JP2681572B2 - Handwritten character input method and device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for inputting handwritten characters using an information processing device, a Japanese word processor, etc. and a storage device for input.

[0002]

2. Description of the Related Art Conventionally, handwritten character readers and optical character readers have been used to calculate slips, statistical data, etc., but the typeface of handwritten characters is greatly affected by variations in personal characteristics. Since they are different, they cannot be completely decoded by an information processing device such as a computer. According to experiments, an average handwritten character can recognize only about 3,000 to several hundred out of about 6,300 kanji up to the JIS second level. With the advent of 12,000 kanji, including supplementary kanji, many word processors of the original language are in widespread use today. However, there are still some people who are still unfamiliar with the keyboard of a personal computer and create handwritten sentences and have the operator input the sentences again to the word processor.

[0003]

It is a double labor and inefficiency that a costly operator, apart from the writer, has to input a prepared sentence by the keyboard after that. Further, since people who dislike keyboards still want to use word processors and personal computers, the advent of a handwriting character recognition device with a high decoding rate that enables such people to use word processors and personal computers has been desired. The present invention has been made to solve the above-mentioned problems, and prioritizes handwriting input without relying on keyboard input, and allows handwritten characters to be recognized so that the drawn characters can be directly input to a word processor or the like. It is an object to provide a recognition method and its device.

[0004]

According to the first aspect of the present invention, a step of decomposing a kanji to extract basic elements constituting the kanji,
A step of organizing these basic elements as individual two-dimensional image data to create a list for retrieval, a step of registering this list in a storage device, and a collection of kanji having the registered basic elements as constituent elements. As a group of Chinese and Chinese characters, creating this group of Chinese and Chinese characters for each basic element, registering multiple Chinese and Chinese character groups in a storage device, and regarding each Chinese and Chinese character group, set the basic elements in the stroke order of Chinese characters. , Collecting the first kanji to start writing into the first stroke order small kanji group, collecting the middle kanji to start writing into the second stroke order small kanji group, and collecting the last kanji starting to write 3 stroke order small Kanji character groups, creating these stroke order small Kanji character groups for each middle Kanji character group, registering a plurality of stroke order small Kanji character groups in a storage device, and the Kanji to be drawn Of handwriting only the basic elements of And converting the basic elements in a two-dimensional image data writing, and selecting similar basic elements in the converted image data from the registered basic elements,
A step of designating one or a plurality of the middle kanji group, the first stroke order small kanji group, the second order small and small kanji group, and the third stroke order small kanji group, which are related to the selected basic element; In the case of two, the step of displaying the specified Chinese character group or the stroke-order small Chinese character group on the display device, and in the case of multiple, the logical product of the specified multiple Chinese character groups or the stroke-order small Chinese character group is displayed and the result is displayed on the display device. It consists of a step and a step of selecting a desired kanji from the displayed middle kanji group or stroke-order small kanji group and inputting handwritten kanji into the information processing device. Further, in the second invention, the step of investigating the position of the basic element in the handwriting arrangement of the Chinese character for each Chinese character group is performed, and a plurality of position small numbers are set for each of the Chinese character group according to the position information. Forming a Chinese character group, creating these small Chinese character groups for each middle Chinese character group, registering multiple multiple small Chinese character groups in a storage device, and drawing only the basic elements of that Chinese character And the step of handwriting
A step of converting into a three-dimensional image data, a step of selecting a basic element similar to the converted image data from the registered basic elements, and a step of relating to the selected basic element from a plurality of position small Kanji character groups. The step of specifying one or more, the step of displaying the specified position kanji group on the display device in the case of one, and the logical product of the specified position kanji character groups in the case of multiple, and the result is displayed on the display device. And a step of selecting a target kanji from the displayed position kanji group and inputting handwritten kanji into the information processing device.

[0005]

[Function] A basic element which is a target kanji part is input by handwriting, image data of the basic element is compared with image data of the basic element, and the basic element of the handwriting is recognized in correspondence with the registered basic element. The middle Chinese character group, the stroke order small Chinese character group, and the position Chinese character group Chinese character related to the determined basic element are displayed on the CRT. A desired kanji is selected from the displayed kanji and input to the information processing device.

[0006]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings. FIG. 1 shows the overall configuration of a handwritten character recognition device using a desktop computer or the like, which includes an input panel 12, a central processing unit (CPU) 13, a baer memory 14, a random access memory (RAM) 15, a pen 18, a magnetic recording device. 19
And a display device (CRT) 20. ROMs 16, 17, and 22 are provided in the magnetic recording device 19 as read-only areas.
7 and 22 are designated as write-protected, but the magnetic recording device 1
9 itself is readable and writable. The ROM 16 stores two-dimensional image data of basic elements such as a radical described later in a list form. ROM17 is a ROM1 for storing Chinese characters
7A, a ROM 17B for storing a stroke order small Kanji character group, and a ROM 17C for storing a position small Kanji character group.
Reference numeral 2 is a well-known character data dictionary that stores up to JIS level 2 and the largest group of large kanji including auxiliary kanji. These ROMs 16, 17, 22 may be formed as an integrated circuit.

