Browse Prior Art Database

Method and system for online stroke order insensitive hand-written input method

IP.com Disclosure Number: IPCOM000226411D
Publication Date: 2013-Apr-03
Document File: 3 page(s) / 50K

Publishing Venue

The IP.com Prior Art Database

Abstract

Many of the current hand-written input methods are sensitive to the inputting stroke order especially for the eastern countries' language, such as Chinese, Korean and Japanese, etc. When the user does not obey the standard stroke order, many current inputting engine will not generate the correct candidates for user’s confirmation This disclosure proposes a method and system to solve this problem by dynamically recognizing the nonstandard stroke sequence online, inferencing the real-intention of the user behind the sequence, and update the stroke order template by adding the personalized order sample effectively.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 3

Method and system for online stroke order insensitive hand-written input method

Usually in order to improve the recognition accuracy, many input method engines strictly obey or have a little flexibility in handling non-standard stroke order inputting sequence. As a result, they are sensitive to the disorder or non-standard stroke order sequence by hand-written. To solve this problem, some methods relax the strict matching of the designated sequence order that are stored in a template database, while at the side effects including decreased efficiency. And it cannot handle the large disorder

fundamentally.

This disclosure proposes a method and system for tagging and updating the stroke order template database dynamically using the contextual information from user's daily input which is frequent and abundant.

The non-standard order inputting sequence can be detected by the following observations and techniques:


(1) When for the non-standard input case, the deletion will usually happen on the latest ONE character, either by cancel or delete,


(2) Compare the stroke sequence between two consecutive typing and if the order is a little different , indicating probably pointing to the same character. (3) Apply the stroke order free template matching (by graph matching) to measure the similarity of the two characters A & B (or other information and techniques can be applied like image recognition). Here A means the previously

unrecognized input sequence and B is the user-intended character.

To make the compuation resource intensive online non-standard order stroke sequence recognition, matching and updating effectively and feasible in application especially on devices, an embodiement can be implementing and running this procedure not immediately. For e...