Multi-sensor image registration using edge-enhanced maximally stable extremal region

Image registration is an important topic in computer vision and many techniques have been developed. But when it comes to the multi-sensor images, intensity-based global methods often achieve poor results. In this paper, we propose an edge-enhanced maximally stable extremal region method (E-MSER) to improve the registration performance of multi-sensor images. It is obtained by combining edge enhancement with maximally stable extremal region (MSER) detector. Thus the region features can maintain stability through the changing image intensities. Criteria including matching score, repeatability, recall, precision and root-mean-square error (RMSE) are used for evaluation. Experiment results show that E-MSER outperforms other detectors with its registration accuracy reaching pixel-level.