Image processing

Image analysis – Pattern recognition – Classification

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

Reexamination Certificate

active

06788819

ABSTRACT:

FIELD OF THE INVENTION
This invention relates to a method and apparatus for automated processing of images. It is particularly useful with images of documents for identifying the types of document shown in the image, such as may be used in an automated mail processing system.
BACKGROUND OF THE INVENTION
The complexity of document images is highly variable. This is particularly apparent in mail processing systems where envelopes can have many forms. The simplest type would be a plain white envelope with only an address indicated on it. More complex envelopes may have other material printed on the envelope such as a return address. There are also, e.g., different coloured envelopes, envelopes with windows which enable the address to be seen, plastic envelopes with material printed on them and which also have partial transparency thereby allowing some of the documents within to be seen.
It has been appreciated that it is often most effective to apply different image processing methods to images of different degrees of complexity when performing subsequent analysis. For example, when locating and attempting to read an address on an envelope, different processing methods will be required for the different types of envelope. A plain white envelope would be most simple as the address can be located very easily and then optical character recognition applied to it. A more complex envelope such as a plastic and partially transparent envelope will require additional processing to identify the address prior to optical character recognition.
Previous methods for automated mail processing have used the same type of image processing for all types of envelopes. We have appreciated that this is computationally inefficient as some documents are subjected to processing which is not appropriate to them because of the simplicity of their images.
SUMMARY OF THE INVENTION
A preferred embodiment of the present invention therefore seeks to provide a method and apparatus for identifying images of different types of document so that subsequent processing may be performed on each image in dependence on its document type. By doing this, it is possible to ensure that the most appropriate subsequent analysis process is applied to each document.
It is important that this analysis of document type is computationally efficient since the intention is to reduce the overall computational overhead of processing.


REFERENCES:
patent: 4514816 (1985-04-01), Ollus et al.
patent: 4577235 (1986-03-01), Kannapell et al.
patent: 5072291 (1991-12-01), Sekizawa
patent: 5200831 (1993-04-01), Tai
patent: 5642442 (1997-06-01), Morton et al.
patent: 5978519 (1999-11-01), Bollman et al.
patent: 6118552 (2000-09-01), Suzuki et al.
patent: 6163624 (2000-12-01), Kobayashi
patent: 6529634 (2003-03-01), Thyagarajan et al.
patent: 2002/0044691 (2002-04-01), Matsugu
patent: 0 195 925 (1986-10-01), None
patent: 2 246 230 (1992-01-01), None
patent: WO 97/09133 (1997-08-01), None
Search Report for GB 0016216.4, dated Jan. 19, 2004.
International Search Report dated Nov. 11, 2001 in PCT International Appl. No. PCT/GB01/02915 (3 pages).
English translation of Description and Claims of JP 11-513023T2, published Oct. 19, 1999 (5 pages).

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Image processing does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Image processing, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-3209824

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.