The main goal of this software tool is to automatically recognise and allocate visual patterns (such as words, drawings and seals) in digitised manuscripts.

Please pay attention to the following limitations:

The required computational resources depend on your data. Detecting several patterns concurrently can exhaust your computer memory, and searching in huge datasets is limited by your storage capacity and your browser upload settings.
This version provides best results of pattern detection when applied to datasets with relatively similar image qualities, such as resolution, degradation level, rotation, etc. an example of such datasets are manuscripts that have been digitised under similar conditions and using similar equipment. Future releases will provide higher tolerance levels with respect to these aspects.
This version has a limited capacity of benefiting from a large number of examples for the same pattern. Future releases will generate pattern models adaptively so that it gets better with every additional pattern example.

Acknowledgement:

The development of this software was sponsored by the Cluster of Excellence 2176 ‘Understanding Written Artefacts’, generously funded by the German Research Foundation (DFG), within the scope of the work conducted at Centre for the Study of Manuscript Cultures (CSMC).

This software has been developed in close collaboration with Dr. Giovanni Ciotti (CSMC), Dr. Volker Märgner (Technische Universität Braunschweig), Dr. Agnieszka Helman-Wazny (CSMC), Dr. Isabelle Marthot-Santaniello (Universität Basel). I would like to thank them for testing this software tool and validating the results.