Method and processor for stereo cylindrical imaging

• US 6,650,396 B2
• Filed: 11/28/2001
• Issued: 11/18/2003
• Est. Priority Date: 12/04/2000
• Status: Expired due to Fees

• Current Assignee: Hytechnology Incorporated
• Original Assignee: Hytechnology Incorporated
• Inventors: Li, Ming
• Primary Examiner(s): Adams, Russell
• Assistant Examiner(s): ESPLIN, DAVID B
• Application Number:
  US09/997,123
• Publication Number:
  US 20020067465A1
• Time in Patent Office: 720 Days
• Field of Search: 352/61, 352/58, 352/57, 359/458, 353/7
• US Class Current: 352/58
• CPC Class Codes: G03B 35/00 Stereoscopic photography
• CPC Class Codes: H04N 13/275 from 3D object models, e.g. computer-generated stereoscopic image signals
• CPC Class Codes: H04N 13/286 having separate monoscopic and stereoscopic modes
• CPC Class Codes: H04N 13/349 Multi-view displays for displaying three or more geometrical viewpoints without viewer tracking for viewing without the aid of special glasses using fly-eye lenses H04N13/307
• CPC Class Codes: H04N 13/363 using image projection screens volumetric display H04N13/388
• CPC Class Codes: H04N 2013/0088 Synthesising a monoscopic image signal from stereoscopic images, e.g. synthesising a panoramic or high resolution monoscopic image

Abstract

Computerized method, processor and computer-readable medium are provided for stereoscopically and seamlessly imaging on a generally cylindrical screen. The method allows to configure imaging data for a plurality of selectable sub-screens encompassing the cylindrical screen. Each sub-screen includes a plurality of selectable strips axially extending relative to a corresponding sub-screen in the cylindrical screen. The method further allows to render two channels imaging data increments for each strip based on a parallax model with a reference line selectively rotatable to be in parallel alignment relative to each respective plane defined by each strip. The method allows to combine each imaging data increment corresponding to each strip in a respective sub-screen to generate stereoscopic imaging data for that respective sub-screen, and further combines each sub-screen imaging data to generate stereoscopic imaging data substantially encompassing the angular span of the cylindrical screen.

Claims

1. A computerized method for stereoscopically and seamlessly imaging on a generally cylindrical screen, the method comprising:

configuring imaging data for a plurality of selectable sub-screens encompassing the cylindrical screen each sub-screen comprising a plurality of selectable strips axially extending relative to a corresponding sub-screen in the cylindrical screen;

rendering two channels of imaging data increments for each strip based on a parallax model with a reference line selectively rotatable to be in parallel alignment relative to each respective plane defined by each strip;

rectifying the imaging data for each corresponding sub-screen by re-rendering said imaging data relative to a computer-synthesized cylinder configured as the cylindrical screen, and texture-mapping the originally rendered sub-screen imaging data on the computer-synthesized cylinder to generate new sub-screen imaging data substantially free of distortion;

combining each imaging data increment corresponding to each strip in a respective sub-screen to generate stereoscopic imaging data for that respective sub-screen; and

combining each sub-screen imaging data to generate stereoscopic imaging data substantially encompassing the annular span of the cylindrical screen.

2. The stereoscopic imaging method of claim 1 further comprising seamlessly blending imaging data between adjacent sub-screens by applying, during an imaging-production stage, a predefined attenuation function over an area selected for overlapping between any two adjacent sub-screens.

3. The stereoscopic imaging method of claim 2 further comprising, during a playback stage, overlapping the two adjacent sub-screens over the selected area so that pixels in that area are selectively attenuated based on attenuation values determined by the applied attenuation function.

4. A computerized stereoscopic imaging method comprising:

configuring a cylindrical screen into N sub-screens, each sub-screen comprising M planar strips axially extending along a corresponding sub-screen;

positioning at each end point of a reference line a respective computer-synthesized camera, each providing a respective field view for each imaging channel;

rendering M increments of imaging data for each imaging channel, each increment of imaging data uniquely corresponding to a respective strip on each sub-screen; and

combining each of the M increments of imaging data for each of the N sub-screens to seamlessly form cylindrical imaging data to be projected on the cylindrical screen, wherein M and N represent respective positive integer numbers.

5. The imaging method of claim 4 wherein the rendering of the M increments of imaging data for each sub-screen is based on a parallax model using the reference line with each computer synthesized camera, said reference line selectively rotatable to be in parallel alignment relative to each respective plane defined by each strip in a corresponding sub-screen, and wherein selective rotation of the reference line comprises incrementally rotating the reference line by an angle based on the following equation:

$\frac{\mathrm{Cylindrical\_Span}}{M*N},$

wherein “

Cylindrical Span”

represents the angular span of the cylindrical screen.

