Particle swarm density for transfer function design

Abstract

This research uses bionic algorithms mapping the number of visits of particles for the design of transfer functions. For a given voxel, its opacity is determined by the number of particles in it at the present and influenced by the number of particles passing it in the past. The proposed system successfully extract features such as bones or issues in the volume data. Initially, the agents scatter around the volume data and are attracted by the featured areas. The movement of agents are governed by global optimization, avoiding trial and error efforts to come up with a good transfer function design for a given volume data.