Wyniki wyszukiwania - BazTech

1

Vision-based positioning of electric buses for assisted docking to charging stations

Nowak Tomasz, Nowicki Michał R., Skrzypczyński Piotr

International Journal of Applied Mathematics and Computer Science

|

2022

|

Vol. 32, no. 4

583--599

EN

We present a novel approach to vision-based localization of electric city buses for assisted docking to a charging station. The method assumes that the charging station is a known object, and employs a monocular camera system for positioning upon carefully selected point features detected on the charging station. While the pose is estimated using a geometric method and taking advantage of the known structure of the feature points, the detection of keypoints themselves and the initial recognition of the charging station are accomplished using neural network models. We propose two novel neural network architectures for the estimation of keypoints. Extensive experiments presented in the paper made it possible to select the MRHKN architecture as the one that outperforms state-of-the-art keypoint detectors in the task considered, and offers the best performance with respect to the estimated translation and rotation of the bus with a low-cost hardware setup and minimal passive markers on the charging station.

2

Vision based persistent localization of a humanoid robot for locomotion tasks

Martínez P. A., Castelán M., Arechavaleta G.

International Journal of Applied Mathematics and Computer Science

|

2016

|

Vol. 26, no. 3

669--682

EN

Typical monocular localization schemes involve a search for matches between reprojected 3D world points and 2D image features in order to estimate the absolute scale transformation between the camera and the world. Successfully calculating such transformation implies the existence of a good number of 3D points uniformly distributed as reprojected pixels around the image plane. This paper presents a method to control the march of a humanoid robot towards directions that are favorable for visual based localization. To this end, orthogonal diagonalization is performed on the covariance matrices of both sets of 3D world points and their 2D image reprojections. Experiments with the NAO humanoid platform show that our method provides persistence of localization, as the robot tends to walk towards directions that are desirable for successful localization. Additional tests demonstrate how the proposed approach can be incorporated into a control scheme that considers reaching a target position.

3

Monocular Vision-based Sensor for Autonomous Mobile Robot Localization by Circular Markers

Lu J., Shi J., Wu S.

Przegląd Elektrotechniczny

|

2013

|

R. 89, nr 1b

131--133

EN

To navigate reliably in indoor environments, a mobile robot must know where it is. This paper is concerned with the design of a monocular vision-based algorithm for on-line estimation of a mobile robot’s location using circular markers. The algorithm is based on 3-D analytic geometry, which is capable of estimating both the orientation and the position of the camera by a single camera image. The method can be used for camerarobot calibration for eye-on-hand systems, and autonomous mobile robot guidance. Laboratory experiments using a moving cylindrical object demonstrate both the accuracy and stability of the method.

PL

Robot mobilny, aby wiarygodnie nawigować we wnętrzu, musi znać swoje położenie. Opracowanie, w celu oszacowania bieżącego położenia robota, koncentruje się na projekcie algorytmu opartego o obraz jedno-okularowy, z zastosowaniem markerów kołowych. Algorytm przeprowadza analizę w geometrii 3D, co umożliwia oszacowanie zarówno orientacji jak i położenia kamery na podstawie danych obrazu z jednej kamery. Metoda może być zastosowana do kalibracji robota-kamery w systemie oko-ręka i do sterowania autonomicznym robotem mobilnym. Zademonstrowano eksperymenty laboratoryjne z wykorzystaniem obiektu cylindrycznego, analizując zarówno dokładność jak i stabilność metody.

4

Straight-lines modelling using planar information for monocular SLAM

Santana A. M., Medeiros A. A. D.

International Journal of Applied Mathematics and Computer Science

|

2012

|

Vol. 22, no. 2

409-421

EN

This work proposes a SLAM (Simultaneous Localization And Mapping) solution based on an Extended Kalman Filter (EKF) in order to enable a robot to navigate along the environment using information from odometry and pre-existing lines on the floor. These lines are recognized by a Hough transform and are mapped into world measurements using a homography matrix. The prediction phase of the EKF is developed using an odometry model of the robot, and the updating makes use of the line parameters in Kalman equations without any intermediate stage for calculating the distance or the position. We show two experiments (indoor and outdoor) dealing with a real robot in order to validate the project.

5

A monocular approach to 3-D reconstruction based on bilateral symmetry

Zhang J. Z., Tsui H. T., Wu Q. M. J.

Machine Graphics and Vision

|

2000

|

Vol. 9, No. 4

763-773

EN

An object with a plane of symmetry is called bilaterally symmetrical. When an arbitrary object and its image in a plane mirror are bith visible in a monocular view, the object and its image in the mirror are also bilaterally symmetrical with respect to the mirror. In this paper, we present a method for 3-D reconstruction from a monocular image obtained by a calibrated camera based on the bilateral symmetry theory. We first verify the algorithm using a calibration block that is bilaterally symmetrical. Then we test the algorithm using more natural object and a plane mirror whose position and orientation are known. Very good reconstruction results have been obtained in the experiments.