JPS6321194B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a character training machine that is a means for helping children learn characters.

Training and learning the correct way to write letters, especially kanji, is a major focus of school education in Japan. The correct way to write letters includes two concepts: accurate (and preferably beautiful) letterforms and correct stroke order. Both are difficult to learn in a short period of time.
Constant and long-term training is required. However, in the current school education situation, it is difficult for teachers to devote long hours to this education, and any lack of learning is left to the efforts of individual students outside of school. Considering these circumstances, it can be said that if an automatic machine were provided that could judge the quality of written characters and give appropriate instructions, it would be a great aid to children's learning of characters.

As described above, the present invention aims to realize and provide a character training machine which is a device that automatically examines the stroke order and character shape of characters written by a student and displays errors if necessary. It is something.

The character training machine according to the present invention includes a display section for presenting a required character, a means for inputting information on the pen point movement of the character written by hand as a time series of features, and a standard glyph shape and regular character shape for the character. a standard pattern storage unit that stores the stroke order as a time series of features and outputs a time series of character features presented on the display unit; a determination unit that associates each feature with a time series of the characteristics of the supplied characters and determines whether or not a predetermined matching condition is satisfied between the associated features; and a means for displaying the determination result in response. It consists of

The principle and device configuration of the present invention will be explained below with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention, and FIG. 2 is a time chart for explaining its operation. Each part in FIG. 1 is a control unit 1.
operates under the control of First, the control section 1 generates a character type designation signal n. This character type designation signal n is changed sequentially as learning progresses, but here we will explain the case where character type n is generally designated. The character pattern storage section 2 stores character shapes of all character types targeted by this device in the form of dot patterns. When the character type designation signal n mentioned above is given, the dot pattern of the character n is displayed on the signal line.
The signal is sent to the display unit 3 via the DSP and visually displayed. When characters are displayed on the display section 3, the trainee writes the characters on the tablet section 4. From the tablet unit 4, the writing point position (x,
y) is generated in the form of a digital signal and sent to the determination unit 10 via the signal line C. This determination section 10 is composed of an input section 5, a matching processing section 7, and a determination processing section 8.

FIG. 3 is a diagram showing an example of the configuration of the input section 5. As shown in FIG.
The writing point position signal C inputted from the tablet section 4 consists of mutually orthogonal coordinate components x and y signals. These signals are written into registers 501 and 502, respectively, in synchronization with timing signal CLK1 given by control section 1. The subtracter 505 subtracts the x signal x'' stored in the register 503 from the x signal x' output from the register 501, and outputs the result as an x difference signal Δx.Similarly, the subtracter 506
Then, the y signal y" output from the register 504 is subtracted from the y signal y' output from the register 502 to generate a y difference signal Δy. These difference signals Δx and Δy are connected to an OR circuit. 507. The difference signals Δx and Δy are each expressed in a 2-bit two's complement format.These 4-bit signals in total are input to the OR circuit 507, so the output The signal q is q=0 when all 4 bits of the input signal are 0, and q=1 at other times.
becomes. The output signal q is input to an AND circuit 508, and the AND with the timing signal CLK1 is calculated and sent to the registers 503 and 504 as a write pulse wp1. After all, the write pulse wp1 is generated when at least one of the differential signals Δx and Δy is not zero. Therefore, when a set of x and y signals is denoted by (x, y), (x', y') is stored in registers 501 and 502, and (x', y') is stored in registers 503 and 504. ``, y''), the write pulse wp1 is generated. This means that a write pulse is generated only when the writing point moves. When this write pulse wp1 is generated, the x signal x' stored in register 501 is written into register 503. Similarly, the y signal y' stored in register 502 is written to register 504. Therefore, registers 501 and 502
Each time there is a change in the pen point position, an x signal and a y signal indicating new coordinates are written.

