Title :
A novel fuzzy Hough transform for shape representation
Author :
Soodamani, R. ; Liu, Z.Q.
Author_Institution :
Dept. of Comput. Sci., Melbourne Univ., Carlton, Vic., Australia
Abstract :
Three techniques of straight line Hough transform are proposed by a generalised mathematical formulation of the Hough transform in terms of aggregation and weight functions. Amongst these techniques, inclusive of the conventional Hough transform, it is shown by experimental results on a set of 3D nonparametric objects, and in a controlled and inherent noise environment, that the fuzzy Hough transform approach performs the best, as a global shape descriptor
Keywords :
Hough transforms; aggregation; fuzzy set theory; noise; nonparametric statistics; object recognition; 3D nonparametric objects; aggregation; fuzzy Hough transform; global shape descriptor; inherent noise environmen; shape representation; straight line Hough transform; weight functions; Australia; Computer science; Computer vision; Convolution; Discrete transforms; Fuzzy control; Machine intelligence; Noise shaping; Shape control; Working environment noise;
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.1998.686359
