JPS6362783B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a character string input method, and in particular to a character string input method that allows character strings of any character size and character string spacing to be input in horizontal or vertical writing from any position on a two-dimensional plane. Regarding column input method.

Conventionally, as a method of inputting a character string at an arbitrary position on a two-dimensional plane such as a cathode ray tube, the character string to be input is inputted using a character input device such as a typewriter, depending on the character size and size.
The character string was entered after specifying the character spacing, vertical writing, and horizontal writing, and the display position of the created character string was specified using a coordinate specifying device such as a cursor on the display, and the character string was input at an arbitrary position. .

However, with this method, input is time-consuming, and when inserting a character string into a closed area, such as in a block diagram or flowchart, if the character size or character spacing is incorrectly specified, the character string may protrude from the closed area. The problem was that it was inserted into a lopsided area.

Therefore, in order to improve the drawbacks of the above-mentioned character string input method, the present invention uses an input device with a two-dimensional coordinate plane, such as a tablet, as a character input device, and inputs information on this two-dimensional coordinate plane. In addition to sequentially handwriting characters and determining the normalized character size and character position from the character string information,
The purpose of the present invention is to provide a character string input method that allows inputting a coded character string at any position on the secondary plane of a cathode ray tube, etc. by recognizing the input characters and generating a character code. shall be. For this purpose, in the character string input method of the present invention, input handwritten characters are a character cutting means for cutting out each character, a character frame detection means for detecting the size of the handwritten character, and an origin detection means for detecting the center point of the first input character to create a character string origin. , vertical/horizontal writing detection means for identifying whether input characters are written vertically or horizontally, character spacing determining means for determining the spacing between characters, and character position normalization for normalizing character positions to be displayed. character recognition means for recognizing input characters and outputting character codes corresponding to the input characters, and obtaining normalized character size, character position, and character code signals from the input character string input. shall be.

First, before describing one embodiment of the present invention in detail, the principle of operation of the present invention will be explained in detail based on FIGS. 1 to 6.

Character strings input by hand onto the two-dimensional coordinate plane of a character input device such as a tablet generally have irregular character sizes, character spacing, character string spacing, etc. Inputting on a dimensional plane is inconvenient because the displayed character strings will be irregular in size and spacing. Therefore, in the present invention, these are normalized using a set reference value and then input into a two-dimensional plane.

Next, these operating principles will be explained in sequence.

(1) Character input and character extraction method FIG. 1 is an explanatory diagram of the character input and character extraction method.

First, character input is performed in stroke units by handwriting with an input pen on a two-dimensional coordinate plane A such as a tablet. For example, the character "a" consists of three strokes, and the character "i" consists of two strokes.

Next, the delimiter information between consecutive characters, for example, the delimiter information between the characters "a" and "i" in FIG. is set, and a pen for character input is brought into contact with this specific coordinate value area B every time one character is input to generate delimiter information. In addition, a stroke with a shape that does not appear when writing characters, for example, a specific stroke CS ₁ or CS ₂ from the bottom to the top as shown in Figure 1, is input for each character input, and this is detected and delimiter information is generated. It is also possible to do this. The latter method has the advantage of requiring less time and effort to input characters than the former method.

(2) Method for determining font size and character string origin coordinates In order to correct and normalize the size and position of characters sequentially cut out using the method described in (1) above, it is necessary to determine the standard font size and character string. It is necessary to set the origin coordinates of the column. FIG. 2 is an explanation of the character size and character string origin coordinate determination method.

In this invention, the reference character size and character string origin coordinates are determined based on the first character cut out. Assuming that the first character just cut out is the character “a” shown in Figure 2 a, the character frame
Detect the size of P ₀ (S ₀ , S ₁ ) and its center position coordinates (X ₀ , Y ₀ ). Then, the larger value of the width S ₁ indicating the size of the character frame F ₀ and the height S ₀ , that is, in the case of “A” shown in the figure, S ₀ > S ₁ , so “S ₀ ” is the value of the character string. The standard character size when normalizing each character is the center position of the character frame F ₀ (X ₀ ,
Y ₀ ) is determined as the origin coordinate of the string.

