An approach to multi-font numeral recognition

Recognition of numerals has been a research area for many years because of its various applications. But there wasn´t much research done for recognition of multi-font numerals. The approaches proposed so far, suffer from larger computation time and training for obtaining feature vectors. They can be extended to recognize many more fonts but the accuracy decreases rapidly. So as to eliminate these drawbacks, in this paper, we propose a method which recognizes 17 multi- fonts of different sizes varying from size 8 to 72, with an accuracy of 99.76% on a database of 2890 numeral images. The method requires less computation time for recognizing a numeral while maintaining the high amount of accuracy. In this method we use Euler number of a numeral to initially characterize the numbers into different groups. And then we use individual distinct features of each numeral for recognizing it.