A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements

Volker Grabe, Heinrich Bulthoff, Paolo Robuffo Giordano

Research output: Chapter in Book/Report/Conference proceedingConference contribution

30 Citations (Scopus)

Abstract

For the purpose of autonomous UAV flight control, cameras are ubiquitously exploited as a cheap and effective onboard sensor for obtaining non-metric position or velocity measurements. Since the metric scale cannot be directly recovered from visual input only, several methods have been proposed in the recent literature to overcome this limitation by exploiting independent 'metric' information from additional onboard sensors. The flexibility of most approaches is, however, often limited by the need of constantly tracking over time a certain set of features in the environment, thus potentially suffering from possible occlusions or loss of tracking during flight. In this respect, in this paper we address the problem of estimating the scale of the observed linear velocity in the UAV body frame from direct measurement of the instantaneous (and non-metric) optical flow, and the integration of an onboard Inertial Measurement Unit (IMU) for providing (metric) acceleration readings. To this end, two different estimation techniques are developed and critically compared: a standard Extended Kalman Filter (EKF) and a novel nonlinear observer stemming from the adaptive control literature. Results based on simulated and real data recorded during a quadrotor UAV flight demonstrate the effectiveness of the approach.

Original languageEnglish
Title of host publicationIEEE International Conference on Intelligent Robots and Systems
Pages5193-5200
Number of pages8
DOIs
Publication statusPublished - 2013 Dec 1
Event2013 26th IEEE/RSJ International Conference on Intelligent Robots and Systems: New Horizon, IROS 2013 - Tokyo, Japan
Duration: 2013 Nov 32013 Nov 8

Other

Other2013 26th IEEE/RSJ International Conference on Intelligent Robots and Systems: New Horizon, IROS 2013
CountryJapan
CityTokyo
Period13/11/313/11/8

Fingerprint

Units of measurement
Optical flows
Unmanned aerial vehicles (UAV)
Position measurement
Sensors
Extended Kalman filters
Velocity measurement
Cameras

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Grabe, V., Bulthoff, H., & Giordano, P. R. (2013). A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements. In IEEE International Conference on Intelligent Robots and Systems (pp. 5193-5200). [6697107] https://doi.org/10.1109/IROS.2013.6697107

A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements. / Grabe, Volker; Bulthoff, Heinrich; Giordano, Paolo Robuffo.

IEEE International Conference on Intelligent Robots and Systems. 2013. p. 5193-5200 6697107.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Grabe, V, Bulthoff, H & Giordano, PR 2013, A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements. in IEEE International Conference on Intelligent Robots and Systems., 6697107, pp. 5193-5200, 2013 26th IEEE/RSJ International Conference on Intelligent Robots and Systems: New Horizon, IROS 2013, Tokyo, Japan, 13/11/3. https://doi.org/10.1109/IROS.2013.6697107
Grabe V, Bulthoff H, Giordano PR. A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements. In IEEE International Conference on Intelligent Robots and Systems. 2013. p. 5193-5200. 6697107 https://doi.org/10.1109/IROS.2013.6697107
Grabe, Volker ; Bulthoff, Heinrich ; Giordano, Paolo Robuffo. / A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements. IEEE International Conference on Intelligent Robots and Systems. 2013. pp. 5193-5200
@inproceedings{ef698d8388f84c419c87022b178e785b,
title = "A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements",
abstract = "For the purpose of autonomous UAV flight control, cameras are ubiquitously exploited as a cheap and effective onboard sensor for obtaining non-metric position or velocity measurements. Since the metric scale cannot be directly recovered from visual input only, several methods have been proposed in the recent literature to overcome this limitation by exploiting independent 'metric' information from additional onboard sensors. The flexibility of most approaches is, however, often limited by the need of constantly tracking over time a certain set of features in the environment, thus potentially suffering from possible occlusions or loss of tracking during flight. In this respect, in this paper we address the problem of estimating the scale of the observed linear velocity in the UAV body frame from direct measurement of the instantaneous (and non-metric) optical flow, and the integration of an onboard Inertial Measurement Unit (IMU) for providing (metric) acceleration readings. To this end, two different estimation techniques are developed and critically compared: a standard Extended Kalman Filter (EKF) and a novel nonlinear observer stemming from the adaptive control literature. Results based on simulated and real data recorded during a quadrotor UAV flight demonstrate the effectiveness of the approach.",
author = "Volker Grabe and Heinrich Bulthoff and Giordano, {Paolo Robuffo}",
year = "2013",
month = "12",
day = "1",
doi = "10.1109/IROS.2013.6697107",
language = "English",
isbn = "9781467363587",
pages = "5193--5200",
booktitle = "IEEE International Conference on Intelligent Robots and Systems",

}

TY - GEN

T1 - A comparison of scale estimation schemes for a quadrotor UAV based on optical flow and IMU measurements

AU - Grabe, Volker

AU - Bulthoff, Heinrich

AU - Giordano, Paolo Robuffo

PY - 2013/12/1

Y1 - 2013/12/1

N2 - For the purpose of autonomous UAV flight control, cameras are ubiquitously exploited as a cheap and effective onboard sensor for obtaining non-metric position or velocity measurements. Since the metric scale cannot be directly recovered from visual input only, several methods have been proposed in the recent literature to overcome this limitation by exploiting independent 'metric' information from additional onboard sensors. The flexibility of most approaches is, however, often limited by the need of constantly tracking over time a certain set of features in the environment, thus potentially suffering from possible occlusions or loss of tracking during flight. In this respect, in this paper we address the problem of estimating the scale of the observed linear velocity in the UAV body frame from direct measurement of the instantaneous (and non-metric) optical flow, and the integration of an onboard Inertial Measurement Unit (IMU) for providing (metric) acceleration readings. To this end, two different estimation techniques are developed and critically compared: a standard Extended Kalman Filter (EKF) and a novel nonlinear observer stemming from the adaptive control literature. Results based on simulated and real data recorded during a quadrotor UAV flight demonstrate the effectiveness of the approach.

AB - For the purpose of autonomous UAV flight control, cameras are ubiquitously exploited as a cheap and effective onboard sensor for obtaining non-metric position or velocity measurements. Since the metric scale cannot be directly recovered from visual input only, several methods have been proposed in the recent literature to overcome this limitation by exploiting independent 'metric' information from additional onboard sensors. The flexibility of most approaches is, however, often limited by the need of constantly tracking over time a certain set of features in the environment, thus potentially suffering from possible occlusions or loss of tracking during flight. In this respect, in this paper we address the problem of estimating the scale of the observed linear velocity in the UAV body frame from direct measurement of the instantaneous (and non-metric) optical flow, and the integration of an onboard Inertial Measurement Unit (IMU) for providing (metric) acceleration readings. To this end, two different estimation techniques are developed and critically compared: a standard Extended Kalman Filter (EKF) and a novel nonlinear observer stemming from the adaptive control literature. Results based on simulated and real data recorded during a quadrotor UAV flight demonstrate the effectiveness of the approach.

UR - http://www.scopus.com/inward/record.url?scp=84893734230&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84893734230&partnerID=8YFLogxK

U2 - 10.1109/IROS.2013.6697107

DO - 10.1109/IROS.2013.6697107

M3 - Conference contribution

AN - SCOPUS:84893734230

SN - 9781467363587

SP - 5193

EP - 5200

BT - IEEE International Conference on Intelligent Robots and Systems

ER -