Image Processing Reference
In-Depth Information
5.2 Acquisition and Display Systems for 3D Holoscopic
Content
3D holoscopic imaging derives from the fundamental concepts of light-field imag-
ing where, by using conventional 2D acquisition and display devices enhanced by
special optical systems, it is possible to capture not only spatial information about a
scene but also angular information, i.e., the “whole observable” (holoscopic) scene.
This section provides an overview of the principles behind 3D holoscopic imaging
and discusses the advantages and current limitations of this approach for acquisition
and display.
5.2.1
3D Holoscopic Imaging Acquisition
The concept behind 3D holoscopic imaging is based on the principle of integral
photography, proposed by Gabriel Lippmman in 1908 [
1
]. Basically, the 3D
holoscopic imaging system comprises a regularly spaced array of small spherical
micro-lenses, known as a “fly
'
s eye” lens array [
2
], which is used for both
acquisition and display of the 3D holoscopic images, as can be seen in Fig.
5.1
.
At the acquisition side (Fig.
5.1a
), the light beams coming from a given object
with various incident angles are firstly refracted through the micro-lens array to be,
then, captured by the 2D image sensor. Hence, each micro-lens works as an
individual small low resolution camera conveying a particular perspective of the
3D object at slightly different angles. As a result, the planar light intensity distribution
representing a 3D holoscopic image corresponds to a 2D array of micro-images, as
illustrated in Fig.
5.2
, where both light intensity and directional information are
recorded.
Apart from some technological limitations that still need to be overcome in this
kind of systems [
2
], there have been several camera setups seeking to improve the
performance of this technology, as can be seen in [
2
-
4
]. Moreover, various advan-
tages of employing a 3D holoscopic system for the acquisition of 3D content have
been identified, notably:
• A single aperture camera system is used, without needing complex calibration
and synchronization procedures between several cameras;
• A high level of display scalability is supported, since it is possible to generate in
a seamless way 2D, multiview, as well as full 3D holoscopic, from a 3D
holoscopic acquisition setup (see Sect.
5.3.4
);
• New degrees of freedom for content production become possible at the post-
processing stage, such as refocusing and changing perspectives (see Sect.
5.3.4
).
