Dall’atramentum al pixel

Abstract

The tradition of humanistic studies is based on the careful analysis of written sources, including ancient manuscripts and printed books. This process can be time-consuming, and logistical challenges may arise when retrieving the originals. Such documents are typically analyzed in detail and on multiple levels, focusing on the material characteristics of the specimen and the writing support, as well as the informative, decorative, and textual elements. The decoding of writing and mise en page/ mise en texte provides valuable information for scholars. This information can be historical, artistic, literary, palaeographic, or diplomatic in nature. Scholars with an attentive eye, educated by study and experience in the field, can perceive even the most minute details. However, when a computer “reads” an ancient manuscript or printed document, what happens? Modern computer systems offer significant computing power and advanced graphical and optical capabilities that support librarians, archivists, and historians. Technical terms such as OCR-Optical Character Recognition, HTR-Handwritten Text Recognition, Document Segmentation, Layout Recognition, Artificial Intelligence, Machine Learning, and Deep Learning are increasingly becoming essential skills in “memory institutions” and humanistic research that involve computer technology.