Organisation internationale de normalisation
Yüklə 2,74 Mb.
|
6Explorations – Free Viewpoint Video/TelevisionThe exploratory work on free-viewpoint video has its roots in the “3DAV” exploration, which was originally started in December 2001, and later led in a first CfP on multiview video compression technology (current MVC development in JVT). As discussed in the previous meeting, FTV can be defined as a compressed representation and associated technologies which enable generating a large number of different views from a sparse view set. This most probably (from technologies currently known) requires implementation of depth/disparity map estimation (non-normative), definition of depth/disparity map representation/compression and interpolation/rendering method (not clear yet whether the latter should be non-normative or normative). All of these elements rely on each other, such that proper technology selection will most probably not be simple. Furthermore, higher distortion may be expected than for MVC (or at least quality may not be measurable in terms of pixel fidelity, geometric distortions may appear that might only be noticeable under certain observation conditions). The amount of distortion most probably would also depend on compactness (density of views) and complexity of the methods. Depending on concrete application, the view number to be generated may range from two for simple stereoscopic up to "many" for almost-free walk-through a scene. In Antalya, more clarification was achieved about the focus of the next phase of the FTV work. Realistic market needs appear to be existing in supporting standardized formats for upcoming 3D (M-view) displays where the number of views M as locally generated influences the quality of visual perception. While currently numbers of approximately M=9 are used in prototypes, while for the future M of up to 40-50 could be expected. Even then, the view angle support will be relatively narrow (max. 20 degrees left-to-right), which is a clear (and implementation-wise realistic) limitation as compared to the “really free” FTV scenario. One additional advantage could be that with such narrow view alteration ranges, co-planar camera setups could still be useful. After extensive discussion, the group agreed that the name of "3D Video" is very well reflecting this subset scenario of FTV (namely, enabling technology for 3D video displays). Related to the last meeting’s call for test sequences and depth/disparity estimation algorithms the following input contributions were brought:
In addition, M15101 reports corrected camera parameters for an existing sequence, M15047 and M15088 relate to more generic applications and requirements of depth map technology and FTV (no need detected to revise the apps & reqs document based on these contributions). M15088 indicated possibilities to perform skipped view encoding when higher-quality depth information is available. Following the focus as described above, it was concluded that co-planar camera setup would be the optimum case for this kind of 3D Video applications, and test sequences should be captured according to this. Nevertheless, slight rectification would most likely be necessary even in the parallel setup, due to slight variation in camera properties and the impossibility for perfect mechanical adjustment. Nevertheless, the original shots should be as close to co-planar as possible to keep rectification artifacts to a minimum. Sequences should be provided in rectified, illumination- and color-compensated version. Following this, a new call for test sequences was produced, which also includes a high-level description of 3DTV and FTV to make the purposes clear for which the new materials should be useful. Again, it is called to provide depth maps, depth estimation and interpolation software packages. To get more evidence about the possible elements of the 3D video chain, an Exploration Experiment was started to find out how depth estimation and interpolation inter-relate, based on the proposals that were brought (and for which software must be made available in this context). The results of this should bring evidence by the next meeting about how we can find out about
Documents reviewed
Output documents:
Yüklə 2,74 Mb. Dostları ilə paylaş:
|