7. The imaging method of claim 4 further comprising seamlessly blending imaging data between adjacent sub-screens by applying, during an imaging-production stage, a predefined attenuation function over an area selected for overlapping between any two adjacent sub-Screens, and, during a playback stage1 overlapping the two adjacent sub-screens over the selected area so that pixels in that area are selectively attenuated based on attenuation values determined by the applied attenuation function.

6. A computerized stereoscopic imaging method comprising:

configuring a cylindrical screen into N sub-screens, each sub-screen comprising M planar strips axially extending along a corresponding sub-screen;

rendering M increments of imaging data for each imaging channel, each increment of imaging data uniquely corresponding to a respective strip on each sub-screen;

re-rendering imaging data originally rendered for each imaging channel, said re-rendering performed relative to a computer-synthesized cylinder configured as the cylindrical screen, and texture-mapping the originally rendered sub-screen imaging data to generate new sub-screen imaging data substantially free of distortion; and

combining each of the M increments of imaging data for each of the N sub-screens to seamlessly form cylindrical imaging data to be projected on the cylindrical screen, wherein M and N represent respective positive integer numbers.

8. A stereoscopic cylindrical screen imaging method comprising:

a) gathering en externally-derived three-dimensional scene model;

b) selecting a plurality of processing parameters selected from the group consisting of angular span of the cylindrical screen, a number of sub-screens for the cylindrical screen, and a number of axially-extending planar strips for each sub-screen;

c) providing a pair of computer-synthesized cameras at the respective end points of a reference line, with each camera providing a respective view of the three-dimensional model for each imaging channel;

d) positioning the mid-point of the reference line to substantially correspond with the center of the cylinder;

e) aligning the reference line to be generally parallel relative to the plane defined by a first strip of the sub-screen;

f) generating a stereoscopic increment of imaging data for the first strip of the sub-screen;

g) rotating the reference line together with the camera pair about the center of the cylinder to be in general parallel alignment relative to the plane defined by the next strip of the sub-screen;

h) generating a stereoscopic increment of imaging data for said next strip of the sub-screen;

i) iteratively continuing with steps g) and h) until imaging data has been rendered for each strip in each of the sub-screens;

j) combining each incremental imaging data to generatel imaging data for each sub-screen; and

k) storing each sub-screen imaging data for additional processing.

9. The imaging method of claim 8 wherein the additional processing comprises re-rendering the stored sub-screen imaging data to generate new sub-Screen imaging data substantially free of distortion.

10. The imaging method of claim 9 wherein said re-rendering comprises:

rendering the three-dimensional scene model relative to a computer-synthesized cylinder configured as the cylindrical screen;

providing N computer-synthesized cameras, wherein each camera has a field of view configured to correspond to the field of view of each projector to be used for the cylindrical screen, and wherein N is an integer indicative of the number of sub-screens;

mapping each frame of the originally rendered imaging data of a first imaging channel for each sub-screen on the computer-synthesized cylinder,

mapping each frame of the originally rendered imaging data of a second imaging channel for each sub-screen on the computer-synthesized cylinder; and

storing each re-rendered sub-screen imaging data for additional processing.

11. The imaging method of claim 8 wherein said additional processing of the re-rendered sub-screen imaging data comprises seamlessly blending imaging data between adjacent sub-screens by applying, during an imaging-production stage, a predefined attenuation function over an area selected for overlapping between any two adjacent sub-screens, and, during a playback stage, overlapping the two adjacent sub-screens over the selected area so that pixels in that area are selectively attenuated based on attenuation values determined by the applied attenuation function.

12. A processor for stereoscopically and seamlessly imaging on a generally cylindrical screen, the processor comprising:

a module for configuring imaging data for a plurality of selectable sub-screens encompassing the cylindrical screen, each sub-screen comprising a plurality of selectable strips axially extending relative to a corresponding sub-screen in the cylindrical screen;

a rendering-module for rendering two channels of imaging data increments for each strip based on a parallax model with a reference line selectively rotatable to be in parallel alignment relative to each respective plane defined by each strip;

a rectifier module configured to rectify the imaging data for each corresponding sub-screen by re-rendering said imaging data relative to a computer-synthesized cylinder configured as the cylindrical screen, and texture-mapping the originally rendered sub-screen imaging data on the computer-synthesized cylinder to generate new sub-screen imaging data substantially free of distortion; and

a combining-module for combining each imaging data increment corresponding to each strip in a respective sub-screen to generate stereoscopic imaging data for that respective sub-screen, the combining module further for combining each sub-screen imaging data to generate stereoscopic Imaging data substantially encompassing the angular span of the cylindrical screen.

