D.B. Megherbi and Y. Iyassu (USA)
- Signals & low level Image Processing, Morphology, Pattern Recognition, Object Recognition, Image Enhancement, Coding, Optical Character Recognition.
In this paper, we introduce a technique for multi-font multi-size printed character recognition. In a previous work [3][4], through a conformal mapping process, we demonstrated the ability to 1) recognize shapes, and 2) concisely represent shape boundaries using a set of polynomial coefficients derived in the mapping process. In this work we illustrate how these previous results can be applied to multi-font multi-size typed character representation and recognition. The proposed method is based on complex variables and conformal mapping methodology and is invariant to transformation such as scaling, rotation and translation. We show and demonstrate the ability of the technique to recognize typed letters and concisely represent these letters' boundaries using a set of polynomial coefficients derived in the mapping process. Experimental results of different typed letters are shown to illustrate the power, versatility and robustness of the proposed techniques.
Important Links:
Go Back
