Method for correcting handwriting on a pen-based computer

A method for correcting textual material displayed on a display screen of a pen based computer system, the method comprising the steps of:

• a) receiving an input with a stylus on a dual function display screen of a pen-based computer system;
• b) recognizing the input, the recognition step including the substep of defining a bounding box that bounds the input, wherein the input includes at least one character;
• c) determining whether the input is a word correction gesture, wherein the word correction gesture determining step includes the substeps of,
• determining whether the bounding box of the input substantially overlaps a displayed textual object having at least one character, wherein when them is not substantially overlap, the input is not considered to be a word correction gesture,
• determining whether more than a predetermined number of characters of said displayed textual object are overlapped by the input, wherein when more than the predetermined number of characters are overlapped by a determined amount the input is not considered to be a word correction gesture,
• determining whether the input is positioned entirely within a multi-space region, wherein when the input is positioned entirely within a multi-space, the input is not considered to be a word correction gesture,
• d) determining an insertion position for the input and a selected word for correction when the input is determined to be a word correction gesture, wherein the insertion determining step includes the substeps of,
• selecting the closest adjacent word to the midpoint of the input as the selected word for correction when the midpoint of the input is over a space character and setting the insert position of the input to an adjacent end of the selected word;
• selecting the position of a most overlapped textual character as the insertion position when the input substantially overlaps textual characters and selecting the word associated with the most overlapped textual character as the selected word for correction;
• e) re-recognizing the input in conjunction with characteristics of the selected word when the input is determined to be a word correction gesture and providing a plurality of potential new interpretations as a result of said re-recognizing;
• f) determining a best potential replacement character result based upon a result of said re-recognizing step, by selecting said best potential replacement character as the highest probability of it being the correct replacement character based upon said plurality of potential new interpretations;
• and
• g) correcting the selected word using the best potential replacement character result and displaying the result.

In a computer system including a processing means, a display screen controlled by the processing means and a stylus for permitting a user to write on the display screen in order to input information into the processing means, an arrangement for correcting textual material displayed on a computer screen comprising:

• means for determining the location of an input written onto the screen by the user;
• a recognizer for recognizing the input as a text object including at least one character;
• means for determining whether the input is a correction gesture intended to correct a word displayed on the display screen based on a determined amount of overlap of the input written onto the screen over the textual material that is displayed on the screen;
• means for causing the recognizer to re-recognize the input in context when the input is determined to be part of a correction gesture, wherein said correction gesture also represents a correction character to replace at least one character of said text object;
• means for selecting a particular re-recognition interpretation from a plurality of potential re-recognition interpretations based in part on whether particular interpretations have already been tried;
• and
• means for replacing said at least one character of said text object with the selected re-recognition interpretation when a recognized input is determined to be part of a correction gesture.

A method of correcting textual material displayed on a display screen of a pen based computer system, the method comprising the steps of:

• a) receiving an input on the display screen, wherein the input includes at least one character;
• b) initially recognizing the input wherein the initial input recognition identifies a set of potential interpretations, each potential interpretation having a probability associated therewith, wherein the set includes at least one potential interpretation;
• c) determining whether the recognized input is a word correction gesture based in part on g determination of whether the recognized input overlaps by a determined mount textual material that is displayed on the screen and if so, determining an insertion position for the recognized input and a selected word for correction;
• d) recognizing the input a second time in conjunction with a context of the selected word for correction when the input is determined to be a word correction gesture;
• e) determining a best potential replacement character result based at least in part upon a result of said step of recognizing the input in conjunction with a context of the selected word for correction;
• and
• f) correcting the selected word using the best potential replacement character result and displaying the result, whereby corrections to textual material displayed on the display screen are made by overwriting the text to be corrected.

