Abstract | ||
---|---|---|
We describe the current state of a system that recognizes printed text of various fonts and sizes for the Roman alphabet. The system combines several techniques in order to improve the overall recognition rate. Thinning and shape extraction are performed directly on a graph of the run-length encoding of a binary image. The resulting strokes and other shapes are mapped, using a shape-clustering approach, into binary features which are then fed into a statistical Bayesian classifier. Large-scale trials have shown better than 97 percent top choice correct performance on mixtures of six dissimilar fonts, and over 99 percent on most single fonts, over a range of point sizes. Certain remaining confusion classes are disambiguated through contour analysis, and characters suspected of being merged are broken and reclassified. Finally, layout and linguistic context are applied. The results are illustrated by sample pages. |
Year | DOI | Venue |
---|---|---|
1987 | 10.1109/TPAMI.1987.4767901 | IEEE Trans. Pattern Anal. Mach. Intell. |
Keywords | Field | DocType |
contour analysis,percent top choice,binary feature,dissimilar font,large-scale trial,correct performance,printed characters,current state,certain remaining confusion class,roman alphabet,binary image,layout,bayesian classifier,shape,manufacturing,feature extraction,pediatrics,bayesian methods,run length encoding,classification algorithms,computer graphics,data mining | Computer vision,Pattern recognition,Naive Bayes classifier,Computer science,Binary image,Font,Feature extraction,Artificial intelligence,Statistical classification,Classifier (linguistics),Computer graphics,Binary number | Journal |
Volume | Issue | ISSN |
9 | 2 | 0162-8828 |
Citations | PageRank | References |
128 | 196.17 | 0 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Simon Kahan | 1 | 279 | 209.26 |
Theo Pavlidis | 2 | 2459 | 1848.41 |
H. S. Baird | 3 | 214 | 479.51 |