On the other hand, at the time when the writing pulse wp1 is generated due to the change in the writing point position as described above, the difference signals Δx and Δy are encoded by the encoder 509 and sent to the input pattern buffer 510. written. The encoder 509 has built-in logic that outputs the code shown in FIG. 4 in accordance with the values of the difference signals Δx and Δy. This encodes the direction of the differential signal vector (Δx, Δy). The output signal d of the encoder 509 is written into the input pattern buffer 510 in synchronization with the write pulse wp1. In parallel with this, the aforementioned x signal x' and y signal y' stored in register 501 and register 502 are also written into this input pattern buffer 510. Now, if the direction code of the difference vector is d(i) and the position of the writing point is (x(i), y(i)), the input pattern buffer 510 has the following formula: 〓(i)=(d(i), ), y(i))) ...(1) The input pattern A is held as a sequence of 3 o'clock signs expressed by a vector. Here, i represents the order in which the above vectors were written into the input pattern buffer 510, and is a storage address designation. When character patterns are displayed in this way, the number of vectors a _i varies depending on the time. When character input from the tablet unit 4 is completed, the input pattern buffer 501 is described by I vectors as follows: A = ₁ , ₂ ,..., _i ,..., _I ... Suppose that it is stored in memory.

In the standard pattern storage unit 6 in FIG. 1, a standard pattern B ⁿ of character n is generally stored as a time series of features B ⁿ =〓 ⁿ ₁ , 〓 ⁿ ₂ , ... 〓 ⁿ _j , ..., 〓 ⁿ _J . . . ...(3) It is stored in the following format.

Here, 〓 ⁿ _j describes the j-th line segment obtained by line segment approximation of this character n as described in the specification of Japanese Patent Application No. 52-23611, 〓 ⁿ _j = (d ⁿ (j), M ⁿ (j), N ⁿ (j) , L ⁿ (j)) ...(4), where d ⁿ (j) is the direction code defined in Figure 4, and M ⁿ (j) is the minimum length of this line segment, N ⁿ
(j) means the maximum length, and L ⁿ (j) means the optimal length of this line segment.

The matching processing section 7 is constituted by a well-known microprocessor, and operates under the control of a built-in program. Here, patent application 52−
Based on the principle described in 23611, matching processing is performed between the input pattern of equation (2) and the standard pattern of equation (3). This matching process is particularly performed between the sequences of direction codes d(i) and d ⁿ (j) of both patterns. The matching performed here can be summarized as the following recurrence formula calculation.

Γ Initial condition g(0,0)=0 ...(5) Γ Recurrence formula g(i,j)=max ^l [g(il,j-1) _l-1 〓 ^x=0 e(ix,j )〕 …(6) However, M ⁿ (j)≦l≦N ⁿ (j) …(7) Γ Calculation range 1≦i≦I, 1≦j≦J …(8) Γ Similarity R ( A, B ⁿ ) = g (I, J) ... (9) The symbol e (i, j) in equation (6) is the signal line a
The direction code d(i) of the input pattern input from
and the direction code d ⁿ (j) of the standard pattern input from the signal line b. According to the principle described in the above-mentioned Japanese Patent Application No. 52-23611, as a result of the recurrence formula calculation of equations (5) to (8) above, the direction code of the standard pattern is added to the series of direction codes d(i) of the input pattern. Comparison between both series is performed by optimally matching d ⁿ (j), and the similarity R(A, B ⁿ ) is obtained by equation (9). This degree of similarity is sent to the determination unit 8 via the signal line n.

If we store the value l^ of l that gives the maximum in equation (6) in the form l(i, j) = l^ ...(10), we can convert the direction code series of the input pattern to the line segment j of the standard pattern. =1,2,...,J
It can be segmented according to the The segmentation points p(j) are sequentially determined by the following recurrence formula calculations.

Initial condition P(J)=I ……(11) Recurrence formula P(j)=P(j+1) −l(p(j+1), j+1) ……(12) j=J−1, J−2, ..., 2, 1 P(j-1) and P(j) are both dividing points (both ends) of the j-th line segment. These division point information are sent to the determination processing section 8 via the signal line h.

