WALK-VIO: Walking-motion-Adaptive Leg Kinematic Constraint Visual-Inertial Odometry for Quadruped Robots

Lim, Hyunjun; Yu, Byeongho; Kim, Yeeun; Byun, Joowoong; Soonpyo, Kwon，; Park, Haewon; Myung, Hyun

doi:10.48550/arxiv.2111.15164

Cited by 1 publication

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1) Pronto [11], which is an EKF-based algorithm using an IMU, leg odometry, and a camera in a loosely coupled manner; (2) VINS-Fusion [24], a factor graph-based multisensor state estimator. In this comparison, a stereo camera and IMU configuration is used for fairness; and (3) WALK-VIO [15], which is the previous version of STEP, utilizing the leg kinematic constraint based on a non-slip assumption, is also compared. To compare the effect of leg kinematic constraints only, the adaptive factor was not considered.…”

Section: Resultsmentioning

confidence: 99%

“…Contrarily, the tightly coupled visual-inertial odometry system, VINS-Fusion [24] showed relatively better performance. However, WALK-VIO [15] degraded on the slippery terrain due to the non-slip assumption, as described in Fig. 4(b).…”

Section: A Gazebo Simulationmentioning

confidence: 97%

“…Our previous work, WALK-VIO [15], tightly fuses an inertial sensor, a camera sensor, and joint encoders. The issue that the generated body motion varies with the different controllers was a motivation for developing WALK-VIO.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

STEP: State Estimator for Legged Robots Using a Preintegrated foot Velocity Factor

Kim¹,

Yu²,

Lee³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

We propose a novel state estimator for legged robots, STEP, achieved through a novel preintegrated foot velocity factor. In the preintegrated foot velocity factor, the usual non-slip assumption is not adopted. Instead, the end effector velocity becomes observable by exploiting the body speed obtained from a stereo camera. In other words, the preintegrated end effector's pose can be estimated. Another advantage of our approach is that it eliminates the necessity for a contact detection step, unlike the typical approaches. The proposed method has also been validated in harsh-environment simulations and real-world experiments containing uneven or slippery terrains.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: A Gazebo Simulationmentioning

confidence: 97%