Recognition of Handwritten Arabic Alphanumeric Characters by Backpropagation Neural Networks

الملخص

This paper presents the results obtained by applying Back
propagation (BP) neural network model, to application of handwritten
Arabic alphanumeric characters (HAAC) recognition. A novel method
for features extraction, based on a shadow projection has been used.
The network is trained using alphanumeric character samples written
by different people (learning set). They are required, after the learning
is over, to recognize alphanumeric characters outside the learning set.
Also, the paper includes an evaluation of the recognition capability of
the BP model for 28 alphanumeric characters.