When writing characters with the pen 18 lightly touching the input panel 12 like writing a character on a sheet of paper with a pencil, the handwriting generates a residual charge due to pressure contact, and the character appears due to polarization or the like. It utilizes phenomena.
This residual charge is two-dimensional image information and is given to the buffer memory 14 as character image data. This image data is stored on the RAM 15 as a work memory by the CUP 13 in the ROM 16 or the ROM 1.
It is calculated together with the data from 7 and intermediate results and final results are stored in the magnetic recording device 19. The final calculation result is displayed on the CRT 20 of the display device via the CUP 13. In FIG. 2, the input panel 12 is composed of an image area 28 for drawing characters by handwriting, a push button type numeric key 25, a radical key 26, an execution key 29, and a logical product key 24 for logical operation. The numeric key 25 is pressed by pressing the numbers 0 to 9, the radical key 26 is pressed by pressing positional information such as the radical, and the execution key 29 is pressed by the input confirmation button. A method for carrying out the present invention using the handwritten character recognition device having the above-described configuration will be described in order.
By the way, each of the currently used educational kanji, common kanji, and personal name kanji is often composed of two or more basic elements due to its structural structure. These basic elements include the radicals of the radicals, which are traditionally used in the Kanji dictionary, bruise, crowns, legs, hangings, gutters, and structures, which are considered as one block in the stroke order as shown in FIG. As one example, "mountain", "word",
Drawing elements such as “large” and “system” correspond to the drawing elements.

Now, a plurality of basic elements are combined to form one kanji. According to the position of each kanji, the kanji is divided into two right and left parts, and the left part is biased.
The right part is divided into upper and lower parts, and the upper part is called a crown, and the lower part is called a leg. In addition to these four parts, the part that hangs down from the top is suspended, the part that goes around the lower part is covered, and the part that wraps the whole is called the frame. A common radical is one obtained by extracting a basic element having a common form from the above seven elements. Further, in addition to these seven classifications, it is also known that the same shape as one of the radicals, such as “uneven” or “near”, is located in the middle of the stroke in the stroke order.

In the present invention, the basic elements such as the radical are
Survey up to JIS level 2 and usable kanji,
As shown in FIG. 3, the vertical is about 20 of A to T, the horizontal is about 30,
20 × 30 = 600 pieces are selected. The selected list of radicals is registered in the magnetic recording device 19 as image data by assigning a code number to each basic element and using a known technique. The part in which this list is registered is set as ROM 16 and RO
M16 is a dictionary of radical data. In this dictionary, each radical or the like is registered as a reference pattern, for example, two-dimensional image information in the form of 5 × 7 dot matrix. Here, you can reduce the number of elements by setting similar basic elements such as "bamboo", "less", "yen" and "tan" as the same element, and display them side by side to make it easier to find the element. You can also create a list of basic elements by devising. Further, in FIG. 3, similar radicals and the like are gathered in the lateral direction and are given the same calling number for convenience of use.

Next, the present invention will be described with reference to the flowchart of FIG. 4 and FIG. 5 as a method of creating a group of various kanji by associating a list of basic elements with JIS1 kanji and JIS2 kanji. First, in step S1 in FIG. 4, the educational kanji, common kanji, and personal name kanji that are currently available are listed as much as possible, and for example, the largest kanji group K up to JIS level 2 is created (in ROM 22). Using letters,
Register additional shortfalls). Next, in step S2, with respect to the collected large Chinese character group K, all the basic elements corresponding to each basic element in the list in FIG. And the Chinese character group 1 is registered in the magnetic storage device 19 in step S3. In step S4, it is checked whether or not the Chinese-Kanji group has been created for all the 600 basic elements in FIG. 5, and steps S2 to S4 are repeated to detect all other basic elements such as the "mountain" and "large" of the radical. Chinese character group 1 ~ n ~ i ~ 60
Create 0. Therefore, by mobilizing surveyors, the Chinese character groups 1 to n to i consisting of approximately 600 groups as shown in FIG.
~ 600 are created. The portion in which this Chinese character group is registered is designated as ROM 17A, and ROM 17A serves as a dictionary of Chinese character group data.

