Autostereoscopic 3D Display with Long Visualization Depth Using Referential Viewing Area-Based Integral Photography

Abstract

We developed an autostereoscopic display for distant viewing of 3D computer graphics (CG) images without using special viewing glasses or tracking devices. The images are created by employing referential viewing area-based CG image generation and pixel distribution algorithm for integral photography (IP) and integral videography (IV) imaging. CG image rendering is used to generate IP/IV elemental images. The images can be viewed from each viewpoint within a referential viewing area and the elemental images are reconstructed from rendered CG images by pixel redistribution and compensation method. The elemental images are projected onto a screen that is placed at the same referential viewing distance from the lens array as in the image rendering. Photographic film is used to record the elemental images through each lens. The method enables 3D images with a long visualization depth to be viewed from relatively long distances without any apparent influence from deviated or distorted lenses in the array. We succeeded in creating an actual autostereoscopic images with an image depth of several meters in front of and behind the display that appear to have 3D even when viewed from a distance.

4: Interactive Volume Rendering Method Using Dynamic Ray Casting for Autostereoscopic Display

“…The rendering kernel first calculates the corresponding subpixel coordinate of the current thread and then maps the subpixel coordinate to viewpoint index and normalized coordinate ( , ) on the imaging plane with conventional integral photography calculation [11,12], as shown in Figure 2. A virtual view frustum array is utilized to cast rays from different viewpoints.…”

Roadmap on 3D integral imaging: sensing, processing, and display

“…With the development of telemedicine and medical education, the 3D medical display is required to present a larger scene with long viewing depth. A computer-generated integral imaging elemental image generation method was proposed to achieve a long visualization depth [119].…”

Real-Time Rendering Method of Depth-Image-Based Multiple Reference Views for Integral Imaging Display

“…Several efficient CGII methods have been presented, such as point retracing rendering (PRR) [6], [7], multiple viewpoint rendering (MVR) [8], [9], parallel group rendering (PGR) [10], [11], viewpoint vector rendering (VVR) [12]- [14], image space parallel processing (ISPC) [15], [16], multiple ray cluster rendering (MRCR) [17], multiple orthographic frustum combing (MOFC) [18], and backward ray-tracing (BRT) [19], [20]. PRR is the simplest elementary image-mapping algorithm, but the heavy calculation required is time-consuming, which makes it unsuitable for real-time processing.…”