Systems and methods of processing scanned data

Patent metadata

Field	Value
Publication number	US-9769354-B2
Application number	US-201514814455-A
Country	US
Kind code	B2
Filing date	Jul 30, 2015
Priority date	Mar 24, 2005
Publication date	Sep 19, 2017
Grant date	Sep 19, 2017

Abstract

An efficient method and system to enhance digital acquisition devices for analog data is presented. The enhancements offered by the method and system are available to the user in local as well as in remote deployments yielding efficiency gains for a large variety of business processes. The quality enhancements of the acquired digital data are achieved efficiently by employing virtual reacquisition. The method of virtual reacquisition renders unnecessary the physical reacquisition of the analog data in case the digital data obtained by the acquisition device are of insufficient quality. The method and system allows multiple users to access the same acquisition device for analog data. In some embodiments, one or more users can virtually reacquire data provided by multiple analog or digital sources. The acquired raw data can be processed by each user according to his personal preferences and/or requirements. The preferred processing settings and attributes are determined interactively in real time as well as non real time, automatically and a combination thereof.

First claim

What is claimed is:

1. A computer-implemented data processing method comprising:
receiving image data from a data acquisition device;
automatically analyzing, using a first analytic engine, at least portions of the image data to determine whether the image data is within a first set of parameters;
generating a first set of processor settings when the image data is not within the first set of parameters;
processing the image data with the first set of processor settings to generate processed data;
automatically analyzing, using a second analytic engine, at least portions of the processed data to determine whether the processed data is within a second set of parameters and if not, generating a second set of processor settings using the second analytic engine, wherein generating the second set of processor settings comprises independently identifying a binarization threshold range for each image feature depicted in the at least portions of the processed data;
reprocessing the image data using the second set of processor settings to generate reprocessed data; and
outputting at least one of the processed data and the reprocessed data;
wherein the second set of parameters is different than the first set of parameters; and
wherein the first set of processor settings and the second set of processor settings each relate to one or more image characteristics selected from: brightness, contrast and gamma.

2. The method as recited in claim 1 , further comprising performing a plurality of thresholding iterations, each thresholding iteration comprising:
the analyzing at least portions of the processed data to determine whether the processed data is within the second set of parameters, and if not, generating the second set of processor settings to reprocess the image data; and
the reprocessing the image data with the second set of processor settings to generate the reprocessed data, and
wherein for each of the plurality of thresholding iterations, the second set of processor settings comprises a different binarization threshold than a binarization threshold employed in an immediately previous thresholding iteration.

3. The method as recited in claim 2 , each thresholding iteration further comprising: extracting image features from one or more of the processed data and the reprocessed data.

4. The method as recited in claim 2 , each thresholding iteration further comprising: extracting textual information from one or more of the processed data and the reprocessed data, and the method further comprising assembling the extracted textual information from two or more of the thresholding iterations.

5. The method as recited in claim 1 , further comprising performing a plurality of thresholding iterations on a per-feature basis.

6. The method as recited in claim 4 , wherein the extracted textual information comprises a plurality of characters, and the method further comprising:
determining a threshold level of each of the plurality of characters from each of the two or more thresholding iterations; and
selecting a set of optimal characters based on the threshold level of each of the plurality of characters; and
wherein the assembling further comprises assembling the set of optimal characters into a composite image.

7. The method as recited in claim 1 , wherein the at least portions of the image comprise one or more trouble regions identified based on a learn-by-example (LBE) training process.

8. The method as recited in claim 4 , further comprising validating the extracted textual information.

9. The method as recited in claim 1 , wherein the processing includes applying a color conversion algorithm, and
wherein applying the color conversion algorithm comprises normalizing intensity values across one or more color channels of the image data, the one or more color channels being selected from R, G and B.

10. The method as recited in claim 1 , wherein at least one of the first set of parameters and the second set of parameters comprise a classification confidence value threshold.

11. The method as recited in claim 1 , wherein each of the first set of processor settings and the second set of processor settings comprise one or more binarization thresholds.

12. The method as recited in claim 1 , wherein analyzing at least portions of the image data to determine whether the image data is within the first set of parameters comprises optical character recognition (OCR).

13. The method as recited in claim 1 , wherein analyzing at least portions of the image data to determine whether the image data is within the first set of parameters comprises a connected components analysis.

14. The method as recited in claim 1 , wherein the at least portions of the image data comprise a plurality of regions of interest each depicting one or more image features, and the method further comprising locating the plurality of regions of interest within the image data based at least in part on a learn-by-example (LBE) training operation; and
wherein analyzing at least portions of the image data to determine whether the image data is within the first set of parameters comprises an image classification operation based at least in part on the learn-by-example (LBE) training operation.

15. The method as recited in claim 1 , further comprising inverting the image data.

16. A computer program product configured to perform data processing, the computer program product comprising a non-transitory computer readable storage medium having stored thereon computer readable program instructions configured to cause one or more processors, upon execution of the instructions, to:
receive image data from a data acquisition device;
analyze at least portions of the image data to determine whether the image data is within a first set of parameters;
generate a first set of processor settings when the image data is not within the first set of parameters;
process the image data with the first set of processor settings to generate processed data;
perform a plurality of thresholding iterations, each thresholding iteration comprising:
analyzing at least portions of the processed data to determine whether the processed data is within a second set of parameters, and if not, generating a second set of processor settings to reprocess the image data;
reprocessing the image data with the second set of processor settings to generate the reprocessed data;
extracting textual information from one or more of the processed data and the reprocessed data; and
wherein for each of the plurality of thresholding iterations, the second set of processor settings comprises a different binarization threshold than a binarization threshold employed in an immediately previous thresholding iteration; and
assemble the extracted textual information from two or more of the thresholding iterations; and
output at least one of the processed data and the reprocessed data,
wherein the second set of parameters is different than the first set of parameters, and
wherein the first set of processor settings and the second set of processor settings each relate to one or more image characteristics selected from: brightness, contrast and gamma.

17. The computer program product as recited in claim 16 , the computer readable program instructions being further configured to cause one or more processors, upon execution of the instructions, to: perform a plurality of thresholding iterations, each thresholding iteration comprising:
analyzing at least portions of the processed data to determine whether the processed data is within the second set of parameters;
gene

Assignees

• Kofax Inc

Inventors

• Thrasher Christopher W
• Shustorovich Alexander
• Thompson Stephen Michael
• Amtrup Jan W
• Macciola Anthony
• Borrey Roland G
• Schmidtler Mauritius A R
• Taylor Robert A
• Fechter Joel S
• Asuri Hari S

Classifications

• G06V10/242

  by image rotation, e.g. by 90 degrees · CPC title

• G06V30/40

  Document-oriented image-based pattern recognition · CPC title

• G06V10/993

  Evaluation of the quality of the acquired pattern · CPC title

• H04N2201/0093

  Facsimile machine (H04N2201/0094 takes precedence) · CPC title

• H04N1/6027

  Correction or control of colour gradation or colour contrast (H04N1/6058 takes precedence) · CPC title