Solving the Occlusion Problem for Three-dimensional Distortion-oriented Displays

Recent research into distortion-oriented displays (DODs) and non-linear magnification techniques has considered extending their application to large three-dimensional datasets. Inherent properties of three-dimensional datasets introduce some difficulties that do not occur in 2 0 environments. This paper considers the Occlusion Problem that of context data hiding, or occluding, some or all of the data within an area of focus, A novel solution to this problem is proposed, namely the use of nongeometric distortions combined with geometric distortions, producing a three-dimensional distortionoriented display that reduces obfuscation of data within the area of focus. An implementation of these techniques is developed, and its application to a small number of 3 0 datasets is examined as proof of concept. These traditional approaches and others like them, however, produce a different, but related problem: that of examining the area of focus while also maintaining context with the rest of the data. In simple zooming, for example, the user can either zoom in on an area of interest, or zoom out to look at the whole of the dataset in reduced detail. If there are multiple areas of the data that the user wishes to view simultaneously, perhaps for comparative purposes, this cannot be done. Likewise with the windowing interfaces: a user wanting to compare some point of interest in one area of the dataset with another similar point in a separate area of the data must navigate somehow through the rest of the data. Focus and context cannot both be maintained in displays using these approaches. For this reason, this problem has often been referred to as the Focus + Context Problem (e.g. [ll]).