Optimization of five-parameter BRDF model based on Hybrid GA-PSO algorithm

Abstract The bidirectional reflection distribution function is usually used to analyze the reflection characteristics of materials. In many cases, the BRDF models are optimized by fitting parameters. We introduced a hybrid particle swarm algorithm (GA-PSO) combining genetic algorithm and particle swarm algorithm to optimize the parameters of the five-parameter model. In order to verify the performance of the hybrid particle swarm optimization algorithm, we measured two different materials of space targets to get the experimental BRDF values. Then we simulated the parameters by using the genetic algorithm, particle swarm algorithm, and hybrid particle swarm algorithm respectively. The fitting results show that the hybrid particle swarm algorithm is better than genetic algorithm and particle swarm algorithm in accuracy and convergence speed under the same condition.