Image Edge Detection Based on Adaptive Fuzzy Morphological Neural Network

Author

Yang, Guo-qing ; Guo, Yan-Ying ; Jiang, Li-Hui

Author_Institution

Civil Aviation of China, Peking

fYear

2006

fDate

13-16 Aug. 2006

Firstpage

3725

Lastpage

3728

Abstract

The fuzzy hit-or-miss transformation is a fuzzy morphological operator, which is a key for the feature extraction in the ambiguous information. In this paper, a neural network implementation for fuzzy morphological operators is proposed, and by means of a training method and differentiable equivalent representations for the operators, efficient adaptive algorithms to optimize structuring elements are derived. The gradient of the fuzzy morphology utilizes a set of structuring elements to detect the edge strength with a view to decrease the spurious edge and suppressed the noise. Results are presented for images in comparison with the others edging detectors

Keywords

edge detection; feature extraction; fuzzy neural nets; fuzzy set theory; image denoising; mathematical morphology; mathematical operators; adaptive fuzzy morphological neural network; feature extraction; fuzzy morphological operator; image edge detection; image processing; noise suppression; Adaptive signal processing; Cybernetics; Detectors; Electronic mail; Feature extraction; Fuzzy neural networks; Fuzzy sets; Image edge detection; Image processing; Machine learning; Morphology; Neural networks; Adaptive fuzzy morphological neural network; Edge detectors; Image processing; Morphological operation;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics, 2006 International Conference on

Conference_Location

Dalian, China

Print_ISBN

1-4244-0061-9

Type

conf

DOI

10.1109/ICMLC.2006.258634

Filename

4028718