JPH1049627A - On-line character recognizing method and device therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an on-line handwritten character recognizing device in which not only a character in which an erroneous stroke order and the erroneous number of strokes are independently generated but also an erroneous stroke order continuing character in which those are simultaneously generated can be recognized, and the scale of a dictionary can be reduced. SOLUTION: A corresponding stroke extracting part 3 carrys out stroke correspondence between an input pattern and a standard pattern. The stroke order in a standard pattern to the erroneous stroke order continuing character obtained from the corresponding stroke extracting part 3 is stored in a stroke order information memory part 9 by writing a character with a correct number of strokes for the erroneous stroke order continuing character. When the erroneous stroke order continuing character in the same stroke order is inputted again, the standard pattern rearranged like the stroke order of the input pattern is prepared by a registered standard pattern preparing part 8 from the stroke order stored in the stroke order information memory part 9.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an on-line character recognition method and apparatus which expresses a handwriting of a character in a stroke coordinate series for each stroke and processes the time series of the stroke coordinates to recognize the character.

[0002]

2. Description of the Related Art As a conventional technique, there is a technique for recognizing characters carefully written by an unspecified writer. There is a problem with a method of recognizing the stroke order. In this regard, strokes are associated between a standard pattern and an input pattern, and the pattern is determined based on the sum of the distances between the associated strokes. Method of Recognition by Obtaining Distance (Transactions of the Institute of Electronics, Information and Communication Engineers (D), Vol. J
65-D, no. 6, pp. 679-686 (Showa 57-
6) described).

[0003] Next, as a method of extending the above method to recognition allowing a change in the number of strokes occurring in a correct stroke order, a process of determining a stroke of a standard pattern corresponding to an input pattern, a process of determining a non-corresponding stroke of the standard pattern, A selection process comprising the steps of generating a combined stroke in which the corresponding stroke and the stroke whose stroke order is mixed before and after, and associating the combined stroke with the input pattern, and calculating the inter-pattern distance from the associated stroke. There is a dynamic stroke coupling method (Japanese Patent No. 1698239).

On the other hand, in online handwritten character recognition for a scribbled character of a specific scribe, all strokes of both a standard pattern and an input pattern are combined in the stroke order to generate a single-stroke pattern expressed in one stroke. Means of recognition by performing DP matching between them (IEICE Transactions (D), Vol. J
68-D, no. 12. pp. 2116-2122 (Showa 60-12).

[0005]

By the way, as a character recognition algorithm, it is required to recognize a character generated in every writing, from a square style to a scribble.
Although the stroke order and the stroke number are different from the correct stroke order and stroke number in the block style writing, the stroke number variation is relatively small. However, in the case of the scribble, the stroke number changes greatly because the user is writing at a high speed, and so-called continuation characters in the wrong stroke order which are continuously written in the wrong stroke order also occur. In order to respond to the above-mentioned demands, there are the following problems when directly using the selective stroke combining method and the one-stroke pattern DP matching method.

[0006] The selective stroke combining method can recognize if the successively written portion has the correct stroke order, but cannot recognize a correct stroke in the continuation of the incorrect stroke order, so that recognition is difficult. In addition, the D
Since the P-matching method is a recognition method in which the stroke order is fixed and the number of strokes is free, a continuous character in which all strokes are written in the correct stroke order can be recognized, but a continuous character in an incorrect stroke order is still difficult to recognize.

That is, in any of the methods, the continuation characters in the erroneous stroke order are not initially recognized. To make this possible, a method of newly registering a pen point coordinate series constituting a character pattern can be considered. When the number increases, there arises a problem that the recognition dictionary scale becomes enormous.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method for recognizing not only characters in which a stroke order error and a stroke number error occur independently in one character, but also a continuation of the erroneous stroke order characters in which these errors occur at the same time. And to provide a device.

