Abstract | ||
---|---|---|
In this paper, a method for off-line writer identification is presented, using the contours of fragmented connected-components in mixed-style handwritten samples of limited size. The writer is considered to be characterized by a stochastic pattern generator, producing a family of character fragments (fraglets). Using a codebook of such fraglets from an independent training set, the probability distribution of fraglet contours was computed for an independent test set. Results revealed a high sensitivity of the fraglet histogram in identifying individual writers on the basis of a paragraph of text. Large-scale experiments on the optimal size of Kohonen maps of fraglet contours were performed, showing usable classification rates within a non-critical range of Kohonen map dimensions. Further validation experiments on variable-sized random subsets from an independent set of 215 writers gives additional support for the proposed method. The proposed automatic approach bridges the gap between image-statistics approaches and manual character-based methods. |
Year | DOI | Venue |
---|---|---|
2004 | 10.1109/IWFHR.2004.22 | IWFHR |
Keywords | Field | DocType |
limited size,independent test set,fraglet histogram,automatic writer identification,kohonen map dimension,manual character-based method,independent training set,fragmented connected-component contours,kohonen map,fraglet contour,individual writer,independent set,image processing,connected component,text analysis,probability distribution,identification | Histogram,Pattern recognition,Computer science,Image processing,Self-organizing map,Independent set,Probability distribution,Artificial intelligence,Connected component,Machine learning,Codebook,Test set | Conference |
ISBN | Citations | PageRank |
0-7695-2187-8 | 27 | 1.78 |
References | Authors | |
9 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Lambert Schomaker Member | 1 | 1309 | 87.50 |
Marius Bulacu | 2 | 514 | 24.17 |
Katrin Franke | 3 | 536 | 52.77 |