VISUAL INPUT FOR PEN-BASED COMPUTERS Student: Mario E. Munich 
Faculty: Pietro Perona 
Support: NSF, NYI, ONR 
Detailed Description

GOAL: Develop a small and portable visual interface that can track and recognize handwriting. 
 
  • We built a new input interface for computers.
  • The handwritten path can be recovered from a sequence of images.
  • The system works in real time at 60 Hz.
 
  • Need to improve the performance of pen up detection.
  • Need to develop a better model of handwriting generation.
 

 

 
Detailed Report

Mario Enrique Munich - mariomu AT vision DOT caltech DOT edu