Automatic computation of best views of objects are very useful. For example, they can be used as the starting point of a scene exploration, or to enrich galleries of objects available through Internet by adding an image a model that may help to decide if it is worth downloading. Recently, a measure to evaluate the quality of a view has been proposed, it is called viewpoint entropy and has a basis on Information Theory. The best view is the one which gives the most information of the object being inspected. For large models, the selection of a set of good views that cover all the faces can aid the user to understand a certain object or scene. For very complex environments, however, a set of images may not suffice. For example, when examining buildings it may be difficult to locate in space the positions where the viewpoints were placed. In these cases, an interactive exploration of the model can better help the user to understand the structure of the scene. In this paper we present an automatic method for the exploration of scenes that uses viewpoint entropy.