JPH01130291A - Data recognizing device - Google Patents

Abstract

PURPOSE:To confirm whether an inputted mathematical expression has been recognized correctly or not by providing a message display means for displaying a question message for asking whether a result of recognition is right or wrong. CONSTITUTION:A handwritten coordinate input device 1 whose input and output are formed unitedly is constituted of a pen 11, a tablet 12 and a liquid crystal display 13. Also, a microcomputer 2 is constituted of a handwritten character recognizing means 21, a mathematical expression recognizing means 22, an arithmetic means 23 and a display control means 24. On the liquid crystal display 13, in order to ask whether an inputted expression has been recognized correctly or not, to an operator, the recognized expression is displayed, and also, a message for asking whether it is satisfactory or not is displayed. In case an answer that said expression has been recognized correctly has been obtained, the arithmetic means 23 executes an operation, and a solution is displayed on the liquid crystal display 13.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention is a data recognition device that can input data by hand, recognize it, and display it.

[Conventional technology]

Conventionally, this type of data recognition device inputs data by hand, captures the input data, extracts the characteristics of that data, searches for corresponding data, and recognizes the retrieved data. It was displayed as data. For example, it is described in Japanese Patent Application No. 194311/1983 filed by the present applicant.

[Problem that the invention seeks to solve]

However, in this case, the handwritten input data may be unclear and there may be gaps due to erroneous recognition. For example “b”
Handwritten data may be mistakenly recognized as "6". In this case, the device displays the most probable data as recognition data and then performs the recognition work on the next handwritten data, so there is a problem in that erroneously recognized data is processed as is.

[Means for solving problems]

In order to solve this problem, the present invention includes a message display means for displaying a question message from the device to the operator asking if the recognition result is correct when there is a recognition error. Become.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of the device. In Figure 1,
1 is a handwriting coordinate input device with integrated input and output ports;
It is composed of a tablet 12 and a liquid crystal display 13. The tablet 12 and the liquid crystal display 13 have an origin at the top left of the screen, an X coordinate of 0 to 255, and an X coordinate of 0 to 255.
It is a coordinate plane with a range of 127, and when an arbitrary position on the tablet 12 is touched with the pen 11, the coordinate value is outputted to the outside. Furthermore, when arbitrary coordinate values are input from the outside, a dot is displayed at that position on the liquid crystal display 13.

Reference numeral 2 denotes a microcomputer, which includes handwritten character recognition means 21, mathematical expression recognition means 22, calculation means 23, and display control means 24.

FIG. 2 shows an example of the hardware configuration of the microcomputer 2. An input unit 25 that receives input from the coordinate input device 1, and a CPU 26 that performs processing according to a program.
, a ROM 27 that stores dictionaries and programs for character recognition and mathematical expression recognition, a RAM 28 for auxiliary storage used for character recognition, mathematical expression recognition, and calculations, and data to the liquid crystal display 13 of the coordinate input device 1. and an output section 29 for outputting.

First, the functions of the handwritten character recognition means 21 will be explained. When characters (including numbers and symbols) are written on the tablet 12 with the pen 11, the coordinate values of each point making up the character are stored in the RAM 28 in the microcomputer 2 in sequence. When the pen 11 is removed from the tablet 12 and a certain period of time has elapsed, it is assumed that input of one character has been completed, and character recognition is started.

First, the code of the character is recognized based on the set of coordinate values stored in the RAM 28. The recognition method is publicly known. In this conventional example, the character shown in Fig. 3J is recognized. .Each character is classified as either a size-determining character or an indeterminate size character.The size-determining characters are the numbers “0” to “9”, the addition sign “+”, the multiplication sign “×”, the division sign “÷”, and the equal sign “=”, each with a length of 1 and a width of 1
A font consisting of 6 dots each (hereinafter referred to as a 16-dot font) and a font consisting of 8 dots in each vertical and horizontal direction (hereinafter referred to as an 8-dot font) are stored in the ROM 27. If the recognized character is a size determining character, the following processing is performed.

Among the set of coordinate values obtained by handwriting mentioned above,
The minimum value of the coordinate is X min, the maximum value of the X' coordinate is X m
ax sy The minimum value of the X coordinate is y mins The maximum value of the Y coordinate is 7 max 0 Here, if at least one of the values of X max Otherwise, select 8 dot font. However, for the equal sign "=", 16 dot fonts are always selected.

