|
|
|
|
| Objectives |
| 2005-02-28 |
| |
Project objectives and completion criteria (pdf)
Looking at the major trends in the display technologies and television (B/W, Colour, HD) it is evident that the next step is the 3D. Even though the numerous developments in display technologies recently, even though the lot of improvements in the image quality, resolution, colour, contrast, etc., the view still looks artificial. The difference is the missing third dimension. Viewers should see a 3D image on the screen, as they would see in reality. There was a boom in using and manipulating 3D data in IT systems; still the weakest chain in the information flow remained the displaying that has not yet been solved properly up to now. 3D displays should provide the same level of functionality and freedom that current 2D displays offer while exceeding their capabilities. Systems that cause any optical discomfort or restrain the viewer will not be broadly accepted on the long term. The proposed system is based on well-proved hologram geometry principles and represents a high-end approach in the 3D displaying. Not limited by roadblocks in the principle and with the continuous technology development it has the potential to reach even the hologram quality displaying. When hanging on the wall, future displays should look like a real window, undistinguishable, except the technology working behind. The technology is here, and true 3D can be the most important development, a paradigm shift, in display technologies for the coming years.
The objective of HOLOVISION project is to develop a next generation holographic 3D display that overcomes the limitations of the current 3D displays, reconstructing natural 3D images to number of viewers in a reasonable field of view, with walk-around possibility without any restrictions. Moreover it will answer today's expectations with regard of resolution, brightness, contrast, highest possible fidelity-fidelity, good depth resolution and size.
The display will have a 16:9 aspect hologram screen with at least 50" diagonal. More than 125 million pixels will be controlled to build up a high-resolution 3D image. To form an idea about this value, however there are no standard terms accepted up to now for defining 3D resolution, it is important to emphasize that the number of addressable voxels (volume pixels) is orders of magnitude higher than the number of pixels since various combinations of 125 million pixels address different voxels. Due to the hologram geometry principles, where 2D is a special subset of 3D, the display will be fully compatible with 2D displays, able to show 2D images without the necessity of any switchover. With 2D terms the targeted image resolution is 1024 x 1920. This target is foreseen as a common longer standing value matching HDTV resolution or those equivalents to WSXGA. This will enable this technology to be among the candidates of potential display technologies for future 3DTV.
The display will provide a large field of view (FOV), where viewers do not have to be positioned. They will experience perfect 3D view with continuous horizontal motion parallax in the whole FOV from the proximity of the screen to far distances widening over 60 degrees.
The field of view is a very characteristic parameter of 3D displays, because wider view requires more data and with many systems it is limited by optical reasons to about only 20 degrees. When views are distributed to a wider range with these systems, the image jumps from a 2D view to the next 2D view and invalid zones are often wedged in.
The use of emerging technologies will be addressed in the development, like LCOS panels with fast switching speed in single panel configuration with unique polarization method, solid-state technology in the illumination based on high brightness LED-s, special micro-optical components, plastic aspheric & diffractive optical elements, redesigned holographic screen, and a high speed electronics based on cluster computing with proper 3D software solutions which help to integrate the 3D display to any kind of system and make it compatible with various formats and standards.
The use of high-speed LCOS microdisplay panels with excellent optical characteristic and LED based illumination promise a brilliant image quality, beyond being 3D, with regard to contrast, colour saturation, low-pixellation coming from colour sequential control, etc. In spite of the significant increase in LED efficiency, there remain still concerns regarding brightness in case of microdisplay-based systems. Due to the parallel character of the optical system in the proposed 3D display, it will be possible to make use of all the advantages offered by LED-s as light sources. Similarly, the distributed principle exploits the advantages of the multi-panel configuration i.e. microdisplays with physically the smallest size pixels and gets around the yield problem of defects that would be almost impossible at 100+ Mpixel systems. The device has a high pixel count (1024 x 1280), excellent optical uniformity and image quality.
