Digital Ink Recognition TRIZ Cases
- Online Handwriting Recognition with Multitask Motion Sensing
A multitask model uses pen motion data to segment strokes and classify characters continuously with fewer sensors and lower system complexity.
- AI Document Classification Using OCR, Barcodes, and Shape Cues
Combining OCR, barcode decoding, and document shape analysis speeds document verification while preserving classification accuracy.
- Pen Data Storage with Embedded Function Identification
Embedded function identification lets pen data be stored and reproduced directly, avoiding file opening and unclear application selection.
- Digital Ink Text Line Extraction for Non-Chronological Handwriting
Strip-based stroke ordering and neural net analysis improve text line extraction when handwriting includes delayed strokes and punctuation.