An efficient 3d registration method using markerless image in AR-Based CNC machining simulation

Authors:
Z. Y. Ma;Y. P. Chen;C. M. Yuan;Z. D. Zhou
Affiliations:
School of Mechanical Engineering, Huazhong University of Science and Technology, China;School of Mechanical Engineering, Huazhong University of Science and Technology, China;School of Mechanical Engineering, Huazhong University of Science and Technology, China;School of Mechanical Engineering, Huazhong University of Science and Technology, China
Venue:
ICAT'06 Proceedings of the 16th international conference on Advances in Artificial Reality and Tele-Existence
Year:
2006

Citing 8
Cited 0

Solution of the simultaneous pose and correspondence problem using Gaussian error model

Computer Vision and Image Understanding
Recent Advances in Augmented Reality

IEEE Computer Graphics and Applications
An Affine Invariant Interest Point Detector

ECCV '02 Proceedings of the 7th European Conference on Computer Vision-Part I
Object Recognition from Local Scale-Invariant Features

ICCV '99 Proceedings of the International Conference on Computer Vision-Volume 2 - Volume 2
Marker-less Tracking for AR: A Learning-Based Approach

ISMAR '02 Proceedings of the 1st International Symposium on Mixed and Augmented Reality
Recognition of 3D Textured Objects by Mixing View-Based and Model-Based Representations

ICPR '00 Proceedings of the International Conference on Pattern Recognition - Volume 1
A real-time tracker for markerless augmented reality

ISMAR '03 Proceedings of the 2nd IEEE/ACM International Symposium on Mixed and Augmented Reality
Fully Automated and Stable Registration for Augmented Reality Applications

ISMAR '03 Proceedings of the 2nd IEEE/ACM International Symposium on Mixed and Augmented Reality

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we present a model-based approach for the real-time recognition and registration for augmented reality applications. It does not need any artificial markers to track the target. What the system need is the 3D model of target and some representative learning images. The most contribution of our work is that we put forward an idea of transferring the construction of correspondent point pairs between model and real scenes to the calculation of the affine matrix between a pair of 2D images. Our method is based on a two-stage process. In the first stage, a set of features is learned from the training images. The second stage matches the learned features with that obtained from the real scenes. If the target is recognized, the final correspondences used for registration are built with the help of the calculated affine matrix. The system is robust to large viewpoint changes and partial occlusions. And in the premise of stability assurance, the system has a good performance in reducing the computation burden.