Application and performance evaluation of a method using block structures for saddle point problems appearing in image reconstruction problems
Shota Ishikawa, Hiroto Tadano, Ayumu Saitoh
Abstract
Open-access reader
Shota Ishikawa, Hiroto Tadano, Ayumu Saitoh
Abstract
Open-access reader
This paper focuses on reducing the computation time for problems in 3D volumetric model reconstruction using the extended CSRBF method. Since the most time-consuming part of the application is solving a saddle point problem with multiple right-hand sides, it is important that this problem be solved in the shortest time possible. Recently, we have proposed an approach for solving saddle point problems using the associated block structure. In this paper, we apply the proposed approach to the problem of 3D volumetric model reconstruction. Through numerical experiments, the performance of the proposed approach is evaluated by comparing it with conventional approaches.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper focuses on reducing the computation time for problems in 3D volumetric model reconstruction using the extended CSRBF method. Since the most time-consuming part of the application is solving a saddle point problem with multiple right-hand sides, it is important that this problem be solved in the shortest time possible. Recently, we have proposed an approach for solving saddle point problems using the associated block structure. In this paper, we apply the proposed approach to the problem of 3D volumetric model reconstruction. Through numerical experiments, the performance of the proposed approach is evaluated by comparing it with conventional approaches.
Key concepts: Saddle point, Saddle, Block (permutation group theory), Computation, Point (geometry), Mathematical optimization, Algorithm, Image (mathematics)