Hopfield Neural Network for Classification of Digits

Automobiles are manufactured at an astonishing rate. It is becoming tedious to control, manage and hold accountable to these assets. As the automotive area is vulnerable to theft, accident, and traffic violations. The registration number plates are unique identifiers in tracking the automotive. In modern days, the role of CCTV helps the authorities in monitoring automotive movements. However, the speed of vehicles can escape the human eye. It is practically impossible to rely on the human-intervention for high-speed capturing. The images may get distorted even with advanced camera equipment and this process comes with a trade-off to high processing time with resource utilization. This paper proposes the optimal solution for automated automotive number plate recognition involving fewer computing resources by leveraging Hopfield neural networks. The accuracy obtained with this proposed method is 70.5%.