Stereoplexing for film and video applications

• US 20080303896A1
• Filed: 06/07/2007
• Published: 12/11/2008
• Est. Priority Date: 06/07/2007
• Status: Active Grant

• Current Assignee: RealD Incorporated
• Original Assignee: Real D
• Inventors: Lipton, Lenny, Akka, Robert
• Granted Patent: US 8,487,982 B2
• Time in Patent Office: Days
• US Class Current: 348/49
• CPC Class Codes: H04N 13/139 Format conversion, e.g. of frame-rate or size
• CPC Class Codes: H04N 13/161 Encoding, multiplexing or demultiplexing different image signal components for multi-view video sequence encoding H04N19/597

Abstract

A method for multiplexing a stream of stereoscopic image source data into a series of left images and a series of right images combinable to form a series of stereoscopic images, both the stereoscopic image source data and series of left images and series of right images conceptually defined to be within frames. The method includes compressing stereoscopic image source data at varying levels across the frame, thereby forming left images and right images, and providing a series of single frames divided into portions, each single frame containing one right image in a first portion and one left image in a second portion. Alternately, single frames may contain two right images in a first two portions of each single frame and two left images in a second two portions of each single frame, wherein each set of right and left images may be processed differently. Multiplexing processes such as staggering, alternating, filtering, variable scaling, and sharpening from original, uncompressed right and left images may be employed.

Claims

1. A method for multiplexing a stream of stereoscopic image source data into a series of left images and a series of right images combinable to form a series of stereoscopic images, both the stereoscopic image source data and series of left images and series of right images conceptually defined to be within frames, the method comprising:

compressing stereoscopic image source data at varying levels across the frame, thereby forming left images and right images; and

providing a series of single frames divided into portions, each single frame containing one right image in a first portion and one left image in a second portion.

2. The method of claim 1, wherein said compressing comprises less compression at a center of the frame of stereoscopic image source data.

3. The method of claim 1, wherein the one left image is associated with the one right image.

4. The method of claim 1, wherein the portions are included within each single frame, and wherein each left image is included in substantially one half of each single frame and each right image is included in substantially another half of each single frame.

5. The method of claim 3, wherein each single frame is divided by a vertically oriented border, with one right image on a first side of the single frame and one left image on a second side of the single frame.

6. The method of claim 1, wherein each left image is based on an unprocessed left image and each right image is based on an unprocessed right image, and the method further comprises:

removing every alternate pixel from every row in the uncompressed right image to form the right image and every alternate pixel from every row in the uncompressed left image to form the left image.

7. The method of claim 1, wherein each left image is based on an unprocessed left image and each right image is based on an unprocessed right image, and the method further comprises:

mapping pixel information from the unprocessed left image to the left image based on information from a plurality of pixels from the unprocessed left image and mapping pixel information from the unprocessed right image to the right image based on information from a plurality of pixels from the unprocessed right image.

8. The method of claim 1, further comprising applying a process to the left image and the right image, the process selected from at least one from a group comprising:

variably applied bicubic scaling;

variable sharpening;

non-variably applied bicubic scaling; and

non-variable sharpening.

9. The method of claim 5, wherein removal of alternate pixels in each row comprises removing a first set of pixels from a first row and a different set of pixels from each row adjacent to the first row.

10. The method of claim 8, wherein removal substantially forms a “

checkerboard”

type pattern of pixels being removed from the unprocessed right image and the unprocessed left image.

11. The method of claim 1, wherein each left image is based on an unprocessed left image and each right image is based on an unprocessed right image, and the method further comprises:

mapping pixel information from the unprocessed left image to the left image by combining more pixels from predetermined areas of the unprocessed left image than other areas of the unprocessed left image, and mapping pixel information from the unprocessed right image to the right image by combining more pixels from predetermined areas of the unprocessed right image than other areas of the unprocessed right image.

12. The method of claim 1, wherein each left image is based on an unprocessed left image having comprising alternating rows and remaining rows of pixels and each right image is based on an unprocessed right image comprising alternating and remaining rows of pixels, and the method further comprises:

mapping pixel information from the unprocessed left image to the left image by combining pixels from alternating rows of the unprocessed left image and selecting pixels from remaining rows of the unprocessed left image, and mapping pixel information from the unprocessed right image to the right image by combining pixels from alternating rows of the unprocessed right image and selecting pixels from remaining rows of the unprocessed right image.