Next, in step S5, the first kanji character group 1-1 is extracted by extracting the kanji whose radical, for example, "Koto" is the first in the stroke order from the middle kanji character group 1, and in step S6 the middle kanji character group 1. From the second stroke order small kanji group 1-2 by extracting the kanji in which "word" is in the middle of the stroke order, and the second kanji in which "word" is the last in the stroke order is extracted from the middle kanji group 1 in step S7. A stroke order small Kanji group 1-3 is created, and in step S8, these first to third
The stroke order small Chinese character groups 1-1 to 1-3 are registered in the magnetic storage device 19 or the like. In step S9, all the Chinese character groups 1 to i in FIG.
About ~ 600 First to third stroke order Small Kanji group 1-1 to i-3
To 600-3 are created, steps S5 to S9 are repeated, and the first to third steps are performed for the basic element "mountain".
Stroke order small kanji groups n-1 to n-3 and "large"
3 stroke order small Kanji character groups i-1 to i-3 and all other basic elements No. 1 to 3 stroke order small Kanji character groups 600-1 to 600-3
Create up to. Therefore, by mobilizing investigators,
Approximately 600 × 3 = 1800 first to third as shown in FIG.
A stroke order small kanji group is created. In the magnetic recording device 19, the portion in which the first to third stroke order small kanji groups are registered is the ROM 17B.
The ROM 17B serves as a dictionary of the first to third stroke order small Kanji group data.

The first to third stroke order small Kanji groups 1-1 to i-
In the process of creating 3 to 600-3, various attributes for each kanji will be clarified. For example, "Take" is a group of 3 or more (actually 4) stroke order small Kanji characters 1-2, n-
1 and i-3 exist, for example, these three types of attributes are added to “Take”, and the Chinese character with the attribute “Take” is registered again in the magnetic storage device 19. Accordingly, individual kanji in the stroke order small kanji groups 1-1 to i-3 to p-3, for example, "Take", are added in the magnetic storage device 19 as shown in FIG. Remembered FIG.
In Japanese, the data packet 29 of the kanji "Take" has the JIS code number 5656 shown in hexadecimal, the first attribute 31 is "word 2", the second attribute 32 is "mountain 1", and the third attribute 33 is "large". 3 ”is added. The number 1 following the basic element of the radical,
2 and 3 indicate that the basic element is in the middle in the stroke order of "Take" in "Koto 2".
Indicates that Kappayama is the first in the stroke order, and "Large 3" is the last in the stroke order. Here, the basic elements of radicals are followed by numbers 1, 2, and 3, but the basic elements of radicals are followed by "2 words,""1mountain," and "3 majors."
It is good.

A plurality of different kanji and code numbers are used for the kanji satisfying the same condition of "word 2" of the first attribute 31, "mountain 1" of the second attribute 32, and "large 3" of the third attribute 33? ? ? ? Exists (???? is an arbitrary number from 1 to 9), and there are a plurality of candidate kanji in the memorable character. These first attribute “Word 2”, second attribute “Mountain 1”, and third attribute “Large 3” are represented by ordinary characters for convenience of description, but they are 2 in the magnetic storage device 19 or the information processing device. It is recorded in hexadecimal or hexadecimal.

Next, a radical in each kanji, for example, the handwriting position of "Koto" is grasped as two-dimensional position information (arrangement information), the middle kanji group is analyzed, and the following position small kanji group is generated. The method for doing so will be described below with reference to the flowchart of FIG. 6 and FIGS. In step S50, all kanji in which the radical "word" is biased at the handwriting position are extracted from the middle kanji group 1 and the first position small kanji group 1-11 is extracted. In step S51, the middle kanji group 1 " The second position small kanji group 1-12 is obtained by extracting the kanji in which "Koto" is a handwriting position.
In step S52, the kanji in which "Koto" is a crown at the handwriting position is extracted from the middle kanji group 1 to create a third position small kanji group 1-13. In step S53, the middle kanji group 1 is extracted. The kanji in which "word" is a handwriting position and is a leg is extracted and the fourth position small kanji character group 1-14 is selected. In step S54, "word" is selected from the middle kanji character group 1 and is "nyo" at the handwriting position. A certain kanji is extracted to create a fifth position small kanji character group 1-15, and in step S55, a kanji character in which "Koto" is in the handwriting position is extracted from the middle kanji character group 1 and the sixth position small kanji character is extracted. Group 1
Create -16.

In step S56, "Koto" is selected from the Chinese character group 1.
Creates the seventh position small Kanji group 1-17 by extracting the upper left Chinese character at the handwriting position, and in step S57 extracts the Chinese character at which the radical "Koto" is at the lower left position at the handwriting position from the middle Chinese character group 1. The 8th position small kanji group 1-18 is extracted from the middle kanji group 1 in step S58 by extracting the kanji whose word position is the upper right at the handwriting position, and the 9th position small kanji group 1-19 is extracted in step S59. From the group 1, the kanji in which the word "Koto" is located at the lower right at the handwriting position is extracted to create the 10th position small kanji group 1-20, and step S
In step 60, the Chinese character in which "Koto" is in the handwriting position is extracted from the middle Chinese character group 1 and the 11th position small Kanji group 1-21 is extracted. The lower kanji is extracted to create a 12th position small kanji group 1-22, and in step S62, the kanji in which "Koto" is in the handwriting position is extracted from the middle kanji character group 1 to the 13th position small kanji group. 1-
23 is created, and these position small Chinese character groups 1-11 to 1-23 are registered in the magnetic recording device 19 in step S63.