If the next cut out character is the character "i" shown in Figure 2b, the size of the character frame is S ₂ and S ₃ , and since S ₃ > S ₂ , the larger S ₃
is the character size of the character "i", and using this _S3 as a reference, it is normalized to the standard character size _S0 , as shown in d. In this case, since S ₃ >S ₀ , normalization results in S ₃ →S ₀ and S ₂ →S ₂₀ , making the whole a little smaller.

If the cut out character is S ₀ ＞S ₃ ′＞ in terms of character size, like the character ``i'' shown in Figure 2C.
If the relationship is S ₂ ′, then S ₀ with S ₃ ′ as the reference
Since it is normalized to "i" as shown in d, the whole becomes a little larger in this case.

Note that the center position (X _C , Y _C ) values of the character frame are not affected during character size normalization.

Also, if the next cut out character is vertically long and S ₂ > S ₃ , S ₂ becomes the character size of that character, and it is standardized to S ₀ based on this S ₂ . .

(3) Determination method for horizontal writing and vertical writing Figure 3 is an explanatory diagram of the determination method for horizontal writing and vertical writing, showing the center position (X ₀ , Y ₀ ) of the first written character L ₀ and the second character L ₁ center position (X ₁ ,
Y ₁ ) and the difference in horizontal coordinates DX ₀ (DX=X ₁ −
The absolute value of |DX _{0 |} is the difference in vertical coordinates DY ₀
When the absolute value (|DY 0 |) of (DY ₀ = Y ₁ − Y ₀ ) is larger than the absolute value (|DY ₀ |), it is determined that the writing is horizontal, and when the opposite is true, it is determined that the writing is vertical.

For example _, in the case of a _and b in Fig. 3, it is determined that the writing is horizontal because |DX _{0 |}
|, so it is determined to be vertical writing.

(4) Method for determining character spacing in character strings Whether written horizontally or vertically, the character spacing in each character string is generally uneven when written by hand. It is recommended that you enter the information.

The character spacing determination method of this invention will be explained with reference to FIG. When writing horizontally as shown in a or b in Figure 3, the first character L ₀ and the second character
The difference (X ₁ −X ₀ ) in the horizontal coordinates of the center position of L ₁ , that is, DX ₀ is determined as the horizontal character interval in the horizontal character string.

In addition, in the case of vertical writing as shown in Figure 3 c or d, the difference (Y ₁ - _{Y 0} ) in the vertical coordinates of the center positions of the first character L ₀ and the second character L ₁ , that is, DY ₀
Determine the vertical character spacing of the vertical string.

In this way, as explained in (3) above, the position of the second character is a factor that determines horizontal writing or vertical writing, and also determines the character spacing in that case.

(5) Method for determining line spacing for character strings Whether writing horizontally or vertically, writing is done on different lines, but when writing by hand, the line spacing between each character string is often uneven.
In this invention, the line spacing is also standardized.

FIG. 4 is an explanatory diagram of the method for determining the line spacing of character strings according to the present invention, and in the same figure, P ₀ ,
P ₁ , P ₂ . . . indicate the center of the character frame of the first, second, third, etc. characters written by hand, and the coordinates are shown in parentheses.

When writing horizontally, the absolute value of the difference between the vertical coordinate of the center of the third and subsequent characters and the vertical coordinate of the center _P0 of the first character becomes larger than a preset reference value DY _S for the first time. , the difference value is determined as the line spacing of the horizontal string. In Figure 4, the center of the fifth letter
The absolute value of the difference in vertical coordinates between P ₄ and the center of the first character P ₀ | Y ₄ − _{Y 0} | is the reference value
When it is determined that the fifth character is larger than DY _S , the fifth character is determined to be a character on the next line, and the difference in vertical coordinates DY _L (DY _L = Y ₄ − Y ₀ ) is the line spacing of the horizontal character string. It is determined that

Similarly for vertical writing (not shown), the absolute value of the difference between the horizontal coordinates of the center of the third and subsequent characters and the horizontal coordinate of the center _P0 of the first character is
When it becomes larger than a preset constant reference value DX _S for the first time, the value of the difference is determined as the line spacing of the vertical character string.

In addition, the standard value when determining the line spacing of each character string
DY _S and DX _S are selected in consideration of the size of the two-dimensional coordinate plane, the size of characters to be written, the spacing between character strings, etc.

