Generic Multi-scale Segmentation and Curve Approximation Method

Authors:
Marielle Mokhtari;Robert Bergevin
Affiliations:
-;-
Venue:
Scale-Space '01 Proceedings of the Third International Conference on Scale-Space and Morphology in Computer Vision
Year:
2001

Citing 6
Cited 4

Three-dimensional object recognition from single two-dimensional images

Artificial Intelligence
Asynchronous perceptual grouping: from contours to relevant 2-D structures

Computer Vision and Image Understanding
Multiscale contour segmentation and approximation: an algorithm based on the geometry of regular inscribed polygons

Computer Vision and Image Understanding
Nonparametric Segmentation of Curves into Various Representations

IEEE Transactions on Pattern Analysis and Machine Intelligence
Finding Parametric Curves in an Image

ECCV '92 Proceedings of the Second European Conference on Computer Vision
Multiscale Compression of Planar Curves Using Constant Curvature Segments

ICPR '98 Proceedings of the 14th International Conference on Pattern Recognition-Volume 1 - Volume 1

Qualitative part-based models in content-based image retrieval

Machine Vision and Applications
Enhancing Boundary Primitives Using a Multiscale Quadtree Segmentation

ISVC '08 Proceedings of the 4th International Symposium on Advances in Visual Computing
Detection of unexpected multi-part objects from segmented contour maps

Pattern Recognition
Enhancing contour primitives by pairwise grouping and relaxation

ICIAR'07 Proceedings of the 4th international conference on Image Analysis and Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

We propose a new complete method to extract significant description(s) of planar curves according to constant curvature segments. This method is based (i) on a multi-scale segmentation and curve approximation algorithm, defined by two grouping processes (polygonal and constant curvature approximations), leading to a multi-scale covering of the curve, and (ii) on an intra- and inter-scale classification of this multi-scale covering guided by heuristically-defined qualitative labels leading to pairs (scale, list of constant curvature segments) that best describe the shape of the curve. Experiments show that the proposed method is able to provide salient segmentation and approximation results which respect shape description and recognition criteria.