In step S64, the first to thirteenth position small Chinese character groups 1 for all the Chinese character groups 1 to n to i to 600 in FIG.
-11 to i-23 to 600-23 are checked, steps S50 to S63 are repeated, and all basic elements such as the basic elements "mountain" and "large" in the list of FIG. 3 are shown in FIGS. 1st-13th position small kanji group 1-11-1-2
3, ~ n-11 to n-23, ~ 600-11 to 600-
Create 23. Therefore, by mobilizing the investigators, it becomes approximately 600 × 13 = 7800 as shown in FIGS.
A group of individual small Kanji characters is created. In the magnetic recording device 19, the portion in which the position small Kanji character group is registered is referred to as a ROM 17C, and the ROM 17C serves as a dictionary of position small Kanji character group data.
Actually, the basic element "word" of the bias does not have all the first to 13th position small Kanji groups 1-11 to 1-23, so the number of position small Kanji groups is less than half, about 3,500. become.

Now, the first to thirteenth position small Chinese character groups 1-11 to 11
1-23 ... n-11 to n-23 ... 600-11
In the process of creating ~ 600-23, various attributes for each kanji are revealed. For example, "Take" is a group of 6 or more small kanji characters, such as "Koto", middle and lower 1-22,
N-13 of "mountain" crown, i-in the lower right of "large" (not shown)
20 and the lower left of "Kemono". These six or more types of attributes are added to "Take" and the Chinese character with the attribute "Take" is registered in the magnetic storage device 19 again. Therefore, the 1st to 13th position small Chinese character groups 1-11 to 1-23 ...
The individual kanji in n-11 to n-23 ... p-11 to p-23, for example, "Take", are shown in FIG.
The attribute is added and stored. In FIG. 10, the data packet 110 for the kanji "Take" is J in hexadecimal.
The code number 5656 of the IS is added with "word pear" of the fourth attribute 111, "mountain force" of the fifth attribute 112, and "large miss" of the sixth attribute 113. Further, “Kemonohishi” as the seventh attribute 114, “Inu Mishi” as the eighth attribute 115, and “Kokuki” as the ninth attribute 116 are added. The katakana "pear" following the basic element of the radical "word" is the basic element "word"
Indicates that it is the handwriting position in the middle and lower part with the character "Take",
In the following, "Yamaka" indicates that "Yama" is the crown of the handwriting position in the character "Take," and "Large Miss" indicates that "Large" is the lower right of the handwriting position.

In addition, "Kemono Hishi" is "Kemono" at the lower left of the handwriting position, "Inu Mishi" is "Dog" at the lower handwriting position at the lower right, and "Kokuki" is "Koku" at the handwriting position as a leg. It shows respectively. Here, the basic elements of the radical are followed by the katakana "pear", "mosquito", "miss", etc., which indicate the position information, but after the katakana "pear", "mosquito", "miss", etc. The basic elements may be continued and used as "Nashi word", "Ka mountain", and "Miss large". Next, a method of inputting handwritten characters will be described with reference to the flowchart of FIG. The image area 2 of the input panel 12 for the text "Take ..."
The case of drawing 8 will be described. It is not necessary for the input person to draw all the strokes, and in step S65 only the easy-to-draw "words" of the basic elements are drawn as if they were shorthand. The residual charge of the handwritten "word" becomes two-dimensional image information and is given to the buffer memory 14 in step S66. Next, in step S67, the CUP 13 sets the RAM 1
5, the radical image data from the ROM 16 and the handwritten image information are compared and compared to select a radical “word” and another candidate radical, for example, “month”, and these candidate radicals are CU.
It is displayed on the CRT 20 of the display device via P13. This comparison method enlarges or reduces the handwritten image information as in the known technique to obtain the same 5 × as the radical data dictionary.
Normalize to a matrix size of 7. Then, the standardized image information and 6 from the ROM 16 of the radical dictionary
00 reference radical patterns are compared, and candidate radicals are numbered and displayed, such as "word" 1, "month" 2, etc., in descending order of similarity.

At step S68, since the inputter intends "word", "1" and "execute" are input from the numerical keypad 25. In step S69, "word" of the component of the desired kanji "Take" is fixed for the first time, and the process proceeds to the flowchart of FIG. In step S10 of FIG. 12, the target kanji “Take” is sought,
Only the confirmed “word” is displayed on the screen of the CRT 18. In step S11, the instruction to the CPU 13 is “outline”
If it is a search, the process proceeds to step S13 and step S13.
So, a group of Chinese characters that includes all the Chinese characters that have radical words 1
Is displayed on the CRT 20.

