We introduce in this paper a new dataset of annotated pages from books of hours, a type of handwritten prayer books owned and used by rich lay people in the late middle ages. The dataset was created for conducting historical research on the evolution of the religious mindset in Europe at this period since the book of hours represent one of the major sources of information thanks both to their rich illustrations and the different types of religious sources they contain. We first describe how the corpus was collected and manually annotated then present the evaluation of a state-of-the-art system for text line detection and for zone detection and typing. The corpus is freely available for research.