[0009]

In order to achieve the above object, an on-line character recognition method according to the present invention provides a method for inputting a character whose stroke order and stroke number are correctly written or a character in which a stroke order error or a stroke error occurs independently. Stroke correspondence between the pattern and the standard pattern is recognized, and for those characters that occur at the same time in the wrong stroke order, they correspond to the stroke order of the incorrect stroke order obtained by writing the correct number of strokes. The stroke order of the standard pattern to be stored is stored, and when a continuation character of the same stroke order is input again, a continuation character standard pattern is generated by rearranging from the previously stored stroke order according to the stroke order of the input pattern. Recognize by doing.

The online character recognition apparatus of the present invention further comprises a corresponding stroke extracting means for associating a stroke between an input pattern and a standard pattern, and a stroke order of the standard pattern obtained from the corresponding stroke extracting means. A stroke order information memory means for storing stroke order conversion information for rearranging the pattern in the stroke order, and a standard pattern rearranged in the stroke order of the input pattern using the stroke order conversion information of the stroke order information memory means. Registration standard pattern creation means.

According to the present invention, a character in which a stroke order error and a stroke number error occur independently is recognized by a selective stroke combination method, and a character having a correct stroke number is separately written for a continuation character in a wrong stroke order. The stroke order of the standard pattern corresponding to the stroke order of the continuation of the erroneous stroke obtained from the corresponding stroke extracting means is stored in the stroke order information memory means. The recognition is performed by generating a continuation character standard pattern rearranged from the stroke order set in the same order as the stroke order of the input pattern. This allows
Recognition of characters whose stroke order or stroke number fluctuates becomes possible,
In addition, it is also possible to recognize characters that are in the wrong stroke order.

According to an embodiment of the present invention, the writer has means for confirming the stroke order conversion information, making corrections if necessary, and storing it in the stroke order information memory means.

According to the embodiment of the present invention, the stroke order information memory means can pre-register or additionally register the stroke order conversion information predetermined by the writer, and can correct or delete the registered stroke order conversion information. It is configured.

According to an embodiment of the present invention, the stroke order information memory means has dedicated stroke order conversion information for each writer.

[0015]

Next, an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an on-line character recognition apparatus according to an embodiment of the present invention. The input display section 1, a preprocessing section 2, a corresponding stroke extraction section 3, and a standard pattern memory section 4 are shown.
An identification scale calculation unit 5, an identification result memory unit 6, a registration processing unit 7, a registered standard pattern creation unit 8, a stroke order information memory unit 9, and a control unit 10. In the figure, the solid line indicates the flow of data between the units, and the broken line indicates the flow of signals of the control unit 10.

Hereinafter, the operation of each processing unit will be specifically described.

An input display unit 1 is composed of an existing tablet and a display, and a pre-processing unit 2 uses a brush coordinate series of characters written on the tablet with a dedicated pen as an input pattern.
To send to.

The pre-processing unit 2 removes noise from the input pattern sent from the input display unit 1 by a known technique,
The input pattern that has undergone processing such as position / size normalization and feature point extraction is sent to the corresponding stroke extraction unit 3.

The corresponding stroke extracting unit 3 is means for associating strokes between the input pattern and the standard pattern, and for example, a well-known selective stroke combining method can be applied.

FIG. 2 is a block diagram of the corresponding stroke extracting unit 3, which comprises an inter-stroke distance matrix calculating circuit 31, a stroke primary correspondence determining circuit 32, a stroke secondary correspondence determining circuit 33, and a continuous character standard pattern generating circuit 34. Have been.

Now, N strokes of the input pattern are S _n (n = 1, 2, ‥‥, N), and M strokes of the standard pattern are T _m (m = 1, 2, ‥‥, M). Let's write it. When it is detected that the input pattern has arrived at the corresponding stroke extracting unit 3, first, the standard pattern of the first category C ₁ is read from the standard pattern memory unit 4.

The inter-stroke distance matrix calculating circuit 31 creates an M × N matrix having a distance d (S _n , T _m ) between two strokes S _n and T _m as elements.

