This paper presents a novel method for edge detection based on multiscale wavelet features and fuzzy c-means clustering. Firstly, an effective feature extraction algorithm using multiscale wavelet transform was proposed to extract classification features, thus the feature vector for each pixel was gained, which contained the gradient information in various directions; and then, these vectors were used as inputs for the fuzzy c-means clustering algorithm, which resulted in an automatic classification. In this way, the edge map can be obtained adaptively. Some comparisons with traditional edge detection algorithms were given in this paper. Experimental results demonstrated that the proposed method had a more satisfying performance.
Multiscale edge detection based on fuzzy c-means clustering
2006-01-01
3398824 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Fuzzy Multiscale Edge Detection (FMED) Applied to Automatic Left Ventricle Boundary Extraction
British Library Conference Proceedings | 1999
|Morphological Pyramids for Multiscale Edge Detection
British Library Conference Proceedings | 1998
|Circle detection based on multiscale edge curvature estimation
British Library Online Contents | 2011
|