Postal Codes Recognition System for Myanmar Post Office using Neural Network Architecture

Abstract

Pattern recognition plays an important role in the modern world. It can solve more complex problems and makes human’s job easier. In this paper, we intend to develop a handwritten Postal Codes recognition system that can be employed in mail sorting for Myanmar Post Office. Postal Code consists of English characters and digits. This system can help humans to sort mails with Postal Codes that are difficult to identify. Two techniques applied in this system are Image processing and Artificial Neural Network (ANN). Image processing steps such as binarilization, cropping, resizing and thinning are used in this system. Multilayer feed forward neural network is used to adapt the weights that are needed to train the network using back propagation algorithm. Image processing is an important area for the application of ANN pattern recognition techniques. Simulation results are also included to show the accuracy of this system.