Subscribe Now Subscribe Today
Science Alert
Curve Top
Journal of Applied Sciences
  Year: 2002 | Volume: 2 | Issue: 3 | Page No.: 281-287
DOI: 10.3923/jas.2002.281.287
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail

Object Recognition Using ANN with Backpropogation Algorithm

I.M. Qureshi and A. Jalil

An object recognition procedure using artificial neutral network is proposed in this paper. This algorithm will recognize objects in the images which are invariant to rotation, translation and scaling of objects. The Sobel operators are proposed to detect the boundary of the object. A Fourier descriptor pattern classifier is able to classify object without regard to rotation, translation and scale variation. Fourier descriptors of boundary image generate Feature vectors by truncation the high frequency components. A back-propagation neural network is proposed to recognize the object based on these Feature vectors.
PDF References Citation Report Citation
  •    Classifier Design Algorithms Aimed at Overlapping Characteristics
  •    Wavelet Packet and Percent of Energy Distribution with Neural Networks Based Gender Identification System
  •    Modular Arithmetic and Wavelets for Speaker Verification
How to cite this article:

I.M. Qureshi and A. Jalil , 2002. Object Recognition Using ANN with Backpropogation Algorithm. Journal of Applied Sciences, 2: 281-287.

DOI: 10.3923/jas.2002.281.287






Curve Bottom