Deployment of Handwriting Recognition System Using Artificial Neural Network (ANN)

Over the last few years, we have observed a big dramatic change in the handwriting recognition systems in terms of the security protocols. In this poster, we focus on the most common and efficient recognition system, which is the Optical Character Recognition (OCR). The need for some forms of automated or semi-automated optical character recognition ( OCR ) has been recognized for decades. Today there are numerous algorithms that perform this task, each with its own strengths and weaknesses. The basic idea of the poster is essentially to run the identical data through the two different algorithms and note the differences in each run along the way. Thus comparisons between the two handwriting and typed OCR detection methods are not influenced by unfair differences in the data or unbalanced procedures that would favor one method over the other.  OCR is used in many documenting systems such as MS word and PDF. It can be used even in transferring the document’s formatting. There are several steps that need to be taken to get a proper recognition such as classification, segmentation and extraction. The most important step is the symbol segmentation because it is based on this step; we can determine the right character or meaning. It fragments the input to a number of separate characters. This process includes other sub-steps which are word, line, and character segmentation. Our algorithm combines pattern recognition with help of a computer algorithm. It is considered as a multitask software, wherein the security is required; for example: police department, banks, and airports. Hence, we can use it as a different classifier at the same time. The static and dynamic approaches are the main categories of OCR which is the basis on its input. There are differences between the static and dynamic categories. Dynamic is more efficient in such multimedia recognition (e.g.: speech). OCR’s technique is almost considered in many applications. It has witnessed a huge deployment in many systems especially in the human computer interaction. We have used in this work four different classifications for deploying of the Artificial Neural Network for optical hand writing recognition system. We have used C sharp programming language to implement the application.  Our validated results show high improvement in the hand written recognition.