Napredna pretraga

Pregled bibliografske jedinice broj: 890534

SOFT-SLAM: Computationally Efficient Stereo Visual SLAM for Autonomous UAVs

Cvišić, Igor; Ćesić, Josip; Marković, Ivan; Petrović, Ivan
SOFT-SLAM: Computationally Efficient Stereo Visual SLAM for Autonomous UAVs // Journal of field robotics, 35 (2018), 4; 578-595 doi:10.1002/rob.21762 (međunarodna recenzija, članak, znanstveni)

SOFT-SLAM: Computationally Efficient Stereo Visual SLAM for Autonomous UAVs

Cvišić, Igor ; Ćesić, Josip ; Marković, Ivan ; Petrović, Ivan

Journal of field robotics (1556-4959) 35 (2018), 4; 578-595

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
Stereo vision ; SLAM ; Visual odometry ; UAVs

Autonomous navigation of unmanned aerial vehicles (UAVs) in GPS-denied environments is a challenging problem, especially for small-scale UAVs characterized by small payload and limited battery autonomy. The solution to the aforementioned problem is vision-based simultaneous localization and mapping (SLAM), since cameras, due to their dimensions, low weight, availability, and large information bandwidth, fulfill all the constraints of UAVs. In this paper we propose a stereo vision SLAM yielding very accurate localization and a dense map of the environment developed with the aim to compete in the European Robotics Challenges (EuRoC) targeting airborne inspection of industrial facilities with small-scale UAVs. The proposed approach consists of a novel stereo odometry with feature tracking (SOFT) currently ranking first among all the stereo methods on the KITTI dataset. Relying on SOFT for pose estimation, we build a feature-based pose graph SLAM solution, which we dub SOFT- SLAM. SOFT-SLAM has a completely separated odometry and mapping thread with large loop- closing and global consistency and achieves a constant-time execution rate of 20 Hz with deterministic results using only two threads of an onboard computer used in the challenge. The UAV running our SLAM algorithm obtained the highest localization score in the EuRoC Challenge 3, Stage IIa–Benchmarking, Task 2. Furthermore, we also present an exhaustive evaluation of SOFT-SLAM on two popular public datasets and compare it to other state-of-the- art approaches ; namely, ORB-SLAM2 and LSD- SLAM. The results show that SOFT-SLAM obtains better localization accuracy on the majority of the datasets sequences, while also having lower runtime.

Izvorni jezik

Znanstvena područja
Elektrotehnika, Računarstvo, Temeljne tehničke znanosti


Fakultet elektrotehnike i računarstva, Zagreb

Časopis indeksira:

  • Current Contents Connect (CCC)
  • Web of Science Core Collection (WoSCC)
    • Science Citation Index Expanded (SCI-EXP)
    • SCI-EXP, SSCI i/ili A&HCI
  • Scopus