2017/02/09 by Soumyadip Sengupta, Sengupta, Soumyadip, Tal Amir +11 · 2 citations
Computer Science · Engineering · #Advanced Image Processing Techniques #Advanced Vision and Imaging #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Sparse and Compressive Sensing Techniques
paper · pdf · doi:10.48550/arxiv.1702.03023
openalex publication_date 2017/02/09 · openalex created_date 2022/10/04 · openalex updated_date 2026/07/28
Accurate estimation of camera matrices is an important step in structure from\nmotion algorithms. In this paper we introduce a novel rank constraint on\ncollections of fundamental matrices in multi-view settings. We show that in\ngeneral, with the selection of proper scale factors, a matrix formed by\nstacking fundamental matrices between pairs of images has rank 6. Moreover,\nthis matrix forms the symmetric part of a rank 3 matrix whose factors relate\ndirectly to the corresponding camera matrices. We use this new characterization\nto produce better estimations of fundamental matrices by optimizing an L1-cost\nfunction using Iterative Re-weighted Least Squares and Alternate Direction\nMethod of Multiplier. We further show that this procedure can improve the\nrecovery of camera locations, particularly in multi-view settings in which\nfewer images are available.\n