Image processing with segmentation using directionally-accumulated difference-image pixel values

• US 9,648,339 B2
• Filed: 05/28/2014
• Issued: 05/09/2017
• Est. Priority Date: 05/28/2013
• Status: Expired due to Fees

• Current Assignee: Grass Valley USA Incorporated
• Original Assignee: Snell Advanced Media Limited (Belden Inc.)
• Inventors: Piroddi, Roberta
• Primary Examiner(s): Patel, Jay
• Assistant Examiner(s): Suh, Joseph
• Application Number:
  US14/289,179
• Publication Number:
  US 20140355682A1
• Time in Patent Office: 1,077 Days
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/20224 Image subtraction
• CPC Class Codes: G06T 7/254 involving subtraction of images
• CPC Class Codes: H04N 19/48 using compressed domain processing techniques other than decoding, e.g. modification of transform coefficients, variable length coding [VLC] data or run-length data motion estimation in a transform domain H04N19/547; processing of decoded motion vectors H04N19/513
• CPC Class Codes: H04N 19/577 Motion compensation with bidirectional frame interpolation, i.e. using B-pictures

Abstract

In the segmentation of an image into spatial regions, an image region is identified by forming difference-image pixel values from the chosen image and at least one other image in the image sequence and accumulating those difference-image pixel values horizontally and vertically. Co-located horizontal and vertical accumulations are combined and utilized in identifying the region. Difference-image pixel values can also be accumulated along other lines which pass through a point of interest.

Claims

1. A method of identifying a region within a chosen image that is part of an image sequence, the method comprising:

forming difference-image pixel values from the chosen image and at least one other image in the image sequence;

forming first directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a first image direction;

forming second directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a second image direction;

combining the respective co-located first directionally-accumulated difference-image pixel values and the second directionally-accumulated difference-image pixel values to form combined directionally-accumulated difference-image pixel values; and

utilizing the combined directionally-accumulated difference-image pixel values in identifying said region;

further comprising;

forming third directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a third image direction; and

forming fourth directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a fourth image direction,

wherein the third and fourth directions are opposing.

2. A method according to claim 1, wherein difference-image pixel values are accumulated along straight lines of pixels.

3. A method according to claim 1, wherein forming difference-image pixel values from the chosen image and at least one other image in the image sequence comprises receiving a displaced frame difference (DFD) comprising displaced frame difference values of pixels derived from a motion estimation process applied to images in the sequence.

4. A method according to claim 1, wherein the said at least one other image comprises at least an image earlier in the sequence than the said chosen image and at least an image later in the sequence than the said chosen image.

5. A method according to claim 1, wherein the first and second directions are opposing.

6. A method according to claim 1, wherein the first and second directions are intersecting lines.

7. A method according to claim 1, wherein combining co-located directionally-accumulated difference-image pixel values comprises:

combining the first and second directionally-accumulated difference-image pixel values to form first combined directionally-accumulated difference-image pixel values; and

combining the third and fourth directionally-accumulated difference-image pixel values to form second combined directionally-accumulated difference-image pixel values.

8. A method according to claim 1, wherein combining co-located directionally-accumulated difference-image pixel values comprises multiplying co-located directionally-accumulated difference-image pixel values together to form a product of accumulated pixel values.

9. A method according to claim 8, wherein the magnitude of a product of accumulated pixel values for a pixel is used to identify that pixel as part of an image region.

10. A method according to claim 9, wherein the magnitude of a weighted sum of product values for a pixel is compared with a threshold and that pixel is considered to form part of an image region when the threshold is exceeded.

11. A method according to claim 10, wherein a product of pixel values is modified by a non-linear function prior to comparison.

12. Apparatus for identifying a region within a chosen image that is part of an image sequence, comprising:

a difference-image store for holding difference-image pixel values;

a first accumulator forming first directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a first image direction;

a second accumulator forming second directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a second image direction; and

a multiplier forming products of directionally-accumulated difference-image pixel values,

wherein co-located pixel values resulting from accumulation in opposing directions are multiplied together.

13. Apparatus according to claim 12, wherein co-located pixel values derived from values accumulated along intersecting lines are multiplied together.

14. Apparatus according to claim 12, wherein the magnitude of a product of accumulated pixel values for a pixel is used to identify that pixel as part of an image region.

15. Apparatus according to claim 14, wherein the magnitude of a weighted sum of product values for a pixel is compared with a threshold and that pixel is considered to form part of an image region when the threshold is exceeded.

16. A non-transitory computer readable medium storing a computer program product having instructions thereon that, when executed, cause one or more programmable apparatuses to implement a method of identifying a region within a chosen image that is part of an image sequence, the method comprising:

forming difference-image pixel values from the chosen image and at least one other image in the image sequence;

forming first directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a first image direction;

forming second directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a second image direction;

combining the respective co-located first directionally-accumulated difference-image pixel values and the second directionally-accumulated difference-image pixel values to form combined directionally-accumulated difference-image pixel values;

utilizing the combined directionally-accumulated difference-image pixel values in identifying said region;

forming third directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a third image direction; and

forming fourth directionally-accumulated difference-image pixel values by accumulating difference-image pixel values in a fourth image direction,

wherein the third and fourth directions are opposing.

17. The non-transitory computer readable medium of claim 16 having instructions thereon that, when executed, cause one or more programmable apparatuses to implement a method of identifying a region within a chosen image that is part of an image sequence, wherein combining co-located directionally-accumulated difference-image pixel values comprises:

combining the first and second directionally-accumulated difference-image pixel values to form first combined directionally-accumulated difference-image pixel values; and

combining the third and fourth directionally-accumulated difference-image pixel values to form second combined directionally-accumulated difference-image pixel values.