(6) Normalization method for input character size and position Standard character size S ₀ and character string origin position (X ₀ ,
Y ₀ ), string character spacing (DX ₀ or DY ₀ ),
Normalization of the character size and center position of input characters using the line spacing (DY _L or DX _L ) of a character string will be explained with reference to FIG. 5 with reference to other figures.

(i) Normalization method for the size of input characters The normalization method for the size of input characters is as explained in (2) above. In other words, as shown in Figure 2, the reference character whose width and height are both S ₀ based on the larger of the width and height (S ₀ ) of the character frame F ₀ of the first character L ₀ written by hand. The size is set. For the second and subsequent characters, the larger width or height of each character frame is determined as the character size, and the entire character is enlarged or reduced so that the character size is equal to the standard character size _S0 . conversion is carried out.

(ii) Normalization method for character center position An example in the case of horizontal writing will be explained with reference to FIG. 5, but the same applies to the case of vertical writing.

In the case of horizontal writing, the coordinate origin (X 0 , Y ₀ ) is determined by the coordinates (X ₀ , Y ₀ ) of the center P ₀ of the first character.
Y ₀ ) is _the character interval _{DX 0 ( DX 0}
= X ₁ - X ₀ ) is the line spacing DY _L of the string from the coordinates (X ₄ , Y ₄ ) of the center P ₄ of the fifth character.
(DY _L = Y ₄ − Y ₀ ) is determined, so the third and subsequent characters are DX ₀ in the X direction and DY _L in the Y direction.
Normalization is performed by rearranging the coordinates to one of the grid points (K ₀ , K ₁ , K _{2 .} . . ), which are intersection points of coordinates separated at equal intervals.

During normalization, each character is relocated to the grid point closest to the center of the character. In Figure 5, P ₁ becomes K ₁ and P ₂ becomes K ₂
Then, P ₃ is normalized to K ₃ ...P ₇ is normalized to K ₇ .

This normalization operation can be performed according to the following arithmetic expression.

_{II = ( X C −} When the signs of X _C −X ₀ ) and DX ₀ are equal -1: Otherwise II: Indicates the integer value part of the operation result on the right side.

Furthermore, the arithmetic device that executes the calculation of formula (or ') is configured to truncate the decimal part of the calculation result without rounding it off, and generate only the integer part as output. This simplifies the configuration of the arithmetic device.

Next, the operation of calculating the above equation will be explained.

When the character center P _C is located between the grid points K _C+1 and K _C (where the X coordinates are X(K _C+1 ) and X(K _C ), respectively), the nearest grid point In order _to normalize the X coordinate, the character center P _C

By the way, arithmetic devices do not have a rounding function and simply output only the integer value part, but in order to have such arithmetic devices perform an operation equivalent to rounding, the first As shown in the second term in the calculation result of the term
An operation of adding 0.5 (or 1/2) DX ₀ is performed. By doing this, when X(K _C )≦X _C <X(K _C )+DX ₀ *0.5, that is, when P _C is close to the K _C point, X(K _C )≦X _C +DX ₀ *0.5 <X (K _C ) + DX ₀ , _that is, X (K _C )≦X _C + _{DX 0 * 0.5 <} indicates the X coordinate X(K _C ) of the K _C point.

If X(K _C )+DX ₀ *0.5≦X _C <X(K _C+1 )
When P _C is close to the K _C+1 point, X (K _C ) + DX ₀ ≦ X _C + DX ₀ *0.5 _<
+DX ₀ *0.5, and the integer part of the value obtained by dividing this X _C +DX ₀ *0.5 by DX ₀ is the X coordinate X of point K _C+1 (K _C+1 )
will be shown.

By the way, when the sign of DX ₀ and the sign of X _C −X ₀ are different, that is, in the case of point P ₂ in FIG. 5, DX ₀ >0 and X _C −X ₀ <0, so P ₂
To normalize X _C − to the nearest grid point K ₂
The value of DX ₀ *0.5 (>0) must be subtracted from X ₀ (=X ₂ −X ₀ <0). SN in the second term on the right side of the above equation is an arithmetic operation performed for this purpose.

The same applies to the second and subsequent lines of the character string, and the normalized positions for characters after the character for which the line spacing (DYL) of the character string has been determined are as follows.