13. The processor of claim 12 further comprising a module for seamlessly blending imaging data between adjacent sub-screens by applying, during an imaging-production stage, a predefined attenuation function over an area selected for overlapping between any two adjacent sub-screens.

14. The processor of claim 13 further comprising a playback processor configured to overlap during a playback stage the two adjacent sub-screens over the selected area so that pixels in that area are selectively attenuated based on attenuation values determined by the applied attenuation function.

15. A computer-readable medium including instructions for causing a computer to process stereoscopic seamless imaging data for a cylindrical screen by:

arranging cylindrical screen into N sub-screens each sub-screen comprising M planar strips axially extending along a corresponding sub-screen;

positioning at each end point of a reference line a respective computer-synthesized camera, each providing a respective field view for each Imaging channel;

rendering M increments of imaging data for each imaging channel each increment of imaging data uniquely corresponding to a respective strip on each sub-screen; and

combining each of the M increments of imaging data for each of the N sub-screens to seamlessly form cylindrical imaging data to be projected on the cylindrical screen, wherein M and N represent respective positive integer numbers.

16. The computer-readable medium of claim 15 wherein the rendering of the M increments of imaging data for each sub-screen is based on a parallax model using the reference line with each computer synthesized camera, said reference line selectively rotatable to be in parallel alignment relative to each respective plane defined by each strip in a corresponding sub-screen, and wherein selective rotation of the reference line comprises incrementally rotating the reference line by an angle based on the following equation:

$\frac{\mathrm{Cylindrical\_Span}}{M*N},$

wherein “

Cylindrical Span”

represents the angular span of the cylindrical screen.

18. The computer-readable medium of claim 15 further comprising seamlessly blending imaging data between adjacent sub-screens by applying, during an imaging-production stage, a predefined attenuation function over an area selected for overlapping between any two adjacent sub-screens, and, during a playback stage, overlapping the two adjacent sub-screens over the selected area so that pixels in that area are selectively attenuated based on attenuation values determined by the applied attenuation function.

17. A computer-readable medium including instructions for causing a computer to process stereoscopic seamless imaging data for a cylindrical screen by:

arranging a cylindrical screen into N sub-screens, each sub-screen comprising M planar strips axially extending along a corresponding sub-screen;

rendering M increments of imaging data for each imaging channel, each increment of imaging data uniquely corresponding to a respective strip on each sub-screen;

re-rendering imaging data originally rendered for each imaging channel, said re-rendering performed relative to a computer-synthesized cylinder configured as the cylindrical screen, and texture-mapping the originally rendered sub-screen imaging data to generate new sub-screen imaging data substantially free of distortion; and

combining each of the M increments of imaging data for each of the N sub-screens to seamlessly form cylindrical imaging data to be projected on the cylindrical screen, wherein M and N represent respective positive integer numbers.

19. A computerized stereoscopic imaging method comprising:

configuring a cylindrical screen into N sub-screens, each sub-screen comprising M planar strips axially extending along a corresponding sub-screen;

rendering M increments of imaging data for each imaging channel, each increment of imaging data uniquely corresponding to a respective strip on each sub-screen;

re-rendering imaging data originally rendered for each imaging channel, said re-rendering performed relative to a computer-synthesized cylinder configured as the cylindrical screen, and texture-mapping the originally rendered sub-screen imaging data to generate new sub-screen imaging data substantially free of distortion;

blending imaging data between adjacent sub-screens by applying, during an imaging-production stage, a predefined attenuation function over an area selected for overlapping between any two adjacent sub-screens, and, during a playback stage, overlapping the two adjacent sub-screens over the selected area so that pixels in that area are selectively attenuated based on attenuation values determined by the applied attenuation function; and

combining each of the M increments of imaging data for each of the N sub-screens to seamlessly form cylindrical imaging data to be projected on the cylindrical screen, wherein M and N represent respective positive integer numbers.

20. A computer-readable medium including instructions far causing a computer to process stereoscopic seamless imaging data for a cylindrical screen by:

arranging a cylindrical screen into N sub-screens, each sub-screen comprising M planar strips axially extending along a corresponding sub-screen;

positioning at each end point of a reference line a respective computer-synthesized camera, each providing a respective field view far each imaging channel;

rendering M increments of imaging data for each imaging channel, each increment of imaging data uniquely corresponding to a respective strip on each sub-screen;

blending imaging data between adjacent sub-screens by applying, during an imaging-production stage, a predefined attenuation function over an area selected for overlapping between any two adjacent sub-screens, and, during a playback stage, overlapping the two adjacent sub-screens over the selected area so that pixels in that area are selectively attenuated based on attenuation values determined by the applied attenuation function; and

combining each of the M increments of imaging data for each of the N sub-screens to seamlessly form cylindrical imaging data to be projected on the cylindrical screen, wherein M and N represent respective positive integer numbers.