Recognition and analysis of Tamil inscriptions and mapping using image processing techniques

Abstract

The most important necessity in image processing lies in the conversion of ancient Tamil characters to modern text. The ancient Tamil stone inscriptions are the source for these ancient Tamil characters. Analysing and Recognising the ancient Tamil characters from the scripts called inscriptions is a difficult task for the present generation who learn to educate, read and write only through the modern Tamil characters. Learning the elaboration of Modern Tamil from ancient Tamil inscriptions is a time devouring process and thus we need a recognition system that helps to teach, understand and also to research the ancient cultures and heritages. To design a good comprehend and better analysing system we propose a method called noise removal which is also called as pre-processing which removes the entire disturbance in the input image. Next we use a method called morphological operation to perform dilation and erosion operations. Next step is to use the connected component to find the letters which are present in the binary image. Finally we will segment each and every character and match it with our current Tamil language using a process called correlation matching to produce the result of matching characters.