In this paper, we proposed a methodology to optimize the parametric window of a liquid crystal display (LCD) system, whose visual performance was deteriorated by the pixel moiré arising in between multiple periodic structures. Conventional analysis and minimization of moiré patterns are limited by few parameters. With the proposed image-based particle swarm optimization (PSO), we enable a multivariable optimization at the same time. A series of experiments was conducted to validate the methodology. Due to its versatility, the proposed technique will certainly have a promising impact on the fast optimization in LCD design with more complex configuration.