible, is to use illumination optimization to automate the design process. Illumination optimization provides the ability to explore a wider range of design options while also providing improved performance. Since Monte Carlo simulations are often used to calculate the system performance but those predictions have statistical uncertainty, the use of noise tolerant optimization algorithms is important. The use of noise tolerant illumination optimization is demonstrated by considering display device designs that extract light using 2D paint patterns as well as 3D textured surfaces. A hybrid optimization approach that combines a mesh feedback optimization with a classical optimizer is demonstrated. Displays with LED sources and cold cathode fluorescent lamps are considered.

Paper Details

Date Published: 10 February 2005
PDF: 14 pages
Proc. SPIE 5638, Optical Design and Testing II, (10 February 2005); doi: 10.1117/12.580179
Show Author Affiliations
William J. Cassarly, Optical Research Associates (United States)
Bruce Irving, Optical Research Associates (United States)

Published in SPIE Proceedings Vol. 5638:
Optical Design and Testing II
Yongtian Wang; Zhicheng Weng; Shenghua Ye; Jose M. Sasian, Editor(s)

© SPIE. Terms of Use
Back to Top