Abstract: Automatic Number Plate Recognition or ANPR , is the most important needs of automatic traffic control system. The main objective of this article is The design of a car plate recognition system with optimization a new method. This system is a combination of Neural network, Image processing , fuzzy logic And Means algorithm related to its structure Can accurately ANPR of Iranian cars . The presented method in addition to high accuracy , Has received good response tests time . Other features in this system, is Section that denomination rule of law that Increase the accuracy of the system by MLP Neural network. Also Using fuzzy logic in carplate Separation time is reduced the sensitivity of the system to rotate the image. The example of Activities is Traffic control , ANPR , process of entry and exit of cars . The results of tests on the sample images , Show the superiority of the proposed method is compared to other methods