Boundary Shape Recognition Using Accumulated Length and Angle Information
In this paper we present a method to recognize shapes by analyzing a polygonal approximation of their boundaries. The method is independent of the used approximation method since its recognition strategy does not rely on the number of segments composing the shape. Length and turning angle information are extracted from the chain of segments. The comparison method is invariant to scale, translation and some occlusions of the extracted contour. A simple pre-processing method, also based on arc-length features, is presented to be used as a coarse fitting method to determine angle rotation and as a first filter to eliminate non pertinent candidates.
KeywordsDynamic Time Warping Content Base Image Retrieval Zernike Moment Turning Angle Normalize Cross Correlation
Unable to display preview. Download preview PDF.
- 1.Latecki, L.J., Lakamper, R., Eckhardt, T.: Shape descriptors for non-rigid shapes with a single closed contour. IEEE Conference on Computer Vision and Pattern Recognition, Proceedings 1, 424–429 (2000)Google Scholar
- 4.Sankoff, D., Kruskal, J.B.: Time Warps, String Edits, and Macromolecules: The Theory and Practice of Sequence Comparison. Addison-Wesley, London (1983)Google Scholar