Hand printed characters recognition using wavelet features and neural networks

Abstract EN

Simplifying automatic recognition algorithms for hand-printed characters attracted immense research efforts [1-3].

Character recognition systems can improve the interaction between man and machine in many applications, including office automation, business and data entry applications.

This paper introduces the use of bi-dimensional wavelet as features extractor that is feed to Artificial Neural Networks (ANNs) for recognition Latin hand-printed characters.

An experiment to verify the efficiency of the system was performed.

The proposed technique can be divided into three major steps: the first step is pre-processing in which the original image is transformed into a digitized image utilizing a 300 dpi scanner.

Second, feature extraction using wavelets .Finally, multilayer artificial neural network is used for characters recognition.