An Improved PSO Method for Detecting Feature Points of Large-Scale Point-Based Models

An improved particle swarm optimization (PSO) method for detecting feature points of large-scale point-based models is presented in this paper. By redefining the particle, fitness, initial and ending conditions, local optimum and global optimum, iterative equations of PSO, this method can search multi-regions for the feature points in an adaptive random and parallel manner. The fitness is defined as local surface variation. The global search and two different local search methods are combined to detect the feature points quickly. This method can realize the fast displaying of the characteristic of large-scale models. The effectiveness of the algorithm has been proved by the experiments.