Interactive segmentation for character extraction in stone inscriptions

Indian stone inscriptions play an important role in the reconstruction of the history of India. Kannada is considered the oldest language next to Sanskrit, Prakrit, and Tamil. According to linguists Tamil and Kannada branched off simultaneously from the Dravidian language of South India before the Christian Era. Indian epigraphy becomes more widespread over the 1st millennium, engraved on the faces of cliffs, on pillars, on tablets of stone, drawn in caves and on rocks, some gouged into the bedrock. Later they were also inscribed on palm leaves, coins, copper plates, and on temple walls. Preserving the heritage of ancient civilizations furthers the education and understanding of world cultures that may help to promote world peace. Without digital preservation of important cultural sites, future generations may never know the achievements of past civilizations, our heritage. As pollution, wars, and natural disasters take their toll on these ancient structures, it is increasingly important to accomplish preservation of those historical monuments for cultural heritage of a nation. Here in the paper, we took different images of stone inscriptions of Kannada characters from various regions in Karnataka. The images are processed such that its character is preprocessed, matched and extracted. Computer vision is concerned with the theory for building artificial systems that obtain information from images. They also document the development and use of written languages in India.