JP2016157314A - Character recognition device, character recognition program, character recognition system, and character recognition method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a novel character recognition technique.SOLUTION: A character recognition device displays a plurality of determination areas on a display screen, and a user inputs a track by referring to the displayed determination areas. When the input track input by the user is obtained, the character recognition device calculates a determination value for each determination area by determining whether the input track hits the determination areas. The character recognition device determines a character corresponding to the input track on the basis of a position of each determination area, the order in which the input track hits the determination areas, and the determination value for each determination area, and outputs the determined character.SELECTED DRAWING: Figure 21

Description

  The present invention relates to a character recognition device that recognizes handwritten characters.

  Conventionally, character recognition devices that recognize handwritten characters have been disclosed. For example, Patent Document 1 discloses a technique for recognizing characters by dividing characters written in succession and cutting out each character.

JP 2014-130510 A

  However, in the conventional technique, the character is recognized by determining whether or not the shape of the handwritten character matches a predetermined shape.

  One of the objects of the present invention is to provide a novel character recognition technique that has not been found in the past.

  In an example of the present invention, the following configuration is adopted in order to solve the above-described problem.

  An example of the present invention is a character recognition device that recognizes a character from an input made by a user, an acquisition unit that acquires an input locus input by the user, an area setting unit that sets a plurality of determination areas, Based on the positional relationship between the input locus and the determination region, a calculation unit that calculates a determination value of the determination region for each determination region, and a determination value for each determination region calculated by the calculation unit, Character determination means for determining characters.

  According to the above, the character indicated by the input trajectory can be determined by calculating the determination value of each determination region based on the input trajectory input by the user.

  In another configuration, the calculation unit may calculate the determination value based on whether or not the input locus hits the determination region.

  According to the above, it is possible to calculate a determination value based on whether or not the input locus hits the determination region, and to determine a character based on the calculated determination value.

  Further, in another configuration, the calculating means determines, as the determination value of the determination region, a first determination value based on whether the input locus hits the determination region, and the input locus defines a boundary of the determination region. You may calculate the 2nd determination value based on whether it passed.

  According to the above, it is possible to calculate a determination value based on whether or not the input trajectory hits the determination area and whether or not the input trajectory has passed the boundary of the determination area, so that more various characters can be determined. it can.

  Further, in another configuration, position information is set in each of the determination areas, and the character determination unit is configured to detect characters based on the determination values of the plurality of determination areas and the position information of each determination area. You may judge.

  Based on the above, it is possible to determine a character based on the position information of the determination area and the determination value of the determination area.

  In another configuration, the character determination unit is configured to determine a character based on the determination values of the plurality of determination areas, the position information of each determination area, and the order in which the input locus hits the determination areas. May be determined.

  According to the above, it is possible to determine a character based on the order in which the input locus hits the determination area. Thereby, a character can be correctly recognized even in an input locus input in an arbitrary direction. For example, even if a shorthand character is input in an arbitrary direction, the character can be correctly determined.

  In another configuration, when the character determination unit determines a character based on the determination values of the plurality of determination regions, the character determination unit sets the last determination region among the plurality of determination regions corresponding to the character as the next character. The next character may be determined as the first determination region.

  Based on the above, it is possible to correctly determine the connected characters.

  In another configuration, the image processing apparatus further includes display control means for displaying the plurality of determination areas on a display means, and the acquisition means is an input input on an input surface including the plurality of determination areas displayed on the display means. A trajectory may be acquired.

  Based on the above, the user can input while viewing the determination area displayed on the display means.

  In another configuration, the plurality of determination areas may include a first determination area and a second determination area smaller than the first determination area.

  In another configuration, the input trajectory may be a shorthand character.

  In another form, a program that causes a computer of a character recognition apparatus to function as each of the above means may be used. Moreover, in another form, the character recognition system comprised by 1 or several apparatus (element) provided with said each means may be sufficient. The system may be realized by connecting a plurality of devices via a network.

  According to an example of the present invention, a character can be recognized by calculating determination values of a plurality of determination areas.

The block diagram which shows the function structure of the character recognition apparatus 1 of this embodiment. The figure which shows an example of V type shorthand character The figure which shows the determination area | region for discriminating the input locus | trajectory input Diagram for explaining definition of judgment value for judgment area A diagram showing the pattern of each shorthand character on the line Figure showing the pattern of each shorthand character Showing the pattern of each shorthand character in a line Illustration to explain the characters written in concatenation A figure showing an example of muffled characters Figure showing an example of semi-voiced characters The figure which shows an example of the sound A figure showing an example of a long sound Figure showing an example of stuttering Diagram showing an example of sound repellent A figure showing an example of a “ku” sound A figure showing an example of a “tsu” sound A figure to explain the “best point” in V-style shorthand The figure which shows the definition of the "wo point" in the character recognition apparatus 1 of this embodiment The figure which shows an example of the shorthand character containing the worst point shown in FIG. The figure which shows an example of the determination pattern table which matched the pattern of the determination value with respect to a determination area | region, and a character The flowchart which shows the detail of the character recognition process which concerns on this embodiment Diagram showing an example of the input trajectory The figure which shows the outline | summary of the character determination process when the input locus | trajectory shown in FIG. 22 is acquired. Figure for explaining character independence and connection determination The figure which shows the portable electronic device 100 as one Example of the character recognition apparatus 1. FIG.

(Configuration of character recognition device)
Hereinafter, a character recognition device according to an embodiment of the present invention will be described. FIG. 1 is a block diagram showing a functional configuration of the character recognition device 1 of the present embodiment.

  As shown in FIG. 1, the character recognition device 1 includes an input unit 2, an input locus acquisition unit 3, a determination area setting unit 4, a determination value calculation unit 5, a character determination unit 6, a database 7, and a display. Means 8. The character recognition device 1 includes a storage device such as a CPU, a RAM, a hard disk, and a nonvolatile memory, and the CPU functions as each unit shown in FIG. 1 when the CPU executes a predetermined character recognition program.

