In this article we investigate the following problem: Find an appropriate test which based on a spatially restricted observation of a (generalized) germ-grain model allows to decide whether interpretration of the grains of this set is possible or not. For two special hypotheses different test statistics are proposed which are useful under particular interpretations of the problem. Some of their properties are given. A conditional test based on the generalized likelihood ratio principle leads to a very simple determination of the critical values and of the power function.