The 8-dot font is used to represent an exponent in a power.

Next, the X coordinate and X coordinate of the center of the font determine the coordinates at which the font is displayed, and the display control means
The selected font in M27 is displayed on the LCD display 13.
to be displayed. On the other hand, size-unstable characters are the decimal point ",", the horizontal line "-", the radical 'f', and the parentheses "(" and ")", and do not have a fixed size.The horizontal line "-" is a negative sign,
It is also used as a subtraction symbol and a fraction line. If the recognized character is an undefined size character, the following processing is performed. Among the set of coordinates obtained by handwriting, the minimum value of the X coordinate is Xm1ns, and the maximum value of the X coordinate is x entry,
The minimum value of the coordinates is ylT11nThe maximum value of the Sy coordinates is yma
Let it be t. For characters of undefined size, vectors are stored in the ROM 27 instead of fonts, and the display control means 24
The coordinates of the left edge, right edge, top edge, and bottom edge of the displayed character are X min and X max + ymin, respectively.
, 7mm! A character is constructed by vectors so as to match the character and displayed on the liquid crystal display 13. In this way, when the code, position, and size of the character to be displayed in the size-determined character or the undefined size character are determined, the character code code (i), the X coordinate of the left end of the character i), the number of dots in the X direction of the character dx (1), and the number of dots in the y direction of the character dy (i
) is stored in RAM2B. Each time the above operation is repeated and each character is input from the tablet 12, the recognized character is displayed on the liquid crystal display 13, and the R
AM28 with code (i), x (i), y (
i), dx (i), and dy (i) as x (
The data is stored in the format shown in Figure 4 while rearranging the data of the characters with the smallest value of i), that is, the characters written on the left. moves to the formula recognition means 22.Next, the processing of the formula recognition means 22 will be described.In this embodiment, an integer expressed by a number, a plus/minus sign, and a decimal point is treated as a number.

Further, as shown in FIG. 5, addition, subtraction, multiplication, division, fraction, exponentiation, and square root operations are handled according to the notation normally used in mathematics. Parentheses can also be used to change the order of operations.

First, we will discuss addition, subtraction, multiplication, and division. Addition, subtraction, multiplication, division, and mixed calculations of integers and decimals are usually written on one line when written by hand. If the priorities are the same, calculations are made from the left.In such a formula written in one line, by inputting the letters to the calculation means 23 in order from the left, the calculation means 23 determines the priority and performs the calculation. This method is known, for example, using a computer language BASIC interpreter.

Therefore, in mixed calculations such as addition, subtraction, multiplication, and division, it is only necessary to line up the handwritten characters in order from the left and output them. Since the code (i) stored in the RAM 28 by the handwritten character recognition means 21 is arranged in descending order of X·(i), the code (i) can be taken out in order and output as a code string. However, when the values of dx (i) and dy (i) in the size determining character described above are 8, it is an 8 dot font and represents an exponent, so the exponentiation must be calculated by the processing described later. Further, since the minus sign and subtraction symbol are also used as a fraction line, it must be confirmed that they are minus signs or subtraction symbols through the processing described later. Note that the plus sign and addition sign, and the minus sign and subtraction sign are output without distinction.

Next, let's talk about fractions. Fractions are another notation for division, and the following conversions can be performed.

1 → (A') ÷ (D') Here, A and B are arbitrary handwritten mathematical formulas, and A
' and B' are code strings generated from A and B, respectively.

First, formula A is constructed by extracting only the characters that exist at the position of the numerator, that is, above the fraction line. The method for generating the code string A' from the formula A when the formula A is a mixed calculation of addition, subtraction, multiplication, and division has already been described. Further, if the formula A includes a fraction, processing related to the fraction is executed recursively. Next, in the same way, only the characters that exist in the denominator position, that is, below the fraction line, are extracted to form formula B, and the code string B'
generate. A code string representing a fraction can be generated by combining code string A' and code string B' in parentheses (with the division symbol "÷"). Note that the numerator and denominator positions relative to the horizontal line are If there is no character, the horizontal line is a fraction line (distinguished as a negative sign or subtraction sign).Next, we will discuss exponentiation.In this conventional example, 8-dot font is used for numbers and operators in the exponent part, and other numbers and 16-dot font is used for the operator.As mentioned above, when writing large letters by hand, 16-dot font is displayed, and when writing small letters, 8-dot font is displayed.16-dot font characters are dx (i) =dy (i) =1
6, and the 8 dot font character is dx (i) =
dy (i)=8. In addition, the following rules apply to characters of undefined size. In other words, for characters of indefinite size other than the right parenthesis ")", if the first size-determining character extracted from now on is an 8-dot font, that character belongs to the exponent part, and if it is a 16-dot font, that character belongs to the exponent part. Belongs outside. Also the right parenthesis “)”
has the same attributes as its left parenthesis. In this conventional example, a power operator "△" is introduced to express the power as a code string, and the following conversion is performed.

