Structural Feature Extraction to recognize some of the Offline isolated Handwritten Gujarati Characters using Decision Tree Classifier

Abstract

Large amount of information is prevailing on paper and in an era of digital technology it requires it to store this information in electronic format. Using scanner this information can be digitized. Later any modification in terms of add, editing, removing and searching to it requires a technique or methodology which will identify text from image and convert into ASCII or Unicode. This paper presents recognition of offline handwritten character for Gujarati script using structural features. For proposed experimental work five characters of Gujarati script are considered. Decision tree classifier is proposed for classification. Various phases of character recognition are implemented such as collecting handwritten input, digitization, preprocessing, extracting structural features and recognition. By proposed work authors are able to achieve 88.78% of average success rate for defined five characters.