Hardware-Software Implementation of Rigid Image Registration with Particle Swarm Optimization Algorithm

Abstract

In this paper, SoC based hardware-software implementation is proposed for image registration process using Particle Swarm Optimization (PSO) algorithm. Zynq 7000 SoC platform is used to perform image registration of the standard test images under test.Particle Swarm Optimization (PSO) algorithm is developed in FPGA hardware and is considered as an optimizer that maximizes the Symmetric Index (SI) of the registered images. OpenCV functions are used in SoC hardware for image processing and histogram matching. The similarity of two images is checked using fourdifferent histogram matching functions: (1) Correlation, (2) Chi-square, (3) Intersections and, (4) Bhattacharya distance. Implementation results show that correlation function performs better than other three histogram functions.