| K. Kutulakos and J. Vallino. Calibration-Free Augmented Reality. IEEE Trans. on VCG, 4(1):1--20, Jan. 1998. |
....and studios. More recently Computer Vision techniques have been developed to enable 3 D camera motion and position relative to the scene to be obtained from arbitrary sequences, in some cases without the use of calibrated cameras. A number of these have been utilised for AR, see for example [8, 9, 13]. The most well known techniques are those based on epipolar geometry and the trilinear tensor [5] and those based on recursive estimation such as the Kalman filter [1, 7] Recently, impressive results have been achieved using the former technique, notably in the work described by Pollefeys et al. ....
K. N. Kutulakos and J. R. Vallino. Calibration-free augmented reality. IEEE Trans. Visualization and Computer Graphics, 4(1):1--20, 1998.
....original footage, giving the impression that the virtual objects are rigidly attached to the structure in the scene. Furthermore, recent advances in structure and motion recovery have enabled the 3D structure and cameras for each frame to be recovered automatically directly from the image sequence [1, 2, 5, 14, 15, 21]. However, although the camera motion can be recovered to sub pixel accuracy, the registration accuracy of the augmentation must equal or surpass this. The errors resulting from inaccurate registration fall into two main categories: a high frequency (frame rate) BMVC99 295 British Machine ....
....identified over two widely separated views, and a reference point on the intersection of two of the lines, we solve for the Euclidean transformation that maps the reference point to the origin and aligns the lines with the coordinate axes. A second method is to define an affine coordinate system [13, 14], and represent the virtual object in that frame. In [14] the affine coordinates are obtained by tracking markers placed in the scene, and a parallel projection approximation is used. In this work, there is no such approximation: we use a perspective camera and track the vanishing points of three ....
[Article contains additional citation context not shown here]
K. N. Kutulakos and J. R. Vallino. Calibration-free augmented reality. IEEE Trans. on Visualization and Computer Graphics, 4(1):1--20, 1998.
....they also explored added connected 3D structures (a rectangular solid) by tracking the vanishing points of three sets of parallel lines to de ne an ane systems and then estimate the camera matrices, subject again to constraints such as parallelness and coplanarity. Kutalakos and Vallino [8] demonstrated direct, but ane, mapping of 3D objects using four accurately tracked non coplanar control points to determine the mapping of the remainder of the 3D object. The research presented here extends this previous work by using perspective projection, avoiding dependence on accurate ....
K. N. Kutalakos, J. R. Vallino. Calibration-free augmented reality. IEEE. Trans. Visualization and Comp. Graphics, 4(1), pp 1-20, 1998.
....1995 Augmented Reality using Uncalibrated Video Sequences 163 Discussion 1. Kostas Daniilidis, University of Pennsylvania: When you have a simple task, like in your case inserting a cube, it is not necessary to compute a Euclidean reconstruction. There is other work, see Kutulakos and Vallino [1], which describes systems that just assume scaled orthographic projections. Kurt Cornelis: I think that s true, but we are also aiming at building a Euclidean reconstruction with which we can finally interact in a way that we are used to in the real world. We might want to compute a trajectory of ....
K. Kutulakos and J. Vallino. Calibration-free augmented reality. IEEE Transactions on Computer Graphics, 4(1):1--20, 1998.
....selfcalibrated) approaches. Self calibration approaches [17] provide a metric reconstruction from multiple views with an accuracy which is suitable only for restricted augmented reality applications like video manipulation where the quality of depth is not relevant. Weakly calibrated approaches [11] provide real time performance and are suitable for augmenting scenes only with synthetic objects. Recently, gaining insight from the afore mentioned projective geometric work, the paradigm of view generation became popular. Instead of calibrating the cameras with respect to a world coordinate ....
K. Kutulakos and J. Vallino. Calibration-free augmented reality. IEEE Trans. on Visualization and Computer Graphics, 4(1):1--20, 1998.
....Menace,etc.Forthis purpose, the camera pose must first be estimated. Typical approaches rely on 3 D position tracking devices, precise calibration, fiducials, or painstakingly manual work. Recently, affine invariant has been used to provide calibration free augmented reality by Kutulakos et al. [50]. However, the affine representation makes it difficult to describe the synthetic objects in a traditional way (i.e. using positions, distances or angles, etc) Another aspect of the same problem, which has been largely ignored, is the mutual occlusion (or other geometrical interactions) between ....
K. N. Kutulakos and S. M. Seitz, "Calibration free Augmented Reality," IEEE Trans. Visualization and Computer Graphics, vol. 4, no. 1, pp. 1-20, 1998.
....original footage, giving the impression that the virtual objects are rigidly attached to the structure in the scene. Furthermore, recent advances in structure and motion recovery have enabled the 3D structure and cameras for each frame to be recovered automatically directly from the image sequence [1, 2, 5, 14, 15, 21]. However, although the camera motion can be recovered to sub pixel accuracy, the registration accuracy of the augmentation must equal or surpass this. The errors resulting from inaccurate registration fall into two main categories: a high frequency (frame rate) British Machine Vision Conference ....
