Identification and classification of multilingual document using maximized mutual information

Document consisting of more than one language is known as multilingual document. This paper is addressing the problem of detecting document that consists of more than one language by using maximized mutual information technique; identified languages are classified by implementing KNN and SVM classification model. Indian languages have their own characteristics and they can be distinguished with the help of visual discrimination methods and statistical methods. To identify these differences through machine we are making use of edge direction based feature. To capture the differences present in languages Edge Direction Histogram (EDH) is used. These techniques are implemented on a dataset of 400 images consisting of documents with Kannada and Hindi languages. The accuracy of proposed method by using KNN classifier is 99.8% and SVM classifier is 100%.