Research On Image Registration Based On Particle Swarm Optimization Algorithm

Image registration is a basic problem in image processing,Its main purpose is to find the best registration relationship between two or more images based on the difference in resolution,grayscale properties and position.Now the technology has been widely used in a number of areas,such as computer vision,medical diagnosis and adjuvant therapy,remote sensing images.Image registration technology can be divided into three categories according to the different methods of image information processing,They are image processing based on feature processing,image registration based on gray processing and image registration based on transform domain processing.In this paper,we use the image registration technique based on gray scale processing and use the normalized mutual information as the measure function of image registration.The hybrid algorithm is used as the optimization algorithm of registration.An algorithm based on improved particle swarm optimization and fireworks algorithm to achieve the registration of two images.In this paper,we first introduce the significance of image registration technology research,background and domestic and foreign research status,and introduced the image registration technology related theory.Then we introduce the optimization problem and optimization algorithm in image registration,and propose an improved particle swarm optimization algorithm.Firstly,the inertia weight in the particle swarm algorithm is improved according to the evolution factor in the optimization process.At the same time,in order to increase the diversity of the particle swarm,the Gaussian mutation operator in the fireworks algorithm is used to deal with the particle swarm to prevent the particle swarm precocious.Experiments show that the improved particle swarm algorithm proposed in this paper can improve the searching ability of the particles effectively,and it is effective and superior in solving the complex function optimization problem.Finally,the algorithm is applied to the image registration and compared with other registration algorithms.The experimental results show that the algorithm can improve the registration accuracy and time-consuming,and also meet the requirements of image registration.