On-Line and Off-Line Handwriting Recognition: A Comprehensive Survey
IEEE Transactions on Pattern Analysis and Machine Intelligence
Moving markup: repositioning freeform annotations
Proceedings of the 15th annual ACM symposium on User interface software and technology
Recognition of freehand sketches using mean shift
Proceedings of the 8th international conference on Intelligent user interfaces
MathPad2: a system for the creation and exploration of mathematical sketches
ACM SIGGRAPH 2004 Papers
Distinguishing Text from Graphics in On-Line Handwritten Ink
IWFHR '04 Proceedings of the Ninth International Workshop on Frontiers in Handwriting Recognition
A Pen-Based Interface for Real-Time Document Edition
ICDAR '07 Proceedings of the Ninth International Conference on Document Analysis and Recognition - Volume 02
A freehand-sketch environment for architectural design supported by a multi-agent system
Computers and Graphics
Automatic interpretation of proofreading sketches
SBM'06 Proceedings of the Third Eurographics conference on Sketch-Based Interfaces and Modeling
Parsing ink annotations on heterogeneous documents
SBM'06 Proceedings of the Third Eurographics conference on Sketch-Based Interfaces and Modeling
Hi-index | 0.00 |
In sketch-based interfaces, the separation of text and graphic elements can be essential when a system has to react to different kinds of input. Even if the interaction with the interface consists in drawing graphic elements, text input may be considered for some purposes, such as annotation, labelling, or input of recognizable text. This work deals with the detection of textual patterns in a set of digital ink elements. The main idea is that text needs a special hand behaviour to be produced, different from the behaviour employed to draw symbols or other graphic elements. Inspired by the models that describe handwriting as a system of coupled oscillations, we believe that the frequencies of these oscillations contain some information about the symbol nature. Therefore, we employ a descriptor that works in the Fourier space. Results show that this representation leads to distinguished patterns for text and graphic elements. The performance of our system is close to the performance one would obtain by using a handwriting recognition engine tuned for this task, while being much faster. Some benefits are also present when both approaches - the proposed and the engine - are combined.