Method for detecting blank pages, and corresponding multifunctional machine

• US 8,724,182 B2
• Filed: 09/10/2010
• Issued: 05/13/2014
• Est. Priority Date: 09/10/2009
• Status: Active Grant

• Current Assignee: Sagemcom Documents SAS (Sagemcom Holding SAS)
• Original Assignee: Sagemcom Documents SAS (Sagemcom Holding SAS)
• Inventors: Labois, Olivier, Sabatier, Pierre, Berger, Jrme
• Primary Examiner(s): Worku, Negussie
• Application Number:
  US13/395,101
• Publication Number:
  US 20120170082A1
• Time in Patent Office: 1,341 Days
• Field of Search: 358/474, 358/475, 358/498, 358/486, 358/501, 358/505
• US Class Current: 358/474
• CPC Class Codes: G06V 10/28 Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
• CPC Class Codes: G06V 30/422 Technical drawings; Geographical maps
• CPC Class Codes: H04N 1/00795 Reading arrangements details of scanning heads H04N1/024; scanning arrangements therefor H04N1/04
• CPC Class Codes: H04N 1/00803 Presence or absence of information
• CPC Class Codes: H04N 2201/0094 Multifunctional device, i.e. a device capable of all of reading, reproducing, copying, facsimile transception, file transception

Abstract

The invention relates to a method for detecting blank pages among sheets of a document scanned on a multifunctional machine comprising a scanner, wherein said scanner: scans (S0) two pages of at least one sheet of a document to be scanned; carries out a basic detection (S6) of a blank page on each scanned sheet; selects (S7) each sheet not comprising at least one page detected to be blank during the detection (S6); compares (S9) the two pages of said selected sheet; and classifies (S9) a page as blank when it is similar to the other page of said sheet. The invention also relates to a multifunctional machine, the machine comprising a scanner for implementing a method according to the invention.

Claims

1. A method for detecting blank pages in the sheets of a document scanned on a machine including a scanner, wherein the scanner

scans two pages of at least one sheet of a document to be scanned;

performs at least one blank page detection on each scanned sheet;

selects each sheet which does not have at least one page which was detected as blank in this blank page detection;

compares the content of the two pages of each selected sheet, said comparison comprising comparing at least a characteristic vector which models the content of each page of each selected sheet, and

classifies a page as blank when this comparison is above a similarity threshold.

2. A method according to claim 1, wherein the scanner

scans the pages to transform each page into a set of pixels belonging to ^é

the space of real numbers, and transposes each pixel so that it has a gray level, and

also performs, prior to the selection of each sheet not having at least one page detected as blank, a binarisation of the pixels of the pages of each sheet, the binarisation consisting of assigning

an active value to each pixel the gray level whereof exceeds a threshold t, and

an inactive value to a pixel the gray level whereof is below the threshold t,

where t is defined dynamically according to the neighborhood of the pixel.

3. A method according to claim 2, wherein,

the scanner assigns subsequent to the binarisation, the front page R of a sheet to the page having the greater probability of being information-bearing, the scanner assigning the hack page V to the other page of the sheet and wherein, to perform the step of comparison between the two pages,

the scanner constructs first of all a local dissimilarities map CDL, made up of pixels CDL(x) having a position x, the value of each pixel CDL(x) being given by the formula:

∀

xε

^é

,CDL(x)=( R(x)·

V(x))·

d(x,R)

where

x represents the position of the pixel on a page;

V(x) is the value of the pixel of the back page at the position x;

R(x) is the inverse of the value of the front page pixel at the position x;

( R(x)·

V(x)), is equivalent to a logical AND operator

d(x,R) is the minimum distance calculated between the position x and the pixels of the front page having an active value;

the scanner then constructs a histogram collecting all the pixels CDL(x) having the same value, for each value of CDL(x), and

the scanner considers that the front R and back V pages are similar when the number of pixels of the local dissimilarities map CDL, the value whereof is less than a predetermined distance value d, is greater than a proportion H, and thus classifies the corresponding back page V as a blank page.

4. A method according to claim 3, wherein the scanner calculates d(x,R) using the so-called Manhattan distance.

5. A method according to claim 3, wherein the predetermined distance value d corresponds to a reference distance d(x,R) of three pixels, the proportion 11 being 70% of the pixels of the local dissimilarities map.

6. A method according to claim 3, wherein, during the calculation of the distance d(x,R), the scanner directly assigns the value (d+1) to the pixel CDL(x) whenever the distance d(x,R) strictly exceeds the reference distance at the point x, and stops the corresponding distance calculation.

7. A method according to claim 3, wherein the step consisting of at least one blank page detection on each scanned sheet comprises, subsequent to the binarisation, as step whereby the scanner performs a second blank page or information-bearing page detection on each binarised page, the step of assigning a front page and an back page being then carried out on each sheet not having at least one page detected as blank during the second detection.

8. A method according to claim 1, wherein the step of at least one blank page detection on each scanned sheet comprises steps whereby the scanner

performs a first blank page or information-bearing page detection on each scanned page, in gray levels;

and

selects each sheet every page whereof has not been detected as blank or information-bearing during the first detection.

9. A method according to claim 1, wherein the scanner performs a step of edge suppression and/or size readjustment and/or transposition into gray levels on the scanned pages.

10. A machine, comprising a scanner implementing a method according to claim 1.