Employing pixel density to detect a spam image

• US 7,882,177 B2
• Filed: 08/06/2007
• Issued: 02/01/2011
• Est. Priority Date: 08/06/2007
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Wei, Ke, Pujara, Jay, Zheng, Hao
• Primary Examiner(s): Patel; Ashok B
• Assistant Examiner(s): GOLDBERG, ANDREW C
• Application Number:
  US11/834,529
• Publication Number:
  US 20090043853A1
• Time in Patent Office: 1,275 Days
• Field of Search: 709/202
• US Class Current: 709/204
• CPC Class Codes: G06T 7/44 using image operators, e.g. filters, edge density metrics or local histograms
• CPC Class Codes: H04L 51/212 using filtering or selective blocking

Abstract

A network device and method are directed towards detecting and blocking image spam within a message by performing statistical analysis on differences in edge pixel distribution patterns. An image spam detection component receives a message with an image attachment. Physical characteristics of the image are examined to determine whether the image is a candidate for further analysis. If so, then the image may be converted to a grayscale image, and then performing edge detection, followed by the elimination of non-maxima and thresholding of weak edges. Edge pixels and then employed to determine a normalized pixel density distribution (PDD). Various statistical analyses are applied to the resulting normalized PDD to determine a likelihood that the image is spam. A signature based exemption may be applied to images improperly identified as spam, based on trusted user feedback.

Claims

1. A method for use in managing delivery of content over a network, comprising:

receiving a message, wherein the message includes an image file;

extracting the image file from the message;

converting the image file to a grayscale image;

performing pixel edge detection within the grayscale image that includes at least eliminating weak edges by converting a pixel to black if its threshold chrominance is below a defined threshold, and eliminating weak edges by converting a contrast of a pixel within a determined weak edge to a higher contrast;

determining a pixel density distribution (PDD) for the edge detected image;

if the PDD statistically indicates that the image file is an image spam file, performing an action to inhibit at least the image file from being forwarded to a destination associated with the message; and

if the PDD statistically indicates that the image file is not image spam, forwarding the message with the image file towards a destination such that the image file is viewable by a recipient at a display;

collecting feedback based, in part, on a number of false-positives for the statistical indications; and

employing the feedback to modify a PDD statistical threshold useable in indicating if the image file is an image spam file.

2. The method of claim 1, the method further comprising:

determining at least one physical characteristic associated with the extracted image file, including at least one of a file type, file size, or file dimension; and

if the at least one physical characteristic statistically indicates that the image file is unlikely to be image spam, then forwarding the message with the image file towards the destination such that the image file is viewable by the recipient at the display.

3. The method of claim 1, wherein performing pixel edge detection further comprises:

employing a Sobel operator to detect the edges within the image.

4. The method of claim 1, wherein determining the PDD further comprises:

for each horizontal line within the image, determining a number of edge pixels; and

determining the PDD based on the number of edge pixels for each line.

5. The method of claim 1, wherein if the PDD statistically indicates further comprises:

determining a statistical threshold value for which to indicate the image file is an image spam file based at least in part on an acceptable false-positive result.

6. The method of claim 1, wherein if the PDD statistically indicates that the image file is, further comprises:

determining at least one physical characteristic of the image file; and

employing the at least one physical characteristic to determine a statistical threshold value;

comparing the resulting PDD for the image file to the statistical threshold value;

if the PDD is above the statistical threshold value, identifying the image file as statistically likely to be image spam; and

if the PDD is at or below the statistical threshold value, identifying the image file as statistically likely to not be image spam.

