Granted Patent B2

US 12,217,405 · App. 17/291,035 · Granted Feb 4, 2025

Inventors: Benjamin Neeman (Raanana, IL); Yael Moscovitz (Mevaseret Zion, IL); Sergey Markin (Herzeliya, IL); Evgeny Ostrovsky (Petah Tikva, IL)

Assignee: AU10TIX LTD.

G06T7/0002G06T7/40G06T2207/10008G06T2207/20056G06T2207/30176

Abstract

Claims (42)

1. A method for processing images generated by unsupervised end-users, the method comprising:

determining whether at least one digital image of a document was generated by scanning an original document or was not generated by scanning the original document, including using a processor for applying a resolution analysis test to a textured region of interest; and

sending at least one output indication of the determining operation,

wherein said image is captured by a client camera device, and wherein the resolution analysis test is performed, by a server deployed remotely vis-à-vis the client camera device, on an image I of the region of interest which is sent to the server and which, to conserve bandwidth, includes less than all of the original document

wherein the determining is performed for plural images of plural documents belonging to different known classes of documents for which the region of interest appears at respective different locations, known to the server, within the document, and

wherein the client camera device first sends the image at a first resolution to the server

which determines which class c from among said classes the image belongs to,

then the server commands the client camera device to send a region appearing at a location within the image,

which is known to the server to include the known location of the region of interest for class c,

and then the client camera device sends only the region and not the entire image, at a second resolution higher than said first resolution, to the server which applies said test.

2. The method according to claim 1 wherein the textured region of interest appears at a known location for each document belonging to a known class of documents.

3. The method according to claim 1 , wherein the resolution analysis test yields results which are compared to at least one threshold value separating test results yielded by images of documents generated by scanning the original document, from test results yielded by images of documents generated by scanning a copy of the original document.

4. The method according to claim 3 , further comprising pre-testing to establish said at least one threshold value, including applying the resolution analysis test, for each of a first plurality of images each known to have been generated from an original instance of the class of documents, and for each of a second plurality of images each known to have been generated from a copy of the original instance of the class of documents, thereby to accumulate results of the resolution analysis test for each of the first and second pluralities of images, identifying at least one value V separating at least one range of results yielded by the first plurality of images from at least one range of results yielded by the second plurality of images, and setting said at least one threshold value to V.

5. The method according to claim 1 , wherein said textured region of interest, to which the resolution analysis test is applied, comprises a repeating pattern.

6. The method according to claim 1 , wherein said textured region of interest comprises a repeating pattern which is subjected to the resolution analysis test which analyzes the texture's resolution, since an image generated from a copy of the original document, is typically smoother and/or has less high resolution, than an image generated from the original document itself.

7. The method according to claim 1 , wherein the resolution analysis test comprises a Fourier-transform based spectrum analysis test.

8. The method according to claim 1 , wherein said determining includes determining whether or not the digital image was obtained by scanning a copy of the original document.

9. A computer program product, comprising a non-transitory tangible computer readable medium having computer readable program code embodied therein, said computer readable program code adapted to be executed to implement a method for processing images generated by unsupervised end-users, the method comprising:

determining whether at least one digital image of a document was generated by scanning an original document or said at least one digital image of the document was generated by scanning a copy, including using a processor for applying a resolution analysis test to a textured region of interest in said digital image; and

sending at least one output indication of the determining operation,

wherein said image is captured by a client camera device, and wherein the resolution analysis test is performed, by a server deployed remotely vis-à-vis the client camera device, on an image I of the region of interest which is sent to the server and which, to conserve bandwidth, includes less than all of the original document

wherein the determining is performed for plural images of plural documents belonging to different known classes of documents for which the region of interest appears at respective different locations, known to the server, within the document, and

wherein the client camera device first sends the image at a first resolution to the server

which determines which class c from among said classes the image belongs to,

then the server commands the client camera device to send a region appearing at a location within the image,

which is known to the server to include the known location of the region of interest for class c,

and then the client camera device sends only the region and not the entire image, at a second resolution higher than said first resolution, to the server which applies said test.

10. A system for processing images generated by unsupervised end-users, the system comprising:

a processor configured for applying a resolution analysis test to a textured region of interest in a digital image and, accordingly, determining whether said digital image was generated by scanning an original document or said digital image was not generated by scanning the original document; and

an output device in data communication with said processor for sending at least one output indication of the determining operation,

wherein said image is captured by a client camera device, and wherein the resolution analysis test is performed, by a server deployed remotely vis-à-vis the client camera device, on an image I of the region of interest which is sent to the server and which, to conserve bandwidth, includes less than all of the original document

wherein the determining is performed for plural images of plural documents belonging to different known classes of documents for which the region of interest appears at respective different locations, known to the server, within the document, and

wherein the client camera device first sends the image at a first resolution to the server

which determines which class c from among said classes the image belongs to,

then the server commands the client camera device to send a region appearing at a location within the image,

which is known to the server to include the known location of the region of interest for class c,

and then the client camera device sends only the region and not the entire image, at a second resolution higher than said first resolution, to the server which applies said test.

11. The method according to claim 1 , wherein the output indication comprises information indicating that the document was scanned from a copy, not from the original.

12. The method according to claim 1 , wherein the textured region of interest where the resolution analysis test is applied comprises a background portion of the document.

13. The system according to claim 10 , wherein the textured region of interest where the resolution analysis test is applied comprises a background portion of the document.

14. The method according to claim 3 , wherein the resolution analysis test comprises a Fourier-transform based spectrum analysis test.

15. The method according to claim 1 wherein the image I is provided by the client camera device to the remote server.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (30)