Classification of Graffiti digits by using Computational Intelligence: Several architectures and techniques to optimize the performance of the Neural Networks in the Pattern Recognition

The technological advances and the massive flood of papers have motivated many researchers and companies to innovate new methods and technologies. They build automatic readers to recognize handwritten documents. In particular, handwriting recognition is very useful technology to support applications like electronic books (eBooks), postcode readers (that sort the mail in post offices), and some bank’s applications. This book proposed systems to discriminate handwritten graffiti digits and some commands with different architectures and abilities. It introduced three classifiers, namely single neural network (SNN) classifier, parallel neural networks (PNN) classifier and tree-structured (TS) classifier. The three classifiers have been designed through adopting feed-forward neural networks. The back-propagation algorithm has been used to optimize the network’s parameters (connection weights). Several architectures are applied and examined to present a comparative study about the three systems from different perspectives.