《True Multi-Image Alignment and its Application to Mosaicing and Lens》.pdf

True Multi-Image Alignment and its Application to Mosaicing and Lens Distortion Correction Harpreet S. Sawhney Rakesh Kumar Sarno Corp oration CN5300, Princeton, NJ 08530 Ph: 609-734-2866 FAX: 609-734-2662 fhsawhney,rkumarg@sarno.com Keywords : Image Sequence Analysis, Video Mosaics, Lens Distortion Correction. True Multi-Image Alignment and its Application to Mosaicing and Lens Distortion Correction Keywords : Image Sequence Analysis, Video Mosaics. Abstract Multiple images of a scene are related through 2D/3D view transformations and linear and non-linear camera transformations. In the traditional techniques to compute these transformations, esp ecially the ones relying on direct intensity gradients, one image and its co ordinate system have b een assumed to b e ideal and distortion free. In this pap er, we present an algorithm for true multi- image alignment that do es not rely on the measurements of a reference image b eing distortion free. The algorithm is develop ed to sp ecically align and mosaic images using parametric transformations in the presence of lens distortion. When lens distortion is present none of the images can b e assumed to b e ideal. In our formulation, all the images are mo deled as intensity measurements represented in their resp ective co ordinate systems, each of which is related to an ideal co ordinate system through an interior camera transformation and an exterior view transformation. The goal of the accompanying algorithm is to compute an image in the ideal co ordinate system while solving for the transformations that relate the id