Automated tests to distinguish computers from humans

• US 8,925,057 B1
• Filed: 02/08/2010
• Issued: 12/30/2014
• Est. Priority Date: 02/06/2009
• Status: Active Grant

• Current Assignee: New Jersey Institute of Technology
• Original Assignee: New Jersey Institute of Technology
• Inventors: Ansari, Nirwan, Neylan, Christopher Andrew, Shevtekar, Amey Bhaskar
• Primary Examiner(s): Nalven, Andrew
• Assistant Examiner(s): FIELDS, COURTNEY D
• Application Number:
  US12/701,949
• Time in Patent Office: 1,786 Days
• Field of Search: 726/7, 726/21, 726/2, 726/26, 382/181, 382/209
• US Class Current: 726/7
• CPC Class Codes: G06F 21/31 User authentication
• CPC Class Codes: G06F 21/36 by graphic or iconic representation
• CPC Class Codes: G06F 2221/2103 Challenge-response
• CPC Class Codes: G06F 2221/2133 Verifying human interaction, e.g., Captcha

Abstract

Completely automated tests that exploit capabilities of human vision to tell humans apart from automated entities are disclosed herein. Persistence of vision and simultaneous contrasts are some of the properties of human vision that can be used in these tests. A video of an image is generated in colors that are distinguishable to the human eye but are not easily distinguished numerically. The image includes text manipulated such that positive image data and negative whitespace data occur at equal rates along with a noise component included in each of the video frames. Thus, raw data is made ambiguous while qualities of human visual interpretation are relied upon for extracting relevant meaning from the video.

Claims

1. A system for user verification, comprising:

a security component comprising a test generation component for generating one or more automated tests that rely on a plurality of human vision properties comprising at least persistence of vision and simultaneous contrast wherein simultaneous contrast comprises enhancement or diminishment, relative to normal, of perception, cognition and related performance as a result of immediately previous or simultaneous exposure to a stimulus of lesser or greater value in the same dimension, to differentiate a human user from an automated entity, wherein the test generation component further comprises a segmenting component which segments a target image into a grid of segments of predetermined sizes, and a frameset generation component for generating at least one frameset for at least one of the automated test, wherein the frameset generation component further comprises one or more of a color mapping component that determines which pixels of the target image are positive and where to put negative whitespace data prior to generating the frameset or a noise generation component that introduces noise into each frame of the frameset;

a communication component configured to execute one or more of transmission of the tests to one or more access requestors or reception of responses from the access requestors; and

a nontransitory computer readable medium having stored thereon one or more of the security component or the communication component.

2. The system of claim 1, wherein the test generation component generates a video-based test from a target image, wherein the video-based test relies on the plurality of human vision properties for the differentiation.

3. The system of claim 2, wherein the video-based test generated by the test generation component is a Completely Automated Public Turing-test to tell Computers and Human Apart (CAPTCHA).

4. The system of claim 2, wherein the target image comprises a selection of text.

5. The system of claim 4, wherein the selection of text comprises one or more of one or more randomly selected letters or meaningful words.

6. The system of claim 1, wherein the segments are all of the same size.

7. The system of claim 1, wherein the test generation component further comprises a color generation component to generate or select a set of colors for manipulating the target image based at least on one of the plurality of human vision properties.

8. The system of claim 7, wherein the set of colors comprises at least three colors such that positive and negative data in the target image are visually differentiable while being numerically ambiguous.

9. The system of claim 1, wherein the frameset comprises equal rates of positive data along with the negative whitespace data in addition to negative non-whitespace data.

10. The system of claim 9, wherein the positive data occurs every 0.05 seconds.

11. The system of claim 1, wherein the test generation component generates the target image of size 104×

44 pixels wherein the target image is divided into 26 vertical segments and 11 horizontal segments, resulting in the video being displayed at 60 FPS with 4 FPFS.

12. A method of user verification employing a processor operating to perform actions in response to executing computer program instructions, comprising:

receiving an access request;

generating, by a nontransitory computer readable medium having stored thereon instructions for generating, one or more video-based automated tests that are based on at least two properties of human vision comprising at least persistence of vision and simultaneous contrast wherein simultaneous contrast comprises enhancement or diminishment, relative to normal, of perception, cognition and related performance as a result of immediately previous or simultaneous exposure to a stimulus of lesser or greater value in the same dimension wherein generating the one or more video-based automated tests further comprises generating a video of a target image comprising text data of one or more of randomly selected letters or meaningful words, segmenting the target image into a grid of segments wherein each segment is of a predetermined size, generating a set of colors for inclusion into the target image in the video, and splitting the grid of segments across a plurality of framesets such that occurrence rates of the positive data comprising the target image or the text data and negative data including the whitespace data is the same;

issuing the one or more tests;

receiving a response to the one or more tests from the access requestors; and

verifying that the access request originated from a human user based at least on the response.

