FROM NEURAL NETWORK TO AIRCRAFT RECOGNITION SYSTEM
Somaie, A., Badr, A., Salah, T. (1999). FROM NEURAL NETWORK TO AIRCRAFT RECOGNITION SYSTEM. EKB Journal Management System, 2(2nd International Conference on Electrical Engineering ICEENG 1999), 305-311. doi: 10.21608/iceeng.1999.62514
A. A. Somaie; A. Badr; T. Salah. "FROM NEURAL NETWORK TO AIRCRAFT RECOGNITION SYSTEM". EKB Journal Management System, 2, 2nd International Conference on Electrical Engineering ICEENG 1999, 1999, 305-311. doi: 10.21608/iceeng.1999.62514
Somaie, A., Badr, A., Salah, T. (1999). 'FROM NEURAL NETWORK TO AIRCRAFT RECOGNITION SYSTEM', EKB Journal Management System, 2(2nd International Conference on Electrical Engineering ICEENG 1999), pp. 305-311. doi: 10.21608/iceeng.1999.62514
Somaie, A., Badr, A., Salah, T. FROM NEURAL NETWORK TO AIRCRAFT RECOGNITION SYSTEM. EKB Journal Management System, 1999; 2(2nd International Conference on Electrical Engineering ICEENG 1999): 305-311. doi: 10.21608/iceeng.1999.62514

FROM NEURAL NETWORK TO AIRCRAFT RECOGNITION SYSTEM

The International Conference on Electrical Engineering
Article 32, Volume 2, 2nd International Conference on Electrical Engineering ICEENG 1999, November 1999, Page 305-311  XML PDF (1.21 MB)
Document Type: Original Article
DOI: 10.21608/iceeng.1999.62514
View on SCiNiTO View on SCiNiTO
Authors
A. A. Somaie1; A. Badr2; T. Salah1
1R & D Centre, EAF Cairo
2Comp & Sys Eng. Dep., Faculty of Eng., Ain Shams University, Cairo
Abstract
In this paper, an aircraft recognition system using a neural network is presented. A 2-D perspective view of aircraft models is first normalized through the preprocessing stage using bilinear interpolation and principal component analysis. The new patterns are invariant to translation, dilation, and rotation. Then, the Kohonen and Grossberg neural networks were trained using a small number of normalized patterns. The presented algorithm was tested on partially incomplete, noisy and geometrically distorted images and it was found that the recognition performance is 100% with six referenced aircraft.
Keywords
Image processing; Neural network; Pattern Recognition
Statistics
Article View: 107
PDF Download: 186