Image Processing Reference
In-Depth Information
rendering algorithms. Two examples of these formats are the representation of the
scene through the use of a plenoptic image and the use of layered depth video.
3.2.1 From Stereo to Multiview and Holoscopic Video
Providing a more immersive and closer to reality, multimedia experience to home-
users has attracted the attention of many researchers for a long time. Stereoscopic
3D systems provide an additional level of immersiveness relatively to 2D, due to
the enhanced depth perception obtained from the extra view presented to the user.
However, stereoscopic systems do not provide motion parallax, as the user is
presented always with the same perspective of the scene, independently of his/her
relative position to the display system. This limitation can be partially solved
through MVV systems that try to represent multiple perspectives of the same
scene, allowing the user to see a different perspective depending on his/her relative
position to the display system.
In order to overcome the disadvantages related to human factors of available 3D
video technologies, intense research in new types of 3D content representation has
continued. Due to recent advances in theory, microlens manufacturing, sensor
resolutions and display technologies, 3D holoscopic imaging, also known as inte-
gral imaging, light-field imaging, or plenoptic imaging, is becoming a practical
prospective 3D technology, which is able to create a more realistic depth illusion
than stereoscopic or multiview solutions and, thus, promises to become a popular
imaging technology in the future.
This section intends to review in a comprehensive way the basics of 3D content
representation, from stereoscopic to multiview and holoscopic representations.
3.2.1.1 Stereoscopic 3D Video
Generically, 3D video stands for any video format that supports the ability to
present “slightly” different motion images (views) of the visual scene to each of
the viewers
eyes. The binocular disparity induced by the different images reaching
each eye is used by the brain to provide enhanced depth perception [
3
].
This means that 3D video is usually denoted as a general video representation
conveying multiple views of the visual scene to the receiver. The particular case of
2-view video formats is commonly referred to as stereoscopic 3D video (S3DV or
just S3D) or simply stereoscopic video. Figure
3.1
illustrates a stereo image pair
and the corresponding difference image, where the horizontal disparities can be
easily observed.
S3DV is, thus, the simplest 3D video format both in terms of acquisition and also
in terms of transmission and display. The image separation at the display side can
be done in a variety of ways [
4
], being the active shutters (time multiplexing) and
the polarized glasses (spatial multiplexing) the most common in cinema and home
'
