Image analysis – Pattern recognition – On-line recognition of handwritten characters
Reexamination Certificate
1999-01-29
2001-12-25
Bella, Matthew C. (Department: 2721)
Image analysis
Pattern recognition
On-line recognition of handwritten characters
C382S188000, C382S203000, C382S228000
Reexamination Certificate
active
06333995
ABSTRACT:
BACKGROUND
1. Technical Field
The present invention relates generally to handwriting recording devices for real-time digitization of handwritten text and, more particularly, to system and methods for selecting keywords in a handwriting recording device.
2. Description of Related Art
There are a variety of handwriting recording devices which digitize and electronically capture handwritten text in real-time as the text is written. Typically, these conventional handwriting recording devices utilize a digitizing tablet in conjunction with an electronic stylus (i.e., pen) to record handwriting data. The digitizing tablet generates positional data representing the coordinates of the electronic stylus by detecting, for example, RF (radio frequency) signal emissions generated by the stylus as a user applies the stylus to the surface of the tablet (e.g., when the user is writing on or near the surface of the tablet).
One type of handwriting recording device is a personal digital notepad (PDN) device. The PDN device includes an electronic stylus having an inking tip which allows the user to write on a piece of paper placed over the digitizing tablet while recording handwriting strokes (positional data). The stylus typically includes a pressure sensor which generates a “pen down” signal when the stylus is in contact with the surface of the digitizing tablet. This signal is usually conveyed to the digitizing tablet via a dedicated line or through a secondary RF emission. Accordingly, by monitoring the pen down signal and the path of the radio emission of the stylus as it move across the radio sensitive grid, the digitizing tablet may generate a data stream representing the pen strokes which comprise written text on, or in proximity to, the surface of the digitizing tablet. The term “stroke” used herein refers to a group of handwriting data which is recorded between a “pen-down” signal and a “pen-up” signal.
One embodiment of a PDN device is disclosed in U.S. patent application Ser. No. 08/747,735 entitled: “Methods, Systems and Products Pertaining To A Digitizer For Use In Paper Based Record Making Systems” filed on Nov. 12, 1996, the disclosure of which is incorporated herein by reference. One currently available embodiment of this PDN device is marketed by the A. T. Cross Company and sold under the name CROSSPAD™. The CROSSPAD allows a user to record handwritten information on a standard paper notepad and simultaneously record an electronic carbon copy of the writing.
One novel feature of the CROSSPAD is its ability to record user defined keywords while a user is writing on the device. A keyword can be one word or a group of words which are selected by the user when producing a handwritten record to represent key concepts or ideas for a given handwritten paragraph or page. In particular, a keyword comprises one or more recorded strokes which correspond to the handwritten text selected by the user and are electronically tagged as a single unit. The strokes comprising the keyword are subsequently converted into a text representation of the ink using handwritten recognition technology.
Specifically, the CROSSPAD is currently configured to define a keyword in the following manner. After the user is done writing on the writing medium associated with the PDN, the user actuates a designated switch, button or menu option on the PDN to invoke a “keyword mode.” While in the keyword mode, the user selects a keyword by drawing “bounding strokes” around the handwritten strokes comprising the desired keyword. The bounding strokes may take the form of a circle, a box, or any other shape drawn by the user to delineate an area encompassing the desired (keyword) handwriting strokes. For instance, as shown in
FIG. 1
, a bounding stroke “b” (in the shape of an ellipse) is drawn around the handwritten word “jumps” on page P. As further illustrated in
FIG. 1
, the bounding stroke “b” is processed by the PDN to detect the extremities of the bounding stroke “b”, i.e., the maximum X and Y coordinates (i.e., Xmax and Ymax) and the minimum X and Y coordinates (i.e., Xmin and Ymin) to generate a “bounding box” (denoted as “B”) through which the detected coordinates extend. All recorded strokes having coordinates which fall within the “bounding box” area are used to define the keyword.
There are various disadvantages to the current keyword selection algorithm. For instance, the user may not fully encircle a handwritten stroke. Consequently, it is possible that some or even all of the handwritten strokes that the user desires to select as a keyword will not be included in the keyword. In addition, even if the bounding stroke “b” encloses the desired handwritten strokes, the current algorithm for generating the bounding box “B” (area) may cause the keyword members to include recorded strokes that are outside the actual bounding stroke “b” (i.e., in the areas denoted as “a” in
FIG. 1
) made by the user. These problems can result in user inconvenience and reduce the usability of the PDN with respect to keyword selection.
SUMMARY
The present application is directed to a system and methods for providing robust keyword selection in a handwriting recording device such as a personal digital notepad (PDN) device, which are less prone to user error/variability during user selection of desired keywords. Particularly, novel keyword selection methods for use with handwriting recording devices render user selection of keyword strokes more robust and less prone to user error/variability by making stroke membership and selection more flexible.
In one aspect of the present invention, a method for selecting a keyword in a handwriting recording device having an electronic inking stylus and a digitizing tablet associated therewith, wherein the device records stroke data representing handwritten strokes concurrently produced on a writing medium placed over the digitizing tablet using the electronic inking stylus, comprises the following steps: invoking a keyword selection mode; writing bounding strokes around handwritten strokes selected by a user for designation as a keyword; defining a bounding region as the area contained within the bounding strokes; and determining keyword membership of a recorded stroke based on a relation between the recorded stroke and the defined bounding region.
In another aspect of the present invention, keyword membership is determined by including only those recorded strokes which fall entirely within the bounding region.
In yet another aspect of the present invention, keyword membership is determined by including only those strokes for which the ratio of the length of the stroke inside the bounding region to the total length of the stroke is larger than some predetermined or user-specified threshold.
In another aspect of the present invention, the keyword membership determination process can be implemented to include only those recorded strokes having a beginning, an end, or a beginning and end included within the bounding region, wherein the “beginning” of a stroke refers to the first recorded position of the stroke and the “end” of a stroke refers to the last recorded position of the stroke.
In yet another aspect of the present invention, the keyword membership determination process is implemented by including or excluding as a keyword member a stroke having a cost factor which exceeds a prespecified threshold, wherein the cost factor is based on a distance from the bounding region of a portion of a stroke that is outside of the bounding region and the percentage of the stroke length which is outside the bounding region.
These and other aspects, features and advantages of the present invention will be described and become apparent from the following detailed description of preferred embodiments, which is to be read in connection with the accompanying drawings.
REFERENCES:
patent: 5566248 (1996-10-01), Ulrich
patent: 5613019 (1997-03-01), Altman et al.
patent: 5862256 (1999-01-01), Zetts et al.
patent: 6021218 (2000-02-01), Capps et al.
patent: 6064765 (2000-05-01), Hirayama
Bella Matthew C.
Desire Gregory
F. Chau & Associates LLP
International Business Machines - Corporation
LandOfFree
System and methods for providing robust keyword selection in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and methods for providing robust keyword selection in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and methods for providing robust keyword selection in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2602889