Normalized position coordinates (Xp, Yp) = (X ₀ +DX ₀ *
II, Y ₀ + DYL * JJ) Here, II is the integer obtained using the above formula JJ = (Y _C − _{Y 0} + SN * DYL / 2) / DYL
...However, SN = 1: When the signs of (Y _C - Y ₀ ) and DYL are equal - 1: In other cases, it is given.

(7) Handwritten character recognition and lowercase code conversion Handwritten characters whose character size and center position have been normalized are recognized as known handwritten characters, and the recognition results are output as character category codes.

In addition, lowercase letters (including punctuation marks ".", ",", etc.) are distinguished by the recognition device using the size information of the input characters, and after converting them into codes, the character category code, character size, and character position are determined. Store in the string storage area.

Also, in the case of vertical writing, for example, the lowercase letter "tsu"
Even if it is input by hand as shown in Figure 6a, it is necessary to correct the ``tsu'' to the upper right corner position as shown in Figure 6b, so it will be recognized as a vertically written and lowercase ``tsu''. , it is converted into a character category code as a lowercase letter written in the position shown in FIG.

Note that "pa", "ba", "ka", etc. are converted into character category codes as a whole, so there is no confusion with lowercase letters.

Furthermore, the 8th character in Figure 5 (center P ₇ )
If a character has already been normalized and entered at grid point _K7 when inputting , the already entered character will be automatically deleted and the newly entered 8th character will be entered. It turns out.
This facilitates character correction.

Next, the configuration of an embodiment of the present invention will be explained with reference to FIG.

The input data Si is a stroke series of characters generated from a two-dimensional coordinate plane such as a tablet (not shown), and what is generated as output is a character with position (Xp, Yp) and character size _S0 . The category code is COD.

Below, FIG. 7 will be explained in the order of its operation.

(A) Input data Si is transmitted to the character cutting circuit 1. This character extraction circuit 1 extracts characters by detecting delimiter information from specific strokes (CS ₁ , CS ₂ ) or specific coordinate value area B shown in FIG. and outputs character data DATA.
This character data DATA is cut out input data and is transmitted to the character frame detection circuit 2 and the character recognition circuit 8. The character number N is then transmitted to a character size and character string origin detection circuit 3, a vertical/horizontal writing detection circuit 4, a character spacing determining circuit 5, a character string line spacing determining circuit 6, and the like.

(B) The character frame detection circuit 2 receives character data for each character transmitted from the character cutting circuit 1.
Based on DATA, the center position (X _C , Y _C ) of the input character and the character size (S) of the input character are output for each character.

(C) The character size and character string origin detection circuit 3 determines the center of the first input character (N=1) transmitted from the character frame detection circuit 2 based on the character number N transmitted from the character cutting circuit 1. Coordinate the position and font size to the origin (X ₀ ,
Y ₀ ) and standard character size S ₀ .

(D) When the vertical/horizontal writing detection circuit 4 detects the second character number N=2 output input from the cutting circuit 1, the center of this N=2 character output from the character frame detection circuit 2 is detected. Position (X ₁ , Y ₁ )
By comparing the absolute values of (X ₁ , X ₀ ) and (Y _{1 , Y 0 )} , as explained in FIG. Determine horizontal writing and send vertical/horizontal writing designation signal
Output TY. At this time, for example, TY = "0" for horizontal writing, and TY = "1" for vertical writing.
It is determined that And this vertical writing/horizontal writing designation signal
TY is sent to the character spacing determining circuit 5, character string line spacing determining circuit 6, character recognition circuit 8, etc.

(E) Character spacing determining circuit 5 is the second (N=2)
By comparing the _character center position (X _C = X ₁ , Y _C = Y ₁ ) with the coordinate origin (X ₀ , Y ₀ ), the horizontal character spacing DX ₀ (DX _{0 =} ), and when writing vertically, the vertical character spacing DY ₀
(DY ₀ = Y ₁ − _{Y 0} ) is output. This horizontal character spacing DX ₀ or vertical character spacing DY ₀ is sent to the character string spacing determining circuit 6.

(F) The character string line spacing determining circuit 6 determines the line spacing in the following manner depending on whether the character string is written horizontally or vertically.

