Optic Flow-Based Reactive Collision Prevention for MAVs Using the Fictitious Obstacle Hypothesis

Feng Xiao; Peter Zheng; Julien Di Tria; Basaran Bahadir Kocer; Mirko Kovac

doi:10.1109/LRA.2021.3062317

Optic Flow-Based Reactive Collision Prevention for MAVs Using the Fictitious Obstacle Hypothesis

Feng Xiao^*
, Peter Zheng
, Julien Di Tria
, Basaran Bahadir Kocer
, Mirko Kovac

^*Corresponding author for this work

Research output: Contribution to journal › Article (Academic Journal) › peer-review

24 Citations (Scopus)

Abstract

Optical flow sensors and optical flow divergence (OFD) have offered partial solutions for obstacle avoidance, landing, and perching with micro aerial vehicles. Theoretically, OFD can indicate the risk of collision, providing that the sensors' field of view is bounded within a single flat surface on the obstacle. However, in the real world, directly measuring the risk of collision with OFD generates false alarms due to rapidly changing speeds and irregular surroundings. In this letter, we present a new obstacle detection strategy based on an extended Kalman filter (EKF) combining the OFD with inertial sensing. The introduction of a fictitious obstacle hypothesis and the use of the EKF estimates enable us to differentiate the surrounding-generated OFD from the OFD caused by the actual obstacle. An embedded constant zero-OFD controller is then used for post-detection emergency deceleration. The ultra-light OFD estimation and control system, with a mass of 20g, estimates OFD at 160Hz. The system was validated on a 158g mini quadrotor in both laboratory and field tests. Experimental results illustrate that the presented system can achieve accurate obstacle detection, near-obstacle distance estimation, and controlled deceleration to prevent collisions.11Video attachment: https://youtu.be/yIyYHYN0jOw.

Original language	English
Article number	9363555
Pages (from-to)	3144-3151
Number of pages	8
Journal	IEEE Robotics and Automation Letters
Volume	6
Issue number	2
DOIs	https://doi.org/10.1109/LRA.2021.3062317
Publication status	Published - 25 Feb 2021

Bibliographical note

Funding Information:
Manuscript received October 15, 2020; accepted February 8, 2021. Date of publication February 25, 2021; date of current version March 19, 2021. This letter was recommended for publication by Associate Editor G. Loianno and Editor P. Pounds upon evaluation of the reviewers’ comments. This work was supported in part by EPSRC (Award no. EP/R009953/1, EP/N018494/1, and EP/R026173/1), in part by NERC under Grant NE/L002515/1 and Grant SEARRP, and in part by the EU H2020 AeroTwin Project under Grant 810321. The work of Mirko Kovac was supported by the Royal Society Wolfson Fellowship (RSWF/R1/18003). (Corresponding author: Feng Xiao.) Feng Xiao, Julien di Tria, and Basaran Bahadir Kocer are with the Aerial Robotics Lab (ARL), Department of Aeronautics, Imperial College London, South Kensington Campus, SW7 2AZ London, U.K. (e-mail: feng.xiao16@ imperial.ac.uk; [email protected]; [email protected]).

Publisher Copyright:
© 2016 IEEE.

Keywords

Aerial systems
applications
biologically-inspired robots
collision avoidance

Access to Document

10.1109/LRA.2021.3062317

Handle.net

Persistent link

Cite this

@article{eac2df42dba04bd9919132b8de0fdbdf,

title = "Optic Flow-Based Reactive Collision Prevention for MAVs Using the Fictitious Obstacle Hypothesis",

abstract = "Optical flow sensors and optical flow divergence (OFD) have offered partial solutions for obstacle avoidance, landing, and perching with micro aerial vehicles. Theoretically, OFD can indicate the risk of collision, providing that the sensors' field of view is bounded within a single flat surface on the obstacle. However, in the real world, directly measuring the risk of collision with OFD generates false alarms due to rapidly changing speeds and irregular surroundings. In this letter, we present a new obstacle detection strategy based on an extended Kalman filter (EKF) combining the OFD with inertial sensing. The introduction of a fictitious obstacle hypothesis and the use of the EKF estimates enable us to differentiate the surrounding-generated OFD from the OFD caused by the actual obstacle. An embedded constant zero-OFD controller is then used for post-detection emergency deceleration. The ultra-light OFD estimation and control system, with a mass of 20g, estimates OFD at 160Hz. The system was validated on a 158g mini quadrotor in both laboratory and field tests. Experimental results illustrate that the presented system can achieve accurate obstacle detection, near-obstacle distance estimation, and controlled deceleration to prevent collisions.11Video attachment: https://youtu.be/yIyYHYN0jOw.",

keywords = "Aerial systems, applications, biologically-inspired robots, collision avoidance",

author = "Feng Xiao and Peter Zheng and Tria, \{Julien Di\} and Kocer, \{Basaran Bahadir\} and Mirko Kovac",