cD → C'Δ(D') Here, C is an arbitrary mathematical expression, and D is an arbitrary mathematical expression that does not include exponentiation. C' and D' are each C9
This is a code string generated from D.

When a character belonging to the exponent part comes next to a character belonging to outside the exponent part while generating a code string, the power symbol "△"
and left parenthesis "(n is inserted. Also, when a character belonging to the exponent part is followed by a character belonging to outside the exponent part, a right parenthesis ")
'' is inserted. In the arithmetic means 23, there is a known method for performing mixed calculations with addition, subtraction, multiplication, and division by using this exponentiation symbol "△" in the same way as other operators. In this case, the priority of operations is shown in parentheses. The order is middle, exponentiation, multiplication/division, addition/subtraction, and if the priorities are the same, they are calculated from the left.

Next, let's talk about square roots. In this conventional example, the following conversion is performed to represent the square root as a code string.

Journey → 4(E') Here, E is an arbitrary mathematical expression, and E' is a code string generated from E. Formula E is constructed by extracting only the characters existing in the internal position of the radical, and code string E' is generated from formula E by the method described above.

If the formula E includes a square root, processing regarding the square root is executed recursively. By combining the code string E' in parentheses (() after the radical "f", a code string representing a square root can be generated. Methods for performing mixed calculations are well known.The priority of operations in this case is the square root in parentheses.

If the priorities are the same, calculations are performed from the left in the order of exponentiation, division, addition, and subtraction.

As described above, when a mathematical formula is written by hand on the tablet 12 using the pen 11, the handwritten character recognition means 21 obtains data in the format shown in FIG. Based on this, the mathematical expression recognition means 22 outputs a code string obtained by the conversion shown in FIG. 5 to the RAM 28. The calculation means 23 inputs the code string, performs calculations while determining the priority order of calculations, and outputs a solution. The solution is displayed on the liquid crystal display 13 by the display control means, thereby functioning as a computer.

Here, the processing of the formula recognition means 22 will be described in more detail. The formula recognition means 22 further has a function of outputting a formula in a one-to-one correspondence with the code string and in a format commonly used in mathematics.

The function of recognizing a mathematical formula based on the code of each character constituting the mathematical formula, the position of the character, and the size of the character and outputting it as a code string is almost the same as that of the conventional example described above. The difference is that in this embodiment, it is necessary to distinguish parentheses inserted by the mathematical formula recognition means when recognizing fractions, powers, and square roots from parentheses "(" and ")" input by hand. Therefore, the latter parentheses are represented by curly braces “(” and “)”. The calculations handled in this embodiment are the same as those shown in FIG.

In this embodiment, a formula in a format normally used in mathematics is output based on a code string obtained by formula recognition. This formula satisfies the following conditions in order to correspond to the code string and l to l.

The X coordinates of each character representing addition, subtraction, multiplication, and division of 1, integers, and decimal numbers are equally spaced, and the X coordinates are all equal.

2. The left end of the fraction line is to the left of the left ends of the numerator and denominator. Furthermore, the right end of the fraction line is to the right of the right ends of the numerator and denominator.

3. Align the right end of the radical with the right end within the radical. Also, align the bottom of the radical with the bottom of the radical.

4. Numbers, operators, decimal points, and parentheses outside the exponent part are represented in 16-dot fonts, and numbers, operators, decimal points, and parentheses inside the exponent part are represented in 8-dot fonts.

A method for generating a mathematical expression in a format that satisfies the above conditions will be explained in detail with reference to flowcharts shown in FIGS. 6 to 13.

FIG. 6 shows the main routine, starting with step 5101.
In, k, I, X, Y, DX, DY
Assign values to each variable. Here, k is the code string S (
1).

