Title :
Object recognition using fuzzy modelling and fuzzy matching
Author :
Soodamani, R. ; Liu, Z.Q.
Author_Institution :
Dept. of Comput. Sci., Melbourne Univ., Parkville, Vic., Australia
Abstract :
In this paper, a new method is proposed for learning continuous valued features and mapping them into fuzzy concepts. Further, an efficient method of updating fuzzy evidences is proposed which enhances recognition of approximate shapes. The techniques are tested in a model based 3D object recognition system that employs a fuzzy Hough transform approach for global shape description
Keywords :
Hough transforms; computer vision; feature extraction; fuzzy set theory; image matching; inference mechanisms; object recognition; stereo image processing; 3D object recognition; Hough transform; computer vision; feature extraction; fuzzy conceptual graphs; fuzzy evidences; fuzzy matching; fuzzy modelling; fuzzy reasoning; fuzzy set theory; learning; shape description; Computer vision; Fuzzy reasoning; Fuzzy systems; Image databases; Image segmentation; Object recognition; Shape; Spatial databases; System testing; Uncertainty;
Conference_Titel :
Fuzzy Systems Proceedings, 1998. IEEE World Congress on Computational Intelligence., The 1998 IEEE International Conference on
Conference_Location :
Anchorage, AK
Print_ISBN :
0-7803-4863-X
DOI :
10.1109/FUZZY.1998.687477
