Göm meny

Ground Truth for Rolling Shutter Visual-Inertial SLAM and Camera-IMU Calibration

Hannes Ovrén, Per-Erik Forssén
Proceedings of SSBA 2016 IAPR
March 2016


In this paper we describe a method to construct realistic ground truth datasets that can be used for evaluation of visual-inertial SLAM and camera-IMU calibration. The ground truth datasets consist of landmark data and a continuous camera trajectory. By feeding a dataset to a simulator, synthetic inertial and visual measurements can be generated for later use as input to an algorithm under evaluation. The datasets are created using inertial data and structure from motion. We show how to simply combine the data to a realistic dataset which includes landmark visibility information, as well as methods and heuristics to handle the scaling problem and do data coordinate frame alignment.

Bibtex entry

  author = 	 {Hannes Ovr\'en and Per-Erik Forss\'en},
  title = 	 {Ground Truth for Rolling Shutter Visual-Inertial SLAM and Camera-IMU Calibration},
  booktitle =    {Proceedings of {SSBA} 2016 {IAPR}},
  year = 	 {2016},
  month = 	 {March},
  organization = {{IAPR}},
  note =         {Non-reviewed workshop},
  publisher =    {{SSBA}}

Per-Erik Forssén

Per-Erik Forssén


Computer Vision Laboratory
Department of Electrical Engineering
Building B
Room 2D:521
SE-581 83 Linköping, Sweden
+46(0)13 285654

< >

My pages:

Informationsansvarig: Per-Erik Forss&eacute;n
Senast uppdaterad: 2023-09-06