13. The method of claim 12, wherein the set of colors is generated such that positive and negative data in the target image are visually differentiable while being numerically ambiguous.

14. The method of claim 12, wherein generating a set of colors comprises generating a set of three or more colors that exploits simultaneous contrast characteristic of human vision.

15. The method of claim 12, further comprising adding a noise component to each segment in a deterministic way which leads to reduced visibility of the text in each frame, but identifiable by a human user when combined in a frameset and shown as a video CAPTCHA.

16. A nontransitory computer readable medium having stored thereon instructions for:

generating one or more video-based automated tests that are based on at least two properties of human vision comprising at least persistence of vision and simultaneous contrast wherein simultaneous contrast comprises enhancement or diminishment, relative to normal, of perception, cognition and related performance as a result of immediately previous or simultaneous exposure to a stimulus of lesser or greater value in the same dimension

generating a video of a target image comprising text data of one or more of randomly selected letters or meaningful words;

segmenting the target image into a grid of segments of predetermined sizes;

splitting the grid of segments across a plurality of framesets such that occurrence rates of the positive data comprising the target image or the text data and negative data including the whitespace data is the same;

receiving a response to the one or more tests; and

verifying that the response is from a human user.

17. The computer readable medium of claim 16, further comprising instructions for generating a set of colors for inclusion into the target image in the video.

18. The computer readable medium of claim 17, further comprising instructions for generating the set of colors such that positive and negative data in the target image are visually differentiable while being numerically ambiguous.

19. The computer readable medium of claim 17, further comprising instructions for generating the set with three or more colors that exploit simultaneous contrast characteristic of human vision.

20. The computer readable medium of claim 16, further comprising instructions for including the positive data within the frameset such that the positive data occurs every 0.05 seconds in the video of the target image.

21. The computer readable medium of claim 16, further comprising instructions for adding a noise component to each segment in a deterministic way which leads to reduced visibility of the text in each frame, but identifiable by a human user when combined in a frameset and shown as a video CAPTCHA.

22. A system for user verification, comprising:

a security component comprising software operable to confirm that an access request is received from a human rather than an automated program, the security component comprising a test generation component comprising software operable to generate one or more automated tests that rely on a plurality of human vision properties to differentiate a human user from an automated entity, the test generation component comprising a segmenting component which segments a target image into a grid of segments of predetermined sizes and a frameset generation component for generating at least one frameset for at least one of the automated tests, wherein the frameset generation component comprises one or more of a color mapping component that determines which pixels of the target image are positive and where to put negative whitespace data prior to generating the frameset or a noise generation component that introduces noise into each frame of the frameset;

a communication component comprising software configured to execute one or more of transmission of the tests to one or more access requestors or reception of responses from the access requestors; and

a nontransitory computer readable medium having stored thereon one or more of the security component or the communication component.

23. A method of user verification employing a processor operating to perform actions in response to executing computer program instructions, comprising:

receiving an access request;

generating, by a nontransitory computer readable medium having stored thereon instructions for generating, one or more automated tests, at least one of the tests comprising a video-based automated test comprising a video of a target image comprising text data of one or more of randomly selected letters or meaningful words, further comprising segmenting the target image into a grid of segments wherein each segment is of a predetermined size, generating a set of colors for inclusion into the target image in the video, and splitting the grid of segments across a plurality of framesets such that occurrence rates of positive data comprising the target image or the text data and negative data including whitespace data is the same;

issuing the one or more tests;

receiving a response to the one or more tests from the access requestors; and

verifying that the access request originated from a human user based at least on the response.

24. A nontransitory computer readable medium having stored thereon instructions for:

generating one or more video-based automated tests comprising a video of a target image comprising text data of one or more of randomly selected letters or meaningful words, segmenting the target image into a grid of segments of predetermined sizes, and splitting the grid of segments across a plurality of framesets such that occurrence rates of positive data comprising the target image or the text data and negative data including whitespace data is the same;

receiving response to the one or more tests; and

verifying that the response is from a human user.