Abstract
Two experiments in ancient Hebraic handwriting identification were conducted. Features based on run-length histograms measured on groups of characters were used in the first experiment, yielding 91. 3 percent correct handwriting classification. In the second experiment the run-length histograms were measured on one selected letter, yielding completely correct handwriting classification.
Original language | English |
---|---|
Pages (from-to) | 405-409 |
Number of pages | 5 |
Journal | IEEE Transactions on Systems, Man and Cybernetics |
Volume | SMC-12 |
Issue number | 3 |
DOIs | |
State | Published - 1 Jan 1982 |
ASJC Scopus subject areas
- General Engineering