Method and system for processing image data

• US 8,406,472 B2
• Filed: 02/04/2011
• Issued: 03/26/2013
• Est. Priority Date: 03/16/2010
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Unruh, Christian, Harres, Dennis
• Primary Examiner(s): AZARIAN, SEYED H
• Application Number:
  US13/021,153
• Publication Number:
  US 20110228981A1
• Time in Patent Office: 781 Days
• Field of Search: 382/100, 382103-107, 382/154, 382/155, 382/162, 382/168, 382/173, 382/181, 382/199, 382/214, 382/232, 382/254, 382/274, 382/276, 382285-294, 382/305, 382/312, 375/240.21, 359/267, 340/436, 348/43
• US Class Current: 382/106
• CPC Class Codes: G06T 2207/30252 Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06T 7/75 involving models

Abstract

A method for processing image data representing a segmentation mask, comprises generating two-dimensional shape representations of a three-dimensional object on the basis of a plurality of parameter sets; and matching motion blocks of the segmentation mask with the two-dimensional shape representations to obtain a best fit parameter set. Thereby, for example, a distance between the three-dimensional object and a camera position may be determined.

Claims

1. A method for processing image data representing a segmentation mask, wherein the image data are generated from a sequence of images, comprising:

generating two-dimensional shape representations of a three-dimensional object on the basis of a plurality of parameter sets;

matching motion blocks of the segmentation mask with the two-dimensional shape representations to obtain a best fit parameter set;

segmenting one of the images into a plurality of segments; and

applying a scaling factor to the image data that are disposed in a specific segment of the image, wherein different scaling factors are assigned to different segments of the image.

2. The method of claim 1, further comprising:

determining a distance between the three-dimensional object and a camera position from the best fit parameter set.

3. The method of claim 1, wherein the parameter set comprises distance and perspective of a camera, that has generated the image data, with respect to the three-dimensional object.

4. The method of claim 1, wherein the segmentation mask is generated from a sequence of images by a method comprising:

partitioning each image of the sequence of images into blocks;

identifying a moving block;

determining direction and distance of movement of the moving block;

grouping adjacent moving blocks with directions and distances within predetermined intervals to form motion blocks.

5. The method of claim 4, wherein moving blocks are identified by supplying image data of the sequence of images concurrently to

a first motion estimator, the first motion estimator being adapted to identify moving items which have performed a movement between consecutive images of a first sequence, the first sequence having a first frame rate;

and

a second motion estimator, the second motion estimator being adapted to identify moving items which have performed a movement between consecutive images of a second sequence, the second sequence having a second frame rate.

6. The method of claim 1, wherein the segments are disposed along one axis.

7. The method of claim 1, wherein the segments are disposed along two axes, the two axes being orthogonal to each other.

8. The method of claim 1, wherein the scaling factor decreases from a value larger than 1 to a value smaller than 1 along an axis.

9. The method of claim 1, wherein the image data are generated from a sequence of images, the method further comprising supplying image data of the sequence of images concurrently to

a first motion estimator, the first motion estimator being adapted to identify moving blocks which have performed a movement between consecutive images of a first sequence, the first sequence having a first frame rate;

and

a second motion estimator, the second motion estimator being adapted to identify moving blocks which have performed a movement between consecutive images of a second sequence, the second sequence having a second frame rate.

10. The method of claim 9, wherein the first frame rate is a multiple integer of the second frame rate.

11. The method of claim 1, further comprising:

detecting a front most object from the motion block; and

subtracting the front most object from the motion block.

12. A method for detecting a moving object, comprising

generating image data by taking a sequence of images;

and

performing the method of claim 1.

13. A non-transitory computer readable medium having stored thereon a computer program that when executed by a computer causes the computer to implement the method according to claim 1 and the steps thereof when it is carried out on a data processing device.

14. A system for processing image data representing a segmentation mask, wherein the image data are generated from a sequence of images, the system comprising components that are configured

to generate two-dimensional shape representations of a three-dimensional object on the basis of a plurality of parameter sets;

to match motion blocks of the segmentation mask with the two-dimensional shape representations to obtain a best fit parameter set;

to segment one of the images into a plurality of segments; and

to apply a scaling factor to the image data that are disposed in a specific segment of the image, wherein different scaling factors are assigned to different segments of the image.

15. The system of claim 14, the system further comprising a component that is adapted to determine a distance between the three-dimensional object and a camera position.

16. The system of claim 14, wherein the image data are generated from a sequence of images, the system further comprising

a first motion estimator, the first motion estimator being adapted to identify moving blocks which have performed a movement between consecutive images of a first sequence, the first sequence having a first frame rate;

and

a second motion estimator, the second motion estimator being adapted to identify moving blocks which have performed a movement between consecutive images of a second sequence, the second sequence having a second frame rate,

the first and the second motion estimators being disposed in such a manner that the image data may be concurrently fed to the first and second motion estimators.

17. The system of claim 14, further comprising components that are adapted

to detect a front most object from the motion block;

and

to subtract the front most object from the motion block.

18. A system for detecting a moving object, comprising

a device that is adapted to generate image data by taking a sequence of images;

and

the system of claim 14.