  The input unit 2 is an input device for a user to input characters (shorthand characters), and for example, a position input device such as a touch panel or a mouse is used. The input trajectory acquisition unit 3 acquires the trajectory input by the input unit 2. The determination area setting unit 4 sets a plurality of determination areas, which will be described later, and displays the plurality of determination areas on the display unit 8. The determination value calculation unit 5 calculates a determination value for each determination region based on the input trajectory acquired by the input trajectory acquisition unit 3 and the determination region set by the determination region setting unit 4. The character determination means 6 determines the character indicated by the input locus based on the determination value calculated by the determination value calculation means 5 and the pattern stored in the database 7 and outputs the determined character to the display means 8. To do.

(Outline of character recognition processing)
The character recognition device 1 of the present embodiment converts a shorthand character input by a stenographer into a normal character (such as hiragana or katakana). The character recognition device 1 according to the present embodiment converts a shorthand character input by an arbitrary shorthand method (an arbitrary shorthand method such as V-type, House of Representatives, House of Councilors, Nakane-type, Waseda-type, Tachain-type, etc.) into a normal character. Although it may convert, the case where the shorthand character input based on the V-type shorthand method is converted is demonstrated below.

  FIG. 2 is a diagram illustrating an example of a V-type shorthand character. As shown in FIG. 2, shorthand characters are associated with 50-character characters according to the length and angle of the line. For example, “A” is a straight line that goes to the upper right and is inclined by 30 degrees from the horizontal, and “I” is a line that goes to the lower left and is inclined by 60 degrees from the horizontal (30 degrees from the vertical). Represented by a line. Also, “ka” is represented by a counterclockwise curve toward the upper right. Further, the “ta” line is represented by a line that is half the length of the “a” line.

  FIG. 3 is a diagram illustrating a determination area for determining the input trajectory. In the present embodiment, for example, a stenographer inputs a stenographic character on a screen on which a plurality of determination areas as shown in FIG. 3 are arranged and displayed, and the inputted stenographic character is converted into a normal character. .

  As shown in FIG. 3, the plurality of determination areas include a main determination area M, a sub determination area A disposed on the right side of the main determination area M, and a sub determination area B disposed below the main determination area M. Including. The sub determination area A and the sub determination area B are circles having the same radius and are smaller than the main determination area M. The shape of each determination area is not limited to a circle, and may be a polygon such as a quadrangle. Moreover, the size of all the determination areas may be the same. As shown in FIG. 3, a set of three determination areas, a main determination area M, a sub determination area A on the right side thereof, and a sub determination area B on the lower side, are set in a vertical and horizontal direction. Provided at equal intervals.

  Specifically, the sub determination area A11 is provided on the right side of the main determination area M11, the main determination area M12 is provided on the right side thereof, and the sub determination area A12 is provided on the right side thereof. A sub determination area B11 is provided below the main determination area M11, a main determination area M21 is provided below the main determination area M11, and a sub determination area B21 is provided below the main determination area M11. Thus, the main determination area M and its sub determination areas A and B are arranged at equal intervals in the vertical direction (row) and the horizontal direction (column), respectively.

  Hereinafter, the main determination area located in the i-th row and the j-th column from the main determination area M11 at the upper left corner is denoted as “Mi, j” (i and j are respectively positive integers). Further, the sub determination area on the right side of the main determination area Mi, j is expressed as “Ai, j”, and the sub determination area below the main determination area Mi, j is expressed as “Bi, j”.

  The character recognition device 1 according to the present embodiment determines the determination value for each determination region depending on how the line indicated by the input shorthand character passes through each of the determination regions shown in FIG. And a normal character corresponding to the shorthand character is determined based on the determination value of each determination region.

  FIG. 4 is a diagram for explaining the definition of the determination value for the determination region. In the present embodiment, a “present” value and a “communication” value are defined as determination values. The “to” value increases by “1” when the line or the point enters (hits) the determination region. More specifically, when at least a part of the input trajectory input by the user exists in the determination area, the input trajectory hits the determination area. The “communication” value increases by “1” when the line passes the boundary of the determination region (when the line enters the determination region from outside the determination region, or when the line goes out of the determination region from the determination region). .

  As shown in FIG. 4A, when the line passes from the outside of the determination area to the outside of the determination area, the line hits the determination area only once according to the above definition. The value is “1”. Further, according to the above definition, the line increases by “1” when entering the determination area from outside the determination area, and further increases by “1” when moving from the determination area to outside the determination area. The value is “2”. In the following, these two values will be referred to as a judgment value “this one 2”.

  Here, in V-type shorthand, “cut line” representing muddy sound or semi-turbid sound, “crossing” or “parallel” representing prompt sound, “sink” representing long sound (or sound), and “n” “Circle (small circle)”, “large circle” representing “ku, ki”, and “oval” representing “tsu” are defined. In the character recognition apparatus 1 of this embodiment, these are expressed as follows.

  For example, as shown in FIG. 4B, when two lines pass through one determination area, it is determined as “cut line”. Specifically, in the “cut line”, two lines pass from the outside of the determination area to the outside of the determination area. The determination value at this time is “this 2 4” according to the above definition.

  In addition, as shown in FIG. 4C, when two lines enter one determination area and do not exit the determination area, it is determined as “intersection”. Specifically, in the “intersection”, two lines enter from outside (or inside) the determination area to inside (or outside) the determination area, intersect within the determination area, and do not exit the determination area. The determination value at this time is “this two-way two”. Further, as shown in FIG. 4D, when two lines enter one determination area and do not exit the determination area, it is determined as “parallel”. Specifically, in “parallel”, two lines enter from the outside (or inside) of the determination area to the inside (or outside) of the determination area, and do not cross within the determination area and do not exit from the determination area. The determination value at this time is “this two-way two”.

  As shown in FIGS. 4C and 4D, the two determination values are both “two 2”, and the above definition cannot distinguish the two. However, in the V-type shorthand, “crossing” and “parallel” mean the same prompting sound, so it is not necessary to distinguish between them.

  Further, as shown in FIG. 4E, when one line enters one determination area and does not exit the determination area, it is determined as “sink”. Specifically, in the “sink”, since one line enters the determination area from outside the determination area and does not come out of the determination area, the determination value is “this one 1”.