S (2), . . . is a pointer pointing to the character currently being processed. S (k) represents one character code in the code string. Further, I is a pointer pointing to the position where new character data is added to the mathematical formula data being generated. Also, X, Y, DX, and DY are each next (
represents the X coordinate, X coordinate, number of dots in the X direction, and number of dots in the y direction of the character. Note that X(1) and y(1) are the X coordinate and the X coordinate of the first character of the handwritten mathematical formula, respectively.

Next, in step 5102, a mathematical expression existing within the area is generated. In the main routine, the entire mathematical formula is included in the area, so the entire mathematical formula is generated in this step.

FIG. 7 shows a processing routine for intra-area generation.b The intra-area generation routine is called in the main routine, and is also called for the generation of formulas for numerators, denominators, radicals, and powers.

First, in step S201, S (k) is a right curly brace "
If it is, the area is finished, and the generation within the area ends.

On the other hand, in steps 5202 to 5204, S (k) is classified into the left curly brace "(", the radical "f", the power operator "△", and other characters, respectively, in steps 5207 and 3208). Step 52
1)9. The process branches to step 8206. Step 520
7. Step 3208. After completing each process in step 5209, the process returns to 5201 to process the next character. In step 5205, S (k) is the number "0" to "9" and the operators "+" and ".

It is either "x", "-", decimal point ",", or parentheses "(", ")". For these characters, the character code C0DE (I) in the formula generated here, the X coordinate of the character x (■), the y coordinate Y (I), the number of dots in the X direction DX (I), the dot in the y direction The numbers DY (I) are respectively determined. That is, C0DE (I) =S
(k).

X (I) =X, Y (1) =Y, DX
(I)=DX, DY(1)=DY. Then, in step 8206, if S (k) = "=", the intra-region generation ends. If S (k) is not “=”, the process advances to step 5210. In step 5210, preparations are made for processing the next character. That is, k is increased by 1, ■ is also increased by 1, and X is increased by DX. Then, the process returns to step 5201. As a result, X increases by DX for each character, and Y remains unchanged, so the above-mentioned condition 1 is satisfied.

FIG. 8 shows the fraction routine of step 5207 in FIG. In step 5301, the first value of I is stored in variable 1o. IO becomes a pointer pointing to the fraction line. Next, in step 5302, the character code C0DE (Io) of the fraction line and the x coordinate X (Io).

y coordinate Y (Io), number of dots in y direction DY (Io
) are determined respectively. Number of dots in the X direction DX (Io
) will be determined later. Next, in step 5303, in order to process the next character, k is incremented by 1 and S (k)
="(" is skipped. ■ is also incremented by 1. Then, in step 5304, the molecule is processed. The molecule processing routine is shown in FIG. 9. In step 5401,
By setting X (Io) +8, the X coordinate of the left end of the molecule is determined to be 8 bits to the right of the left end of the fraction line. Then, in step 5402, intra-area generation is performed. When the right curly brace ")" is found, the intra-region generation ends. Next, in step 5403, in order to align the X coordinate of the lower end of the molecule with the fractional line, the X coordinates of all characters constituting the molecule are decreased by a fixed amount.

Then, in step 5404, in order to process the next character, k is increased by 3 and S (k) = ")".

S (k+1) = “÷”, S (k+2) = “(
”. Furthermore, save the X coordinate of the right end of the numerator in the variable X1. This completes the numerator processing routine. Next, in step 5305 of FIG. The denominator processing routine is shown in FIG. 10. In step 5501,
By setting it to +8, the X coordinate of the left end of the denominator is determined to be 8 dots to the right of the left end of the fraction line. and step 55
In step 02, intra-area generation is performed. When the right curly brace ")" is found, the intra-region generation ends.

Next, in step 503, the X coordinates of all characters forming the denominator are increased by a fixed amount in order to align the X coordinate of the upper end of the denominator with the fraction line. Then, in step 5504, in order to process the next character, k is incremented by 1 and S (k)
= Skip one character “)”. Furthermore, the X coordinate of the right end of the denominator is stored in variable X2. This completes the denominator processing routine. Next, in step 5306 of FIG. 8, if xl>x2, it indicates that the molecule is to the right of the right end, and the process proceeds to step 5307. In step 5307, in order to align the X coordinate of the center of the denominator with the center of the numerator, the X coordinates of all characters constituting the denominator are increased by a fixed amount.

