Résumé: Scientific illustrations (schemes, board, maps, geometrical draws, diagrams) edited in non- scientific productions (literature, law, theology, philosophy) published in medieval and early modern manuscripts or printed matter, designed as ornamentation in engravings or drawings or even on clothing, textiles and artefacts are frequently found in European libraries and museum but remain untapped whereas they are still digitalized. Individual collections donated to these libraries or museum reveal a visual literacy and the consistency of the knowledge of the donator but also the “making-of” of our European culture through a diversity of books edited in many European locations, disciplines or legacies. To expose this European cultural consistency and to study these illustrations, specific corpus located in representative parts of European libraries or museum (Belgium, France, Portugal, Romania, United-Kingdom) are selected by our team. Each collection of illustrated books, objects or textiles will generate amounts of datas (digitalisation and database with their shapes, patterns, lines, colors) and will be classified in order to elaborate a history of European data visualisation. This classification will be organized and promoted with both literary or historian studies, and new computing tools and search engines’ reports. These tools will be configured to discover similar or specific patterns, shapes and lines inside these pictures but will also deal with many challenges: manuscripted character recognition, automatic exploration of non figurative pictures and annoted images, big data on digital humanities produced in several languages and quality of impression. Our main purpose is to identify European way of thinking with scientific imagery and epistemic pictures in a non-scientific framework: as a European visual literacy could be useful nowadays, we will try to experiment these early modern models and patterns to collect, synthetize and visualize nowadays datas, included lost and immaterial datas. Ethical issues, misinterpretations and audience ‘s manipulation will be taken into account to select patterns and figurative models.