  In addition, as shown in FIG. 4F, when one line enters one determination area and exits from the determination area and enters the determination area again, it is determined as a “small circle”. The determination value at this time is “this two times three” according to the above definition because the line enters and exits after entering the determination region. Also, when a single line enters one determination area, exits from the determination area, enters the determination area again, and goes out of the determination area, it is determined as a “small circle”. The determination value at this time is “this 2 4” according to the above definition. Note that “small circle” and “cut line” have the same judgment value, but “cut line” is not used for the first or last judgment area of a character, and “small circle” is always a character. Therefore, even if these determination values are the same, “small circle” and “cut line” can be distinguished.

  Further, as shown in FIG. 4G, the “large circle” can take the same determination value as the “small circle” for the last determination region, but the sub determination adjacent to the last main determination region. A circle is drawn including the area. Such a pattern is recognized based on the determination value of the last main determination area and the determination values of the sub determination areas adjacent to it (up / down / left / right sub determination areas).

  Further, as shown in FIG. 4H, the “oval” can take the same determination value as the above “small circle” for the last determination area, but the diagonal sub-direction of the last main determination area. A circle is drawn including the judgment area. Such a pattern is recognized based on the determination value of the last main determination region and the determination value of the sub determination region in the diagonal direction.

  Next, the passage of each determination area and the determination value at that time will be described by taking a part of a specific shorthand character as an example. FIG. 5 is a diagram showing a pattern of each shorthand character in the line. FIG. 6 is a diagram showing patterns of shorthand characters in each line. FIG. 7 is a diagram showing patterns of each shorthand character in the line. In this embodiment, it is assumed that characters are always started to be written from the main determination area M, but characters may be started from the sub determination area. Further, in the present embodiment, it is assumed that characters are written in the main determination area M, but characters may be written in the sub determination area.

  As shown in FIG. 5, for example, the line representing “A” passes through the main determination area M42, passes through the upper right sub-determination area B33, passes through the right sub-determination area B34, and passes through the upper right. Passes through the main determination area M35. At this time, the determination value of each determination region is “this one 2”.

  For example, the line representing “I” passes through the main determination area M24, passes through the lower left sub determination area A33 of M24, passes through the lower sub determination area A43, and passes through the lower left main determination area M53. pass.

  The line representing “U” passes through the main determination area M32, passes through the lower right sub-determination area B33 of M32, passes through the right sub-determination area B34, and lower right in the main determination area. Pass through M45.

  For example, the line representing “e” passes through the main determination area M24, passes through the sub determination area B24 below M24, passes through the main determination area M34 below, and passes through the sub determination area B34 below. Passes through the main determination area M44 below.

  For example, the line representing “o” passes through the main determination region M32, passes through the right sub determination region A32, passes through the right main determination region M33, and passes through the right sub determination region A33. Then, it passes through the right main determination area M34.

  As shown in FIG. 6, the line representing “ka” passes through, for example, the main determination area M32, the sub determination area B32, the main determination area M33, and the main determination area M24. The line representing “ki” passes through, for example, the main determination area M24, the sub determination area A33, the sub determination area A43, and the main determination area M44. Further, the line representing “ku” passes through, for example, the main determination area M32, the sub determination area A42, the sub determination area B43, and the main determination area M54. Further, the line representing “ke” passes through, for example, the main determination area M24, the sub determination area A33, the sub determination area A43, and the main determination area M54. Further, the line representing “ko” passes through, for example, the main determination area M32, the sub determination area B33, the sub determination area B34, and the main determination area M35.

  Further, as shown in FIG. 7, the line representing “ta” passes through, for example, the main determination area M43, the sub determination area B34, and the main determination area M35. Further, the line representing “chi” passes through, for example, the main determination area M34, the sub determination area A43, and the main determination area M53. Further, the line representing “tsu” passes through, for example, the main determination region M33, the sub determination region B34, and the main determination region M45. The line representing “te” passes through, for example, the main determination region M34, the sub determination region B34, and the main determination region M44. Further, the line representing “to” passes through, for example, the main determination region M33, the sub determination region A33, and the main determination region M34.

  Note that the positions of the determination areas at the beginning of writing shown in FIGS. 5 to 7 are examples, and each line may be started from any area. The character recognition device 1 can determine which character the input line represents based on the relative position from the determination area (or the final determination area) at the beginning of writing. For example, if “A” shown in FIG. 5 is started to be written from the main determination area Mi, j in the i-th row and j-th column, the lines indicating “A” are Mi, j, Bi-1, j + 1, It will pass through Bi-1, j + 2 and Mi-1, j + 3.

  In this way, the hiragana of each Japanese syllabary may be a line with a shape and size in accordance with the V-type shorthand (a line with a shape or size slightly modified in accordance with the V-type shorthand, It may be a line having a shape or size that completely matches the expression shorthand), and it is defined which judgment area passes for each character. In a normal Japanese character, since the line passes through each determination area, the determination value is “this one 2”. In the character recognition device 1 according to the present embodiment, a pattern of determination values in each determination area and a character are stored in association with each other. The character recognition device 1 discriminates a normal character corresponding to the inputted shorthand character by calculating which judgment area is passed in which order with respect to the input line and calculating a judgment value of each judgment area. be able to.

  Next, the case where the shorthand characters defined as described above are connected and written will be described. FIG. 8 is a diagram for explaining the characters written in a linked manner.

  FIG. 8 shows a case where three characters “CHI”, “KE”, and “I” are connected and written. When “chikei” is written concatenated, a line indicating “chi” passes through the main determination area M14, the sub determination area A23, and the main determination area M33, and the line indicates “ke”. It is drawn continuously as a part, passes through the sub determination areas A42 and A52 next to M33, and passes through the main determination area M63. Further, the line is continuously drawn as a line indicating “yes”, passes through the sub determination areas A72 and A82 next to M63, and finally passes through the main determination area M92. In this way, when characters are connected and drawn (with a single stroke), the last determination region of a character and the first determination region of the next character to be drawn are common.

  Next, muddy sound and semi-muddy sound will be described. FIG. 9 is a diagram illustrating an example of muffled characters. FIG. 10 is a diagram illustrating an example of a semi-voice character.

