The understanding of any image (that is its labelling) involves the matching of features extracted from the image with pre-stored models. The closer and more complete the match then the better understood the scene. The development and use of these models is then critical to the effectiveness of any computer vision system. The production of a high-level symbolic model requires the representation of knowledge about the objects to be modelled, their relationships, and how and when to use the information stored within the model. In the whole field of computer science, the data representation schema chosen is very important to ensure a good solution to the problem being tackled. In the field of artificial intelligence, and model based computer vision in particular, the choice of knowledge representation schema is even more important since the system's ability to know, update, perceive and understand is restricted by the content and structure of its knowledge base [1].