The determination processing section 8 is composed of a well-known microprocessor, and operates under the control of a built-in program. As the input signal h from the matching processing section 7, the similarity R(A,
B ⁿ ) is sent, the standard pattern B ⁿ of character n is sent.
A comparison with a threshold value θ ⁿ stored in association with is performed. When R(A, B ⁿ )<θ ⁿ , the shape or stroke order of the input character is clearly inappropriate, and this fact is displayed on the result display section 9.

Next, when the division point information P(j) is input, the following two types of determination processing are performed for each line segment j=1, 2, . . . , J. The first determination process is related to stroke order. Standard direction code d ⁿ of the jth line segment
(j) is input from the standard pattern storage section 6 through the signal line lx. On the other hand, from the input pattern buffer 510, i=P associated with the j-th line segment
A direction code d(j) from (j-1) to i=P(j) is input and compared with the direction code d ⁿ (j).
As a result, in the direction code d(j) of the input pattern, d ⁿ
When there are very few lines that match (j), it is determined that this j-th line segment is different in direction from the standard line segment, and this fact is displayed on the result display section 9. The second determination process is related to character shapes. Based on the segmentation point signals P(j-1) and P(j), the coordinates of both ends of the j-th line segment are (x(P(j-1)), y(P(j-1)), (x (P(j)),y(P(j))...(13) is input from the input pattern buffer 510.The determination processing section uses the coordinates of both ends as the distance between the two ends. The length l(j) of the j line segment is calculated.On the other hand, the optimum length L ⁿ (j) of the above equation (4) is input from the standard pattern storage section 6 via the signal line lx. This optimal length has been normalized in advance so that the sum of these lengths is constant.
That is, _J 〓 ^j=1 L ⁿ (j)=C ……(14) Therefore, the line segment length l(j) is also The sum of these values is converted to a constant value C by the following calculation. This is because character shape determination is not influenced by the similar size relationships of characters. Next, a comparison is made between this l'(j) and the optimum length L ⁿ (j), and if a difference beyond a certain limit is found, a message indicating that the character shape is inappropriate is displayed on the result display section 9. Ru. If the character shape and stroke order are appropriate as a whole, a confirmation signal PP is sent to the control unit 1. When this confirmation signal PP is input to the control section 1, the character type designation signal n is incremented by 1, and the training for the next character is started.

Although the principle of the present invention has been explained above by showing examples, these descriptions do not limit the scope of the present invention. In particular, various configurations of the determination section are conceivable. For example, as shown in Japanese Patent Application No. 49-113815, a configuration is also possible in which the input pattern is represented as a two-dimensional figure and the matching process is performed. Furthermore, the input section 5, matching processing section 7, and determination processing section 8 shown in FIG. 1 may be configured by one microprocessor. Furthermore, the display section 3 and the result display section 9
It is also possible to share the same.

As is clear from the above description, according to the character training machine of the present invention, children can efficiently learn correct stroke order and beautiful character shapes without bothering teachers or other personnel.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention;
FIG. 2 is a time chart showing an example of the operation of the apparatus shown in FIG. 1, FIG. 3 is a diagram showing an example of the configuration of an input section, and FIG. 4 is a diagram for explaining the definition of a direction code. 1...Control unit, 2...Character pattern storage unit, 3...
...display section, 4...tablet section, 5...input section,
6... Standard pattern storage section, 7... Matching processing section, 8... Judgment processing section, 9... Result display section, 10
...Determination section, 501, 502, 503, 504...
...Register, 505, 506...Subtractor, 507
...OR circuit, 508 ...AND circuit, 509
... Encoder, 510 ... Input pattern buffer.

Claims (1)

[Claims] 1. A display unit that presents the required characters, a means for inputting information on the stroke movement of the characters written by hand as a time series of features, and a time series of features that displays the standard glyph shape and normal stroke order of the characters. a standard pattern storage section that stores a time series of character features presented on the display section; Character training consisting of a determination unit that associates each feature with a time series and determines whether a predetermined matching condition is satisfied between the associated features, and a means for displaying the determination result in response. Machine.