Using vanishing points to improve visual-inertial odometry

Camposeco, Federico; Pollefeys, Marc

doi:10.1109/icra.2015.7139926

Cited by 42 publications

(23 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The vanishing points from parallel lines on the images relates to the camera orientation directly. It has shown that using vanishing points can improve visual SLAM [20] [21] and visual-inertial odometry [22]. However, in those methods the line features are used as only intermediate results for extracting vanishing points.…”

Section: Related Workmentioning

confidence: 99%

“…The process however becomes much simpler if an IMU is available, since the accelerometer renders the vertical direction observable because of gravity. We adopt a similar approach [22] to detect new Manhattan worlds in the image. We start Manhattan world detection whenever vertical lines have been identified as described in Section V-B.…”

Section: F Detecting and Merging Manhattan Worldsmentioning

confidence: 99%

See 1 more Smart Citation

StructVIO: Visual-Inertial Odometry With Structural Regularity of Man-Made Environments

et al. 2019

View full text Add to dashboard Cite

We propose a novel visual-inertial odometry approach that adopts structural regularity in man-made environments. Instead of using Manhattan world assumption, we use Atlanta world model to describe such regularity. An Atlanta world is a world that contains multiple local Manhattan worlds with different heading directions. Each local Manhattan world is detected on-the-fly, and their headings are gradually refined by the state estimator when new observations are coming. With fully exploration of structural lines that aligned with each local Manhattan worlds, our visual-inertial odometry method becomes more accurate and robust, as well as more flexible to different kinds of complex man-made environments. Through extensive benchmark tests and real-world tests, the results show that the proposed approach outperforms existing visual-inertial systems in large-scale man-made environments.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: F Detecting and Merging Manhattan Worldsmentioning

confidence: 99%

StructVIO: Visual-Inertial Odometry With Structural Regularity of Man-Made Environments

et al. 2019

View full text Add to dashboard Cite

show abstract

“…A group of parallel lines project to image plane will converge to a vanishing point (VP). For example, Camposeco et al [20] deal VP as a measurement within an EKF-based visual-inertial odometry (VIO) system to improve the localization accuracy. Reference [21] use VP as a high-level landmark in a multilayer feature graph to directly calculate line landmarks direction in 3D space.…”

Section: A Line Segment Based Slammentioning

confidence: 99%

“…Two VPs of different groups of parallel lines can be used to estimate camera rotation [44]. VP has also been used for relative pose estimation [45] and as a measurement within an EKF-based visual-inertial odometry (VIO) system [20]. In [21], VPs are used as high-level landmarks in a multilayer feature graph, and the direction of such landmarks are represented by corresponding VPs.…”

Section: Geometry For Line Segments and Vanishing Pointsmentioning

confidence: 99%

Line-Based Stereo SLAM by Junction Matching and Vanishing Point Alignment

Wang

et al. 2019

IEEE Access

View full text Add to dashboard Cite

In this paper, we propose a stereo simultaneous localization and mapping (SLAM) method based on line segments. For the front-end module of SLAM, we designed a novel method based on the coplanar junction detection, description, and matching. Then the junctions along with their multi-scale rotated BRIEF descriptors are used in other SLAM modules, including line tracking, mapping, and loop closure. The line extraction and matching thread runs at 20 ∼ 40Hz for stereo image sequences on a laptop, making it a practical front-end for line-based SLAM system. For the back-end module, a cost function is designed to minimize both of the reprojection error of line segments and alignment error of the vanishing points. The experimental results demonstrate that the proposed method exhibits more accurate localization and reconstruction than state-of-the-art line-based SLAM systems in line-rich environments. INDEX TERMS Simultaneous localization and mapping (SLAM), line detection, vanishing point, visual odometry, loop closure.

show abstract

“…An alternative method using only two dominant directions is presented by [ 16 ] based on the Rodrigues’ formula. The vanishing point-based attitude estimation method has been investigated to help indoor pedestrian navigation or improve VO performance in hallway environments [ 17 , 18 , 19 ] and UAV navigation in urban areas with structured buildings [ 20 , 21 ].…”

Section: Introductionmentioning

confidence: 99%

Motion Constraints and Vanishing Point Aided Land Vehicle Navigation

Liu

El‐Sheimy

et al. 2018

Micromachines

View full text Add to dashboard Cite

In the typical Inertial Navigation System (INS)/ Global Navigation Satellite System (GNSS) setup for ground vehicle navigation, measures should be taken to maintain the performance when there are GNSS signal outages. Usually, aiding sensors are utilized to reduce the INS drift. A full motion constraint model is developed allowing the online calibration of INS frame with respect to (w.r.t) the motion frame. To obtain better heading and lateral positioning performance, we propose to use of vanishing point (VP) observations of parallel lane markings from a single forward-looking camera to aid the INS. In the VP module, the relative attitude of the camera w.r.t the road frame is derived from the VP coordinates. The state-space model is developed with augmented vertical attitude error state. Finally, the VP module is added to a modified motion constrains module in the Extended Kalman filter (EKF) framework. Simulations and real-world experiments have shown the validity of VP-based method and improved heading and cross-track position accuracy compared with the solution without VP. The proposed method can work jointly with conventional visual odometry to aid INS for better accuracy and robustness.

show abstract

Using vanishing points to improve visual-inertial odometry

Cited by 42 publications

References 22 publications

StructVIO: Visual-Inertial Odometry With Structural Regularity of Man-Made Environments

StructVIO: Visual-Inertial Odometry With Structural Regularity of Man-Made Environments

Line-Based Stereo SLAM by Junction Matching and Vanishing Point Alignment

Motion Constraints and Vanishing Point Aided Land Vehicle Navigation

Contact Info

Product

Resources

About