Template Matching Method for Recognition of Stone Inscripted Kannada Characters of Different Time Frames Based on Correlation Analysis

Abstract

Stone inscripted literature speaks about the history, language of different regions of the world. Preservation of such document through digitalization process is become very important. To stop degradation and missing further, the analysis of the same will through light on historical events of that region. In this connection present work proposes a simple method of digitization using ordinary digital camera further, the pre-processing algorithm is implemented to enhance the image and improve the readability. Here it recognizes the Kannada characters based on template matching. In this method is normally implemented by first picking template and then it  call the search image, then by simply comparing the  template over each point in the search image and it calculate the sum of products between the coefficient. Based on this calculated product value it recognizes the character.  Cross correlation technique is implemented in matching the characters coefficient. Experimental results shows, it demonstrates relatively high accuracy in recognizing Stone inscriptions characters of both Hoysala, Ganga time frames and with better time efficiency when compared to previous methods.

 DOI:http://dx.doi.org/10.11591/ijece.v4i5.6333