公开(公告)号：GB2494976B

公开(公告)日：2015-06-10

申请号：GB201216437

申请日：2012-09-14

Applicant: IBM

Inventor： GOODMAN BRIAN DANIEL

IPC: G06K9/22 ,  G06F3/0488

Abstract: An embodiment of the invention provides a method including receiving input from a user, which includes a handwritten symbol. The input is compared to prototype symbols to determine whether the input includes a threshold degree of similarity with a prototype symbol. If the input does not include a threshold degree of similarity with a prototype symbol, the input is stored as a prototype symbol. If the input includes a threshold degree of similarity with a prototype symbol, it is determined whether the input represents a text character. If the input represents a text character, the text character is identified and a prototype text character is identified. The input is mapped to the identified text character and the identified prototype text character. If the input does not represent a text character, the input is mapped to a prototype shape, and the input is mapped to the prototype shape.

公开(公告)号：GB2497206B

公开(公告)日：2014-01-08

申请号：GB201221690

申请日：2012-12-03

Applicant: IBM

Inventor： JANG SOOBAEK ,  GOODMAN BRIAN DANIEL ,  DELUCA LISA SEACAT

IPC: G06F3/01

Abstract: A tool for detecting potential unintentional user input. Eye tracking technology is used to keep a record of where on a display a user is looking or if the user is even looking at the display. When input, such as a mouse selection or a tap on a touch screen, is received, the location of the selection is compared to a location of the user's gaze around when the selection was made. If the gaze location is outside of an acceptable range from the selection location, it is determined that the selection may have been in error and the selection is disregarded or a confirmation is requested of the user.

公开(公告)号：GB2494976A

公开(公告)日：2013-03-27

申请号：GB201216437

申请日：2012-09-14

Applicant: IBM

Inventor： GOODMAN BRIAN DANIEL

IPC: G06K9/22 ,  G06F3/0488

Abstract: The invention provides a method of handwriting recognition including receiving input from a user, which includes a handwritten symbol. The input is compared 220 to prototype symbols to determine whether the input includes a threshold degree of similarity with a prototype symbol. If it does not, the input is stored as a prototype symbol 230. If the input does include a threshold degree of similarity with a prototype symbol 240, it is determined whether the input represents a text character. If the input represents a text character, the text character is identified 250 and a prototype text character is identified 252. The input is then mapped to, i.e associated with, the identified text character and the identified prototype text character 254. If the input does not represent a text character, the input is matched to a prototype shape 262, and the input is mapped to the prototype shape 264. Stored prototypes may be replaced by new prototypes, or by an averaged prototype.