Transactional video marking system

• US 9,262,794 B2
• Filed: 03/14/2014
• Issued: 02/16/2016
• Est. Priority Date: 03/14/2013
• Status: Active Grant

• Current Assignee: Vobile Incorporated (Vobile Group Ltd.)
• Original Assignee: Verance Corporation
• Inventors: Zhao, Jian, Mercier, Guillaume
• Primary Examiner(s): Yentrapati, Avinash
• Application Number:
  US14/214,393
• Publication Number:
  US 20140270338A1
• Time in Patent Office: 704 Days
• US Class Current: 1/1
• CPC Class Codes: G06T 1/0021 Image watermarking
• CPC Class Codes: G06T 1/0028 Adaptive watermarking, e.g. Human Visual System [HVS]-based watermarking
• CPC Class Codes: G06T 1/0071 using multiple or alternating watermarks
• CPC Class Codes: G06T 2201/0051 Embedding of the watermark in the spatial domain
• CPC Class Codes: G06T 2201/0053 Embedding of the watermark in the coding stream, possibly without decoding; Embedding of the watermark in the compressed domain
• CPC Class Codes: G06T 2201/0202 whereby the quality of watermarked images is measured; Measuring quality or performance of watermarking methods; Balancing between quality and robustness

Abstract

Methods, devices, systems and computer program products facilitate embedding and extraction of transactional watermarks into and from a video content. One method for selecting an area for watermark insertion includes selecting a candidate area within a frame of the video content that is smaller in size that the frame itself, adding a first and a second watermark value to obtain corresponding first and second versions of the area, add noise to each of the first and second versions, and perform watermark extraction operations to obtain corresponding correlation values for the selected area. The above operations are repeated using different areas that are selected at pseudo-random offset values from the previously selected areas until the entire video frame is exhausted. One or more areas with the highest correlation value(s) are selected and designated as areas that are suitable for watermark insertion.

Claims

1. A method for selecting an area within a video content for insertion of watermarks, the method comprising:

(a) selecting a candidate area within a frame of the video content that is smaller in size that the frame itself;

(b) adding a first watermark value to the selected candidate area to obtain a first version of the area;

(c) adding one or more noise components to the first version of the area to obtain a noisy first version of the area;

(d) conducting a watermark extraction operation on the noisy first version of the area to obtain a first correlation value;

(e) adding a second watermark signal to the selected candidate area to obtain a second version of the area;

(f) adding one or more noise components to the second version of the area to obtain a noisy second version of the area;

(g) conducting a watermark extraction operation on the noisy second version of the area to obtain a second correlation value;

(h) selecting one or more pseudo-random offset values to select a different candidate area within the frame for adding the first watermark value or for adding the second watermark value;

(i) repeating operations (a) through (g) until all pixels of the frame have been selected as part of one or more of the selected candidate areas, wherein in each iteration of operations (a) through (g) the values obtained for the first correlation value and the second correlated value are stored; and

(j) designating one or more of the selected areas that produced the largest first correlation value or the largest second correlation value for the insertion of watermarks.

2. The method of claim 1, further comprising conducting a spatial psychovisual analysis on one or more frame frames that include the designated selected area(s), the spatial psychovisual analysis comprising:

applying a non-linear inverse gamma curve to luminance components of the one or more video frames corresponding to the human visual system response in dark areas;

performing a Laplacian or a Gaussian pyramid operation multiple times at multiple resolutions on the one or more video frames subsequent to the applying of the non-linear inverse gamma curve to detect highly textured areas; and

performing an edge detection operation to detect edge features within the one or more frames.

3. The method of claim 1, further comprising:

embedding the first watermark value in the designated area of the frame to produce a first embedded version of the frame;

embedding the first watermark value in the designated area of the frame to produce a first embedded version of the frame;

re-encoding the first and the second versions of the frame to obtain a first and a second encoded version of the frame, respectively;

decoding the first and second versions of the encoded frames;

comparing one or more measures of quality associated with the decoded first and second versions with corresponding one or more reference measures of quality; and

based on the comparing, making a determination to accept or reject the frame as a frame that is suitable for watermark insertion.

