en-es
en
0.25
0.5
0.75
1.25
1.5
1.75
2
Spline Fusion: A continuous-time representation for visual-inertial fusion with application to rolling shutter cameras
Published on Apr 03, 20142969 Views
This paper describes a general continuous-time framework for visual-inertial simultaneous localization and mapping and calibration. We show how to use a spline parameterization that closely matches
Related categories
Chapter list
Spline fusion: A continuous-time representation for visual-inertial fusion with application to rolling shutter cameras00:00
Motivation: Flexible slam / Calibration00:11
Traditional visual SLAM / SFM00:57
What about rolling shutter cameras?01:28
High-rate / Unsyncrhonized devices?02:01
A solution: Continuous time slam02:51
Previous research03:39
What’s the right representation?03:56
Quaternion interpolation05:12
B-splines and control poses07:15
Cumulative B-splines09:16
Cumulative B-splines ∈ SE310:36
Closed form 1st and 2nd derivatives11:49
Objective function12:02
Simulated visual-inertial SLAM - 113:07
Simulated visual-inertial SLAM - 213:37
Rolling shutter and IMU - 114:13
Rolling shutter and IMU - 215:18
Monocular rolling shutter SLAM15:28
Prediction into rolling shutter15:53
Improved monocular odometry16:40
Thanks!17:19