Then, in step 5308, by setting X to X1+8, the X coordinate of the right end of the fraction line is determined to be 8 dots to the right of the right end of the numerator. Then, in step 5309, the number of dots DX (Io) in the X direction of the fraction line is
Io). On the other hand, in step 5306
If X2, this indicates that the right end of the denominator is to the right of the right end of the numerator, and the process advances to step 5310. Step 531
0, the X coordinates of all the characters constituting the numerator are increased by a fixed amount in order to align the X coordinate of the center of the numerator with the center of the denominator. Then, in step 5311, by setting X to X2+8, the X coordinate of the right end of the fraction line is determined to be 8 dots to the right of the right end of the denominator. Then, in step 5309, the number of dots DX (Io) in the X direction of the fraction line is
(Io) Decide. The above-mentioned processing of fractions satisfies the above-mentioned condition 2.

FIG. 11 shows the square root routine of step 8208 in FIG. In step 5601, the first value of I is stored in the variable ■0. Io becomes a pointer pointing to a radical. Next, in step 5602, the radical character code CO'DE (Io) and the x coordinate X (Io) are determined. The y coordinate Y, (Io), the number of dots in the X direction DX (Io), and the number of dots DY (Io) in the y direction are determined later. Next, in step 5603, in order to process the next character, k is incremented by 2, and S (k
) = "△", S (k+1) = "(" skips the two characters. I increases by l. Then, in step 5604, processing within the radical is performed. The processing routine within the radical is the first
Shown in Figure 2. In step 57Q1, convert X to
o) By setting it to +16, the X coordinate of the left end within the radical is determined to be 1.6 dots to the right of the left end of the radical.

Then, in step 5702, intra-area generation is performed. When the right curly brace ")" is found, the intra-region generation ends. Next, in step 5703, in order to process the next character, k is increased by 1 and one character of S (k) = ")" is skipped. Furthermore, the X coordinate of the upper end of the radical symbol is stored in a variable Y, and the X coordinate of the lower end of the radical symbol is stored in a variable Y2. This completes the processing routine within the radical. Next, in step 5605 of FIG. 11, the y coordinate Y of the radical
(Io) is determined to be 4 dots above the upper end in the radical. Further, the number of dots DX (Io) in the X direction of the radical is determined to be the value obtained by subtracting the X coordinate X (Io) of the left end of the radical from the right end X within the radical. Also, the number of dots in the y direction of the radical DY (I
o) is the y-coordinate Y (
The value obtained by subtracting Io) is determined. The above-mentioned condition 3 is satisfied by the above square root processing.

FIG. 13 shows the exponentiation routine of step 5209 in FIG. In step 5801, in order to process the next character, k is incremented by 2 and S (k)=“△”.

Skip the two characters S (k+1)=“(”. Next, in step 5802, set DX=8 and DY=8 to make numbers, operators, decimal points, and parentheses in the next area 8 dot fonts. .And step 5803
In-region generation is performed in . When the right curly brace ")" is found, the intra-region generation ends.

Next, in step 5804, in order to process the next character, k is incremented by 1 and one character of S (k) = ")" is skipped. Then, in step 5805, DX=1
By setting 6 and DY=16, the following numbers, operators, decimal points, and parentheses are set to 16 dot fonts. The above-described exponentiation process satisfies the above-mentioned condition 4.

By the method described above, code string 5(t) is generated.

A mathematical expression corresponding to S (2), . . . on a one-to-l basis is generated.

Character code C0DE(I) for each character that makes up this entire formula
, x coordinate X (I), y coordinate Y (I), number of dots in the X direction DX (I), number of dots in the Y direction DY (I)
Based on each data, a mathematical formula is displayed on the liquid crystal display 13 by the display control means 24. In addition, the calculation means 23
The solution obtained by is also displayed at the same time.

As mentioned above, by displaying a formula that corresponds one-to-one with the code string obtained by formula recognition, the operator can check whether the handwritten formula has been recognized correctly, and can ensure that the formula is correct. input can be made. In addition, the displayed mathematical formulas are displayed using the notation normally used in mathematics, just like handwritten formulas, so they are very easy to understand, and because there are no variations in the position of each character due to handwriting, the displayed mathematical formulas are visually appealing. can do.