In step S14, the desired "Take" is found in the candidate Kanji character list of the middle Kanji group 1 and is confirmed by the "Execute" key 29, the search process is terminated, and "Take" in step S90.
Enter the following characters by handwriting. If the kanji cannot be determined because there are too many candidate characters in step S14, “Zaku” is searched for from the radical other than “Koto” according to arrow Z.
Return to the flowchart of 1. This time, for example, the basic element “mountain” is drawn on the panel 12, steps S65 to S69 are performed again, radical recognition is performed, and only the “mountain” confirmed in step S15 is CR according to the arrow X this time.
Display it on the screen of T18. "Mountain" in step S16
The Chinese character group n of is displayed, and the Chinese character group n is displayed in step S17.
When the target "Take" is obtained and confirmed in the candidate kanji list of, the search process ends and the characters after "Take" are input by handwriting.

If the kanji cannot be determined in step S17,
In order to search for “Take” from radicals other than “mountain” according to arrow Z, the process returns to the flowchart of FIG. This time, for example,
Draw the basic element "large", repeat steps S65 to S69, and follow arrow X this time to step S69.
At 18, only the confirmed “large” is displayed on the screen of the CRT 20. In step S19, the Chinese character group i of "Large" is further displayed, and in step S20, when the target "Take" is obtained and confirmed in the candidate kanji list of the Chinese character group i, the search process ends and the characters under "Take" are selected. Enter by handwriting. Even in the "outline" search, most kanji are read and searched with about three radicals.
Input can be completed. In addition, the radical to draw is "mountain", "word",
You can go first from "large", and you can do whatever you want with ease.

In step S11 of FIG. 12, the CPU 13
If the instruction to search is "detailed" instead of "summary", arrow A
In accordance with this, the process proceeds to the flowchart of FIG. 13 (CPU13
Input to the normal keyboard). In step S22, it is selected whether to use the first to third stroke order small Kanji groups in FIG. 5 or to use the first to 13th position small Kanji groups in FIGS. Now, in step S22, when "1" is input by the instruction of the pen 18, the search target is expanded to the 1800 sets of the first to third stroke order small Kanji groups 1-1 to i to 3 in FIG. If the basic element determined in step S23 is "word", the process proceeds to step S24. This step S24
When you press "1" of the key 25 with the pen 18 and specify "Word 1" to write first, in step S25 the radical "Word"
The first stroke-order small Kanji group 1-1 of FIG. 5 showing all the Kanji that are first in the stroke order is displayed.

In step S26, the first stroke order small Chinese character group 1-1
When the target "Take" is obtained and confirmed from the candidate kanji list of,
The search process ends, and the characters under "Take" are input by handwriting. Here, since "Take" does not exist in the first stroke order small Kanji group 1-1 and the Kanji can be determined, the process proceeds to step S27. Here, since the pen 18 presses the "2" of the key 25 to specify the "word 2" to be written in the middle, in step S28 the radical "word" shows all the kanji in the middle of the stroke order. The second stroke order small Kanji group 1-2 of 5 is displayed. Step S29
Then, the target "Take" is obtained and confirmed in the candidate kanji list of the second stroke order small kanji group 1-2, so the search processing ends,
Hand-write the characters after "Take". For the sake of explanation, if the kanji cannot be confirmed in step S29, step S29 is performed.
Go to 30. Here, since the pen 18 is used to press the “3” of the key 25 to specify the “word 3” to be written in the middle, at step S31, the radical “word” shows all the kanji whose stroke order is the last. The second stroke order small Kanji group 1-3 of No. 5 is displayed.
If the candidate character is not found in step S32 and the kanji cannot be determined, the process proceeds to step S33 in which the desired kanji "Take" is JI.
Consider another measure assuming that it is not registered in the S2 level range. In step S30, if "others" is indicated instead of "word 3," an error is displayed.

Further, the “word 1” in step S24, the “word 2” in step S27, and the “word 3” in step S30 may be searched first. Step S2
If the input of the handwritten character is "mountain" instead of "word" in 3, the process proceeds to step S34. About "mountain", "mountain 1",
The search is continued by advancing to "mountain 2" and "mountain 3" and following the same operations as in steps S24 to S33. Furthermore,
If the input of the handwritten character is "large" instead of "mountain" in step S34, the process proceeds to step S35. With respect to "large", the process is advanced to "large 1", "large 2", "large 3", and the same operation as in steps S24 to S33 is followed to continue the search. Further, the "word" in step S23, the "mountain" in step S34, and the "large" in step S35 may be drawn first.

An actual use example of the present invention will be described below by using a kanji group. For example, regarding "words", the one that this basic element starts to write first is "1 word", the one that starts writing in the middle is "2 words", and the last to start writing is "3 words". Here, for example, "1 word (148)" indicates that there are 148 Kanji characters that start writing word bias or Straw, and the Kanji characters under ":" represent candidate characters to be displayed.

