Browse Prior Art Database

Mark Recognition Method

IP.com Disclosure Number: IPCOM000015493D
Original Publication Date: 2002-Jul-01
Included in the Prior Art Database: 2003-Jun-20
Document File: 3 page(s) / 81K

Publishing Venue

IBM

Abstract

[Title]

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 55% of the total text.

Page 1 of 3

Mark Recognition Method

[Title]

Mark Recognition Method

[Technical Area]

The technical area of this method is mark recognition. Marks are written over fixed preprinted figures, such as Fig. 1. The method realizes high speed and high accuracy recognition on binary images. It can be applied to OCR (Optical Character Recognition), OMR (Optical Mark Recognition), and other applications.

ยทยทยทยท

(a) Original Form (b) Marked Sample

Fig. 1

[Problem to be solved]

In general, to recognize a mark that is written over a fixed preprinted figure, the figure has to be registered beforehand and used for eliminating or ignoring preprinted pattern in a marked image. The mark recognition process determines the presence of the mark mainly by counting the rest of the black dots.

One of the basic problems of this technique is that the registered figure may not completely match with the marked image due to any deviation factors in scanning or position adjustment, such as skew, distortion, density, and so on.

There is a generally-known method that dilates the registered shape(line) to absorb the image deviation. The dilating size is equal or larger than the size of maximum of estimated deviation.

But the dilated also may include the mark's black dots if the mark is written close to the preprinted image. So there is a risk that such marks are not recognized. Fig. 2 shows such case. Fig. 2 (a) is the preprinted figure. Fig. 2 (b) is the marked image. Fig. 2 (c) is the remained black dots after elimination by the conventional method. In this case, the number of remained black dots is not enough to determine the mark existence, since it is possibly noise.

(a) (b) (c)

Fig. 2

[Method description]

The disclosed method also uses dilated preprinted figure like the conventional method aforementioned. However the dilating size is one dot smaller in each direction than the conventional method. The direction means the four aspects: up, down, left and right. In addition, after the preprinted image is eliminated, it is passed through a filter. The filter is a

1

[This page contains 5 pictures or other non-text objects]

Page 2 of 3

simple one that eliminates one-dot-width-noise. It is based on the following reasons:

If no mark is written, the remaining black dots are of the preprinted image and its horizontal or vertical...