Real-Time Scale Selection in Hybrid Multi-scale Representations
Local scale information extracted from visual data in a bottom-up manner constitutes an important cue for a large number of visual tasks. This article presents a framework for how the computation of such scale descriptors can be performed in real time on a standard computer.
The proposed scale selection framework is expressed within a novel type of multi-scale representation, referred to as hybrid multi-scale representation, which aims at integrating and providing variable trade-offs between the relative advantages of pyramids and scale-space representation, in terms of computational efficiency and computational accuracy. Starting from binomial scale-space kernels of different widths, we describe a family pyramid representations, in which the regular pyramid concept and the regular scale-space representation constitute limiting cases. In particular, the steepness of the pyramid as well as the sampling density in the scale direction can be varied.
It is shown how the definition of γ-normalized derivative operators underlying the automatic scale selection mechanism can be transferred from a regular scale-space to a hybrid pyramid, and two alternative definitions are studied in detail, referred to as variance normalization and lp-normalization. The computational accuracy of these two schemes is evaluated, and it is shown how the choice of sub-sampling rate provides a trade-off between the computational efficiency and the accuracy of the scale descriptors. Experimental evaluations are presented for both synthetic and real data. In a simplified form, this scale selection mechanism has been running for two years, in a real-time computer vision system.
Unable to display preview. Download preview PDF.
- Bretzner, L., Laptev, I. & Lindeberg, T. (2002), Hand-gesture recognition using multiscale colour features, hierarchical features and particle filtering, Face and Gesture’02, 63–74.Google Scholar
- Chomat, O., de Verdiere, V., Hall, D. & Crowley, J. (2000), Local scale selection for Gaussian based description techniques, ECCV’00, Springer LNCS 1842, 117–133.Google Scholar
- Comaniciu, D., Ramesh, V. & Meer, P. (2001), The variable bandwidth mean shift and data-driven scale selection, ICCV’01, 438–445.Google Scholar
- Crowley, J. L. & Parker, A. C. (1984), ‘A representation for shape based on peaks and ridges in the Difference of Low-Pass Transform’, IEEE-PAMI 6(2), 156–170.Google Scholar
- Elder, J. H. & Zucker, S. W. (1996), Local scale control for edge detection and blur estimation, in ‘ECCV’96’, 57–69..Google Scholar
- Florack, L. M. J. (1997), Image Structure, Kluwer, Netherlands.Google Scholar
- Frangi, A. F., Niessen, W. J., Hoogeveen, R. M., van Walsum, T. & Viergever, M. A. (1999), Quantitation of vessel morphology from 3D MRI, ‘MICCAI, 358–367.Google Scholar
- Hadjidemetriou, E., Grossberg, M. D. & Nayar, S. K. (2002), Resolution selection using generalized entropies of multiresolution histograms, ECCV’02, Springer LNCS 2350, 220–235.Google Scholar
- Hall, D., de Verdiere, V. & Crowley, J. (2000), Object recognition using coloured receptive fields, ECCV’00, Springer LNCS 1842, 164–177.Google Scholar
- Jägersand, M. (1995), Saliency maps and attention selection in scale and spatial coordinates: An information theoretic approach, ICCV’95, 195–202.Google Scholar
- Jähne, B. (1995), Digital Image Processing, Springer-Verlag.Google Scholar
- Koller, T. M., Gerig, G., Szèkely, G. & Dettwiler, D. (1995), Multiscale detection of curvilinear structures in 2-D and 3-D image data, ICCV’95, 864–869.Google Scholar
- Laptev, I. & Lindeberg, T. (2001), Tracking of multi-state hand models using particle filtering and a hierarchy of multi-scale image features, Scale-Space’01, Springer LNCS 2106, 63–74.Google Scholar
- Lindeberg, T. (1993b), On scale selection for differential operators, SCIA’93, 857–866.Google Scholar
- Lindeberg, T. (1994), Scale-Space Theory in Computer Vision, Kluwer, Netherlands.Google Scholar
- Lindeberg, T. (1998b), ‘Feature detection with automatic scale selection’, IJCV 30(2), 77–116.Google Scholar
- Lindeberg, T. & Bretzner, L (2003), Real-time scale selection in hybrid multi-scale representations, Technical report, KTH, Stockholm, Sweden.Google Scholar
- Lowe, D. (1999), Object recognition from local scale-invariant features, ICCV’99, 1150–1157.Google Scholar
- Majer, P. (2001), The influence of the γ-parameter on feature detection with automatic scale selection, Scale-Space’01, Springer LNCS 2106, 245–254.Google Scholar
- Mikolajczyk, K. & Schmid, C. (2002), An affine invariant interest point detector, ECCV’02, Springer LNCS 2350, 128–142.Google Scholar
- Nielsen, M. & Lillholm, M. (2001), What do features tell about images, Scale-Space’01, Springer LNCS 2106, 39–50.Google Scholar
- Niessen, W. & Maas, R. (1996), Optic flow and stereo, in J. Sporring et al (eds) Gaussian Scale-Space Theory, Kluwer.Google Scholar
- Pedersen, K. S. & Nielsen, M. (2001), Computing optic flow by scale-space integration of normal flow, Scale-Space’01, Springer LNCS 2106, 14–25.Google Scholar
- Simoncelli, E. P. & Freeman, W. T. (1995), The steerable pyramid: A flexible architecture for multi-scale derivative computation, ICIP’95, 444–447.Google Scholar
- Staal, J., Kalitzin, S., ter Haar Romeny, B. & Viergever, M. (1999), Detection of critical structures in scale-space, Scale-Space’99, Springer LNCS 1682, 105–116.Google Scholar
- Witkin, A. P. (1983), Scale-space filtering, 8th IJCAI, pp. 1019–1022.Google Scholar
- Yacoob, Y. & Davis, L. S. (1997), Estimating image motion using temporal multi-scale models of flow and acceleration. In: Motion-Based Recognition, Kluwer.Google Scholar