A. Campilho, M. Kamel (Eds.): ICIAR 2004, LNCS 3212, pp. 489–496, 2004. © Springer-Verlag Berlin Heidelberg 2004 Automatic Recognition of Impact Craters on the Surface of Mars Teresa Barata 1 , E. Ivo Alves 2 , José Saraiva 1 , and Pedro Pina 1 1 CVRM / Centro de Geo-Sistemas, Instituto Superior Técnico Av. Rovisco Pais, 1049-001 Lisboa, Portugal {tbarata, jsaraiva, ppina}@alfa.ist.utl.pt 2 Centro de Geofísica da Universidade de Coimbra Av. Dias da Silva 3000-134 Coimbra, Portugal livo@ci.uc.pt Abstract. This paper presents a methodology to automatically recognise impact craters on the surface of Mars. It consists of three main phases: in the first one the images are segmented through a PCA of statistical texture measures, fol- lowed by the enhancement of the selected contours; in a second phase craters are recognised through a template matching approach; in a third phase the rims of the plotted craters are locally fitted through the watershed transform. 1 Introduction Mars is currently the target of intensive exploration programs, with no less than three probes in orbit, and more to come in the future. Craters stand out visually among the features on any planetary surface, but their true importance stems from the kind of information that a detailed analysis of their number and morphology can bring forth. Evaluating the density of craters on differ- ent areas of the planet has led to the establishment of a large-scale stratigraphy for Mars [1], a matter still under refinement, as coverage of the entire surface with better resolving power continues to become available. The study of craters can also improve our knowledge of the cratering mechanism itself, as of the characteristics of the mate- rials targeted; furthermore, we can search for clues about the exogenous processes which cause the degradation of craters (with ejecta removal, ruin of walls and filling of floors) and play such an important role in defining the present character of the surface of Mars. Thus, it is only to be expected that craters are among the most studied of subjects when it comes to analysing planetary surfaces, and that the question of their auto- matic identification in images has been tackled in several studies. In many instances methods from one field are combined with others, in the search for the best answer. Though not an automated procedure, a word should be said about a NASA project, known as clickworkers, where laypeople were asked on the internet to mark the loca- tion of craters on grayscale images of the martian surface [2].