In the following, regarding "word" and "kemonohen", all the first to third stroke order small Kanji groups 1-1 to 1-3, p-
With respect to 1 to p-3, a kanji group collected and displayed according to whether to start writing first, start writing in the middle, or start writing last is displayed. The small kanji group for the radical with each attribute is as follows. 1 word (148): So-called poetry ... ・ Sanyo Satoshi 2 words (25): Word ... Utaki ... Yasutoba 3 words (25): Police ... People ... : Monkey ... Peaceful 2 Kemonohen (9): 荻 嶽 轴 蕕 藐 葸 逖 霾 3 Kemonohen (1): Yemen

As described above, the radical or the like of the desired Chinese character is input by handwriting, the list of the basic elements is used as a dictionary to recognize the characters of the radical or the like, and the Kanji is searched for using each basic element. However, it is possible to select the detailed mode in which the stroke order of the elements is specified from the basic elements in the table, or the general mode in which the elements are simply specified without specifying the stroke order of the elements. In the detailed mode, there is also a method of instructing a plurality of stroke orders to obtain the logical product thereof, and in the general mode, instructing a plurality of basic elements to obtain the logical product thereof and displaying the conditioned Kanji on the CRT. .

By using the general mode and the logical product,
The case of searching for will be described. Handwriting input is "word"
In that case, after recognition, the Kanji (1 word + 2 words + 3 words) having all “words” are displayed on the CRT 20. If the input of handwritten characters is "Kemonohen", then (1 Kemonohen + 2 Kemonohen + 3 Kemonohen) will be instructed.
AND of "Kemonohen" is obtained and as a result
You can get "Take" and "Prison". It is easy to choose "chaud" from these. Of course, the handwritten characters may be input in any order. This operation is C in FIG.
The PU 13 reads out the Chinese character group of "Koto" and the Chinese character group of "Kemonohen" from the POM 17A. A package equivalent to that shown in FIG. 9 is examined for these groups of middle Chinese characters, and only the Chinese characters having two common attributes (ignoring the stroke order) and two Chinese characters having the same radical are displayed on the CRT 20.

Next, description will be given of a case where the "custom-character" is searched for by logical product using the detailed mode. Enter the radical "word" or "kemonohen" of the desired kanji by handwriting, use the list of basic elements as a dictionary to recognize the radical characters, and use this for each basic element to search for kanji. To search for Kanji using this, "word" was recognized by handwriting input. At this time, by giving the command "start writing first", "1 word" can be selected, and on the CRT "1 word" candidates are displayed in. Next, "Kemonohen" was recognized by handwriting input. At this time, it is possible to select "2 Kemonohen" by giving the command "Start writing on the way", and the logic of "1 word" and "2 Kemonohen" The product is sought, and as a result, only """is displayed on the CRT 20. It can be seen that the selectivity is superior to that in the general mode. Of course, the order in which the basic elements are selected from the list may be selected in any manner.

If the instruction of the pen 18 is "2" in step S22, the process proceeds to the flowchart of FIG. 14 using the handwriting position small Chinese character group of FIGS. This is because the search target is the 1800 pairs of the 1st to 13th position small Chinese character groups 1-11 to 1-23 ... n-11 to n-23 in FIGS.
... i-11 to i-23 ... 600-11 to 600
It is expanded to -23. Since the "word" is recognized by the handwriting input in step S70, the process proceeds to step S71. Since the position information is given after the radical "word" is confirmed, in step S71, if the "lower left" (not shown) of the key 26 is pressed with the pen 18 to specify "word hit" to be written in the lower left,
In step S72, the radical "Koto" shows all the kanji in the lower left at the handwriting position. The eighth position small kanji group 1-18 in FIGS.
Is displayed.

In step S73, the eighth position small Chinese character group 1-1
When the target “Take” is obtained and confirmed from the candidate kanji list of No. 8, the search processing ends, but since “Take” does not exist in the eighth position small Kanji character group 1-18 and the character can be confirmed, step S74. Proceed to. Here, since the pen 18 is used to press another position information "middle lower" (not shown) of the key 26 to specify "no word" written in the lower middle, the radical "word" is changed in step S75. FIG. 7, which shows all the kanji in the lower middle in the handwriting position,
Eighth 12th position small Chinese character group 1-22 is displayed. In step S76, the target “Take” is obtained and confirmed in the candidate kanji list of the 12th position small Kanji group 1-22, and thus the search process ends, and the characters below “Take” are input by handwriting in step S90.
Proceed to. For the sake of explanation, if the Chinese character cannot be determined in step S76, the process proceeds to step S77. Here, since the "center" of the key 26 is pressed with the pen 18 to specify "Kona" to be written in the middle of the character, in step S78 all the kanji whose radical "Koto" is in the middle of the handwriting position. The 13th position small Kanji character group 1-23 of FIGS. When the target “Take” is obtained and confirmed in the candidate kanji list, the search process ends.

If the target character is not found and the kanji cannot be determined in step S79, other position information and the biased "word"
All the kanji whose radical "word" is biased at the handwriting position is displayed. Similarly, it is possible to promptly determine the character by freely specifying a search from among 13 types of position information. Further, the "word speech" (lower left) of step S71, the "word speech" (lower middle) of step S74, and the "language speech" (middle) of step S77 may be designated first.

