Finger print Feature Extraction Using Hybrid Approach: QPSO and Bees Algorithms Based on 3D Logistic Map

Abstract

Particle swarm optimization algorithm is easy to access premature convergence in the solution process, and also fall into the local optimal solution. The Bees algorithm is an inference optimization based on the foraging behavior of honey bees. It has been assured that this algorithm is able to search for global optimum, but there is one defect, it’s the fact that the global best solution is not used in a direct manner, but the Bees algorithm stores it of each iteration. We propose a new hybrid approach in order to address these problems, it is between Quantum particle swarm optimization and Bees algorithm based on 3D logistic map. On one hand, the 3D logistic map is employed for initializing uniform distributed particles so as to enhance the initial population quality, which is a very efficient yet simple method for improving the quality of initial population. On the other hand, the essence of this approach is the use Quantum particle swarm optimization for optimum fitness value of population in Bees algorithm. After determining the starting point in the new algorithm, the form of distribution is of circles with random angles and random diagonal where 3D logistic map generate the random numbers. The algorithm was applied for extracting the characteristics of the fingerprint, and the results when compared to the traditional particle swarm optimization algorithm were excellent.