Resumen: This paper presents a dense monocular mapping algorithm that improves the accuracy of the state-of-the-art variational and multiview stereo methods by incorporat- ing scene priors into its formulation. Most of the improvement of our proposal is in low- textured image regions and for low-parallax camera motions; two typical failure cases of multiview mapping.
The specific priors we model are the pla- narity of homogeneous color regions, the re- peating geometric primitives of the scene –that can be learned from data– and the Manhat- tan structure of indoor rooms. We evaluate the performance of our method in our own sequences and in the publicly available NYU dataset, emphasizing its strengths and weak- nesses in different cases. Idioma: Inglés DOI: 10.1016/j.amc.2015.06.042 Año: 2015 Publicado en: Applied Mathematics and Computation 268 (2015), 227-245 ISSN: 0096-3003 Factor impacto JCR: 1.345 (2015) Categ. JCR: MATHEMATICS, APPLIED rank: 54 / 254 = 0.213 (2015) - Q1 - T1 Factor impacto SCIMAGO: 0.95 - Computational Mathematics (Q2) - Applied Mathematics (Q2)