In step S70, if the input of the handwritten character is not "word" but "mountain", the process proceeds to step S80.
Similarly, for “mountain”, “yamahishi”, “yamanashi”, and “yamana” are advanced, and the same operation as in steps 71 to S79 is followed to continue the search. Furthermore, in step S80, if the input of the handwritten character is not "mountain" but "large", the process proceeds to step S81. Similarly, with respect to "large", the position information which can be thought of as "large pear", "large pear", and "large pear" is input, and the search is continued by following the same operation as in steps S71 to S79. Further, the “word” in step S70, the “mountain” in step S80, and the “large” in step S81 may be drawn first.

Next, an example of searching for the "custom character" by AND using the detailed mode will be described. Enter the radical "word" or "kemonohen" of the desired kanji by handwriting, use the list of basic elements as a dictionary to recognize the radical characters, and use this for each basic element to search for kanji. To search for Kanji using this, "Word" was selected from the list, so at this time, give the command "Write on the left" and "Word" and then the AND key 24. Press, and then continue handwriting input to select "Kemonohen" from the list,
At this time, the command "write in" is given to input "Kemonohenna" from the panel 12. The logical product of "Kotohi" and "Kemonohenna" is obtained, and as a result, only "裑" is displayed on the CRT 20. It can be seen that the selectivity is superior to that in the general mode. Of course, the order of handwriting input of the basic elements may be selected so that the characters on the radical are recognized.

Selected by the above method of logical product,
There may be multiple kanji displayed on the screen, but the kanji displayed are quite narrowed down, so even if there are multiple kanji, the number is small, and the desired kanji can be specified from among them. It is extremely easy to do. In specifying the character, in addition to the selection by the pen 18, the kanji can be selected by inputting the code attached to each kanji from an ordinary keyboard. In the above description, "Kemonohen"
It is expressed in Hiragana, but in practice, these are registered and used as basic elements of Kanji.

[0036]

As described above, according to the present invention, according to the present invention, it is possible to write a kanji sentence in a notebook by simply writing an abbreviated character without typing in the keyboard. Can be easily input with a word processor. This is very useful when inputting the contents of an interview while face-to-face with a customer, as in shorthand. Since the radical of the basic element is a simple pattern, character recognition is instantly and surely performed. Therefore, even with the current character recognition device, high-accuracy kanji recognition can be realized.

[Brief description of the drawings]

FIG. 1 is a block diagram of an information device in which the present invention is implemented.

FIG. 2 is a plan view of a main part of an information device in which the present invention is implemented.

FIG. 3 is a diagram showing a list of basic elements according to an embodiment of the present invention.

FIG. 4 is a flow chart showing the operation of the method for creating one storage device of the present invention.

FIG. 5 is a diagram showing a configuration of a Chinese character group including a stroke order according to an embodiment of the present invention.

FIG. 6 is a flowchart showing the operation of the other storage device creating method of the present invention.

FIG. 7 is a diagram showing a partial configuration of a Chinese character group including a handwriting position according to another embodiment of the present invention.

FIG. 8 is a diagram showing an overall structure of a Chinese character group including a handwriting position according to another embodiment of the present invention.

FIG. 9 is a diagram showing a memory configuration of one of the kanji of the present invention.

FIG. 10 is a diagram showing another storage configuration of Chinese characters of the present invention.

FIG. 11 is a flowchart showing the operation of the handwritten character recognition method of the present invention.

FIG. 12 is a flowchart showing a schematic mode of a search stage of the present invention.

FIG. 13 is a flowchart showing one detailed mode of the search stage of the present invention.

FIG. 14 is a flowchart showing one detailed mode of the search stage of the present invention.

[Explanation of symbols]

 1, n, i Chinese character group 1-1, n-1, i-1 First stroke order small Chinese character group 1-2, n-2, i-2 Second stroke order small Chinese character group 1-3, n-3, i-3 3rd stroke order small kanji group 1-11, n-11, i-11 1st position small kanji group 1-12, n-12, i-12 2nd position small kanji group 1-23, n-23 , I-23 13th position small kanji group K large kanji group 12 input panel 13 CPU 14 buffer memory 15 RAM 16 17 22 ROM 17 CPU 18 pen 19 magnetic storage device 20 CRT 24 logical key 25 numeric key 26 position key 29 execute key

Claims (4)