A method of correcting textual material displayed on a display screen of a pen based computer system, the method comprising the steps of:

• a) receiving an input on the display screen;
• b) recognizing the input, wherein the input recognition identifies a set of potential interpretations, each potential interpretation having a probability associated therewith, wherein the set includes at least one potential interpretation, the recognition step further including the substep of defining a bounding box that bounds the input;
• c) determining whether the input is a word correction gesture, wherein the word correction gesture determining step includes the substeps of,
• determining whether the bounding box of the input substantially overlaps a displayed textual object by a determined amount, wherein when there is no substantial overlap, the input is not considered to be a word correction gesture,
• determining whether more than a predetermined number of characters are overlapped by the input, wherein when more than the predetermined number of characters are overlapped, the input is not considered to be a word correction gesture,
• determining whether the input is positioned entirely within a multi-space region, wherein when the input is positioned entirely within a multi-space, the input is not considered to be a word correction gesture,
• d) determining an insertion position for the input and a selected word for correction when the input is determined to be a word correction gesture, wherein the insertion determining step includes the substeps of,
• selecting the closest adjacent word to the midpoint of the input as the selected word for correction when the midpoint of the input is over a space character and setting the insert position of the input to an adjacent end of the selected word;
• selecting the position of a most overlapped textual character as the insertion position when the input substantially overlaps textual characters and selecting the word associated with the most overlapped textual character as the selected word for correction;
• e) re-recognizing the input in conjunction with characteristics of the selected word when the input is determined to be a word correction gesture, the re-recognizing step being arranged to reassign probabilities to the potential interpretations when the probabilities calculated in the re-recognizing step for each potential interpretation differ from the probabilities associated with the potential interpretations of the recognizing step;
• f) using the re-recognition results to create at least one resultant word based upon the potential interpretations and comparing the resultant word to words in a dictionary to further facilitate recognition of the input;
• g) determining a best potential replacement character result based at least in part upon a result of said resultant word comparing step;
• and
• h) correcting the selected word using the best potential replacement character result and displaying the result.

In a computer system including a processing means, a display screen controlled by the processing means and a stylus for permitting a user to write on the display screen in order to input information into the processing means, an arrangement for correcting textual material displayed on a computer screen comprising:

• means for determining the location of an input written onto the screen by the user, wherein the input includes at least one character;
• a recognizer for recognizing the input;
• means for determining whether the recognized input is a correction gesture intended to correct a word displayed on the display screen based in part on a determined amount of relative overlap of the input written onto the semen and the location where textual material is displayed on the screen;
• means for causing the recognizer to re-recognize the input in context with the present textual material, when the input is determined to be part of a correction gesture;
• means for selecting a particular re-recognition interpretation from a plurality of potential re-recognition interpretations based in pan on whether particular interpretations have already been tried;
• and
• means for replacing an overwritten character with the selected re-recognition interpretation when a recognized input is determined to be part of a correction gesture.

A method of correcting existing textual material displayed on a display screen of a pen based computer system, the method comprising the steps of:

• a) receiving a handwritten input on the display screen, wherein the input includes at least one character;
• b) recognizing the input, wherein the input recognition identifies a set of potential interpretations, each potential interpretation having a probability associated therewith, wherein the set includes at least one potential interpretation;
• c) determining whether the recognized handwritten input is a word correction gesture based in part on a determination of whether the input overlaps by a determined amount the existing textual material that is displayed on the screen and if so, determining an insertion position for the input and a selected word for correction;
• d) re-recognizing the input taking into consideration characteristics of the selected word for correction in conjunction with the set of potential characters when the input is determined to be a word correction gesture to facilitate improved recognition of the input, the re-recognizing step being arranged to reassign probabilities for the potential interpretations when appropriate;
• e) determining a best potential replacement character result from the set of potential interpretation based at least in part upon a result of said re-recognizing step;
• and
• f) correcting the selected word using the best potential replacement character result and displaying the result, whereby corrections to existing textual material displayed on the display screen we made by overwriting the text to be corrected.

A method of correcting textual material displayed on a display screen of a pen based computer system, the method comprising the steps of:

• a) receiving a handwritten input on the display screen, wherein the input includes at least one character;
• b) recognizing the input, wherein the input recognition identifies a set of potential interpretations, each potential interpretation having a probability associated therewith, wherein the set includes at least one potential interpretation;
• c) determining whether the recognized input is a word correction gesture based in part on a determination of whether the input overlaps by a determined mount textual material that is displayed on the screen and if so, determining an insertion position for the input and a selected word for correction;
• d) using the potential interpretations to create a set of at least one resultant word and comparing the resultant word set to words in a dictionary when the input is determined to be a word correction gesture to facilitate improved recognition of the input;
• e) determining a best potential replacement character result based upon a result of the resultant word comparing step;
• and
• f) correcting the selected word using the best potential replacement character result and displaying the result, whereby corrections to textual material displayed on the display screen are made by overwriting the text to be corrected.