Abstract:We show that surface normal information allows to significantly improve the accuracy of a spatio-temporal multi-view reconstruction. On one hand, normal information can improve the quality of photometric matching scores. On the other hand, the same normal information can be employed to drive an adaptive anisotropic surface regularization process which better preserves fine details and elongated structures than its isotropic counterpart. We demonstrate how normal information can be used and estimated and explai… Show more
“…We proposed the approach which is based on the direct integration scheme of Wu and Li [5]. It shows better results for real image surfaces in addition to convex space-time multi-view approach proposed by Oswald and Cremers [10] and Kameras [11]. By break the normals field onto local areas of interest according to Schwartz equations the integration process become more robust in local areas of interest.…”
The method of integrating normal’s field is improved by using the proposed algorithm for breaking local areas based on Schwartz equations. It was proposed local criteria of splitting normal’s onto local areas, which is crucial for the next step of building the sequence chain. The process of classification in case of reducing the training set to select regions of integration is investigated. The reason for the effect of retraining is conditioned with a minimal number of errors on the training sample. It is shown that stratification of algorithms by mistakes and increasing their similarities reduce the likelihood of the retraining. Proposed approaches are implemented as a software and are suitable for the broad class of real surfaces.
“…We proposed the approach which is based on the direct integration scheme of Wu and Li [5]. It shows better results for real image surfaces in addition to convex space-time multi-view approach proposed by Oswald and Cremers [10] and Kameras [11]. By break the normals field onto local areas of interest according to Schwartz equations the integration process become more robust in local areas of interest.…”
The method of integrating normal’s field is improved by using the proposed algorithm for breaking local areas based on Schwartz equations. It was proposed local criteria of splitting normal’s onto local areas, which is crucial for the next step of building the sequence chain. The process of classification in case of reducing the training set to select regions of integration is investigated. The reason for the effect of retraining is conditioned with a minimal number of errors on the training sample. It is shown that stratification of algorithms by mistakes and increasing their similarities reduce the likelihood of the retraining. Proposed approaches are implemented as a software and are suitable for the broad class of real surfaces.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.