[0024] Stroke primary correspondence determination circuit 32 is the distance becomes minimum when viewed from the stroke _{S n d (S n, T} m) (n =
_{Let Tm} of (1, 2,..., N) be a stroke corresponding to Sn. If the input pattern is a continuation character, since N <M,
(MN) strokes are left as unsupported strokes in the standard pattern.

The stroke secondary correspondence determination circuit 33 converts the uncorresponding stroke _Tm into the corresponding stroke T before and after the stroke order.
Two strokes T generated by combining _m-1 and _{Tm + 1
m-1_m,} T and _{m_m + 1,} S _n which corresponds to the T _m-1, and S
_The distance d calculated between each of _{n and} the next S _{n + 1}
A stroke pair providing the smaller distance among (S _n , T _{m- 1 — m} ) and d (S _{n + 1} , T _{m — m + 1} ) is determined as a corresponding stroke pair. Thereby, one-to-one stroke correspondence is performed between the input pattern and the standard pattern including the combined stroke.

The continuation character standard pattern generation circuit 34 determines the standard pattern obtained by rearranging the stroke order of the standard pattern in the same order as the stroke order of the input pattern based on the stroke correspondence information obtained by the stroke secondary correspondence determination circuit 33. It is sent to the calculation unit 5. Performing the above processing for all of the categories of the second category C ₂ or later.

The above processing will be described using an example.

FIG. 3 shows a standard pattern of the category "male", in which strokes _Tm (m = 1, 2,..., 7) are written in the correct stroke order and stroke number in the order of m.

FIG. 4 shows an input pattern of the category "male".
Strokes S _n (n = 1, 2, ‥‥, 6) are written in the order of n. S ₄ is intended is T _4, T ₅ of the standard pattern is written continued, also, S _5, S ₆ the stroke order variation occurs.

The 7 × 6 matrix shown in FIG. 5 is the inter-stroke distance matrix obtained by the inter-stroke distance matrix calculation circuit 31, and the circles represent the stroke pairs that have been established by the stroke primary correspondence determination circuit 32. . At this point, the stroke order variations of S ₅ and S ₆ are relieved. Stroke secondary correspondence determination circuit 3
3 generates a T _{4_5,} T _{5_6} as binding stroke associated with unsupported stroke _{T 5, d (S 4,} T 4_5) <d (S
_5, T _{5_6)} to give the corresponding the S ₄ and T _{4_5} is taken, strokes fluctuation is absorbed at this point. Cursive standard pattern generating circuit 34 is T ₁ the stroke order of the standard _{pattern, T 2, T 3, T} 4_5, T
_7, rearrange the order of T _6.

The discrimination scale calculation unit 5 calculates a pattern-to-pattern between a standard pattern in which the stroke order of K categories C _k (k = 1, 2,）, K) to be recognized is rearranged and an input pattern. distance seek D (C _k), and sends the C _k corresponding to the smallest D (C _k) of all the k to the identified result memory section 6 as an identification result. As a method for obtaining the inter-pattern distance D (C _k ), for example, a distance based on a known DP matching technique, a corresponding inter-stroke distance d
(S _n , T _m ).

The identification result memory unit 6 receives the identification result from the identification scale calculation unit 5, and displays the identification result in print in a character pattern arranged in a handwritten input pattern on the display of the input display unit 1.

Thus, an input pattern having a variation in the stroke order or the number of strokes is recognized.

Next, a method for recognizing a continuation character in erroneous stroke order, which is a main element of the present invention, will be described in detail using examples.

FIG. 6 shows a continuation character input in the erroneous stroke order of the category "man".
It is an example of a pattern, stroke X _n(N = 1, 2, ‥
‥, 6) are written in order of n, and X_ThreeIs the standard
T of the third and fifth strokes of the turn_Three, T_FiveIs written in the wrong stroke order
ing.

When the erroneous stroke order continuous character input pattern shown in FIG. 6 is input, the pre-processing unit 2 and the corresponding stroke extracting unit 3 sequentially perform the processing. Therefore, the reliability of the classification result in the classification scale calculation unit 5 is low.

