Identifying Steganographic Data in an Image

• US 20090129670A1
• Filed: 11/19/2007
• Published: 05/21/2009
• Est. Priority Date: 11/19/2007
• Status: Active Grant

• Current Assignee: Seiko Epson Corporation (Seiko Group)
• Original Assignee: Seiko Epson Corporation (Seiko Group)
• Inventors: Zandifar, Ali, Xiao, Jing
• Granted Patent: US 7,974,437 B2
• Time in Patent Office: Days
• US Class Current: 382/171
• CPC Class Codes: G06T 1/0028 Adaptive watermarking, e.g. Human Visual System [HVS]-based watermarking
• CPC Class Codes: G06T 1/005 Robust watermarking, e.g. average attack or collusion attack resistant
• CPC Class Codes: G06T 2201/0051 Embedding of the watermark in the spatial domain
• CPC Class Codes: G06T 2201/0065 Extraction of an embedded watermark; Reliable detection
• CPC Class Codes: H04N 1/32208 involving changing the magnitude of selected pixels, e.g. overlay of information or super-imposition H04N1/32219, H04N1/32224 take precedence
• CPC Class Codes: H04N 1/32229 with selective or adaptive application of the additional information, e.g. in selected regions of the image H04N1/32213 takes precedence
• CPC Class Codes: H04N 1/32251 in multilevel data, e.g. greyscale or continuous tone data

Abstract

Disclosed are embodiments of systems and methods for embedding and/or extracting data from images. In embodiments, an image may be segmented into regions, and characters or other image groups within a segmented region may be determined to be embedding sites. A data vector may be embedded into a set of corresponding ordered embedding sites by representing each data element as different intensity values assigned to the pixels within one portion of an embedding site relative to the pixels in another portion of the embedding site. In embodiments, embedded data may be extracted from an image by extracting and decoding a set of bit values from a set of identified and ordered embedding sites.

Claims

1. A method for identifying a robust average intensity of a portion of an embedding site comprising a plurality of pixels of a common class to extract a data value represented as a difference between a first robust average intensity value of a first portion of the plurality of pixels in the embedding site and a second robust average intensity value of a second portion of the plurality of pixels in the embedding site, the method comprising:

selecting reliable intensity value pixels from a first portion of the plurality of pixels to form a first set of reliable pixels;

iteratively pruning, from the first set of reliable pixels, reliable intensity value pixels that reside in an outer region of a histogram of the first set of reliable pixels until a difference between successive average intensity values of the first set of reliable pixels is less than a difference threshold; and

setting a first robust average intensity value of the first set of reliable pixels based upon at least one of the successive average intensity values that had less than a difference threshold.

2. The method of claim 1 wherein selecting reliable intensity value pixels from a first portion of the plurality of pixels to form a first set of reliable pixels comprises excluding pixels from at least the first portion of the plurality of pixels that have at least one neighboring pixel that is not a member of the plurality of pixels.

3. The method of claim 1 wherein the outer region of the histogram comprises a first set of intensity values that are less than a lower bounding value and a second set of intensity values that are greater than an upper bounding value.

4. The method of claim 3 wherein the lower bounding value is an average intensity value of the first set of reliable pixels minus a first value multiplied by a standard deviation of the first set of reliable pixels, and the upper bounding value is the average intensity value of the first set of reliable pixels plus a second value multiplied by the standard deviation.

5. The method of claim 1 further comprising extracting a data value from the embedding site by:

calculating a second robust average intensity value of a second set of reliable pixels from the embedding site;

responsive to a difference between the first robust average intensity and the second robust average intensity value being greater than a threshold, extracting an embed pattern by determining which of the first and second robust average intensity values is greater than the other robust average intensity value; and

correlating the embed pattern with the data value.

6. The method of claim 5 further comprising extracting a data vector from an image comprising pixels by extracting a set of data values from a set of ordered embedding sites in the image.

7. A computer readable medium having instructions for performing the method of claim 1.

8. A method for calculating a robust average intensity of a portion of an embedding site comprising a plurality of pixels wherein an embedded data value represents a difference between a first robust average intensity value of a first portion of the plurality of pixels in the embedding site and a second robust average intensity value of a second portion of the plurality of pixels in the embedding site, the method comprising:

[a] responsive to a pixel from the plurality of pixels belonging to a neighborhood of pixels from the plurality of pixels, adding the pixel to a set of reliable pixels;

[b] identifying a portion set of reliable pixels comprised of reliable pixels from the set of reliable pixels that reside within a portion of the embedding site;

[c] calculating a first average intensity value of the portion set of reliable pixels;

[d] calculating a standard deviation of the portion set of reliable pixels;

