STEREOSCOPIC IMAGE DISPLAY SYSTEM, DISPARITY CONVERSION DEVICE, DISPARITY CONVERSION METHOD AND PROGRAM

• US 20120148147A1
• Filed: 05/25/2011
• Published: 06/14/2012
• Est. Priority Date: 06/07/2010
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Ogata, Masami, Morifuji, Takafumi, Ushiki, Suguru
• Granted Patent: US 8,605,994 B2
• Time in Patent Office: Days
• US Class Current: 382/154
• CPC Class Codes: G06T 19/00 Manipulating 3D models or images for computer graphics
• CPC Class Codes: H04N 13/128 Adjusting depth or disparity
• CPC Class Codes: H04N 2013/0081 Depth or disparity estimation from stereoscopic image signals

Abstract

Disparity in a stereoscopic image is converted, according to features of a configuration element of an image that influences depth perception of a stereoscopic image. A disparity detecting unit 110 detects disparity from a left image L and right image R of an input image, and generates a disparity map dM. A disparity correction unit 150 corrects the disparity in the disparity map dM and generates a corrected disparity map dM′. A correction feature setting unit 130 sets the correction features in the event of performing disparity correction in the disparity correction unit 150. The image synthesizing unit 160 synthesizes the left image L and right image R of the stereoscopic image based on the corrected disparity map dM′ and outputs the stereoscopic image made up of a left image L′ and right image R′ as an output image. Thus, a stereoscopic image having disparity according to the set correction features is output. The correction features therein are set according to the degree that the configuration element features such as size of disparity, position on screen, manner of special change and so forth influence the depth perception.

Claims

1. A disparity conversion device comprising:

a disparity detecting unit that detects disparity from a left image and right image of an input image, and generates a disparity map that holds the disparity for each pixel or pixel group;

a correction feature setting unit that sets correction features in the event of correcting the disparity in said disparity map;

a disparity correction unit that corrects the disparity in said disparity map according to said set correction features and generates a corrected disparity map; and

an image synthesizing unit that synthesizes a left image and right image of an output image from the left image and right image of said input image according to said corrected disparity map.

2. The disparity conversion device according to claim 1, further comprising:

a region setting unit that sets at least two regions in said disparity map;

wherein said correction feature setting unit sets said correction features by each of said at least two regions;

and wherein said disparity correction unit corrects the disparity in said disparity map according to said correction feature according to said at least two regions in said disparity map.

3. The disparity conversion device according to claim 2, wherein said disparity correction unit corrects the disparity in said disparity map according to a weighted sum of said correction features according to adjacent regions, in predetermined buffer regions of which said at least two regions are adjacent.

4. The disparity conversion device according to claim 2, wherein said correction feature setting unit sets said correction features so that, in a region corresponding to the left and right edges of said disparity map of said at least two region, depth in the forward direction is suppressed.

5. The disparity conversion device according to claim 1, wherein said correction feature setting unit sets said correction features for each of at least two components in said disparity map;

said disparity correction unit including

a component dividing unit that divides said disparity map into each of said component,

a component disparity correction unit that corrects the disparity by component in said disparity map according to said correction feature set for each of said components, and

a component synthesizing unit that synthesizes the disparity map which has been corrected for each of said components and generates said corrected disparity map.

6. The disparity conversion device according to claim 5, wherein said components set with said correction features setting unit are components according to the disparity frequency in said disparity map.

7. The disparity conversion device according to claim 1, wherein said correction features set with said correction feature setting unit are non-linear features wherein the slope is 1 where the disparity in said disparity map nears zero, and converges into a predetermined value as the disparity in said disparity map increases.

8. The disparity conversion device according to claim 7, wherein said correction feature set with said correction features setting unit is defined based on a sigmoid function.

9. The disparity conversion device according to claim 1, wherein said image synthesizing unit comprises:

a center monocular image synthesizing unit that synthesizes a center monocular image from the left image and right image of said input image based on disparity of said disparity map;

a disparity/depth conversion unit that converts from disparity in said corrected disparity map into depth and generates a corrected depth map;

a three-dimensional mapping unit that generates a corrected three-dimensional map by mapping said center monocular image in a three-dimensional space based on said corrected depth map; and

a stereoscopic image synthesizing unit that synthesizes the right image and left image of said output image by projecting said corrected three-dimensional map on a display plane.

10. The disparity conversion device according to claim 1, wherein said image synthesizing unit comprises:

a direct three-dimensional mapping unit that generates a three-dimensional map by mapping the left image and right image of said input image in a three-dimensional space based on the disparity in said disparity map;

a center monocular image synthesizing unit that synthesizes the center monocular image by projecting said three-dimensional map on a display plane facing a virtual center eye;

a disparity/depth converting unit that converts from the disparity in said corrected disparity map into depth and generates a corrected depth map;

a three-dimensional mapping unit that generates a corrected three-dimensional map by mapping said center monocular image in a three-dimensional space based on said corrected depth map; and

a stereoscopic image synthesizing unit that synthesizes the right image and left image of said output image by projecting said corrected three-dimensional map onto a display plane.

11. The disparity conversion device according to claim 1, further comprising:

a disparity analyzing unit that analyzes the disparity in said disparity map and generates a cumulative frequency distribution of said disparity;

wherein

said disparity correction unit corrects the disparity in said disparity map based on said cumulative frequency distribution and said correcting features.

12. A stereoscopic image display system comprising:

an input image supplying unit that supplies an input image having a left image and right image as a pair of stereoscopic images;

a disparity detecting unit that detects disparity from a left image and right image of said input image, and generates a disparity map that holds the disparity for each pixel or pixel group;

a correction feature setting unit that sets correction features in the event of correcting the disparity in said disparity map;

a disparity correction unit that corrects the disparity in said disparity map according to said set correction features and generates a corrected disparity map;

an image synthesizing unit that synthesizes a left image and right image of an output image from the left image and right image of said input image according to said corrected disparity map; and

an image display device that displays said output image.

13. A disparity conversion method comprising:

correction feature setting procedures to set correction features in the event of correcting the disparity in the disparity map holding the disparity for each pixel or pixel group;

disparity detecting procedures that detect disparity from a left image and right image of an input image, and generates said disparity map;

correction feature setting procedures that set correction features in the event of correcting the disparity in said disparity map;

disparity correction procedures that correct the disparity in said disparity map according to said set correction features and generate a corrected disparity map;

image synthesizing procedures that synthesize a left image and right image of an output image from the left image and right image of said input image according to said corrected disparity map; and

image display procedures to display said output image on a display device.

14. A program for causing a computer to execute

correction feature setting procedures to set correction features in the event of correcting the disparity in the disparity map holding the disparity for each pixel or pixel group;

disparity detecting procedures that detect disparity from a left image and right image of an input image, and generates said disparity map;

disparity correction procedures that correct the disparity in said disparity map according to said set correction features and generate a corrected disparity map;

image synthesizing procedures that synthesize a left image and right image of an output image from the left image and right image of said input image according to said corrected disparity map; and

image display procedures to display said output image on a display device.