FIG. 7 is a plan view of the display of the input display unit 1, which includes a writing character frame 101 and an identification result display frame 10.
2. Rewriting character frame 103, rewriting pattern identification result display frame 104, registration button 105, EXIT button 10
6. Stroke order conversion information display frame 107, increase / decrease button 108, continuous character standard pattern call button 109, erase button 1
10.

In the figure, the erroneous stroke order continued character input pattern of FIG. 6 is displayed in the writing character frame 101, and the identification result of the erroneous stroke order continued character input pattern of FIG. In this example, the identification is incorrect.

The writer performs the following operation to make the continuation of the incorrect stroke order character input pattern of FIG. 6 readable.

The writer rewrites a character pattern in the rewriting character frame 103 in the same writing order as the continuation character input pattern in the erroneous writing order shown in FIG. When the re-writing pattern is input, the pre-processing unit 2 and the corresponding stroke extracting unit 3 sequentially perform the processing. However, since the stroke number is correctly written in the wrong stroke order, the corresponding stroke extracting unit 3 Can correctly associate strokes. As a result, the discrimination scale calculator 5 obtains a highly reliable discrimination result. The discrimination result is displayed in the rewriting pattern discrimination result display frame 104, and the stroke order conversion information is displayed in the stroke order conversion information display frame 107.
Are displayed respectively.

FIG. 8 is an example of a rewriting pattern, in which strokes Y _n (n = 1, 2,..., 7) are written in the order of n, and Y ₄ and Y ₅ are standardized in FIG. The fourth of the pattern
Although the stroke order of five strokes can be seen, the number of strokes is written correctly.

FIG. 9 shows an inter-stroke distance matrix obtained from the rewritten pattern and the standard pattern, and the circles indicate the corresponding stroke pairs. From this, the stroke order conversion information for rearranging the stroke order of the standard pattern according to the stroke order of the input pattern is [1, 2, 3, 5, 4, 6, 6].
7].

The writer sets the rewriting pattern identification result display frame 1
Confirm that the identification result of 04 is correct. If the identification result is incorrect, the user repeatedly writes in the rewriting character frame 103 until a correct identification result is obtained. Next, the writer checks the stroke order conversion information displayed in the stroke order conversion information display frame 107. If the user does not need to make any correction, the user immediately touches the registration button 105 with a pen. If it is necessary to correct the stroke order, the user touches the corresponding stroke order number in the stroke order conversion information display frame 107 with a pen, adjusts the number with the increase / decrease button 108, corrects the correct stroke order, and touches the registration button 105 with a pen. When receiving the pen touch signal, the control unit 10 transmits an activation signal to the registration processing unit 7. Registration processing unit 7 accesses the corresponding stroke extracting unit 3 and receives signals from the control unit 10, and stores together with the category C _k is the correct identification result to obtain the stroke order conversion information on the stroke order information memory unit 9.

[0044] Figure 10 is an example of a stroke order information memory section 9, and a set of stroke order conversion information table 91 of category C _k.

[0045] Figure 11 is a stroke order conversion information table 91 of category C _k - a configuration example, the address 900 is the type number L of the stroke order conversion information, address 901 to (900+
L) stores L types of stroke order conversion information. In the case of the rewriting pattern shown in FIG. 8, the registration is the first registration, and if L = 1, the writing order conversion information [1, 2, 3, 3] is added to the address 901.
5, 4, 6, 7].

Next, the writer writes the second incorrect stroke order continuous character input pattern in which the third and fifth strokes of the standard pattern are continuously written in the same manner as the erroneous stroke order continuous character input pattern of FIG. Suppose you wrote. The second incorrect stroke order continuous character input pattern is processed by the pre-processing unit 2 and the corresponding stroke extracting unit 3. No response can be taken.

