An adaptive observer framework for accurate feature depth estimation using an uncalibrated monocular camera

Wang

2023

Sensors

Aiming at the time-varying uncertainties of robot and camera models in IBUVS (image-based uncalibrated visual servo) systems, a finite-time adaptive controller is proposed based on the depth-independent Jacobian matrix. Firstly, the adaptive law of depth parameters, kinematic parameters, and dynamic parameters is proposed for the uncertainty of a robot model and a camera model. Secondly, a finite-time adaptive controller is designed by using a nonlinear proportional differential plus a dynamic feedforward compensation structure. By applying a continuous non-smooth nonlinear function to the feedback error, the control quality of the closed-loop system is improved, and the desired trajectory of the image is tracked in finite time. Finally, using the Lyapunov stability theory and the finite-time stability theory, the global finite-time stability of the closed-loop system is proven. The experimental results show that the proposed controller can not only adapt to the changes in the EIH and ETH visual configurations but also adapt to the changes in the relative pose of feature points and the camera’s relative pose parameters. At the same time, the convergence rate near the equilibrium point is improved, and the controller has good dynamic stability.

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Design of A Finite-Time Adaptive Controller for Image-Based Uncalibrated Visual Servo Systems with Uncertainties in Robot and Camera Models

Wang

2023

Sensors

2018 IEEE Conference on Decision and Control (CDC)

“…e-mail: {sean.obrien, jochen.trumpf, viorela.ila, rob.mahony}@anu.edu.au More recent work includes the development of a dynamic filtering algorithm for the computation of dense optical flow in real-time [13]. In [14], an observer for sparse depth estimation using monocular cameras is proposed. In [15] an observer for tracking the depth of a given object using from perspective vision data (such as monocular camera data) is formulated that exponentially converges to the object's coordinates.…”

Section: Introductionmentioning

confidence: 99%

A Geometric Observer for Scene Reconstruction Using Plenoptic Cameras

O'Brien

Trumpf

Ila

et al. 2018

This paper proposes an observer for generating depth maps of a scene from a sequence of measurements acquired by a two-plane light-field (plenoptic) camera. The observer is based on a gradient-descent methodology. The use of motion allows for estimation of depth maps where the scene contains insufficient texture for static estimation methods to work. A rigourous analysis of stability of the observer error is provided, and the observer is tested in simulation, demonstrating convergence behaviour. arXiv:1809.07931v1 [math.OC]

2017 2nd International Conference on Mechatronics and Information Technology (ICMIT 2017)

“…In monocular SLAM problems, the depth estimation has always been a difficult problem and people's exploration of this problem has not been stopping [8,9,10]. The existing methods can be divided into three categories [11,12,13]: feature-based, gradient-based and optical-based.…”

Section: Introductionmentioning

confidence: 99%

Delayed Initialization of Monocular SLAM Based on Improved Least Squares

Huang¹,

Fu²,

Ming³

2017

Abstract:In order to solve the problem of lacking depth information of monocular SLAM, this paper proposes a method of delay depth initialization based on improved least squares. According to the principle of camera imaging, the robot performs two different observations on the same calibration point respectively, and the triangulation method is used to solve the initial depth information of the target pixel. Aiming at the problem of depth solution caused by the error of monocular wheeled robot, this paper improves the least squares method to estimate the optimal depth of target point, and the simulation experiment is carried out by using Matlab software. The experimental results show that the method proposed in this paper is simple and feasible for monocular SLAM. The method proposed in this paper is very accurate to initialize the depth information of monocular robot and the improved algorithm is superior to the least squares algorithm. Experiments show that the method proposed in this paper can solve the problem of lacking deepth information of monocular SLAM, and having the feasibility of applying to monocular SLAM.