Extraction of 3D Data Through Single-Lens SLR Camera System, based on photostereosynthesis

Garnet Hertz, October 2008


This project involves extracting 3D data through a single-lens SLR camera system. For information on this camera technology, see http://www.conceptlab.com/photostereosynthesis/cameracontroller.html. The single-lens camera system and custom camera controller would be used to extract 3D data through multiple still images shot at a sequence of increasing focal lengths at extremely large apertures. Because of the short (1cm) region of focus, multiple exposures have the effect of scanning the depth of a scene or object, detecting the distance of surfaces by analyzing regions of sharpness.



In addition to its merit as a "special effects" camera, the goal of this research would be to develop a low cost system to acquire 3D data with consumer digital cameras. I propose that post-processing (edge detection, etc.) and deletion of non-focused areas at large aperture through incremental focal lengths could be used to construct simple 3D models. This would be similar in some ways to Sauter & L├╝sebrink's http://www.artcom.de/images/stories/2_pro_invisibleshape/invisibleshape_lrg.mov - but would have a stationary camera and not "cut out" a square, but edge-detected shapes.



In addition to dimensional data, shooting automated frames at varying focal lengths has potential to be explored in the construction of 2D images in a similar manner to High Dynamic Range (HDR) photographs.

See http://www.conceptlab.com/photostereosynthesis/ for background, sketches, and other proposals related to this project



[Garnet Hertz, 2008]