Interactive projection images generation for swept-based 3D display

Abstract

Unifying different types of graphics file formats is one basic requirement for widespread adoption of volumetric 3D displays. In this paper a 3D data visualization and processing framework is presented which is capable of interfacing with legacy 3D software packages and providing compatibility to a wide range of true 3D display applications. Exploiting OpenGL technologies and the performance of graphics processing units, the proposal allows the user to visualize true 3D volume from different types of graphic file formats. The method can process constructed 3D models of different formats and decompose the models into a series of image slices with different thickness, angle of rotations and rotating axis in real-time to meet the projection requirements of the swept-volume display. Boolean operations and local view redrawing methods are employed to review the shape of the interior parts of the 3D objects, which can provide users with a better understanding of the complexity and spatial relationship inside the objects. The user can browse and interact with volumetric images in an intuitive manner. Approaches are also proposed to export the image slices that can be projected onto the rotating flat screen in sequence to appear in the volumetric 3D display. Two different kinds of image files are generated, and the results show that a vector image has better quality than a bitmap image in terms of file size and time-consumption. Finally, by developing a simulation platform, the procedure of forming geometric slices into a 3D spatial image can be realized to simulate the presentation effect of a volumetric 3D display.