Skip to main content

Geometry of two, three, or four cameras

Event Details

Date
Wednesday, May 10, 2023
Time
12:30 p.m.
Location
Orchard View Room, Third Floor, Discovery Building
Description
Abstract: Estimating the relative orientation of multiple cameras is a classical problem of computer vision, and a key part of modern systems that reconstruct 3D scenes from images. I will briefly survey well-known and commonly-used methods for two perspective cameras that estimate the so-called fundamental or essential matrix. These methods require solving systems of polynomial equations with (very) special structure.
Cost
Free

Tags