However, the control unit 10 sends an activation signal to the registered standard pattern creating unit 8 this time. When the registered standard pattern creation unit 8 receives the activation signal from the control unit 10, it reads out the stroke order conversion information from the stroke order conversion information table 91 of the category C _k stored in the stroke order information memory unit 9 and stores it in the standard pattern memory unit 4. A standard pattern in which the strokes of the standard pattern of the category C _k are rearranged is created and sent to the corresponding stroke extracting unit 3. The corresponding stroke extracting unit 3 takes a stroke correspondence between the standard pattern in which the strokes are rearranged and the second incorrect stroke order continuous character input pattern. In the correspondence, the problem of the continuation character in the wrong stroke order is solved, and the problem is merely a variation in the stroke order, so that a correct stroke correspondence can be obtained.

The corresponding stroke extraction unit 3 sends the stroke correspondence results of the K categories stored in the standard pattern memory unit 4 and the categories generated by the registered standard pattern creation unit 8 to the discrimination scale calculation unit 5. . In the discrimination scale calculation unit 5, the reliability of the discrimination result for the K categories is obtained as low as the first time, but the reliability of the discrimination result for the category generated by the registered standard pattern creation unit 8 is high. Can be expected.

In the writing character frame 101 shown in FIG.
The identification result of the continuation character input pattern in the erroneous stroke order is displayed each time, and the character is correctly identified in the second time.

In this way, the characters in the erroneous stroke order are recognized.
In the present embodiment, since only the information on the erroneous stroke order is stored, the dictionary scale can be significantly reduced as compared with the method of storing the stroke coordinate sequence itself of the erroneous stroke order pattern.

In the above processing, the writer sets the registration button 105
When the EXIT button 106 is touched with a pen at the place where the pen is touched, the stroke order conversion information of the erroneous stroke order continued character input pattern is returned to the initial screen without being registered. EXIT
The pen touch of the button 106 is used when avoiding the registration of the temporary continuation of the erroneous stroke order which occurs in a situation different from a situation where writing is usually performed.

[0052] Also the scribe in the process, and displays the cursive standard pattern re writing character frame 103 standard patterns of category C _k by pen touch the call button 109, is stored in the stroke order information memory section 9 The contents of the stroke order conversion information table 91 of the category Ck in which it is present can be displayed in the stroke order conversion information display frame 107. Each time the pen touch is repeated, the registered stroke order conversion information can be displayed one after another. If the user wants to change the stroke order conversion information, he or she touches the number indicating the corresponding stroke order in the stroke order conversion information display frame 107 with a pen, and then increases or decreases the size of the number indicating the stroke order with the increase / decrease button 108. Is input, and the registration button 105 is touched with a pen, whereby the correction of the stroke order conversion information can be performed. Further, after displaying the L pieces of stroke order conversion information registered in the category C _k , the user touches the pen once more, manually inputs new stroke order conversion information in the stroke order conversion information display frame 107, and touches the registration button 105 with a pen. Then, additional registration can be performed manually. further,
When the erase button 110 is touched with a pen while the stroke order conversion information is displayed, the stroke order conversion information already registered in the stroke order conversion information table 91 can be deleted. The pen touch of the erase button 110 is used as a correction means when an erroneous registration is made.

In this embodiment, an example is described in which the stroke order conversion information is additionally registered. However, the stroke order conversion information may be created in advance by a designer and registered in advance.

Although the present embodiment has been described with reference to a case where the number of erroneous stroke order information memory sections 9 is one, the writer has a plurality of erroneous stroke order information memory sections 9 so that the writer can specify and use a specific one. There may be.

In this embodiment, all the categories are processed by the corresponding stroke extracting unit 3. However, after the preprocessing unit 2, candidates are narrowed down by a large classification, and the narrowed candidates are processed by the corresponding stroke extracting unit 3. Processing may be performed.

[0056]

As described above, according to the present invention, a character in which a stroke order error and a stroke number error occur independently is recognized by the selective stroke combining method. By writing a correct character, the stroke order information of the continuation character of the erroneous stroke order is stored, and again, when a continuation character of the same stroke order is input, the stroke order of the input pattern is changed from the previously stored stroke order. Recognition by generating continuous character standard patterns rearranged as described above enables recognition in a wide variety of writings from square typeface characters to scribbled characters, and registers stroke order information only for continuation characters in incorrect stroke order. Therefore, there is an effect that the apparatus can be configured with a dictionary size much smaller than that of a method of registering a brush point coordinate series of a character pattern conventionally used.

