This project is in collaboration with IMAJ (Institut Mondial des Arts de la Jeunesse, supported by UNESCO), which organises an international drawing competition every year in the French city of Troyes. IMAJ now has an exceptional digital database of more than 120,000 drawings by young artists aged 3 to 25 from more than 150 countries. The ambition of the project is to carry out the first image processing analysis of the drawing database. By computing image features (e.g. colour palette, frequency information, multi-scale analysis, etc.), the student will perform a statistical analysis in order to cluster the drawings based on the most relevant drawing features. We will seek to assess whether the metadata associated with the drawings, such as the age or the geographical origin of the authors, can be inferred on the basis of the clustering process. The aim is to estimate the predictive capacity of clustering methods on a single database of children's drawings, focusing on cultural invariants.