A Generalized Tool for Accurate and Efficient Image Registration of UAV Multi-lens Multispectral Cameras by N-SURF Matching

The original multispectral (MS) images obtained from multi-lens cameras (MSCs) have significant misregistration errors, which require image registration for precise spectral measurement. However, due to the nonlinearity intensity differences among MS images, performing matching is difficult find sufficient correct matches (CMs) registration, and results in a complex coarse-to-fine solution. Based on modification of speed-up robust feature (SURF), we proposed normalized SURF (N-SURF) that can significantly increase amount CMs different pairs make one-step possible. In this study, first introduce N-SURF adopt datasets acquired three representative MSCs (MCA-12, Altum, Sequoia) evaluate its ability. Meanwhile, utilized transformation models-affine transform (AT), projective (PT), an extended (EPT) errors their co-registration correctness. show obtain 6-20 times more than successfully match all while failed cases differences. Moreover, visual comparison, accuracy assessment, residual analysis EPT accurately viewpoint lens distortion AT PT, it 0.2-0.4 pixels. Subsequently, using successful model, developed automatic tool suitable various multilens MSCs.