Un innovativo Graphic Matching System per la ricerca in database di manoscritti antichi

Abstract

The paper outlines a pattern recognition process, which uses a graphic matching algorithm based on a shape contour recognition function without the need to apply segmentation. The process starts from the identification of a Region of Interest (ROI) within the image. The ROI is then managed for the creation of the shape model then used to perform searches for similar models in one or more target images. The system has been developed and tested with the aim of proposing an innovative approach in the research and retrieval of information in digital libraries related to ancient manuscript and printed documentation. This approach is based on the application to the data humanities of the fourth knowledge paradigm that underlies data science. Following this approach, the algorithm is used to deduce new research hypotheses through the discovery of models directly inferred from large digital libraries.