(B) Difference between the Y coordinate of the input character and the origin when writing horizontally (TY = "0", DX ₀ ≠ 0) (Y _C - Y ₀ )
When the absolute value of
=Y _C −Y ₀ and output it.
Otherwise, the values of DX ₀ and DY ₀ remain unchanged.

(b) When writing vertically (TY = "1", DY ₀ ≠ 0), the difference between the X coordinate of the input character and the origin (X _C -
When the _absolute value of
Determine DXL=X _C −X ₀ and output this. Otherwise, the values of DX ₀ and DY ₀ remain unchanged.

Then, by operating (a) or (b), character spacing DX ₀ and line spacing DYL are output for horizontal writing (TY = "0"), and for vertical writing (TY = "1")
In this case, the character spacing DY ₀ and the line spacing DXL are output and sent to the character position normalization circuit 7.

(G) The character position normalization circuit 7 has the above-mentioned (C), (E), and (F).
Parameters determined by (X ₀ , Y ₀ , DX ₀ , DY ₀ ,
DYL, DXL) to calculate the normalized position (Xp, Yp) of the input character. For example, when writing horizontally, calculations such as the above equations and equations are performed. Here, Xp=X ₀ +DX ₀ *II, Yp=Y ₀ +DYL *JJ.

(H) The character recognition circuit 8 recognizes the handwritten input character based on the character data DATA and the character size S transmitted from the character cutting circuit 1, converts it into a character category code COD, and outputs this character category code COD. However, when writing vertically (TY = 1) and lowercase letters, this is converted to a character category code for vertical writing.

In this way, the finally output character size, the reference character size S ₀ obtained from the character string origin detection circuit 3, the normalized position (Xp, Yp) obtained from the character position normalization circuit 7, and the character recognition circuit The character category codes COD obtained from 8 are stored as a set in a character string storage area (not shown) in character input order.

When displaying this, each input character is normalized to a reference character size _S0 on a display device having a two-dimensional plane (not shown).

If a display device that is not equipped with normalization means is used, the character recognition circuit 8 outputs the standard character size S0 in advance based on the character size and the standard character size _S0 output from the character string origin detection circuit 3. All you have to do is normalize the character category code COD.

As explained above, according to the present invention, by writing character strokes by handwriting ordinary characters on a two-dimensional coordinate plane such as a tablet, characters can be automatically encoded, their sizes normalized, and Since the position can be normalized, it is possible to omit the input of parameters for creating character strings, which were conventionally necessary for inputting character strings, and it is possible to create character strings very easily.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram of character input and cutting, Fig. 2 is an explanatory diagram of character size and character string origin determination, Fig. 3 is an explanatory diagram of horizontal writing, vertical writing determination, and character spacing determination of character strings, and Fig. 4 is an explanatory diagram of character input and cutting state. The figure is an explanatory diagram of the line spacing determination state of a character string, Fig. 5 is an explanatory diagram of the normalization state of the character center position, Fig. 6 is an explanatory diagram of the lowercase character code conversion state, and Fig. 7 is a configuration diagram of an embodiment of the present invention. It is. In the figure, 1 is a character cutting circuit, 2 is a character frame detection circuit, 3 is a character size and character string origin detection circuit,
4 is a vertical writing/horizontal writing detection circuit; 5 is a character spacing determining circuit;
Reference numeral 6 indicates a character string line spacing determination circuit, 7 a character position normalization circuit, and 8 a character recognition circuit.

Claims (1)

[Claims] 1. In a character string input method that inputs a character string of any character size and character string spacing in horizontal or vertical writing from any position on a two-dimensional plane, a character cutting means for cutting out input handwritten characters character by character; , a character frame detection means for detecting the size of this handwritten character, an origin detection means for creating a character string origin by detecting the center point of the first input character, and a character frame detection means for detecting the center point of the first input character to create a character string origin, and a character frame detection means for detecting the center point of the first input character to create a character string origin, and a character frame detection means for detecting the center point of the first input character to create a character string origin. vertical and horizontal writing detection means, character spacing determination means for determining the spacing between characters, character position normalization means for normalizing the character position to be displayed, and 1. A character string input method characterized in that a character recognition means for outputting a corresponding character code is provided, and normalized character size, character position, and character code signals are obtained from an input character string.