Imposters for Particle-Based Datasets

Many particle-based datasets produced by molecular dynamics simulations consist of millions of particles. Current visualization techniques are incapable of representing those large-scale datasets consistently across all scales. They either produce overlysmooth representation or are prone to aliasing due to under-sampling. This work introduces a technique that captures the micro-scale surface features accurately and is able to represent the complex local illumination behavior of hundreds of particles in a single pixel-footprint. This scale-consistent technique allows for an overview that is resistant to aliasing and true to the micro-scale surface. This technique produces visualization of shock waves that could not be seen before without dedicated visualization methods. It can be applied to any opaque particle glyphs and BRDF model.