  As shown in FIG. 9, the voiced character “G” has a “cut line” with respect to the determination region (sub determination region B 34) immediately before the last determination region of “ko” defined in FIG. Added. At this time, the determination value of the sub-determination area B34 is “this two 4”. That is, the determination value of the sub-determination area B34 of “ko” is “this one-way two”, whereas the determination value of the sub-determination area B34 of “go” is “two-time four”. The determination value in the determination area is the same for “ko” and “go”.

  As shown in FIG. 10A, the line indicating “ha” passes through, for example, the main determination area M43, the sub determination area A43, the main determination area M44, the main determination area M35, and the main determination area M26. On the other hand, in the semi-voice character “Pa”, a “cut line” is added to the second determination region (sub-determination region A43) of “ha”. That is, the determination value of the sub-determination area A43 of “ha” is “this 1 2”, whereas the determination value of the sub-determination area A43 of “pa” is “this 2 4”. The determination value of the determination area is the same for “ha” and “pa”.

  Next, the expression of the prompt sound and the long sound will be described. FIG. 11 is a diagram illustrating an example of the prompt sound. FIG. 12 is a diagram illustrating an example of a long sound.

  In FIG. 11, the letters “Kak” including the prompt sound are described. As described above, if “ka” is started to be written from the main determination area M41, the line passes through the sub determination area B41 and the main determination area M42 after the start of writing, and finally passes through the main determination area M33. To do. On the other hand, as shown in FIG. 11, in the case of “Kaku” including a prompt sound, a line indicating “K” is drawn so as to stop within the last main determination region M33. Then, a line indicating “ta” is started from within the main determination region M33. At this time, the determination value of the main determination region M33 is “this two 2” as described above (“parallel” shown in FIG. 4D).

  Further, in FIG. 12, the characters “kuuko” including a long sound are described. As described above, if “ku” is started to be written from the main determination area M31, the line passes through the sub determination area A41 and the sub determination area B42 after the start of writing, and finally passes through the main determination area M53. To do. On the other hand, as shown in FIG. 12, in the case of the long sound “kukou”, the line indicating “ku” is drawn so as to stop in the last main determination area M53, and the line indicating “ko” Writing is started from outside the main determination area M53. At this time, the determination value of the main determination region M53 is “this 1 through 1” as described above (“sink” shown in FIG. 4E). When the end of the character is “sink”, “i” is added after the character in the “a” column. For example, when the last determination area of “A” is “sink”, it becomes “Ai”.

  Next, the expression of stuttering will be described. FIG. 13 is a diagram illustrating an example of stuttering.

  In the V shorthand notation, a roaring sound such as “Kya” is expressed by attaching a small key to “ka”. “Kyu” is expressed by attaching a small key to “ku”, and “Kyo” is expressed by attaching a small key to “ko”.

  In the present embodiment, “Kya” is expressed by attaching a key to “ka” as shown in FIG. Specifically, assuming that “ka” is written from, for example, the main determination area M42, a line indicating “ky” is started to be written from the main determination area M32, and then passes through the main determination area M42. After that, it is the same as “ka”. Also, as shown in FIG. 13B, if “ku” is started to be written from, for example, the main determination area M32, a line indicating “kyu” is started to be written from the main determination area M22, and subsequently the main determination area is displayed. Passes through M32, and thereafter passes through a determination region similar to “ku”. Further, as shown in FIG. 13C, if “ko” is started to be written from, for example, the main determination area M32, a line indicating “cho” is started to be written from the main determination area M22, and subsequently, the main determination area Passes through M32, and thereafter passes through a determination region similar to “ko”. That is, in this embodiment, the stuttering is expressed by adding a key that passes through the upper main determination area M (a main determination area one level higher than the sub determination area one level higher) at the beginning of writing. The

  Next, the expression of sound repellent (“n”) will be described. FIG. 14 is a diagram illustrating an example of sound repellent. In the example shown in FIG. 14, “Kantou” is written. Specifically, when a line indicating “ka” is drawn from the main determination area M51, for example, it passes through the sub determination area B51, the main determination area M52, and the main determination area M43. At this time, the end of the line indicating “ka” passes through the main determination area M43. However, in the case of “kan”, a small circle (see FIG. 4F) is displayed in the last main determination area M43. Add). In this case, the determination value of the main determination region M43 is “this two 4” as shown in FIG. Further, a line indicating “to” is drawn continuously from the small circle, and is flowed at M44 which is the last of “to” to express “to-”. One line drawn in this way is recognized as “Kantou”.

  FIG. 15 is a diagram illustrating an example of a “ku” sound. In the case of expressing the “ku” sound, the above-described great circle (see FIG. 4G) is drawn. For example, as shown in FIG. 15, “ka” is started to be written from the main determination area M21, and a great circle including the last main determination area M13 of “ka” and the sub determination area A12 adjacent to the left is drawn. Then, “ku” is added to the written character “ka” and recognized as “ku”. In FIG. 15, “NI” is further written, and a small circle is drawn in the last main determination area M34 of “NI”, so the line shown in FIG. Be recognized. When a great circle is drawn in this way, the determination value of the last main determination area of the character is “this two-way three” or “this two-way four”, and either the upper, lower, left, or right of the last main determination area Such a sub-determination area has a “this” value (“this one 2”). When a character having such a pattern is recognized, “ku” is added after the character. Note that whether the upper, lower, left, or right sub-determination area is drawn includes a great circle depending on the character. When a great circle is written at the end of the characters in the “e” column, “ki” is added after the characters in the “e” column. For example, when a great circle is added to the last main determination area M of “e”, it is recognized as “eki”.

  FIG. 16 is a diagram illustrating an example of the “tsu” sound. When the “tsu” sound is expressed, the above-described ellipse (see FIG. 4H) is drawn. For example, as shown in FIG. 16, “ka” is started to be written from the main determination area M31, and an ellipse including the last main determination area M23 of “ka” and the sub determination area A13 on the upper right is drawn. Is added to the written character “ka” and recognized as “and”. In FIG. 16, “to” is further written, and the last main determination area M24 of “to” is “sink”. Further, the sub-determination area A23 of the line indicating “to” is “cut line”. For this reason, the line shown in FIG. 16 is recognized as “Katsudo”. When an ellipse is drawn in this way, the determination value of the last main determination area of the character is “this two-way three” or “this two-way four”, and the diagonal sub-determination of the last main determination area The region will have a “this” value (“one per two”). When a character having such a pattern is recognized, “t” is added after the character. It should be noted that whether the ellipse is drawn including which sub-determination area in the oblique direction differs depending on the character.

  Next, the “bad point” process will be described. In this case, the “bad point” in the shorthand is to regard a point as a predetermined character string when a point is attached to a predetermined position from the end of a line indicating a certain character. FIG. 17 is a diagram for explaining the “bad point” in V-type shorthand.

  In the V-type shorthand, as shown in (1) of FIG. 17, the point at the position P1 diagonally right above the end of the line means the character “that”, and at the right position P2 with respect to the end of the line. The point means the character “So”, the point at the position P3 below the end of the line means the character “That”, and the point at the position P4 diagonally lower left to the end of the line means the character “thing” "Means. Further, in the V-type shorthand, as shown in FIG. 17 (2), the point at the position P1 diagonally to the upper right with respect to the end of the line means the character “done”, and the right position with respect to the end of the line. The point at P2 means the character “ori”, the point at the position P3 below the end of the line means the character “is”, and the point at the position P4 diagonally lower left relative to the end of the line is the character “ Means " Further, as shown in (2) of FIG. 17, a point at a position P5 diagonally lower right with respect to the end of the line means the character “do”. Here, as shown in (1) and (2) of FIG. 17, the point at the same position P1 means a different character. However, in shorthand, the pattern of (1) or (2) is usually used. Judging from the context.

  In the character recognition apparatus 1 according to the present embodiment, the “best point” shown in FIG. 17 is defined. FIG. 18 is a diagram showing the definition of “null point” in the character recognition device 1 of the present embodiment. (1) in FIG. 18 shows a point corresponding to the pattern of (1) in FIG. 17, and (2) in FIG. 18 shows a point corresponding to the pattern (2) in FIG.

  As shown in (1) of FIG. 18, when the determination value in the main determination area M14 at the upper right of the main determination area M23 indicating the end of a certain character is “this one 1”, in FIG. It is regarded as a point at position P1 in the pattern, and means “that”. Further, when the determination value in the main determination area M24 on the right of the main determination area M23 indicating the end of a certain character is “this 1 through 1”, it means “no”, and the main determination area below the main determination area M23. When the determination value of the area M33 is “this one 1”, it means “reason”. In addition, when the determination value of the main determination area M32 or the sub determination area A32 diagonally lower left of the main determination area M23 indicating the end of a certain character is “one 1”, it means “thing”.

  In addition, when the determination value of each determination area is “this one 0” (that is, when a point is drawn in the determination area and does not go outside the determination area), the pattern of (2) in FIG. It is determined as a corresponding point. Specifically, as shown in (2) of FIG. 18, when the determination value in the main determination area M14 at the upper right of the main determination area M23 indicating the end of a certain character is “this one 0”, FIG. In the pattern (2), it is regarded as a point at the position P1, and it means “do”. Further, when the determination value in the main determination area M24 on the right of the main determination area M23 indicating the end of a certain character is “this one 0”, it means “origin” and the main determination below the main determination area M23. When the determination value of the area M33 is “this one 0”, it means “is”. Further, when the determination value in the main determination area M32 or the sub determination area A32 diagonally to the left of the main determination area M23 indicating the end of a certain character is “this one 0”, it means “do”. In addition, when the determination value in the main determination area M34 diagonally to the lower right of the main determination area M23 indicating the end of a certain character is “this one 0”, it means “Yes”.

  FIG. 19 is a diagram showing an example of a shorthand character including the worst point shown in FIG. (1) in FIG. 19 shows an example of a shorthand character including the worst point shown in (1) of FIG. 18, and (2) in FIG. 19 shows the shorthand character including the worst point shown in (2) of FIG. An example is shown.

  In (1) of FIG. 19, the lines in the main determination areas M11 to M14 are recognized as normal characters “Y”, and the lines M14 to M43 drawn in succession are recognized as “I”. Since the determination value of the main determination area M34 that is diagonally to the upper right of the main determination area M43, which is the end of this “I”, is “1”, this M34 is recognized as “that” according to the definition of the above point. The Further, a line is drawn from M34 to M26, and the lines M34 to M26 are recognized as the normal character “NA”. And since the judgment value of M27 on the right of M26 which is the end of the line indicating “NA” is “this 1 through 1”, it is recognized as “NO” according to the definition of the above point. Accordingly, the series of trajectories shown in (1) of FIG. 19 is recognized as “because it is good”.

  Further, in (2) of FIG. 19, the lines of the main determination areas M11 to M23 are recognized as ordinary characters “su”, and the lines of M23 to M33 drawn subsequently mean “te”. Since a cut line is added to B23 between M23 and M33 (because the determination value of B23 is “this 2 through 4”), it is recognized as “de”. A line is continuously drawn from M33 to M54, and the line from M33 to M54 is recognized as a normal character “ni”. Since the determination area of M63 diagonally to the left of M54 is “this one 0”, the point drawn in M63 is recognized as “do” according to the definition of the above point. Similarly, since the determination value of M64 on the right side of M63 is “this one 0”, the point drawn in M64 is recognized as “origin” according to the definition of the above point. A line drawn from M64 to M57 is recognized as a normal character “MA”, and a line drawn from M57 to M69 is recognized as a normal character “SU”. Accordingly, the series of trajectories shown in (2) of FIG. 19 is recognized as “already done”.

  Note that the determination value indicating the worst point shown in (1) of FIG. 18 is the same as the determination value at the time of “sink” described above. For this reason, when a line is drawn with the intention of “sink”, it may be erroneously recognized as a critical point. In order to prevent such misrecognition, when the determination value of the last determination area of the pattern (line segment) is “this one 1” (sink), the last determination of the pattern is recognized after the pattern is recognized. The judgment value of the area is changed to “this one 2” (normal). In addition, the judgment value of the last judgment area of the pattern (line corresponding to the character) such as “parallel”, “intersection”, “small circle”, “large circle”, “ellipse”, etc. Similarly, when the value is other than “2”, after the pattern is recognized, the determination value of the last determination area of the pattern is changed to “this one 2” (normal).

  Next, a determination table for character determination stored in advance in the database will be described. FIG. 20 is a diagram illustrating an example of a determination pattern table in which a determination value pattern for a determination region is associated with a character. In FIG. 20, only the above-mentioned “A row”, “Ka row”, “Ta row”, “Ha”, and a part of the roar (“Kya”, “Kyu”, “Kyo”). As shown, patterns for the other characters according to the definition of each character are stored in the table. Specifically, in the determination pattern table, all characters defined by the determination region having the determination value of 1 and 2 are stored, and in addition to the normal hiragana characters, stuttering characters are stored. Has been.

  As shown in FIG. 20, for example, in the pattern 1, the line passes the main determination area “Mi, j” first, passes the sub determination area “Ai, j” second, and the main determination third. The pattern passes through the region “Mi, j + 1”, and this pattern 1 means “to”. In pattern 2, the line is first in the main determination area “Mi, j”, second in the sub determination area “Ai, j”, third in the main determination area “Mi, j + 1”, and fourth in the sub determination. The region “Ai, j + 1” and the fifth main determination region “Mi, j + 2” are passed, and this pattern 2 means “o”. The character recognizing device 1 detects in which order and in which order the trajectory input by the user has hit (passed), and the pattern of the hit (pass) matches any of the patterns shown in FIG. And the character corresponding to the matching pattern is output.

(Details of character recognition processing)
Next, details of character recognition processing in the character recognition device 1 of the present embodiment will be described. FIG. 21 is a flowchart showing details of character recognition processing according to the present embodiment. The process shown in FIG. 21 is performed by the CPU of the character recognition device 1 executing a character recognition program. In the process of the flowchart shown in FIG. 21, for convenience of explanation, it is assumed that characters are determined using the table shown in FIG. 20, and there are no patterns other than the 19 patterns listed in this table. explain.

  As shown in FIG. 21, first, the CPU displays each determination area on the display device (step S1). Specifically, the CPU sets the positions and sizes of the plurality of determination areas, and displays the plurality of determination areas side by side at predetermined positions on the screen of the display device. Next, the CPU accepts an input of a locus indicating a shorthand character (step S2). Specifically, the CPU sequentially acquires positions on the screen (vertical and horizontal coordinate values) input using the input unit. When a trajectory is input by the user, the CPU next acquires input trajectory determination information (step S3). Here, the CPU calculates a determination value (a “current” value and a “communication” value) of each determination region with respect to the acquired input trajectory. Specifically, the CPU determines whether or not the input trajectory hits the determination area and whether or not the determination area has passed based on the coordinate value indicating the input position and the position and size of each determination area. By doing so, the judgment value of each judgment area is calculated.

  Next, the CPU performs character determination processing (step S4). Here, the character determination processing in step S4 will be described with reference to FIGS. FIG. 22 is a diagram illustrating an example of an input trajectory. FIG. 23 is a diagram showing an outline of the character determination process when the input locus shown in FIG. 22 is acquired.

  For example, when the input trajectory as shown in FIG. 22 is acquired, the determination values of the ten determination areas are calculated as “this one 2” in step S3. As shown in FIG. 23A, the calculated determination information of each determination area is stored in the array in the order in which the input trajectory hits. In FIG. 23A, M14, A23, M33, A42, M53, A52, M63, A72, A82, and M92 are arranged in this order. Specifically, the type of determination area (main determination area M, sub-determination areas A and B) and information indicating its position (information indicating what row and what column), and the determination value of the determination area Is stored in each element of the array. In FIG. 23, only the information indicating the type and position of the determination area is shown.

  Here, the CPU retrieves the information stored in the first and second of the array, and whether the pattern that becomes the determination area M14 first and the determination area A32 second is in the determination pattern shown in FIG. Determine whether or not. Since there are patterns 12 to 15 in which M23 is followed by A23 (in terms of relative position, “Mi, j” followed by “Ai + 1, j−1”), the CPU uses these patterns 12-15. Select as a candidate for judgment character. Further, the CPU determines whether or not the patterns arranged in the order of M14, A23, and M33 including the third M33 are among the patterns 12 to 15 selected as candidates. Here, since the pattern 15 is the only matching pattern, the CPU recognizes the pattern arranged in the order of M14, A23, and M33 as the pattern 15. That is, the line passing through the three determination areas M14, A23, and M33 is determined as the character “Chi”. At this stage, since it is only determined whether or not the “present” value is 1 or more, for example, when the determination value of A23 is “present two 4”, it is similarly determined as “chi”. However, after determining each character indicated by the input trajectory, the CPU refers to the determination value of each determination area corresponding to each character to determine, for example, whether or not a cut line has been added. Or “ぢ”. Whether it is an intersection, parallel, flow, or small circle is similarly determined after each character is recognized.

  Subsequently, the CPU starts searching for the second character pattern using the last determination region M33 of the first character as the first determination region. Specifically, as shown in FIG. 23B, the third to the tenth are set as the remaining input trajectories, and a matching pattern is searched from the plurality of patterns shown in FIG. 20 in the same procedure as described above. Specifically, the first and second characters (third and fourth in the array) of the second character are M33 and A42 ("Mi, j" and "Ai + 1, j-1" in relative positions), respectively. Therefore, the CPU selects patterns 12 to 15 as candidates for determination characters by selecting matching patterns with reference to the table shown in FIG. Next, as shown in FIG. 23C, the CPU refers to the fifth (third third character) of the array, and whether there is a pattern arranged in the order of M33, A42, and M53 in the table shown in FIG. Determine whether or not. Such a pattern does not exist in the table. For this reason, the CPU ignores the fifth M53 and then refers to the sixth as the third of the second character. Here, the reason why the fifth M53 is ignored is to consider a writing mistake by the user, and in this embodiment, one writing error is allowed every time the determination area is advanced by one. That is, if the input line passes through the incorrect determination region (second) after passing through the correct determination region (first), and further passes through the correct determination region (third), the incorrect line Character recognition is performed by ignoring the determination region (second) and using the two correct determination regions (first and third). On the other hand, if the input line passes through the wrong judgment area (second and third) after passing through the correct judgment area (first), it is processed as unrecognizable. . It should be noted that the extent to which a user's writing mistake is allowed may be adjusted as appropriate.

  That is, in the present embodiment, with reference to a determination area (N + 1) next to a certain determination area (Nth), if the pattern does not match a predetermined pattern, the determination area (N + 1) is ignored, Further, with reference to the next determination area (N + 2), it is determined whether or not the pattern including the Nth and N + 2th matches a predetermined pattern. Since the sixth is M52, as shown in FIG. 23C, the CPU determines whether the patterns arranged in the order of M33, A42, and A52 are in the patterns 12 to 15 shown in FIG. . Here, since the patterns 12 to 14 remain as candidates, the CPU continues the search. The CPU further refers to the next seventh to determine whether the pattern arranged in the order of M33, A42, A52, and M63 matches any of the patterns 12 to 14 remaining as candidates. Here, since the pattern 14 is the only matching pattern, the CPU recognizes the pattern arranged in the order of M33, A42, A52, and M63 as the pattern 14. That is, the line passing through the three determination areas M33, A42, A52, and M63 is determined as the character “ke” (FIG. 23D).

  The same applies to the remaining input trajectories. That is, as shown in FIG. 23E, the CPU refers to the seventh to tenth arrays, recognizes the pattern arranged in the order of M63, A72, A82, and M92 as the pattern 12, and determines the character “I”. judge.

  As described above, in the character determination process in step S4, an array in which areas having a value of 1 or more are sequentially stored is searched from the beginning, and determination is performed for each character. Note that if the current value is an area of 1 or more and the determination value of the area is not any of those shown in FIG. 4, the determination value of that area is set to “this 1 2” and the character determination process of step S4 is performed. Also good.

  Subsequent to step S4, the CPU outputs the character determined in the character determination process (step S5). And CPU determines whether the process of FIG. 21 is complete | finished (step S6). For example, if termination is instructed by the user, the CPU determines YES in step S6 and terminates the processing shown in FIG. On the other hand, when it determines with NO by step S6, CPU performs the process of step S2 again.

  The output character may differ depending on whether the input locus is an independent character or a plurality of connected characters. Further, the output character may be different depending on whether the connected character string is the first or last character string. For this reason, the CPU may determine whether it is an independent character or a connected character in the character determination processing in step S4.

  FIG. 24 is a diagram for explaining determination of character independence and connection. FIG. 24A shows independent characters, and FIG. 24B shows concatenated characters.

  As shown in FIG. 24A, when characters are drawn independently, there are around the first determination region M42 indicating the characters (eight main determination regions M and twelve sub determination regions). Except for the next determination area B33, there is no area where the “present” value is 1 or more. In addition, there is no area where the “present” value is 1 or more around the last determination area M35 of the character, except for the determination area B34 before the determination area M35. Therefore, the CPU can determine that such a pattern is independent.

  On the other hand, as shown in FIG. 24B, when characters are connected, the first character in the connected character string is surrounded by the first determination region M24 (see FIG. 24B). There is no “To” value in the determination area M24 (except for the next determination area A33), and there is a “To” value around the last determination area M43 of the first character (except A33 before M43). (A52 has a “current” value). In addition, for the last character in the concatenated character string, there is a “current” value around the first determination region M73 (excluding the next A82) among the plurality of determination regions (A62 includes “ There is no “present” value around the last determination area M10,2 (except A92 before M10,2) in the last character. That is, for the first character in the concatenated character string, there is no “this” value around the first determination region M (19 regions excluding the next region) in the first character, and There is a “present” value around the last determination area M in the first character (19 areas excluding the previous area). For the last character in the concatenated character string, there is a “to” value around the first determination area M (19 areas excluding the next area) in the last character, and at the end. There is no “current” value around the last determination area M (19 areas excluding the previous area) in the character of. Therefore, the CPU can determine that such a pattern is a concatenation from the beginning to the end. In this way, whether a character is connected or independent is determined in a plurality of areas excluding the determination area used by the character itself, among the determination areas around the first and last determination areas of the character. Judgment can be made based on whether or not there is a value. That is, when a certain character is connected to another character, the first or last determination area of a certain character is used as the last or first determination area of another character. For this reason, if a certain character has a value around the first determination area of the character (a plurality of determination areas excluding the determination areas used by the character itself among the 20 determination areas), at least the character It can be determined that another character is connected before. Similarly, if a certain character has a value around the last determination area of the character (a plurality of determination areas excluding the determination areas used by the character itself among the 20 determination areas), at least the character It can be determined that another character is connected after. Depending on the definition of the character, it may not be possible to determine whether or not the character is concatenated only by determining whether or not there is a corresponding value in the determination region around the first and last determination regions of the character. However, it is possible to determine whether or not such a character is a concatenated character by making a determination specific to that character.

  By searching for such a feature, when an input trajectory is acquired, the CPU can determine whether it is an independent character or a concatenated character, and recognizes the character in units of concatenated character strings. Can do. The character recognition device 1 of the present embodiment can determine whether or not characters are connected as described above, and determine the start and end of character connection, and can change the character to be output according to the result. . For example, in normal shorthand, there are abbreviations, and the letter “A” is sometimes used as “A”. In normal shorthand, the letter “A” and the abbreviation “Ari” are the same, and these are judged from the context. The character recognition device 1 of the present embodiment determines the beginning and end of character concatenation as described above. For example, if the concatenated character string ends with “a”, the character “a” is the last character. It may be output as “Yes”. On the other hand, an independent “a” or “a” other than the last in a concatenated character string may be output as “a” as it is. Thus, when each character is recognized, the character to be output may be different depending on the position of the independent character, the connected character, or the connected character string.

  Next, a specific embodiment of the character recognition device 1 will be described. For example, as an example of the character recognition device 1, a tablet-type portable electronic device 100 including a touch panel may be used. FIG. 25 is a diagram illustrating a portable electronic device 100 as an example of the character recognition device 1. The portable electronic device 100 includes a display 108 and a touch panel 102 provided on the display. Further, the display 108 is provided with a character display area for displaying the determined character. The display 108 may be provided with an “eraser button”, “kanji conversion button”, “clear button”, and the like.

  As the display 108, for example, an arbitrary display device such as a liquid crystal display device or an organic EL display device may be used. The touch panel 102 is a position input device that detects the touched position when the screen of the display 108 is touched with a touch pen (or a user's finger). May be used.