(57) [Claims] 1. A kanji character is decomposed to extract the basic elements constituting the kanji character, these basic elements are organized as individual two-dimensional image data to create a list for retrieval, and this list is stored in a storage device. The Chinese characters having the registered basic elements as constituent elements are collected to form a Chinese character group, the Chinese character groups are created for each of the basic elements, and a plurality of the Chinese character groups are registered in the storage device. For each of the above-mentioned middle Chinese character groups, the basic element is the first stroke order small Chinese character group by collecting the basic elements in the stroke order of the Chinese character, and the second stroke order small Chinese character group is collected by collecting the intermediate Chinese characters. Then, the last kanji to be written is collected to form a third stroke order small kanji group, these stroke order small kanji groups are created for each of the middle kanji groups, and a plurality of stroke order small kanji groups are registered in the storage device and drawn. Handwrite only the basic elements of the desired Kanji, The handwritten basic elements are converted into two-dimensional image data, basic elements similar to the converted image data are selected from the registered basic elements, and the Chinese character group, the first stroke order small Chinese character group, and the second stroke order small. From the kanji group and the third stroke-order small kanji group, one middle kanji group or stroke-order small kanji group corresponding to the selected basic element is selected, and the designated middle kanji group or stroke-order small kanji group is displayed. Displayed on
A method for inputting handwritten characters in which a desired kanji is selected from the middle kanji group or the stroke order small kanji group and the handwritten kanji is input to the information processing device. 2. A kanji character is decomposed to extract basic elements constituting the kanji character, these basic elements are organized as individual two-dimensional image data to create a list for retrieval, and this list is stored in a storage device. Registered in, to collect the Chinese characters having the registered basic elements as constituent elements into a group of Chinese characters, create this group of Chinese characters for each of the basic elements, and register a plurality of groups of Chinese characters in the storage device, For each of the middle Chinese character groups, the basic element is the first stroke order small Chinese character group in the stroke order of the Chinese character, and the first stroke order small Chinese character group is collected.
No. 1 stroke order small Kanji character group, and the last Kanji character to start writing is collected as a third stroke order small Kanji character group. These stroke order small Kanji character groups are created for each of the middle Chinese character groups, and a plurality of stroke order small Chinese character groups are stored in the storage device. , Hand-write multiple basic elements for the kanji you want to draw, convert each of these hand-written basic elements into two-dimensional image data, and select each basic element similar to these converted image data from the registered basic elements. Then, from the group of Chinese characters, the first group of small stroke characters of the first stroke, the second group of small Chinese characters of the stroke order and the third group of small stroke characters of Chinese stroke, a group of middle Chinese characters or a small group of stroke orders corresponding to each of the selected basic elements is selected. By selecting a plurality of kanji groups, the kanji that are commonly included in the selected middle kanji group or stroke order small kanji group are displayed as candidates on the display device, and the desired kanji is selected from these, and handwriting is performed. Kanji Handwriting input method for inputting to the processing unit. 3. An input device for handwriting input of basic elements constituting a kanji by decomposing the kanji, and a first storage device in which the basic elements are organized to create a list for search and the list is registered. And a second storage device in which a plurality of Chinese character groups having the basic elements as constituent elements are collected to form a Chinese character group, and the Chinese character group is created for each of the basic elements, and the Chinese character group is registered. About the group, in terms of the stroke order of Kanji,
The first kanji to be written is collected into the first stroke order small kanji group, the middle kanji to be written into the second stroke order small kanji group, and the last kanji to be written is collected into the third stroke order small kanji group. , A third storage device in which a plurality of stroke order small kanji groups created for each of the middle kanji groups are registered, and a handwriting image in which the basic elements of the kanji to be drawn are manually input from the input device A recognition device for confirming the basic element by comparing with the table, a display device for displaying the confirmed basic element, a group of Chinese characters in the second storage device and a first group of stroke order small Chinese characters in the third storage device. Instructing means for instructing any one of the second small stroke order Chinese character group and the third small stroke order Chinese character group, and the second character instructed the Chinese character having the decided basic element.
A handwritten character input device comprising a central processing unit which is extracted from a storage device or a third storage device as target candidate Kanji characters and is displayed on the display device. 4. An input device for handwriting input of basic elements constituting Kanji by decomposing Chinese characters, and a first storage device in which the basic elements are organized to create a list for search and the list is registered. And a second storage device in which a plurality of Chinese characters including the basic elements are collected to form a Chinese character group, and the Chinese character group is created for each of the basic elements, and the Chinese character group is registered. About the group, in terms of the stroke order of Kanji,
The first kanji to be written is collected into the first stroke order small kanji group, the middle kanji to be written into the second stroke order small kanji group, and the last kanji to be written is collected into the third stroke order small kanji group. , A third storage device in which a plurality of stroke order small Kanji groups created for each of the above middle Chinese character groups are registered, and a handwriting image in which a plurality of basic elements for the Kanji to be drawn are input from the input device A recognition device for confirming each basic element by collating with the list, a display device for displaying each of the confirmed basic elements, a Chinese character group for the second storage device, and a first Chinese character for the third storage device. Instructing means for instructing a plurality of stroke order small Kanji groups, second stroke order small Kanji groups, and third stroke order small Kanji groups, and a medium Chinese character group or a stroke order small Kanji group corresponding to each of the determined basic elements. By instructing a plurality of instructing means, Thus, a handwritten character input device comprising a central processing unit for extracting, as a candidate, a kanji that is commonly included in a plurality of instructed middle kanji groups and a stroke-order small kanji group, and displaying it on the display device.