[e] generating a pruned portion set of the reliable pixels by removing reliable pixels from the portion set of reliable pixel that do not have an intensity within a range defined by the average intensity value minus a first value multiplied by the standard deviation and the average intensity value plus a second value multiplied by the standard deviation;

[f] calculating a second average intensity value of the pruned portion set of the reliable pixels;

[g] responsive to a difference between the first average intensity value and the second average intensity value being greater than a difference threshold value;

setting the pruned portion set of the reliable pixels as the portion set of reliable pixels;

setting the second average intensity as the first average intensity; and

repeating steps [d] and [g];

[h] responsive to a difference between the first average intensity value and the second average intensity value being less than a difference threshold value, selecting the second average intensity value as a first robust average intensity value of the portion set of reliable pixels.

9. The method of claim 8 further comprising extracting a data value from the embedding site by:

calculating a second robust average intensity value of a second portion set of reliable pixels comprised of reliable pixels from the set of reliable pixels that reside within a second portion of the embedding site;

responsive to a difference between the first robust average intensity value and the second robust average intensity value being greater than a threshold, extracting an embed pattern by determining which of the first and second robust average intensity values is greater than the other robust average intensity value; and

correlating the embed pattern with the data value.

10. The method of claim 9 further comprising extracting at least one data vector from an image comprising pixels by extracting a set of data values from a set of ordered embedding sites in the image.

11. The method of claim 10 wherein the at least one data vector is a bit vector, and decoding the at least one data vector comprises:

removing an indicator prefix from the bit vector; and

extracting data by decoding the bit vector.

12. The method of claim 11 further comprising the step of applying an error-correcting code decoder to the bit vector.

13. A computer readable medium having instructions for performing the method of claim 8.

14. A system for extracting a set of data values from a set of ordered embedding sites, an embedding site comprising pixels, the system comprising:

a bitstream extractor, coupled to receive the set of ordered embedding sites, that extracts a set of data values wherein a data value is extracted from an embedding site from the set of ordered embedding sites by;

selecting reliable intensity value pixels from a first portion of pixels from the embedding site to form a first set of reliable pixels and selecting reliable intensity value pixels from a second portion of the pixels from the embedding site to form a second set of reliable pixels;

iteratively pruning, from the first set of reliable pixels, reliable intensity value pixels that reside in an outer region of a histogram of the first set of reliable pixels until a difference between successive average intensity values of the first set of reliable pixels is less than a difference threshold and setting a first robust average intensity value of the first set of reliable pixels based upon at least one of the successive average intensity values that had less than a difference threshold;

iteratively pruning, from the second set of reliable pixels, reliable intensity value pixels that reside in an outer region of a histogram of the second set of reliable pixels until a difference between successive average intensity values of the second set of reliable pixels is less than a difference threshold and setting a second robust average intensity value of the second set of reliable pixels based upon at least one of the successive average intensity values that had less than a difference threshold;

responsive to a difference between the first robust average intensity value and the second robust average intensity value being greater than a threshold, extracting an embed pattern by determining which of the first and second robust average intensity values is greater than the other robust average intensity value; and

correlating the embed pattern with the data value.

15. The system of claim 14 wherein selecting reliable intensity value pixels from a first portion of pixels from the embedding site to form a first set of reliable pixels and selecting reliable intensity value pixels from a second portion of the pixels from the embedding site to form a second set of reliable pixels comprises:

excluding pixels from the first portion of the pixels of an embedding site that have at least one neighboring pixel that is not a member of the pixels; and

excluding pixels from the second portion of the pixels of the embedding site that have at least one neighboring pixel that is not a member of the pixels.

16. The system of claim 14 further comprising:

an embedding site selector, coupled to receive an image, that identifies a set of embedding sites in the image; and

an embedding site ordering processor, coupled to receive the set of embedding sites, that orders the set of embedding sites to obtain the set of ordered embedding sites.

17. The system of claim 16 wherein the embedding site selector identifies a set of ordered embedding sites in the image by performing the steps comprises:

segmenting the image into regions;

identifying a plurality of connected components comprising pixels from one of the regions;

identifying the plurality of embedding sites within the plurality of connected components; and

assigning an order to the plurality of embedding sites.

18. The system of claim 17 wherein assigning an order to the plurality of embedding sites comprises using at least one lines sweep to order the embedding sites from the plurality of embedding sites.

19. The system of claim 14 further comprising a bitstream segmenter, coupled to receive the set of data values, that extracts at least one data vector from the set of data values by segmenting the set of data values according to a marker.

20. The system of claim 19 wherein the at least one data vector is a bit vector, and decoding the at least one data vector comprises:

removing the marker from the bit vector; and

extracting the data by decoding the bit vector.