F. Roli and S. Vitulano (Eds.): ICIAP 2005, LNCS 3617, pp. 1125 – 1132, 2005. © Springer-Verlag Berlin Heidelberg 2005 Using Strings for On-line Handwriting Shape Matching: A New Weighted Edit Distance Claudio De Stefano 1 , Marco Garruto 2 , Luis Lapresa 3 , and Angelo Marcelli 2 1 Dipartimento di Automazione, Elettromagnetismo Ingegneria dell’Informazione e Matematica Industriale, Università di Cassino, Via Di Biasio, 43 03043 Cassino (FR), Italy destefano@unicas.it 2 Dipartimento di Ingegneria dell’Informazione ed Ingegneria Elettrica, Università di Salerno, Via Ponte Don Melillo, 84084 Fisciano (SA), Italy marcounisa@libero.it, amarcelli@unisa.it 3 Departament de Ciències Matemàtiques i Informàtica, Universitat de les Illes Balears, Carretera de Valldemossa, Km 7,5 07122 Palma (Illes Balears), Spain lapresa@gmail.com Abstract. Edit Distance has been widely studied and successfully applied in a large variety of application domains and many techniques based on this concept have been proposed in the literature. These techniques share the property that, in case of patterns having different lengths, a number of symbols are introduced in the shortest one, or deleted from the longest one, until both patterns have the same length. In case of applications in which strings are used for shape descrip- tion, however, this property may introduce distortions in the shape, resulting in a distance measure not reflecting the perceived similarity between the shapes to compare. Moving from this consideration, we propose a new edit distance, called Weighted Edit Distance that does not require the introduction or the dele- tion of any symbol. Preliminary experiments performed by comparing our tech- nique with the Normalized Edit Distance and the Markov Edit Distance have shown very encouraging results. 1 Introduction Edit Distance has been widely studied and successfully applied in a large variety of application domains. In fact, in the applications in which matching, detection or recog- nition of patterns are of primary interest, a key role is played by the way in which the similarity or the dissimilarity between patterns is measured: in this context, edit dis- tance techniques offer an effective and computationally efficient way of performing such a measure, and it has been demonstrated that their applicability is not limited to alphabet-based strings in text processing, but they can be profitably used in a multitude of different applications. Examples include genome representation in bioinformatics [1], message codes in information theory [2] and sound information in speech process- ing [3]. Moreover, the concept of edit distance has been widely used in many research disciplines of pattern recognition, image processing and computer vision [4, 5].