Viewpoint Selection – Planning Optimal Sequences of Views for Object Recognition
In the past decades most object recognition systems were based on passive approaches. But in the last few years a lot of research was done in the field of active object recognition. In this context there are several unique problems to be solved, like the fusion of several views and the selection of the best next viewpoint.
In this paper we present an approach to solve the problem of choosing optimal views (viewpoint selection) and the fusion of these for an optimal 3D object recognition (viewpoint fusion). We formally define the selection of additional views as an optimization problem and we show how to use reinforcement learning for viewpoint training and selection in continuous state spaces without user interaction. We also present an approach for the fusion of multiple views based on recursive density propagation.
The experimental results show that our viewpoint selection is able to select a minimal number of views and perform an optimal object recognition with respect to the classification.
KeywordsBayesian Network Object Recognition Recognition Rate Camera Movement Multiple View
Unable to display preview. Download preview PDF.
- 4.Deinzer, F., Denzler, J., Niemann, H.: Improving Object Recognition By Fusion Of Multiple Views. In: 3rd Indian Conference on Computer Vision Graphics and Image Processing, Ahmedabad, Indien, pp. 161–166. Allied Publishers Pvt. Ltd (2002)Google Scholar
- 5.Denzler, J., Brown, C.M.: Information theoretic sensor data selection for active object recognition and state estimation. PAMI 24(2) (2002)Google Scholar
- 6.Grässl, C., Deinzer, F., Niemann, H.: Continuous Parametrization of Normal Distributions for Improving the Discrete Statistical Eigenspace Approach for Object Recognition. In: PRIP 2003 (May 2003) (submitted)Google Scholar
- 7.Isard, M., Andrew, B.: CONDENSATION – Conditional Density Propagation for Visual Tracking. In: IJCV 1998, vol. 29(1), pp. 5–28 (1998)Google Scholar
- 8.Kalman, R.E.: A New Approach to Linear Filtering and Prediction Problems. Journal of Basic Engineering, 35–44 (1960)Google Scholar
- 10.Madsen, C.B., Christensen, H.I.: A Viewpoint Planning Strategy for Determining True Angles on Polyhedral Objects by Camera Alignment. PAMI 19(2) (1997)Google Scholar
- 11.Lehel, P., Hemayed, E.E., Farag, A.A.: Sensor Planning for a Trinocular Active Vision System. In: CVPR, vol. II, pp. 306–312 (1999)Google Scholar
- 12.Roy, S.D., Chaudhury, S., Banerjee, S.: Recognizing Large 3-D Objects through Next View Planning using an Uncalibrated Camera. In: ICCV 2001, Vancouver, Canada, vol. II, pp. 276–281. IEEE Computer Press, Los Alamitos (2001)Google Scholar
- 13.Schiele, B., Crowley, J.L.: Transinformation for Active Object Recognition. In: ICCV 1998, Bombay, India, pp. 249–254 (1998)Google Scholar
- 14.Sutton, R.S., Barto, A.G.: Reinforcement Learning. A Bradford Book. Cambridge, London (1998)Google Scholar