Autor/Autorin: Mutter Kussay Nugamesh
Kussay Nugamesh Mutter is a lecturer at the University of AlMustansiryah-Baghdad. Studied the B.Sc. in Physics and M.Sc. in Image Processing at University of AlMustansiryah-Baghdad. The Ph.D. was in Optical Computing from the Universiti Sains Malaysia. The research interest is in field of Optical Artificial Neural Networks.
With the advent of advanced computer technology in the past few decades, there are different techniques have been proposed for computer classification and identification of fingerprint patterns. The Region-of-Interest (ROI) in the fingerprint images are learned, converged and identified using Hopfield neural network. The latter is affected by fingerprint image variations such as noises and image coordinates. We used the correlation operation to realign the image coordinates in the pre-processing stage. Furthermore, the net is boosted by the genetic algorithm to rebuild fingerprints images with more accurate than Hopfield net alone. Six experiments were implemented involving learning, identifying, and reconstructing processes. The experimental results showed the success of the proposed system in identifying and reconstructing the fingerprint images with more accurate and reducing the running time.
Über den Autor
Kussay Nugamesh Mutter is a lecturer at the University of AlMustansiryah-Baghdad. Studied the B.Sc. in Physics and M.Sc. in Image Processing at University of AlMustansiryah-Baghdad. The Ph.D. was in Optical Computing from the Universiti Sains Malaysia. The research interest is in field of Optical Artificial Neural Networks.
Klappentext
With the advent of advanced computer technology in the past few decades, there are different techniques have been proposed for computer classification and identification of fingerprint patterns. The Region-of-Interest (ROI) in the fingerprint images are learned, converged and identified using Hopfield neural network. The latter is affected by fingerprint image variations such as noises and image coordinates. We used the correlation operation to realign the image coordinates in the pre-processing stage. Furthermore, the net is boosted by the genetic algorithm to rebuild fingerprints images with more accurate than Hopfield net alone. Six experiments were implemented involving learning, identifying, and reconstructing processes. The experimental results showed the success of the proposed system in identifying and reconstructing the fingerprint images with more accurate and reducing the running time.
