Implementing an Improved Stereoscopic Camera Model

Abstract

The usable perceived depth range of all stereoscopic 3D displays is limited by human factors considerations to a bounded range around the plane of the display. To account for this our Three Region stereoscopic camera model is able to control the depth mapping from scene to display while allowing a defined region of interest in scene depth to have an improved perceived depth representation compared to other regions of the scene. This can be categorized as a focus+context algorithm that manipulates stereoscopic depth representation along the viewing axis of the camera. We present a new implementation of the Three Region stereoscopic camera model as a Utility plug-in for the popular modelling and rendering package 3ds max. We describe our user interface, designed to incorporate stereoscopic image generation into the user s natural work flow. The implementation required us to overcome a number of technical challenges including; accurately measuring scene depth range, simulating asymmetric camera frustum in a system only supporting symmetric frustum, merging multiple renderings and managing anti-aliasing in layered images. We conclude from our implementation that it is possible to incorporate high quality stereoscopic camera models into standard graphics packages.