Abstract | ||
---|---|---|
This paper presents an integrated approach to parsingtextual structure in freeform handwritten notes. Text-graphicsclassification and text layout analysis areclassical problems in printed document analysis, but theirregularity in handwriting and content in freeform notesreveals limitations in existing approaches. We advocatean integrated technique that solves the layout analysisand classification problems simultaneously: the problemsare so tightly coupled that it is not possible to solve onewithout the other for real user notes. We tune andevaluate our approach on a large corpus of unscripteduser files and reflect on the difficult recognition scenariosthat we have encountered in practice. |
Year | DOI | Venue |
---|---|---|
2003 | 10.1109/ICDAR.2003.1227628 | ICDAR-1 |
Keywords | Field | DocType |
text layout analysis areclassical,freeform notesreveals limitation,integrated approach,integrated technique,printed document analysis,freeform handwritten note,freeform handwritten notes,difficult recognition scenariosthat,layout analysisand classification problem,real user note,discerning structure,large corpus,image recognition,text analysis,feature extraction | Document analysis,Handwriting,Computer science,Document layout analysis,Feature extraction,Natural language processing,Artificial intelligence,Parsing | Conference |
ISSN | ISBN | Citations |
1520-5363 | 0-7695-1960-1 | 35 |
PageRank | References | Authors |
2.48 | 5 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Michael Shilman | 1 | 325 | 22.43 |
Zile Wei | 2 | 136 | 8.32 |
Sashi Raghupathy | 3 | 120 | 9.89 |
Patrice Simard | 4 | 1268 | 621.43 |
David Jones | 5 | 545 | 25.90 |