Classification and identification of Telugu handwritten characters extracted from palm leaves using decision tree approach

Research in character recognition is very popular for various application potentials in banks, post offices, defense organizations, reading aid for the blind, library automation, language processing and multi-media design. Even though Epigraphical work dealing with stone inscriptions have been analyzed, these have been done largely manually and also on 2D traces. A large collection of these are available in the classical Indian languages like Sanskrit, Tamil, Pali etc as well as in more modern languages like Telugu. These characters on the palm leaf have the additional properties like depth, an added feature which can be gainfully exploited in character recognition. In this paper, we explore how these 3D features can be extracted and how they can be used in the recognition and classification process. This paper describes a system to identify and classify Telugu (a south Indian language) characters extracted from the palm leaves, using Decision Tree approach. The decision tree is developed using SEE5 algorithm, which is an improvement from the predecessor ID3 and C4.5 algorithm. The identification accuracy obtained is 93.10% using this method.