[Brief description of the drawings]

FIG. 1 is a block diagram of an online character recognition device according to an embodiment of the present invention.

FIG. 2 is a block diagram of a corresponding stroke extracting unit 3;

FIG. 3 is a diagram showing an example of a standard pattern.

FIG. 4 is a diagram illustrating an example of an input pattern.

FIG. 5 is a diagram showing an inter-stroke distance matrix obtained from the patterns of FIGS. 3 and 4;

FIG. 6 is a diagram illustrating an example of an incorrect stroke order continuous character input pattern.

FIG. 7 is a plan view of a display of the input display unit 1.

FIG. 8 is a diagram showing an example of a rewriting pattern.

9 is a diagram showing an inter-stroke distance matrix obtained from the patterns of FIGS. 3 and 8; FIG.

FIG. 10 is a diagram illustrating an example of a stroke order information memory unit 9;

FIG. 11 is a diagram illustrating a configuration example of stroke order conversion information 91 of a category C _k .

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Input display part 2 Preprocessing part 3 Corresponding stroke extraction part 4 Standard pattern memory part 5 Identification scale calculation part 6 Identification result memory part 7 Registration processing part 8 Registration standard pattern creation part 9 Stroke order information memory part 10 Control part 31 Distance between strokes Matrix calculation circuit 32 stroke primary correspondence determination circuit 33 stroke secondary correspondence determination circuit 34 continuation character standard pattern generation circuit 91 stroke order conversion information table 900 address 901 address:: (900 + L) address 101 writing character frame 102 identification result display frame 103 re Writing character frame 104 Rewriting pattern identification result display frame 105 Registration button 106 EXIT button 107 Stroke order conversion information display frame 108 Increase / decrease button 109 Continuous character standard pattern call button 110 Erase button T _{1 to} T ₇ Standard pattern stroke S _{1 to} S ₆ strokes and Y ₁ to Y ₆ re writing pattern of the stroke X ₁ to X ₆ false stroke order cursive input pattern of the input pattern

Claims (5)

[Claims] 1. An online character recognition method for expressing a handwriting of a character by a stroke coordinate series for each stroke and processing the time series of the stroke coordinates to recognize the character. This is a method of recognizing characters by taking a stroke correspondence.In the case of erroneous stroke order continuation characters, the stroke order of the standard pattern corresponding to the stroke order of erroneous stroke order continuation characters obtained by writing characters with the correct number of strokes If a continuation character in the same stroke order is input again, it is recognized by generating a standard continuation character pattern rearranged from the previously stored stroke order as the stroke order of the input pattern. An online character recognition method comprising: 2. An on-line character recognition apparatus for expressing a handwriting of a character by a stroke coordinate series for each stroke and processing a time series of the stroke coordinates to recognize an erroneous character. A corresponding stroke extracting means for performing stroke correspondence in; a stroke order information memory means for storing stroke order conversion information for rearranging the stroke order of the standard pattern obtained from the corresponding stroke extracting means in accordance with the stroke order of the input pattern; An on-line character recognizing device, comprising: registered standard pattern creating means for creating a standard pattern rearranged according to the stroke order of the input pattern using the stroke order conversion information of the stroke order information memory means. 3. The online character recognition apparatus according to claim 1, further comprising means for allowing a writer to check the stroke order conversion information, modify the stroke order information if necessary, and store the corrected stroke order information in the stroke order information memory means. 4. The writing order information memory means is configured so that a writer can pre-register or additionally register predetermined stroke order conversion information, and can also modify or delete registered stroke order conversion information. Item 7. The online character recognition device according to Item 1. 5. The online character recognition apparatus according to claim 1, wherein said stroke order information memory means has dedicated stroke order conversion information for each writer.