Interactive design of smooth objects with probabilistic point constraints
ACM Transactions on Graphics (TOG) - Special issue on interactive sculpting
On 3D model construction by fusing heterogeneous sensor data
CVGIP: Image Understanding
The NURBS book
Surface fitting with hierarchical splines
ACM Transactions on Graphics (TOG)
Advanced surface fitting techniques
Computer Aided Geometric Design
Scattered Data Interpolation with Multilevel B-Splines
IEEE Transactions on Visualization and Computer Graphics
An Introduction to the Kalman Filter
An Introduction to the Kalman Filter
Recursive Estimation of 3D Motion and Surface Structure from Local Affine Flow Parameters
IEEE Transactions on Pattern Analysis and Machine Intelligence
Convection-Driven Dynamic Surface Reconstruction
SMI '05 Proceedings of the International Conference on Shape Modeling and Applications 2005
A recursive approach to the design of adjustable linear models for complex motion analysis
SPPRA '07 Proceedings of the Fourth IASTED International Conference on Signal Processing, Pattern Recognition, and Applications
Multi-sensor calibration through iterative registration and fusion
Computer-Aided Design
| Hi-index | 0.00 |
In this paper, we present a new B-spline surface reconstruction approach, called dynamic surface reconstruction, aiming to close the sensing-and-modeling loop in 3D digitization. At its core, this approach uses a recursive least squares method, the Kalman filter, to dynamically reconstruct the B-spline surface as the surface data are acquired. That is, the acquired data are dynamically incorporated into the surface model and the updated surface model is then used to dynamically guide further data acquisition. It thus enables a closed-loop shape sensing-and-modeling methodology for 3D digitization. Our technical contribution lies on the exploitation of the recursive nature of the Kalman filter for B-spline surface reconstruction. This enables dynamic parameterization of data points, dynamic determination of next optimal sensing locations, and low-discrepancy based efficient sensing and reconstruction. Experiments demonstrate that such dynamic surface reconstruction leads to more efficient data acquisition and better surface reconstruction.