الفهرس 
Microarray image preprocessingOnly 14 pages are availabe for public view
 |  | from | 104 |  |  |
| | | from | 104 | | |
Abstract
ABSTRACTThe DNA microarray image is considered to be the magic biometric technique for personal identification based on the concurrent identification of millions of genes. In this work we present new methods for human matching using microarray image recognition techniques which is based on artificial neural network (Feed-forward Artificial Neural Network (FFANN) and Radial-basis Artificial Neural Networks (RBANN), and an enhanced method for DNA microarray image segmentation using radial-basis neural network (RBNN). This method starts with spot addresses and gridding, then spot extraction, finally segment each spot using RBNN classifier. Various experimental results show that our method is better than the other segmentation methods (K-means – support vector machine(SVM)) used in the cDNA microarray segmentation. The experimental results of human recognition by cDNA microarray images data in a database of over 346 samples, we have achieved a good recognition and rejection accuracy rate for the recognition of registered persons and rejection of unknown samples respectively. These results show that, the proposed technique is effective personal identification technique.