Applying Mathematical Morphology for the Classification of Iberian Ceramics from the Upper Valley of Guadalquivir River M. Lucena 1 , A.L. Mart´ ınez-Carrillo 2 , J.M. Fuertes 1 , F. Carrascosa 1 , and A. Ruiz 2 1 Department of Computer Science, University of Jaen, Spain mlucena@ujaen.es 2 Research University Institute for Iberian Archaeology, University of Jaen, Spain caai@ujaen.es Abstract. Although the potential of morphometrics for the study of archaeolog- ical artefacts is recognized, quantitative evaluation of the concordance between such methods and traditional typology and the potential of these techniques as supported methodologies in the archaeological analysis is a pending task. We present a new method to characterize and classify wheel-made pottery by its pro- file, using Mathematical Morphology. Each piece is represented as a vector, ob- tained by sampling the so called morphological curves (erosion, dilation, opening and closing), and Euclidean Distance is used as a similarity measure. The pro- posed technique has been studied using a sample of 1133 complete ceramic ves- sels from the Iberian archaeological sites from the upper valley of Guadalquivir River (Andalusia, Spain), showing that it is compatible with the existing corpus, established by experts. Keywords: Ceramic Profiles, Typologies, Shape Matching, Mathematical Morphology. 1 Introduction and Related Work The study and analysis of archaeological ceramics constitutes one of the most frequent activities of the archaeological work, which consists habitually of classifying the thou- sands of ceramic fragments gathered in the interventions and selecting those that con- tribute to deduce forms, functions and chronology [1]. The different criteria used in the elaboration of classifications do not contribute to homogenize the analysis of the pottery shapes, since the election of criteria depends on each researcher and moment [2]. In this sense, Shepard saw three phases in the election of criteria: the study of whole vessels as culture-objects; the study of sherds as dating evidence for stratigraphic sequences; the study of pottery technology as a way of relating more closely to the potter, but she did not try to put dates to them. Chronologically, the most used criteria have been artistically, typological, functional, technological, statistical, and contextual. Finally we can see developing interest in inte- grating ceramics into a wide analysis of finds assemblages. This must be the next step in ceramic studies: having integrated the various aspects of ceramics studies in the con- textual phase (1960 and after), we must now begin to integrate ceramic studies into the wider field of general finds assemblages. J.F. Mart´ ınez-Trinidad et al. (Eds.): MCPR 2014, LNCS 8495, pp. 341–350, 2014. c  Springer International Publishing Switzerland 2014