WANDA: a generic framework applied in forensic handwriting analysis and writer identification

Katrin Franke, Lambert Schomaker, Christian Veenhuis, Christian Taubenheim, Isabelle Guyon, L.G. Vuurpijl, Merijn van Erp, Geertje Zwarts

Radboud Repository (Radboud University) · 2003 · 25 citations · 3 references

Full text

Open access

Concepts

Abstract

This paper presents the WANDA Workbench, which is an open framework\nfor electronic data processing. The framework provides generic interfaces for ’plugin’\napplications for graphical user interfaces (client desktop with client plug-ins) and\nprocessing modules (server with server plug-ins). The applied plug-in concept allows\nfor the functional extension of the workbench without changing the framework. Moreover,\nfor data modeling, messaging and system configuration the eXtensible Markup\nLanguage (XML) is implemented, which supports the interoperability with existing\napplications and allows for continuous adaptations to the evolving state of technology.\nThe current application domain of the framework is writer identification and handwriting\nexamination as frequently used in crime investigation and prosecution. Dedicated\nplug-ins using digital image processing and pattern recognition establish objective\nmeasurements and feature extraction to promote reproducible analysis results.\nMoreover, based on domain knowledge, which was collected from forensic document\nexaminers, the data standard WANDAXML was designed that allows for the annotation,\nprocessing, journaling and storage of digitized questioned handwriting documents.\nIn the following, we detail the adopted design philosophy, provide pointers for\nemploying the data model, and sketch the realized framework and current plug-ins.\nThe generic framework with appositive plug-in programmer handbooks as well\nas the data standard will be made available to the research community through the\nInternet http://pentel.ipk.fhg.de/wanda.

References

3