Method of image pre-analyzing of a machine-readable form of non-fixed layout

• US 8,170,371 B2
• Filed: 06/26/2003
• Issued: 05/01/2012
• Est. Priority Date: 03/28/2003
• Status: Active Grant

• Current Assignee: ABBYY Development LLC
• Original Assignee: ABBYY Software
• Inventors: Zuev, Konstantin, Garashchuk, Ruslan
• Primary Examiner(s): Carter, Aaron W
• Application Number:
  US10/603,216
• Publication Number:
  US 20040190790A1
• Time in Patent Office: 3,232 Days
• Field of Search: 382291-292, 382/112, 715505-508, 715900-911, 715221-226, 715243-254
• US Class Current: 382/291
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables
• CPC Class Codes: G06V 30/414 Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text

Abstract

A method for pre-recognition processing of forms with non-fixed fields. One or more objects that are present on a form can be reliably identified after conversion of the form into an electronic state. Objects are preliminarily assigned to act as reference points for spatial binding of data input fields or groups thereof. In the case of a text object as a reference point, text recognition is additionally performed. The spatial location of a reference point may be not fixed. A reference point may be described as alternative. Field identification may be performed either automatically or manually.

Claims

1. A method of pre-processing of a machine-readable form with non-fixed fields layout, the method comprising:

acquiring a bit-mapped image of the machine-readable form filled in with print in one or more non-fixed data input fields;

identifying at least one form model, the at least one form model containing spatial and parametric properties of objects thereof;

preliminarily assigning at least one object of the form as a reference point for spatial binding of at least one non-fixed data input field thereof;

eliminating any skew, distortion and noise in the bit-mapped image;

parsing the bit-mapped image into regions; and

defining a spatial location of at least one non-fixed data input field relative to at least one reference point, wherein said defining the location of the at least one non-fixed data input field comprises;

selecting a non-fixed data input field to search in the at least one form model;

accepting from the at least one form model at least one reference point property for a spatial relative reference of the said data input field;

searching and locating said at least one reference point on the form bit-mapped image;

searching and locating the said data input field on the form bit-mapped image relative to at least one reference point taking into account all spatial and parametrical properties described in the form model; and

identifying each data input field position in the case of multiple data input fields.

2. The method as recited in claim 1, wherein the said reference point is represented by a text region.

3. The method as recited in claim 2, wherein the said text region is additionally subjected to a text recognition.

4. The method as recited in claim 1, wherein in the case of multiple data input fields the said identification of each data input field is performed via setting up and accepting of hypotheses and compliance estimation of the form model.

5. The method as recited in claim 4, wherein an additional information about each of said non-fixed data input field is used.

6. The method as recited in claim 1, wherein a non-fixed data input field may be used as a reference point.

7. The method as recited in claim 1, wherein identifying each data input field in the case of multiple data input fields is performed at least partly manually.

8. The method as recited in claim 1, wherein the spatial location of a reference point is not fixed from a first scan of the machine-readable form to a second scan of the machine-readable form.

9. The method as recited in claim 1, wherein one reference point is used for spatial binding of more than one data input field.

10. The method as recited in claim 1, wherein the spatial binding of one data input field is performed relative to more than one reference point.

11. The method as recited in claim 1, wherein a reference point comprises more than one form object.

12. The method as recited in claim 1, wherein the identifying each data input field position in the case of multiple data input fields is a profound identification.