13. The method of claim 1, wherein each left image is based on an unprocessed left image having comprising alternating rows and remaining rows of pixels and each right image is based on an unprocessed right image comprising alternating and remaining rows of pixels, and the method further comprises:

mapping pixel information from the unprocessed left image to the left image by combining at least two pixels from first positions in alternating rows of the unprocessed left image and combining at least two pixels from second positions in remaining rows of the unprocessed left image, and mapping pixel information from the unprocessed right image to the right image by combining at least two pixels from first positions in alternating rows of the unprocessed right image and combining at least two pixels from second positions in remaining rows of the unprocessed right image.

14. The method of claim 1, wherein each left image and each right image is based on at least one unprocessed image, and the method further comprises:

performing a process on at least one unprocessed image to reduce image size to form at least one resultant image;

dividing each resultant image into a plurality of blocks; and

rearranging the plurality of blocks to fit within a predefined space smaller than each unprocessed image.

15. A method for multiplexing a stream of stereoscopic image source data into a series of left images and a series of right images combinable to form a series of stereoscopic images, the method comprising:

compressing stereoscopic image source data; and

providing a series of single frames divided into portions, each single frame containing two right images in a first two portions of each single frame and two left images in a second two portions of each single frame.

16. The method of claim 15, wherein the each left image in the single frame is associated one right image in the single frame.

17. The method of claim 15, wherein the portions are included within each single frame, and wherein each left image is included in substantially one fourth of each single frame and each right image is included in substantially one fourth of each single frame.

18. The method of claim 17, wherein each single frame is divided by a vertically oriented border and a horizontally oriented border, with two right side images on a first segment of the single frame and two left images on a second segment of the single frame.

19. The method of claim 15, wherein each left image is based on an unprocessed left image and each right image is based on an unprocessed right image, and each right image is created from an associated unprocessed right image using a process different from processing used to create each other right image, and each left image is created from an associated unprocessed left image using a process different from processing used to create each other left image.

20. The method of claim 15, wherein each left image is based on an unprocessed left image and each right image is based on an unprocessed right image, and the method further comprises:

removing every alternate pixel from every row in one uncompressed right image to form one right image and every alternate pixel from every row in one uncompressed left image to form one left image.

21. The method of claim 15, wherein each left image is based on an unprocessed left image and each right image is based on an unprocessed right image, and the method further comprises:

mapping pixel information from one unprocessed left image to one left image based on information from a plurality of pixels from the unprocessed left image and mapping pixel information from the unprocessed right image to the right image based on information from a plurality of pixels from the unprocessed right image.

22. The method of claim 15, further comprising applying a process to one left image and one right image, the process selected from at least one from a group comprising:

variably applied bicubic scaling;

variable sharpening;

non-variably applied bicubic scaling; and

non-variable sharpening.

23. The method of claim 20, wherein removal of alternate pixels in each row comprises removing a first set of pixels from a first row and a different set of pixels from each row adjacent to the first row.

24. The method of claim 23, wherein removal substantially forms a “

checkerboard”

type pattern of pixels being removed from one unprocessed right image and one unprocessed left image.

25. The method of claim 15, wherein each left image is based on an unprocessed left image and each right image is based on an unprocessed right image, and the method further comprises:

mapping pixel information from one unprocessed left image to one left image by combining more pixels from predetermined areas of one unprocessed left image than other areas of one unprocessed left image, and mapping pixel information from one unprocessed right image to one right image by combining more pixels from predetermined areas of one unprocessed right image than other areas of one unprocessed left image.

26. The method of claim 15, wherein each left image is based on an unprocessed left image having comprising alternating rows and remaining rows of pixels and each right image is based on an unprocessed right image comprising alternating and remaining rows of pixels, and the method further comprises:

mapping pixel information from the unprocessed left image to the left image by combining pixels from alternating rows of the unprocessed left image and selecting pixels from remaining rows of the unprocessed left image, and mapping pixel information from the unprocessed right image to the right image by combining pixels from alternating rows of the unprocessed right image and selecting pixels from remaining rows of the unprocessed right image.

27. The method of claim 15, wherein each left image is based on an unprocessed left image having comprising alternating rows and remaining rows of pixels and each right image is based on an unprocessed right image comprising alternating and remaining rows of pixels, and the method further comprises:

mapping pixel information from the unprocessed left image to the left image by combining at least two pixels from first positions in alternating rows of the unprocessed left image and combining at least two pixels from second positions in remaining rows of the unprocessed left image, and mapping pixel information from the unprocessed right image to the right image by combining at least two pixels from first positions in alternating rows of the unprocessed right image and combining at least two pixels from second positions in remaining rows of the unprocessed right image.