7. A network device for selectively managing delivery of messages over a network, comprising:

a transceiver to send and receive data over the network; and

a processor that is operative to perform actions, including;

receiving an image file associated with a message;

determining at least one physical characteristic of the image file; and

if the physical characteristic indicates that statistically, the image file is unlikely to be associated with image spam, enabling the image file and message to be forwarded to a destination, otherwise performing the following actions, comprising;

converting the image file to a grayscale image;

performing pixel edge detection within the grayscale image and eliminating weak edges by converting a pixel to black if its threshold chrominance is below a defined threshold, and eliminating weak edges by converting a contrast of a pixel within a determined weak edge to a higher contrast;

determining a pixel density distribution (PDD) for the edge detected image;

if the PDD statistically indicates that the image file is an image spam file, performing an action to inhibit at least the image file from being forwarded to a destination associated with the message; and

if the PDD statistically indicates that the image file is not image spam, forwarding the message with the image file towards the destination such that the image file is viewable by a recipient at a display;

collecting feedback based, in part, on a number of false-positives for the statistical indications; and

employing the feedback to modify a PDD statistical threshold useable in indicating if the image file is an image spam file.

8. The network device of claim 7, wherein the physical characteristic comprises at least one of the following:

a dimension, a file type, a file size, or a compression ratio.

9. The network device of claim 7, wherein if the PDD statistically indicates that the image file is, further comprises, using at least in part, the at least one physical characteristic to determine a statistical threshold value useable to compare the PDD to, to determine statistically a likelihood that the image file is or is not image spam.

10. The network device of claim 7, wherein the processor is operative to perform actions, further comprising:

receiving feedback that the image file is incorrectly identified as image spam or incorrectly identified as non-image spam;

determining a digital signature for the incorrectly identified image file; and

if another image file is received that matches the digital signature, processing the image file based on a corrected identity according to the received feedback.

11. The network device of claim 7, wherein performing pixel edge detection further comprises employing a Sobel operator to detect the edges within the image.

12. The network device of claim 7, wherein if the PDD statistically indicates that the image file is an image spam file, further comprises:

determining if the image file is statistically likely to be an image spam file based on a statistically significant amount of detected text within the image file.

13. The network device of claim 7, wherein if the physical characteristic indicates that statistically, further comprises:

using feedback about whether the image file is correctly or incorrectly identified as image spam to modify at least one of a physical characteristic used in determining if the physical characteristic indicates that statistically the image file is unlikely to be associated with image spam.

14. A system for use in selectively enabling delivery of content over a network, comprising:

a message server that is configured and arranged to perform actions, including;

receiving a message; and

if the message includes an image file, providing the image file to an image spam detection component; and

the image spam detection component being configured to perform actions, including;

converting the image file to a grayscale image;

performing pixel edge detection within the grayscale image and eliminating weak edges by converting a pixel to black if its threshold chrominance is below a defined threshold, and eliminating weak edges by converting a contrast of a pixel within a determined weak edge to a higher contrast;

determining a pixel density distribution (PDD) for the edge detected image;

if the PDD statistically indicates that the image file is an image spam file, providing an alert to the message server such that the image file is prevented from being forwarded to a destination associated with the message; and

if the PDD statistically indicates that the image file is not image spam, providing an indicator to the message server such that the image file is forwarded towards the destination such that the image file is viewable by at the destination;

collecting feedback based, in part, on a number of false-positives for the statistical indications; and

employing the feedback to modify a PDD statistical threshold useable in indicating if the image file is an image spam file.

15. The system of claim 14, wherein performing pixel edge detection further comprises employing a Sobel operator to detect the edges within the image file.

16. The system of claim 14, wherein the image spam detection component is configured to perform actions, further including:

determining at least one physical characteristic associated with the image file; and

if the at least one physical characteristic statistically indicates that the image file is unlikely to be image spam, then forwarding the message with the image file towards the destination such that the image file is viewable by the recipient at the display.

17. The system of claim 14, wherein the image spam detection component is configured to perform actions, further including:

determining at least one physical characteristic associated with the image file; and

using the physical characteristic to determine, in part, a threshold value useable in statistically indicating if the image file is an image spam based on the determined PDD for the image file.

18. The system of claim 17, wherein at least one physical characteristic comprises at least one of a file size, a file type, or a file dimension.

19. The system of claim 14, wherein the image spam detection component is configured to perform actions, further including:

receiving feedback that the image file is incorrectly identified as image spam or incorrectly identified as non-image spam;

determining a digital signature for the incorrectly identified image file; and

if another image file is received that matches the digital signature, processing the image file based on a corrected identity according to the received feedback.