Petr Hrubý: Minimal Solvers for Estimating Camera Pose with Various Image Features and Camera Models
Tid: Ti 2025-10-07 kl 10.15 - 11.15
Plats: KTH 3418, Lindstedtsvägen 25 and Zoom
Videolänk: Zoom meeting ID: 655 8335 8144
Medverkande: Petr Hrubý (KTH)
Abstract.
Solutions of systems of polynomial equations are used for the estimation of the camera pose, which is an important step for 3D reconstruction. To quickly solve the systems of polynomial equations of a certain form, so-called minimal solvers are used. Most existing minimal solvers assume a pinhole camera model, and rely only on point correspondences. In this talk, I am going to discuss new minimal solvers for the estimation of the pose for different types of cameras, such as radial cameras and rolling shutter cameras, as well as solvers utilizing image features such as lines and affine correspondences to estimate the pose.
