Particle Swarm Optimisation as a Hardware-Oriented Meta-heuristic for Image Analysis

In this paper we propose a variant of particle swarm optimisation (PSO), oriented at image analysis applications, that is suitable for implementation on hardware chips. The new variant, called HPSO (Hardware PSO), can be mapped easily to field-programmable gate arrays (FPGAs). The modularity of our new architecture permits to take full advantage of the active dynamic partial reconfiguration allowed by modern FPGAs. Experimental results based on simulations of a license plate detection task are presented to evaluate our design for solving real-world problems.