....identified over two widely separated views, and a reference point on the intersection of two of the lines, we solve for the Euclidean transformation that maps the reference point to the origin and aligns the lines with the coordinate axes. A second method is to define an affine coordinate system [13, 14], and represent the virtual object in that frame. In [14] the affine coordinates are obtained by tracking markers placed in the scene, and a parallel projection approximation is used. In this work, there is no such approximation: we use a perspective camera and track the vanishing points of three ....
[Article contains additional citation context not shown here]
K. N. Kutulakos and J. R. Vallino. Calibration-free augmented reality. IEEE Trans. on Visualization and Computer Graphics, 4(1):1--20, 1998.
....Self calibration approaches [Maybank and Faugeras, 1992] provide a metric reconstruction from multiple views with an accuracy which is suitable only for restricted augmented reality applications like video manipulation where the quality of depth is not relevant. Weakly calibrated approaches [Kutulakos and Vallino, 1998] provide real time performance and are suitable for augmenting scenes only with synthetic objects. Our approach is the first that provides an optimal balance between depth accuracy and speed and therefore can be applied in tele immersion. 3. SYSTEM DESCRIPTION AND SCENARIO The tele immersion ....
Kutulakos, K. and Vallino, J. (1998). Calibration-free augmented reality. IEEE Trans. on Visualization and Computer Graphics, 4(1):1--20.
.... scene of the image into which virtual objects will be integrated, the shading and occlusion effects can be easily handled using an existing graph software (e.g. OpenGL) Recently, another class of AR approaches which use affine or projective spaces for the integration of virtual objects (e.g. [9][17] were proposed. They are referred to as calibration free approaches. In our opinion, advantages of these approaches include the allowance of less priori information and easy for implementation. However, they also suffer from some drawbacks. For example, these approaches require users to ....
K. N. Kutulakos, J. R. Vallino, "Calibration-Free Augmented Reality," IEEE Trans. Visualization and Computer Graphics, Vol. 4, pp. 3-20, 1998.
....self calibrated) approaches. Selfcalibration approaches [15] provide a metric reconstruction from multiple views with an accuracy which is suitable only for restricted augmented reality applications like video manipulation where the quality of depth is not relevant. Weakly calibrated approaches [11] provide a real time performance and is suitable for augmenting scenes only with synthetic objects. Our approach is the first that provides an optimal balance between depth accuracy and speed and therefore can be applied in teleimmersion. Immersive Displays A spatially immersive display (SID) is a ....
K. Kutulakos and J. Vallino. Calibration-free augmented reality. IEEE Trans. on Visualization and Computer Graphics, 4(1):1--20, 1998.
No context found.
K. Kutulakos and J. Vallino. Calibration-Free Augmented Reality. IEEE Trans. on VCG, 4(1):1--20, Jan. 1998.
No context found.
K. N. Kutalakos, J. R. Vallino. Calibration-free augmented reality. IEEE. Trans. Visualization and Comp. Graphics, 4(1), pp 1-20, 1998.
No context found.
K. N. Kutalakos, J. R. Vallino. Calibration-free augmented reality. IEEE. Trans. Visualization and Comp. Graphics, 4(1), pp 1-20, 1998.
No context found.
K. Kutulakos, J. Vallino, Calibration-free augmented reality, IEEE Trans. Vis. Comput. Graph. 4 (1) (1998) 1--20.
No context found.
K. Kutulakos and J. Vallino. Calibration-Free Augmented Reality. IEEE Trans. on VCG, 4(1):1--20, Jan. 1998.
No context found.
K. Kutulakos, J. Vallino. Calibration-free augmented reality. IEEE TVCG, 4(1):1--20, 1998.
No context found.
K. N. Kutulakos and J. R. Vallino, "Calibration-free augmented reality", IEEE Trans. Visualization Comput. Graph., vol. 4, pp. 1--20, 1998. 455
No context found.
Kutulakos, K. and Vallino, J. (1998). Calibration-free augmented reality. IEEE Transactions on Visualization and Computer Graphics, 4(1):1--20.
No context found.
Kutulakos, K. and Vallino, J. (1998). Calibration-free augmented reality. IEEE Transactions on Visualization and Computer Graphics, 4(1):1--20.
No context found.
K. Kutulakos and J. Vallino. Calibration-Free Augmented Reality. IEEE Trans. on VCG, 4(1):1--20, Jan. 1998.
No context found.
K. Kutulakos and J. Vallino. Calibration-free augmented reality. IEEE Trans. Visualization and Computer Graphics, 4(1):1--20, 1998.
No context found.
#K.N. Kukulakos, J. Vallino, Calibration-Free Augmented Reality, IEEE Trans. Visualization and Computer Graphics. vol. 4, no. 1, Jan.-Mar. 1998, pp. 1-20.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC