Homography Based Egomotion Estimation with a Common Direction

@article{Saurer2017HomographyBE,
  title={Homography Based Egomotion Estimation with a Common Direction},
  author={Olivier Saurer and Pascal Vasseur and R{\'e}mi Boutteau and C{\'e}dric Demonceaux and Marc Pollefeys and Friedrich Fraundorfer},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year={2017},
  volume={39},
  pages={327-341}
}
In this paper, we explore the different minimal solutions for egomotion estimation of a camera based on homography knowing the gravity vector between calibrated images. These solutions depend on the prior knowledge about the reference plane used by the homography. We then demonstrate that the number of matched points can vary from two to three and that a direct closed-form solution or a Gröbner basis based solution can be derived according to this plane. Many experimental results on synthetic… 

Visual Odometry Using a Homography Formulation with Decoupled Rotation and Translation Estimation Using Minimal Solutions

TLDR
This paper demonstrates that rotation and translation estimation can be decoupled and derives different algorithms based on this decoupling that allow an efficient estimation of two-view relative motion estimation.

Egomotion Estimation Under Planar Motion with an RGB-D Camera

TLDR
The experiments show that the proposed approach is efficient and robust enough for egomotion estimation in the Manhattan-like environments compared with the state-of-the-art methods.

Homography-based camera pose estimation with known gravity direction for UAV navigation

TLDR
This paper statistically optimizes the solution for the homography-based relative pose estimation problem by assuming a known gravity direction and a dominant ground plane, and developing an iterative estimation method of the camera trajectory for visual odometry.

Homography-Based Egomotion Estimation Using Gravity and SIFT Features

TLDR
New minimal solvers that require fewer correspondences than traditional approaches are derived and are compared with the state-of-the-art point-based solvers on both synthetic data and real images, showing comparable accuracy and significant improvement in terms of speed.

Calibrated and Partially Calibrated Semi-Generalized Homographies

TLDR
The first minimal solutions for estimating the semi-generalized homography given a perspective and a generalized camera are proposed, using five 2D-2D image point correspondences induced by a scene plane.

Feature‐based recursive observer design for homography estimation and its application to image stabilization

TLDR
Theoretical analysis and experimental results are provided to demonstrate the robustness of the proposed algorithm and show excellent performance even in the case of very fast camera motion, severe occlusion, and in the presence of specular reflections.

Rotational Alignment of IMU-camera Systems with 1-Point RANSAC

TLDR
A minimal solution for the rotational alignment of IMU-camera systems based on a homography formulation that allows us to cope with feature mismatches efficiently and robustly within a random sample consensus (RANSAC) scheme.

Rolling Shutter Homography and its Applications

TLDR
This article shows that a direct point-to-point relationship on a RS pair can be expressed as a set of 3 to 8 atomic 3x3 matrices depending on the kinematic model used for the instantaneous-motion during image acquisition.

Visual Tilt Correction for Vehicle-mounted Cameras

TLDR
This paper proposes two methods to correct camera images if this parallelity constraint does not hold, one utilizing that the ground is visible in the camera images, therefore its normal gives the gravity vector; the second uses vanishing points to estimate the relative correction homography.

Minimal solutions for the rotational alignment of IMU-camera systems using homography constraints

References

SHOWING 1-10 OF 40 REFERENCES

A Homography Formulation to the 3pt Plus a Common Direction Relative Pose Problem

TLDR
This paper uses the homography constraint to derive a novel formulation for the 3pt problem that allows the computation of the normal vector of the plane defined by the three input points without any additional computation in addition to the standard motion parameters of the camera.

Two Efficient Solutions for Visual Odometry Using Directional Correspondence

TLDR
A simple, closed-form solution and a solution based on algebraic geometry which offers numerical advantages to the two-view, relative pose problem from three image point correspondences and one common reference direction are presented.

1-Point-based monocular motion estimation for computationally-limited micro aerial vehicles

We propose a novel method to estimate the relative motion between two consecutive camera views, which only requires the observation of a single feature in the scene and the knowledge of the angular

A New Solution to the Relative Orientation Problem Using Only 3 Points and the Vertical Direction

TLDR
The elements necessary to build a specific algebraic solver are given in this paper, allowing for a real-time implementation and the results on real and synthetic data show the efficiency of this method.

Rotation estimation and vanishing point extraction by omnidirectional vision in urban environment

TLDR
Experimental results on various data sets for diverse robotic applications have demonstrated that the novel framework is accurate, robust, maintains the orthogonality of the vanishing points and can run in real-time.

2-Point-based outlier rejection for camera-IMU systems with applications to micro aerial vehicles

TLDR
It is shown that in the case of a monocular camera mounted on a quadrotor vehicle, motion priors from IMU can be used to discard wrong estimations in the framework of a 2-point-RANSAC based approach.

A 3 points vision based approach for MAV localization in GPS denied environments

TLDR
A new method to localize a micro aerial vehicle (MAV) in GPS denied environments and without the usage of any known pattern is introduced, based on a closed solution which provides the vehicle pose from a single camera image, once the roll and the pitch angles are obtained by the inertial measurements.

A Minimal Case Solution to the Calibrated Relative Pose Problem for the Case of Two Known Orientation Angles

TLDR
A novel minimal case solution to the calibrated relative pose problem using 3 point correspondences for the case of two known orientation angles is presented and it is shown that the new 3-point algorithm can cope with planes and even collinear points.

1-Point-RANSAC Structure from Motion for Vehicle-Mounted Cameras by Exploiting Non-holonomic Constraints

  • D. Scaramuzza
  • Computer Science
    International Journal of Computer Vision
  • 2011
TLDR
By exploiting the nonholonomic constraints of wheeled vehicles it is possible to use a restrictive motion model which allows us to parameterize the motion with only 1 point correspondence and results in the two most efficient algorithms for removing outliers: 1-point RANSAC and histogram voting.

Indoor robot motion based on monocular images

TLDR
The estimation of the 2D relative motion of an indoor robot using monocular vision is presented and accurate motion, both in rotation and translation angles of 0.4 and 1.7 deg, is successfully derived.