Publications

This study is conducted for the purpose of recognizing numbers printed on the license plates images. The size of different patterns in the Saudi license plates is 20: the Arabic and English numbers. It is hard to classify all the numbers due to the similarity between specific numbers. This paper will propose to use a clustering technique called X-Means in order to regroup the numbers that have the same characteristics. Later develop a specific classification technique for each cluster. The experimentation of the proposed approach is applied on our constructed dataset gave us some limitation in classification. The results are improved by constructing a reference image for each class selected using a specific criteria from the training dataset