Morphological analysis for investigating artistic images

a r t i c l e i n f o This paper describes an approach for automatically analyzing the alterations of an original artwork during its reproduction. The overall deformation of the artwork is modelled by a piecewise linear model, where regions of the artwork that feature similar alterations are automatically inferred and assigned to the different model components. Model complexity, that is, the required number of affine components required for registration, is automatically estimated using a statistical stability analysis. The main challenge is to simultaneously solve three tasks: (i) inferring the correspondences between both shapes, (ii) identifying the groups in the image that share the same transformation, and (iii) estimating the transformation of these groups. Our approach is tested on controlled scenarios as well as on real historical images. Although some stylistic movements in art like impressionism or pointillism define themselves by color, shape has been the predominant way to perceive an artwork. The theory about the primacy of shape can be traced back to Giorgio Vasari (1511–1574) who propagated the line drawing as the predominant technique of all visual arts. His use of the term disegno (conceptual design) can be read as the assignment of ideas to shapes. This " shaped idea " is represented through shapes in preparatory drawings, in the artwork itself as well as in drawn reproductions. Based on this observation, changes in shape between artworks and their reproductions or preparatory drawings can be associated with changes in ideas and concepts that reveal artistic choices and stylistic variations. Thus, the analysis of these changes helps art historians to understand the impact of historical influences in the creation and reproduction of art. However, in many cases, these alterations between shapes are very subtle and thus, it becomes extremely difficult, even for experts, to determine the nature and extent of the deformations suffered by different parts within an artwork. The automatic solution of such a shape analysis poses an ambitious computer vision task, and its solution is the focus of the present paper. The nature of the artwork deformations analyzed in this work arises either due to deliberate alterations or due to geometric errors accumulated during the drawing process. For instance, a typical example for a deliberate alteration between a preparatory drawing and the finished work is a subtle conceptual change that induces small alterations in the relative position of extremities in a human pose. These …