[Other Examples]

FIG. 14 is a flowchart showing the processing of the second embodiment of the present invention. In the first embodiment described above, it was possible to check whether the input formula was correctly recognized. The second embodiment is a development of the first embodiment, in which the operator is asked whether the automatically input formula has been correctly recognized if there is a possibility that it has not been correctly recognized. . If the operator answers that the recognition is correct, the calculation is performed and the solution is displayed. On the other hand, if the answer is that it was not recognized correctly, the recognition is canceled and the handwritten formula is re-edited. In addition, when there is no risk of incorrect recognition, the calculation is performed without asking the operator and the solution is displayed to save the operator's time and effort.

Mathematical formulas may not be recognized correctly in the following cases: In other words, since a certain character only partially overlaps the numerator area, denominator area, radical area, etc., it is difficult to determine whether the character is within the area or outside the area. If there is a difference between the operator and the formula recognition means, the formula may not be recognized correctly.

In other words, if the following four conditions are met, the characters in the numerator, denominator, and radical can be determined and recognized correctly.

■The X coordinate of the left end of the 1-minute fraction line is to the left of the left end of the numerator and denominator.

2. The X locus at the right end of the fraction line is to the right of the right ends of the numerator and denominator.

3. The X coordinate of the right end of the radical is to the right of the right end within the radical.

4. The y-coordinate of the lower end of the radical is below the lower end within the radical.

A method for determining whether the above conditions are satisfied and determining processing will be explained in detail with reference to the flowchart in FIG. 15.

In steps 5901, 5902, 5903, and 5904, it is checked whether the input formula satisfies the four conditions described above, and it is determined whether there is a risk that it will not be recognized correctly.

In step 5901, the X coordinate of the left end of the fraction line is
Check whether it is to the left of the left side of the molecule.

In step 5902, it is checked whether the X coordinate of the right end of the fraction line is to the right of the right ends of the numerator and denominator. In step 5903, it is checked whether the X coordinate of the right end of the radical is to the right of the right end within the radical.

In step 5904, it is checked whether the y-coordinate of the lower end of the radical is below the lower end of the radical. Step 590
A formula that satisfies all the conditions of 1.5902.5903.5904 is considered to be safe from being recognized correctly, proceeds to step 5905, performs calculations using the recognized code string, and displays the solution in step 8906. . Meanwhile, steps 5901, 5902.59
A formula that does not satisfy any of the conditions of 03 and 5904 is considered to be at risk of not being recognized correctly, and the process proceeds to step 5906. Displays the formula and displays a message asking if this is correct. The operator looks at the recognized mathematical formula displayed by the first embodiment and answers whether it has been correctly recognized. If the answer is that the code has been recognized correctly in step 5907, the process proceeds to step 5905, where calculations are performed using the recognized code string, and the solution is displayed in step 5906. On the other hand, if any answer that was not correctly recognized is obtained, the process advances to step 8908, the display of the recognized mathematical formula is erased, the handwritten mathematical formula is redisplayed, and the operator is asked to edit it again.

As mentioned above, if there is a possibility that the input formula is not recognized correctly, the solution can be displayed only when it is recognized correctly by automatically asking the operator whether the input formula was recognized correctly. This allows you to input formulas more reliably. In addition, since it does not ask you every time there is no possibility that it is not recognized correctly, it saves you time and effort.

〔Effect of the invention〕

As explained above, by providing a means for displaying the recognized formula in a format that corresponds one-to-one with the code string to be output, the operator can confirm whether the input formula has been correctly recognized. This has the effect of making it possible to input formulas reliably.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the configuration of the present invention. FIG. 2 is a diagram showing the configuration of an embodiment of the present invention. FIG. 3 is a diagram showing characters handled in the embodiment of the present invention. FIG. 4 is a diagram showing the contents of the storage means. FIG. 5 is a diagram showing calculation formulas used in the embodiment of the present invention. FIGS. 6 to 14 are flowcharts showing control procedures in an embodiment of the present invention.

Claims (1)

[Scope of Claims] A means for inputting handwritten data as coordinate data, a recognition means for recognizing the handwritten data from the coordinate data from the input means, a display means for displaying the handwritten data recognized by the recognition means, If there is a misrecognition by the recognition means, a message display means for displaying a question message as to whether or not there is a misrecognition on the display means, and according to the operator's answer to the question message displayed by the message display means, A data recognition device comprising: control means for processing data recognized by the recognition means.