Image Processing Reference
In-Depth Information
4.2 Full Parallax 3D Content
A light-field representation of a scene can be obtained by using several images
captured from different points/angles of view. We consider here two technologies
that can provide full parallax content: Super Multi-view Video (SMV) and integral
imaging, and both are associated with a specific acquisition and display technology
and with a specific representation.
4.2.1 Acquisition Technology and Resulting Content
4.2.1.1 Camera Arrays
Super multi-view content can be captured using a camera rig as illustrated in
Fig. 4.1 . The cameras can be aligned horizontally (providing horizontal motion
parallax only) or in horizontal and vertical dimensions in the case of full motion
parallax content. The camera system can be arranged in linear, arc, or even sparse
alignment. Each camera captures the scene from a different point of view and the
resulting content consists of several viewpoint images with horizontal and vertical
disparities.
4.2.1.2
Integral Imaging
Integral imaging (or holoscopy) is a technology based on plenoptic photography
[ 7 ]. Integral imaging acquisition uses a lenticular array set in front of a camera
device as illustrated in Fig. 4.2 . This lenticular array is composed of a large number
of micro-lenses, which can have a round, square or hexagonal shape, and can be
aligned in rectangular grid or in quincunx. The resulting captured holoscopic image
consists of an array of Micro-Images (MIs, also called elemental images) as
illustrated in Fig. 4.3 . Each micro-lens produces one MI, and each MI contains
Fig. 4.1 Super Multi-View acquisition using a camera rig
Search WWH ::




Custom Search