Image analysis – Pattern recognition – Feature extraction
Reexamination Certificate
1996-09-05
2001-04-03
Couso, Jose L. (Department: 2621)
Image analysis
Pattern recognition
Feature extraction
Reexamination Certificate
active
06212298
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to a character recognition apparatus, method and computer readable memory for recognizing characters on the basis of a plurality of entered strokes.
In order to realize a character recognition method that is independent of the stroke order of a character in recognition of characters entered using a coordinate input unit such as a digitizer, recognition is carried out by matching each stroke constituting the entered character with a dictionary that has been previously stored in a memory device such as a hard disk. (By way of example, the dictionary stores character codes of standard characters expressed by, say, a table of Chinese characters in common use, and information on the strokes constructing these characters.) Since matching is performed on a stroke-by-stroke basis according to this method, recognition time, which is that required to recognize an entered character, is proportional to the square of the number of strokes of the character.
Accordingly, a method has been proposed, as set forth in the specification of Japanese Patent Publication No. 4-65431, in which the strokes of an entered character are matched with some of the strokes of dictionary strokes (e.g. dictionary strokes corresponding to the number of entered strokes and dictionary strokes over a range of one stroke before to one stroke after a given stroke). Though this method is not perfect, recognition independent of the stroke order is made possible, thereby shortening recognition time.
However, according to the matching method used in character recognition disclosed in Japanese Patent Publication No. 4-65431, basically matching is carried out on the basis of the stroke order of the entered character. Consequently, with regard to a character such as “E” which may be entered by the user in a completely incorrect stroke order, it is not possible to achieve high-speed character recognition that is independent of stroke order.
SUMMARY OF THE INVENTION
An object of the present invention is to provide a character recognition apparatus, method and computer readable memory through which characters can be recognized at high speed and without dependence upon stroke order.
According to the present invention, the foregoing object is attained by providing a character recognition apparatus for recognizing characters based upon an entered plurality of strokes, comprising holding means for holding a dictionary which stores strokes of standard characters in a predetermined order based upon relative positional relationship of the strokes, storage means for storing the plurality of strokes in the predetermined order, comparison means for comparing the plurality of strokes stored in the storage means and each stroke of the standard characters stored in the dictionary, and recognition means for recognizing a character, which is constructed by the plurality of strokes, based upon results of the comparison performed by the comparison means.
In a preferred embodiment, the comparison means compares each stroke of the standard characters stored on the storage medium and a stroke stored in said storage means which corresponds to an order of said each stroke, and an order before and after said order.
In a preferred embodiment, the predetermined order is an order in which prescribed points of a sequence of points constructing each stroke are mapped to a predetermined straight line. By making the mapping to a predetermined straight line, the order of a plurality of strokes entered in two dimensions can be assigned.
In a preferred embodiment, the predetermined order is an order in which the prescribed points are mapped to a prescribed straight line along a predetermined curve.
With regard to an input area in which at least a plurality of strokes are entered, all of a plurality of unit input areas constructing the input area are searched. In a preferred embodiment, the predetermined order is an order in which prescribed points of a sequence of points constructing each stroke of the plurality of strokes are detected.
In a preferred embodiment, the prescribed points include at least one of a starting point, end point, mid-point and centroid of a stroke.
In a preferred embodiment, the predetermined order is an order in which two or more points from a sequence of points constructing each stroke of the plurality of strokes are mapped to a predetermined straight line.
According to the present invention, the foregoing object is attained by providing a character recognition method for recognizing characters based upon an entered plurality of strokes, comprising a holding step of holding a dictionary which stores strokes in a predetermined order based upon relative positional relationship of the strokes, a storage step of storing the plurality of strokes on a storage medium in the predetermined order, a comparison step of comparing the plurality of strokes stored on the storage medium at the storage step and each stroke of the standard characters stored in the dictionary, and a recognition step of recognizing a character, which is constructed by the plurality of strokes, based upon results of the comparison performed by at the comparison step.
According to the present invention, the foregoing object is attained by providing a computer readable memory storing program codes for character recognition processing, the memory comprising a program code of a holding step of holding a dictionary which stores strokes of standard characters in a predetermined order based upon relative positional relationship of the strokes, a program code of a storage step of storing the plurality of strokes on a storage tedium in the predetermined order, a program code of a comparison step of comparing the plurality of strokes stored on the storage medium at the storage step and each stroke of the standard characters stored in the dictionary, and a program code of a recognition step of recognizing a character, which is constructed by the plurality of strokes, based upon results of the comparison performed by at the comparison step.
In accordance with the present invention, as described above, it is possible to provide a character recognition apparatus, method and computer readable memory through which characters can be recognized at high speed independently of stroke order.
Other features and advantages of the present invention will be apparent from the following description taken in conjunction with the accompanying drawings, in which like reference characters designate the same or similar parts throughout the figures thereof.
REFERENCES:
patent: 3996557 (1976-12-01), Donahey
patent: 4829583 (1989-05-01), Monroe et al.
patent: 5438631 (1995-08-01), Dai
patent: 5734750 (1998-03-01), Arai et al.
patent: 61-086881 (1994-10-01), None
Arai Tsunekazu
Takasu Eiji
Yoshii Hiroto
Canon Kabushiki Kaisha
Couso Jose L.
Fitzpatrick ,Cella, Harper & Scinto
Werner Brian P.
LandOfFree
Character recognition apparatus, method and computer... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Character recognition apparatus, method and computer..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Character recognition apparatus, method and computer... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2443759