Estimating the Pen Trajectories of Static Signatures Using Hidden Markov Models
IEEE Transactions on Pattern Analysis and Machine Intelligence
Verification of dynamic curves extracted from static handwritten scripts
Pattern Recognition
Techniques for static handwriting trajectory recovery: a survey
DAS '10 Proceedings of the 9th IAPR International Workshop on Document Analysis Systems
| Hi-index | 0.00 |
This paper proposes a model for tracking handwriting based on curvature minimization. Aiming at recovering stroke sequences from words written in the past, the problem is formulated as a graph theoretical question. The solution does not require mathematical functions. In particular, loops are considered elementary basic strokes and are not approximated with functions. Recovering stroke sequences is equivalent to ordering the edges of a graph, which can be derived in a straightforward manner from the scanned binary word image. This paper does not deal with any special recognition method. However, an important aim is to provide a mechanism for deriving temporal information to help to improve off-line recognition methods. The most important advantages of this method over methods proposed so far are: a single global principle (global optimization of curvature), implicit modeling of retraced strokes and simplicity.