A New Approach For Off-Line Signature Verification Using Fuzzy ARTMAP

  • 발행 : 1995.12.01

초록

This paper delas with the detection of freehand forgeries of signatures based on the averaged directional amplitudes of gradient vetor which are related to the overall shape of the handwritten signature and fuzzy ARTMAP neural network classifier. In the first step, signature images are extracted from the background by a process involving noise reduction and automatic thresholding. Next, twelve directional amplitudes of gradient vector for each pixel on the signature line are measure and averaged through the entire signature image. With these twelve averaged directional gradient amplitudes, the fuzzy ARTMAP neural network is trained and tested for the detection of freehand forgeries of singatures. The experimental results show that the fuzzy ARTMAP neural network cna lcassify a signature whether genuine or forged with greater than 95% overall accuracy.

키워드

참고문헌

  1. Pattern Recognition v.22 no.2 Automatic signature verification and writer identification-the state of the art R.Plamondon;G.Lorette
  2. IEEE Trans. Syst. Man. Cybernetics v.4 Experimental investigation of automatic signature verification W.F.Nemcek;W.C.Lin
  3. IEEE Trans.Conputers v.C-26 no.9 Computer detection of freehand forgeries R.N.Nagel;A.Rosenfeld
  4. Proc. 8th int. conf. on Pattern Recognition A new effective approach for off-line verification of signatures by using pressure features M.Ammar;Y.Yoshida;T.Fukumura
  5. Proc. 11th IAPR International conference on Pattern Recognition v.2 Off-line signature verification using directional PDF and neural networks R.Sabourin;J.P.Drouhard
  6. Tech. Rep. Computer screening of handwritten signatures:A proposal R.N.Nagel
  7. Computer vision, Graphics, and Image Processing v.29 A new method for grey-level picture threshold using the entropy of the histogram J.N.Kapur;P.K.Sahoo;A.K.C.Wong
  8. Digital Image Processing(2nd ed) W.K.Pratt
  9. Neural Networks v.4 Fuzzy ART:Fast stable learning and categorization of analog patterns by an Adaptive Resonance System G.A.Carpenter;S.Grossberg;D.B.Rosen
  10. Proc. of IJCNN v.Ⅱ Fuzzy ART:An adaptive resonance algorithm for rapid, stable classification of analog patterns G.A.Carpenter;S.Grossberg;D.B.Rosen
  11. IEEE Trans. Neural Networks v.3 no.5 Fuzzy ARTMAP: A neural network architecture for incremental supervised learning of analog multidimensional maps G.A.Carpenter;S.Grossberg;N.Markuzon;J.H.Reynolds;D.B.Rosen
  12. 1993 INNS Annual Meeting(World Congress on Neural Networks) v.Ⅰ Quantitative Study of the QRS Complex Using Fuzzy ARTMAP and the MIT/BIH Arrhythmia Database F.M.Ham;Soowhan Han
  13. IEEE Eng. in Med. and Bio. Soc., 15th Annual International Conference v.Ⅰ Cardiac Arrthymia Classification Using Fuzzy ARTMAP Soowhan Han;F.M.Ham