Dismiss
InnovationQ/InnovationQ Plus content will be updated on Sunday, June 25, 10am ET, with new patent and non-patent literature collections. Click here to learn more.
Browse Prior Art Database

MIXED-MODE ALPHANUMERIC/FAX MESSAGING

IP.com Disclosure Number: IPCOM000007253D
Original Publication Date: 1994-Jun-01
Included in the Prior Art Database: 2002-Mar-07
Document File: 3 page(s) / 186K

Publishing Venue

Motorola

Related People

Jyh-Han Lin: AUTHOR

Abstract

The problem of sending alphanumeric messages to the paging terminal is a long-standing problem in alphanumeric paging. Two approaches that employ fax machines have been proposed to alleviate this problem: 1. Fax paging combined with binary image compres- sion. This approach is highly demanding in terms of capacity.

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 45% of the total text.

Page 1 of 3

MO-LA Technical Developments Volume22 June 1994

MIXED-MODE ALPHANUMERIC/FAX MESSAGING

by Jyh-Han Lin

  The problem of sending alphanumeric messages to the paging terminal is a long-standing problem in alphanumeric paging. Two approaches that employ fax machines have been proposed to alleviate this problem:

1. Fax paging combined with binary image compres- sion. This approach is highly demanding in terms of capacity.

2. Fax paging combined with off-line handwriting recognition technology. The problem with this approach is that even if we constrain the user to write in boxes, the recognition rates of numbers range between 90% to 95% and the recognition rates of English letters are usually hovering around in the low 80%. Therefore, the system is not relia- ble enough to deliver crucial information.

  An alternative approach is to combine hand- writing recognition with binary image compression to achieve a balance between reliability and cost. We call this approach "mixed-mode alphanumeric/ fax messaging?' This paper proposes a scheme for realizing this type of system, including a new algo- rithm that tItl1y takes advantage of binary image com- pression, a alphanumeric message form, a new "com- pound message" format, and a new way of presenting the "compound message" to the user.

to the paging terminal via a fax machine. Figure 3 also shows mixed-mode messaging in Chinese ide- ogram, where each letter box contains one Chinese ideographic character. We assume that off-line hand- writing recognition software operates at a recogni- tion unit of individual characters. The use of the form simplifies character segmentation. The tech- nique itself, however, can be easily adapted to other forms and recognition units, e.g., words and phases.

  For each character bitmap to be recognized, most handwriting recognition software usually provides two pieces of information after recognition: The first piece of information is a list of probable characters corresponding to the character bitmap. The second piece of information is a set of numbers that indi- cate the degree ofconfidence for each probable char- acters. With no context information, we usually select the character with the greatest confidence value.

  For each character bitmap P, we use the conti- dence value L of the most probable character C as a cue to decide whether to send the recognized char- acter only or to send its raw bitmap as well. That is, given a threshold value T (which can be either be fixed or set dynamically), when L is greater than T, then we use character C and P so the user can view the raw bitmap to see what the character really is. We can use a spell checker and a grammar checker to compliment the thresholding. The raw bitmaps that have to be sent are joined together and then compressed as a single bitmap. A list of positional indices is also sent to identify the positions of char- acters with low confidence values and (implicitly) the position ofits corresponding raw bitmap....