note = "Funding Information: Manuscript received October 15, 2020; accepted February 8, 2021. Date of publication February 25, 2021; date of current version March 19, 2021. This letter was recommended for publication by Associate Editor G. Loianno and Editor P. Pounds upon evaluation of the reviewers{\textquoteright} comments. This work was supported in part by EPSRC (Award no. EP/R009953/1, EP/N018494/1, and EP/R026173/1), in part by NERC under Grant NE/L002515/1 and Grant SEARRP, and in part by the EU H2020 AeroTwin Project under Grant 810321. The work of Mirko Kovac was supported by the Royal Society Wolfson Fellowship (RSWF/R1/18003). (Corresponding author: Feng Xiao.) Feng Xiao, Julien di Tria, and Basaran Bahadir Kocer are with the Aerial Robotics Lab (ARL), Department of Aeronautics, Imperial College London, South Kensington Campus, SW7 2AZ London, U.K. (e-mail: feng.xiao16@ imperial.ac.uk; [email protected]; [email protected]). Publisher Copyright: {\textcopyright} 2016 IEEE.",

year = "2021",

month = feb,

day = "25",

doi = "10.1109/LRA.2021.3062317",

language = "English",

volume = "6",

pages = "3144--3151",

journal = "IEEE Robotics and Automation Letters",

issn = "2377-3766",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

number = "2",

}

TY - JOUR

T1 - Optic Flow-Based Reactive Collision Prevention for MAVs Using the Fictitious Obstacle Hypothesis

AU - Xiao, Feng

AU - Zheng, Peter

AU - Tria, Julien Di

AU - Kocer, Basaran Bahadir

AU - Kovac, Mirko

N1 - Funding Information: Manuscript received October 15, 2020; accepted February 8, 2021. Date of publication February 25, 2021; date of current version March 19, 2021. This letter was recommended for publication by Associate Editor G. Loianno and Editor P. Pounds upon evaluation of the reviewers’ comments. This work was supported in part by EPSRC (Award no. EP/R009953/1, EP/N018494/1, and EP/R026173/1), in part by NERC under Grant NE/L002515/1 and Grant SEARRP, and in part by the EU H2020 AeroTwin Project under Grant 810321. The work of Mirko Kovac was supported by the Royal Society Wolfson Fellowship (RSWF/R1/18003). (Corresponding author: Feng Xiao.) Feng Xiao, Julien di Tria, and Basaran Bahadir Kocer are with the Aerial Robotics Lab (ARL), Department of Aeronautics, Imperial College London, South Kensington Campus, SW7 2AZ London, U.K. (e-mail: feng.xiao16@ imperial.ac.uk; [email protected]; [email protected]). Publisher Copyright: © 2016 IEEE.

PY - 2021/2/25

Y1 - 2021/2/25

N2 - Optical flow sensors and optical flow divergence (OFD) have offered partial solutions for obstacle avoidance, landing, and perching with micro aerial vehicles. Theoretically, OFD can indicate the risk of collision, providing that the sensors' field of view is bounded within a single flat surface on the obstacle. However, in the real world, directly measuring the risk of collision with OFD generates false alarms due to rapidly changing speeds and irregular surroundings. In this letter, we present a new obstacle detection strategy based on an extended Kalman filter (EKF) combining the OFD with inertial sensing. The introduction of a fictitious obstacle hypothesis and the use of the EKF estimates enable us to differentiate the surrounding-generated OFD from the OFD caused by the actual obstacle. An embedded constant zero-OFD controller is then used for post-detection emergency deceleration. The ultra-light OFD estimation and control system, with a mass of 20g, estimates OFD at 160Hz. The system was validated on a 158g mini quadrotor in both laboratory and field tests. Experimental results illustrate that the presented system can achieve accurate obstacle detection, near-obstacle distance estimation, and controlled deceleration to prevent collisions.11Video attachment: https://youtu.be/yIyYHYN0jOw.

AB - Optical flow sensors and optical flow divergence (OFD) have offered partial solutions for obstacle avoidance, landing, and perching with micro aerial vehicles. Theoretically, OFD can indicate the risk of collision, providing that the sensors' field of view is bounded within a single flat surface on the obstacle. However, in the real world, directly measuring the risk of collision with OFD generates false alarms due to rapidly changing speeds and irregular surroundings. In this letter, we present a new obstacle detection strategy based on an extended Kalman filter (EKF) combining the OFD with inertial sensing. The introduction of a fictitious obstacle hypothesis and the use of the EKF estimates enable us to differentiate the surrounding-generated OFD from the OFD caused by the actual obstacle. An embedded constant zero-OFD controller is then used for post-detection emergency deceleration. The ultra-light OFD estimation and control system, with a mass of 20g, estimates OFD at 160Hz. The system was validated on a 158g mini quadrotor in both laboratory and field tests. Experimental results illustrate that the presented system can achieve accurate obstacle detection, near-obstacle distance estimation, and controlled deceleration to prevent collisions.11Video attachment: https://youtu.be/yIyYHYN0jOw.

KW - Aerial systems

KW - applications

KW - biologically-inspired robots

KW - collision avoidance

UR - https://www.scopus.com/pages/publications/85101811753

U2 - 10.1109/LRA.2021.3062317

DO - 10.1109/LRA.2021.3062317

M3 - Article (Academic Journal)

AN - SCOPUS:85101811753

SN - 2377-3766

VL - 6

SP - 3144

EP - 3151

JO - IEEE Robotics and Automation Letters

JF - IEEE Robotics and Automation Letters

IS - 2

M1 - 9363555

ER -

Optic Flow-Based Reactive Collision Prevention for MAVs Using the Fictitious Obstacle Hypothesis

Abstract

Bibliographical note

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

Cite this