In general, the very basic law to compose true 3D images is the use of orders of magnitude higher number of pixels than used for 2D. This amount of pixel/sec (or spot/sec) rate should be present in the 3D display; otherwise the image will be compromised some way. It is possible to generate 3D images by systems controlling large number of pixels simultaneously, theoretically it is also possible to reach this rate by higher speed devices, however we do not have the proper fast devices today.
In the HOLOVISION project we plan to realise a new principle: the combination of the two approaches. The parallel, spatially distributed system, that is the known modular arrangement, will be combined with novel temporal feature, enabling to optimise the system to fully exploit the capability of components and technologies used. The use of special spatial light modulators will be investigated, e.g. the high-speed digital imager FLCOS panel is capable of providing 1500 frame/sec. By using reconfigurable geometry at the illumination with specially designed optics and a control generating intermediate views accordingly, it will be possible to introduce virtual positions in the optical modules where light beams are emitted from, thus to go up with the angular resolution and increase the field of depth (FOD).
There are no standard illumination modules available that can be designed in to this application, so it will be necessary to develop an advanced solid-state custom illumination module. By virtue of their small size, efficient energy conversion, and fast switching speed, light emitting diodes (LEDs) offer the most attractive candidate technology. The objective will be to develop a suitably compact illuminator, which is specifically matched to the optical, electrical, mechanical and thermal requirements of the holo-display system. More specifically, the illuminator needs to be luminous enough for user brightness requirements, capable to offer enhanced colour and angular resolution depending on the control mode, fast enough to cope with the proposed high frame rates, robust and thermally integrated to the extent that the required operating lifetimes are achieved.
By these features it will be possible to realise different operational modes at the holo-display, additional functionalities beyond the normal mode by distributing resources as optimal to the view to be displayed. The display will be able to switch between a high-depth mode using less grey/colour scale steps, operating in a super colour-fidelity mode, or even in a high-brightness fine grey-scale achromatic mode. The optimising for the 3D image content, and a possible intelligent operational mode control for image quality enhancement is among the most exiting challenges in the development project.
The high information content of true 3D images relative to the 2D equivalents raises bandwidth / capacity issues at the transmission, storage, and processing of dynamic 3D content. Comprehensive future 3D standards will be based on effective compression technologies, making possible the integration of 3D hardware and software components into systems, providing interoperability, preserving compatibility with existing 2D/3D formats in 3DAV, Web3D, while open to future extensions. In the frame of the HOLOVISION project the development work will focus on 3D formats based on 3D compression algorithms that fulfil these criteria. We will compress 3D content using standard 2D compression algorithms like MPEGx, DivX, develop multi depth–map (Z-map) based representation, use inherent 3D transformations like 3D-wavelet, 3D model based descriptions (VRML, X3D) and set a frame for these in an open 3D protocol. Different approaches will be compared with regard to the achievable 3D image quality in high-end displaying, data-rate, compatibility, etc., and recommendations will be formulated contributing to future standards.
The consortium has centred the project work plan around continuous and detailed end-user involvement in the research, development, evaluation, and validation activities. The end-user will also play an important role in the dissemination and exploitation strategy. BAES ATC will develop scientific visualisation applications, which support collaborative decisions making for 3D environments. These provide complementary views of 3D data to individual observers. The proposed display provides a realisation of that facility. A common feature of a range of applications in urban and rural planning, geographical visualisation, etc., is the understanding of terrain, and it is the presentation of terrain views that form the test case which is to be developed refined with the feedback from users. The appreciation and understanding of the data will be compared with alternative visualisations using conventional displays.
.
The project will target not only challenging final deliverables, but also useful technology outcomes beyond the prototype phase. As a result of product oriented development, structuring the architecture of the holo-display to match existing product category manufacturing chains, like RPTV, easy replacement of the optical engine vs. 3D optical engine, the projection screen vs. holo-screen, will help to open up potential mass-markets. The research will be conducted against an ambitious, but achievable, 30-month schedule, to guarantee timely delivery, evaluation, and demonstration of tangible results. |
|
|
|
|
|
