Abstract: Script is a set of symbols and rules used to express or convey the information in a graphic form. Script Identification is one of the challenging steps in the Optical Character Recognition system for multi-script documents. In Indian and Non-Indian context some results have been reported, but research in this field is still emerging. This paper presents study on word wise script Identification which is based on scale Invariant Feature Transform, The system is developed and tested for 500 document images representing English, Hindi, Kannada, Bengali and Gurumukhi scripts. The system is developed includes a feature extractor which is based on scale invariant feature transform and for classification nearest neighbour classifier is used. The method is found to be robust and classification accuracy across five scripts is found to be 97.8%.

Keywords: Script Identification, Image Processing, SIFT, KNN.