Ink texture analysis for writer identification

K. Franke1, O. Bunnemeyer1, T. Sy1
1Fraunhofer Institute for Production Systems and Design Technology, Berlin, Germany

Tóm tắt

This paper presents an approach for ink type recognition. Ink type classes will be derived from the physical properties of ink. Ink specific trace morphologies are considered as textures. From these discriminant texture features of the co-occurrence matrix will be derived. The proposed method for automated ink type recognition was tested using 62 different kinds of pens and refills. The achieved recognition result of 99.7% for 600 dpi and 98.4% for 300 dpi handwritings further promotes the study of trace morphologies in particular for application in forensic writer identification.

Từ khóa

#Ink #Writing #Forensics #Handwriting recognition #Morphology #Image processing #Humans #Production systems #Automatic testing #Forgery

Tài liệu tham khảo

10.1515/9783110892413 10.1007/978-1-4613-1553-7 khanmy, 1996, Mogliche ursachen von einfiirbungsmangeln bei kugelschreiberstrichen, Mannheimer Hefte for Schriftenvergleichung, 22 martin, 1980, Gerichtliche schriftuntersuchunhen im makro-und mikrobereich, Micmscopia Acta, 82, 321 hecker, 1993, Forensische handschriftenuntersuchung, Kriminalistik Verlag hilton, 1984, Effects of writing instruments on handwriting details, Journal of Forensic Science, 29, 80, 10.1520/JFS11637J 10.1142/S0218001401001088 10.1109/TSMC.1973.4309314 franke, 2001, A computer-based system to support forensic studies on handwritten documents, International Journal on Document Analysis and Recognition, 3, 218, 10.1007/PL00013565 fryd, 1975, The direction of pen motion and its effects on the written line, Med Sci Law, 15, 167, 10.1177/002580247501500304 10.1007/978-3-642-77281-8_10 10.1023/A:1009715923555 rauber, 1997, Two-dimensional shape description, Pmc 8th Conference of the International Graphonomics Society 10.1016/0031-3203(93)90168-V ammar, 1993, A new effective approach for off-line verification of signatures by using pressure features, Proc 8th International Conference on Pattern Recognition, 566 franke, 1998, The automatic extraction of pseudodynamic information from static images of handwriting based on marked grayvalue segmentation (extended version), J Forensic Document Examination, 11, 17 10.1017/CBO9780511801389 conrad, 1971, Empirische untersuchungen zur differentialdiagnose zwischen verschiedenen unterschriftsgattungen, Zeitschrijt fur Menschenkunde, 35, 195 10.1162/15324430152733142 10.1142/S021800149400036X duda, 2001, Pattern Classification 10.1007/BF01450853