ICCS2004 Application Submission/Review

Abstract for
"Artificial Neural NetworkBased Offline Hand written character recognition for Postal Services"


The most common goal of finding address constants such as city state country from letters is the recognition of offline handwritten character information. However, in order to do this, it is necessary to use a reliable and efficient process not only for recognizing character but also for extracting characters from letters , removing noise in written characters , joining characters to make a word , recognizing the meaning of word . In this paper we present a process for removing noise from extracted information using various filtering technique including orientation checking , recognizing characters using Neural Net technique , word recognition using plug and play learning mechanism. The main advantage of this process as compared with other similar approaches is that these methods easier to implement , requires less memory and time and they are cost effective.