FPGA implementation of the RANSAC based image mosaicing algorithm using the Nios II softcore

摘要

Image mosaicing is a powerful mean for obtaining a larger view of a scene than the available within a human eyes view or compact camera view. Homography matrix is the most suitable transformation model which is needed to create mosaiced images, and the estimation of the parameters of that model using RANSAC (Random Sample Consensus) is a key step in feature matching as it improves the stability of image registration, because it can estimate the parameters of homography matrix with a high degree of accuracy. In this paper a general framework for creating an image mosaic is illustrated and a RANSAC based image mosaicing algorithm will be implemented on FPGA using the configurable softcore processor ¿¿¿Nios II¿¿¿ which is synthesized on the Altera FPGA device.