4. A computer program product stored on a one or more non-transitory computer readable media, comprising:

program code for (a) selecting a candidate area within a frame of the video content that is smaller in size that the frame itself;

program code for (b) adding a first watermark value to selected candidate area to obtain a first version of the area;

program code for (c) adding one or more noise components to the first version of the area to obtain a noisy first version of the area;

program code for (d) conducting a watermark extraction operation on the noisy first version of the area to obtain a first correlation value;

program code for (e) adding a second watermark signal to the selected candidate area to obtain a second version of the area;

program code for (f) adding one or more noise components to the second version of the area to obtain a noisy second version of the area;

program code for (g) conducting a watermark extraction operation on the noisy second version of the area to obtain a second correlation value;

program code for (h) selecting one or more pseudo-random offset values to select a different candidate area within the frame for adding the first watermark value or for adding the second watermark value;

program code for (i) repeating operations (a) through (g) until all pixels of the frame have been selected as part of one or more of the selected candidate areas, wherein in each iteration of operations (a) through (g) the values obtained for the first correlation value and the second correlated value are stored; and

program code for (j) designating one or more of the selected areas that produced the largest first correlation value or the largest second correlation value for the insertion of watermarks.

5. The computer program product of claim 4, further comprising program code for conducting a spatial psychovisual analysis on one or more frame frames that include the designated selected area(s), the spatial psychovisual analysis comprising:

applying a non-linear inverse gamma curve to luminance components of the one or more video frames corresponding to the human visual system response in dark areas;

performing a Laplacian or a Gaussian pyramid operation multiple times at multiple resolutions on the one or more video frames subsequent to the applying of the non-linear inverse gamma curve to detect highly textured areas; and

performing an edge detection operation to detect edge features within the one or more frames.

6. The computer program product of claim 4, further comprising:

program code for embedding the first watermark value in the designated area of the frame to produce a first embedded version of the frame;

program code for embedding the first watermark value in the designated area of the frame to produce a first embedded version of the frame;

program code for re-encoding the first and the second versions of the frame to obtain a first and a second encoded version of the frame, respectively;

program code for decoding the first and second versions of the encoded frames;

program code for comparing one or more measures of quality associated with the decoded first and second versions with corresponding one or more reference measures of quality; and

program code for based on the comparing, making a determination to accept or reject the frame as a frame that is suitable for watermark insertion.

7. A device, comprising:

a processor; and

a memory comprising processor executable code, the processor executable code when executed by the processor configures the device to;

(a) select a candidate area within a frame of the video content that is smaller in size that the frame itself;

(b) add a first watermark value to the selected candidate area to obtain a first version of the area;

(c) add one or more noise components to the first version of the area to obtain a noisy first version of the area;

(d) conduct a watermark extraction operation on the noisy first version of the area to obtain a first correlation value;

(e) add a second watermark signal to the selected candidate area to obtain a second version of the area;

(f) add one or more noise components to the second version of the area to obtain a noisy second version of the area;

(g) conduct a watermark extraction operation on the noisy second version of the area to obtain a second correlation value;

(h) select one or more pseudo-random offset values to select a different candidate area within the frame for adding the first watermark value or for adding the second watermark value;

(i) repeat operations (a) through (g) until all pixels of the frame have been selected as part of one or more of the selected candidate areas, wherein in each iteration of operations (a) through (g) the values obtained for the first correlation value and the second correlated value are stored; and

(j) designate one or more of the selected areas that produced the largest first correlation value or the largest second correlation value for the insertion of watermarks.

8. The device of claim 7, wherein, the processor executable code when executed by the processor, configures the device to:

conduct a spatial psychovisual analysis on one or more frame frames that include the designated selected area(s), including the following;

apply a non-linear inverse gamma curve to luminance components of the one or more video frames corresponding to the human visual system response in dark areas;

perform a Laplacian or a Gaussian pyramid operation multiple times at multiple resolutions on the one or more video frames subsequent to the applying of the non-linear inverse gamma curve to detect highly textured areas; and

perform a edge detection operation to detect edge features within the one or more frames.

9. The device of claim 7, wherein, the processor executable code when executed by the processor, configures the device to:

embed the first watermark value in the designated area of the frame to produce a first embedded version of the frame;

embed the first watermark value in the designated area of the frame to produce a first embedded version of the frame;

re-encode the first and the second versions of the frame to obtain a first and a second encoded version of the frame, respectively;

decode the first and second versions of the encoded frames;

compare one or more measures of quality associated with the decoded first and second versions with corresponding one or more reference measures of quality; and

based on the comparing, make a determination to accept or reject the frame as a frame that is suitable for watermark insertion.