BETTER PERFORMANCE OF ORTHONORMAL-BACKPROPAGATION NEURAL NETWORK ALGORITHM IN IDENTIFICATION OF OBJECT FEATURES IN STEREOSCOPIC IMAGES

CHELLAM H.G.¹*, JOSEPH R.V.^2
1Department of Computer Science, Kamarajar Government College, Surandai- 627 859, TN, India.
²Department of Computer Science and Application, Kamaraj College, Thoothukudi- 628 003, TN, India.
* Corresponding Author : hchellam@gmail.com

Received : 29-08-2012     Accepted : 13-03-2014     Published : 31-03-2014
Volume : 3     Issue : 1       Pages : 39 - 42
Inform Sci Tech 3.1 (2014):39-42

Keywords : Artificial Neural Network, Backpropagation, Orthonormal, Stereoscopic images
Conflict of Interest : None declared

Any 3D object can be registered in stereoscopic camera as 2D image pair similar to that of the left eye view and to that of right eye view seen by human eyes. From the stereo pair of images displayed on a screen, the viewer could achieve 3D perceptron by wearing special viewing aids on eyes. On the other hand volumetric display of sectional image of an object display tank concedes 3D view of the object to the viewer without calling for any special viewing aids. This paper reports the computation of the object from stereo pair of images by identifying object features and placing them in respective depth planes. Backpropagation and Orthonormal-backpropagation algorithm of Neural Network are used for extracting the object features in stereoscopic images. Also the efficiency of the Orthonormal-backpropagation algorithm is tested and reported.

[1] Grimson W.E.L. (1981) From Images to Surfaces: A Computa-tional Study of the Human Early Visual System, 4, MIT Press, Cambridge  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[2] Ohta Y. & Kanade T. (1985) IEEE Transactions on Pattern Analysis and Machine Intelligence, (2), 139-154  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[3] Baker H.H. & Binford T.O. (1982) A System for Automated Ste-reo Mapping, CA, 215-222.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[4] Bishop C.M. (1995) Neural Networks for Pattern Recognition, Oxford University Press.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[5] Ramasami R. & Sivarajan K.N. (2000) Optical Networks: A Practical Perspective, Harcourt  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[6] LiMin Fu (2003) Neural Networks in Computer Intelligence, Tata Mcgraw-Hill  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[7] Ken Calvert (2006) ACM SIGCOMM Computer Communication Review, 36(2), 27-30  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[8] Prema K.V. (2005) Journal of Computer Society of India, 35(1).  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[9] Krishnamurthy V., Mainra V.P. & Arora J.L. (1976) An Introduc-tion to Linear Algebra, East-West Press Private Limited, New Delhi  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[10] Arumugam S. & Isaac A.T. (2006) Modern Algebra, Scitech Publication, Chennai.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[11] Chandresekhara Rao K. (2006) Functional Analysis  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[12] Kush A., Gupta P. & Kumar R. (2005) Journal of Computer Society of India, 35(2).  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus