Method and system for image processing

• US 7,436,979 B2
• Filed: 03/29/2002
• Issued: 10/14/2008
• Est. Priority Date: 03/30/2001
• Status: Expired due to Term

• Current Assignee: Korber Supply Chain LLC
• Original Assignee: Siemens Energy & Automation
• Inventors: Bruce, Ben F., Kessler, Linda J.
• Primary Examiner(s): Lu; Tom Y
• Application Number:
  US10/113,814
• Publication Number:
  US 20020168090A1
• Time in Patent Office: 2,391 Days
• Field of Search: 382/101, 382/171, 382173-231, 382181-190, 382/198, 382/284, 382/321
• US Class Current: 382/101
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/127 with the intervention of an operator
• CPC Class Codes: G06V 30/1444 Selective acquisition, locating or processing of specific regions, e.g. highlighted text, fiducial marks or predetermined fields

Abstract

A method of processing an image containing written information according to the invention includes the steps of scanning a surface of an object to obtain an image of the surface represented by image data, analyzing the image data to determine locations of one or more regions with written indicia on the surface, where indicia may include logos, stamps, meter marks, alphanumeric information, labels or other distinctive marks, creating one or more sub-images substantially limited to the dimensions of the indicia-bearing regions, transmitting the sub-images from a first computer or process used to create the sub-images to a second computer or process, and further analyzing the sub-images at the second computer or process, such as by OCR, to decode the written indicia in the sub-images. By transmitting the sub-images, preferably in combination with position data indicating where each sub-image was on the original image, the scanned written indicia can be decoded automatically or manually without having to transmit the original image from one computer or process to another. This is particularly important when the image data is a grayscale or color image that is much larger than a black and white (binary) image of the same dimensions.

Claims

1. A method of processing an image of a mail piece having containing written delivery information, comprising the steps of:

(a) presenting a series of mail pieces one after another to a scanner;

(b) scanning a surface of each mail piece with the scanner to obtain an image of the surface represented by image data;

(c) analyzing the image data to determine locations of one or more regions with written indicia on the surface, the remainder of the image data comprising background image data;

(d) creating one or more sub-images in a grayscale or color format substantially limited to the dimensions of the indicia-bearing regions, which indicia bearing regions include a delivery address;

(e) electronically transmitting the sub-images substantially free of background image data, and also transmitting coordinates of the sub images indicating where the subimage was located in the surface image whereby each sub-image can be displayed at a position that corresponds to its original location in the image of the mail piece surface allowing re-creation of the surface image without the background image data, from a first computer or a first process used to create the sub-images to a second computer or a second process;

(f) further analyzing the sub-images at the second computer or second process to decode the written indicia in the sub-images by applying character recognition logic to the sub-images, where the written indicia further may include one or more of logos, stamps, meter marks, alphanumeric information, labels or other distinctive marks;

(g) determining the identity of a sub-image from among a predetermined set of possible postal data types based on criteria including the position of the sub-image on the surface; and

(h) determining a delivery address for each mail piece.

2. The method of claim 1, further comprising forming a data packet containing the sub-images and associated coordinates for the image data from each mail piece.

3. The method of claim 1, wherein step (h) further comprises, if applying character recognition logic to a sub-image does not decode a sub-image to a predetermined level of confidence, displaying the sub-image to a human operator for editing of data fields which correspond to information in the sub-images.

4. The method of claim 3, wherein the transmitting step (e) further comprises transmitting image coordinates of the sub-images to the second computer or second process, whereby each sub-image can be displayed at a position that corresponds to its original location in the image, and step (h) further comprises reconstructing a facsimile of the scanned image of the mail piece surface by simultaneous display of the sub-images at positions indicated by the image coordinates.

5. The method of claim 1, wherein the scanner is part of an automated mail sorting machine which sorts the mail pieces based on the delivery address.