Non-rigid object tracking with elastic structure of local patches and hierarchical sampling

2010 ◽  
Author(s):  
Kwang Moo Yi ◽  
Soo Wan Kim ◽  
Hawook Jeong ◽  
Songhwai Oh ◽  
Jin Young Choi
Keyword(s):  
Object Tracking ◽  
Elastic Structure ◽  
Rigid Object ◽  
Hierarchical Sampling

Multi-part Non-rigid Object Tracking Based on Time Model-Space Gradients

2000 ◽  
pp. 72-82 ◽  
Author(s):  
T. Nunomaki ◽  
S. Yonemoto ◽  
D. Arita ◽  
R. Taniguchi ◽  
N. Tsuruta
Keyword(s):  
Object Tracking ◽  
Model Space ◽  
Time Model ◽  
Rigid Object

RIGID OBJECT TRACKING USING MEAN-SHIFT WITH ADAPTIVE KERNEL BANDWIDTH SELECTION

2006 ◽  
Vol 03 (02) ◽  
pp. 101-107
Author(s):  
JIEXIN PU ◽  
NINGSONG PENG ◽  
XINHAN HUANG
Keyword(s):  
Object Tracking ◽  
Mean Shift ◽  
Bandwidth Selection ◽  
Rigid Object ◽  
Mean Shift Algorithm ◽  
Scale Parameters ◽  
Affine Model ◽  
Backward Tracking ◽  
Adaptive Kernel ◽  
Estimate Method

The mean shift algorithm is an efficient way for object tracking. However, there is presently no clear mechanism for selecting kernel bandwidth while the object is changing in size. This paper presents a novel bandwidth selection method for mean shift based rigid object tracking. The kernel bandwidth is updated by discovering the scale parameters of the object's affine model that are estimated by using the correspondences between the corner object in two consecutive frames. The centroid of the object is registered by a special backward tracking method. Therefore, we can not only get translation parameter to simplify affine model but also improve the accuracy of finding corner correspondences. In addition, the M-estimate method is employed to reject mismatched pairs (outliers) so as to get better regression results. We have applied the proposed method to track vehicles changing in size with encouraging results.

Comparison of Natural Feature Descriptors for Rigid-Object Tracking for Real-Time Augmented Reality

2014 ◽  
Author(s):  
Francely Franco Bermudez ◽  
Christian Santana Diaz ◽  
Sheneeka Ward ◽  
Rafael Radkowski ◽  
Timothy Garrett ◽  
...  
Keyword(s):  
Augmented Reality ◽  
Object Tracking ◽  
Real Time ◽  
Virtual Object ◽  
Interest Points ◽  
Rigid Object ◽  
Binary Descriptors ◽  
Rigid Objects ◽  
Natural Feature ◽  
Feature Descriptors

This paper presents a comparison of natural feature descriptors for rigid object tracking for augmented reality (AR) applications. AR relies on object tracking in order to identify a physical object and to superimpose virtual object on an object. Natural feature tracking (NFT) is one approach for computer vision-based object tracking. NFT utilizes interest points of a physcial object, represents them as descriptors, and matches the descriptors against reference descriptors in order to identify a phsical object to track. In this research, we investigate four different natural feature descriptors (SIFT, SURF, FREAK, ORB) and their capability to track rigid objects. Rigid objects need robust descriptors since they need to describe the objects in a 3D space. AR applications are also real-time application, thus, fast feature matching is mandatory. FREAK and ORB are binary descriptors, which promise a higher performance in comparison to SIFT and SURF. We deployed a test in which we match feature descriptors to artificial rigid objects. The results indicate that the SIFT descriptor is the most promising solution in our addressed domain, AR-based assembly training.

scholarly journals Rigid Object Tracking Algorithms for Low-Cost AR Devices

2014 ◽  
Author(s):  
Timothy Garrett ◽  
Saverio Debernardis ◽  
Rafael Radkowski ◽  
Carl K. Chang ◽  
Michele Fiorentino ◽  
...  
Keyword(s):  
Object Tracking ◽  
Low Cost ◽  
Internal Representation ◽  
Object Identification ◽  
Sparse Representations ◽  
Microsoft Kinect ◽  
Future Research ◽  
Rigid Object ◽  
Tracking Method ◽  
Mesh Reconstruction

Augmented reality (AR) applications rely on robust and efficient methods for tracking. Tracking methods use a computer-internal representation of the object to track, which can be either sparse or dense representations. Sparse representations use only a limited set of feature points to represent an object to track, whereas dense representations almost mimic the shape of an object. While algorithms performed on sparse representations are faster, dense representations can distinguish multiple objects. The research presented in this paper investigates the feasibility of a dense tracking method for rigid object tracking, which incorporates the both object identification and object tracking steps. We adopted a tracking method that has been developed for the Microsoft Kinect to support single object tracking. The paper describes this method and presents the results. We also compared two different methods for mesh reconstruction in this algorithm. Since meshes are more informative when identifying a rigid object, this comparison indicates which algorithm shows the best performance for this task and guides our future research efforts.

Sign in / Sign up

Export Citation Format

Share Document