  As shown in FIG. 25, when the user moves the pen while touching the display 108 using the pen, the portable electronic device 100 sequentially detects the touch position at predetermined time intervals, and points are detected at the detected touch position. By drawing, an image showing the input locus is displayed on the display 108 in real time. The inputted input locus is converted into characters at a predetermined timing and displayed in the character display area. For example, character recognition processing may be performed when conversion is instructed by the user, and characters may be displayed in the character display area, or character recognition processing may be performed automatically after a stroke stroke is input. In other words, a character string corresponding to a one-stroke trajectory may be displayed.

  The “eraser button” is for erasing the determination value of the determination area when the user accidentally touches the determination area. In the example described with reference to FIG. 23, it is assumed that the character recognition device 1 automatically detects a user's erroneous input. However, by using the “eraser button”, the user accidentally touches himself / herself. The determination area can be deleted. For example, the user can clear the determination value of the selected area by selecting a determination area where the determination value is to be deleted by performing a predetermined operation and touching an “eraser button”.

  The “kanji conversion button” is used to convert an input character into kanji and output it. The user touches the Kanji conversion button after inputting the shorthand characters, and the characters after Kanji conversion are displayed in the character display area. It should be noted that after being displayed in hiragana, it may be converted into kanji when the kanji conversion button is touched. In addition to these Kanji conversion buttons, hiragana buttons, katakana buttons, alphabetic buttons, numbers / symbol buttons, abbreviation buttons (buttons for outputting abbreviations), etc. are provided. It may be changed. In addition, a button for inputting muddy sound or semi-muddy sound may be provided, and when a locus is inputted when the button is pressed, muddy sound or semi-muddy sound may be added to the character indicated by the inputted locus. .

  The “clear button” is for erasing the input locus. When this clear button is touched, all of the input locus displayed on the display 109 is deleted, and characters displayed in the character display area are also deleted. Note that a button for deleting the input locus and a button for deleting the character output in the character display area may be provided separately.

  In addition to such a function, there may be a function of saving an image indicating an input locus. For example, when the “clear button” is touched, the input trajectory may be saved as an image, or when the “save button” is provided and the save button is touched, the image is stored in a non-volatile memory or a hard disk. You may make it preserve | save in a memory | storage means. As described above, the image indicating the input locus is saved, so that the user can later confirm what character has been input.

  As another specific example of the character recognition device 1, a digital pen that can store an input locus as data may be used. Specifically, the digital pen is a pen capable of writing characters, figures, and the like on paper, and data indicating the drawn characters and figures can be stored in a nonvolatile memory inside the digital pen. For example, the digital pen can grasp and store the writing position by reading a predetermined pattern printed on the dedicated paper with a camera provided on the digital pen. A shorthand character input using such a digital pen can be converted into a normal character by the method described above. For example, an information processing apparatus connected to the digital pen wirelessly or by wire may be configured to execute the above processing. Specifically, data indicating the input locus read with the digital pen is transmitted to the information processing apparatus, and the information processing apparatus performs the above-described processing based on the transmitted data. Or the data memorize | stored in the digital pen may be taken into an information processing apparatus later, and the input shorthand character may be converted into a normal character by performing the process mentioned above based on the said data. Further, for example, the digital pen itself may execute the above-described processing, read an input shorthand character, and convert it into a normal character.

  In addition, the present invention is not limited to the case where an input locus input on dedicated paper is read using a screen provided with a touch panel or a digital pen and converted into characters, and an input locus input by another method is read. It may be converted to characters. For example, a plurality of determination areas may be displayed on the screen, and an input locus may be input while instructing on the screen using a position input device such as a mouse or a joystick. A pointer may be displayed at a position on the screen instructed using a position input device such as a mouse, and the user may input a shorthand character by moving the pointer.

  In addition, when an input locus is input on the screen on which such a determination area is displayed, the determination area may be displayed in a different display mode according to the determination value of the determination area. For example, an area where the input locus is not hit is displayed in a normal color, a determination area of “This 1 through 1” is displayed in blue, and a determination area of “This 2 through 4” is displayed in green. May be.

  In addition, an input trajectory input to a normal paper using a normal pen may be read and converted to a normal character. In this case, the above-described determination area is printed on the paper, and the user can input shorthand characters while looking at the printed determination area. The input locus input on the paper is read by a scanner, a camera, or the like, and converted into characters by performing the character recognition process described above. Specifically, the character recognition device recognizes a determination area printed on the read paper by pattern matching or the like and recognizes an input locus input by the user. Then, the character recognition device can determine whether the input track has hit (passed) each determination region based on the positional relationship between the determination region and the input track. In this case, the character recognition device needs to recognize the direction of the input trajectory (the order in which the input trajectory hits each determination area), but the shorthand characters are basically from left to right and from top to bottom. Using the premise that the input is input, it is possible to estimate from where and in which direction the read input trajectory is written. That is, assuming that the read line is drawn from left to right and from top to bottom, the direction in which the line is drawn is estimated by reading the paper in the same direction as the actual drawn direction. be able to. After estimating the direction in which the line is drawn, it is possible to determine in which order the line hits and in what order by the method described above. In addition, if a predetermined pattern for recognizing the vertical and horizontal directions is printed on the paper, the character recognition device can recognize the direction of the paper by recognizing the predetermined pattern, and the direction of the input locus Can be estimated.

  Various customizations may be possible depending on the user. For example, a user-specific pattern may be registered in a determination pattern table used for character determination. In this case, a pattern uniquely registered by the user (pattern indicating characters) may be preferentially output. In addition, the size, shape, arrangement, number of rows, and number of columns of the determination area may be set by the user, the ratio of the size of the main and sub determination areas, the distance between the main and sub determination areas, etc. It may be set. Further, the color for each determination value in the determination region, the color or thickness of the line for drawing the input locus, and the like may be set by the user.

  Moreover, the process shown in the flowchart of the said embodiment is only an example, and if it is a process which implement | achieves the character recognition method mentioned above, a part or all of the said process may be changed, and a part of said process May not be executed, or other processing may be added. Further, any order of processing may be used. For example, all the determination areas may be examined in order to search for a determination area having a “present” value. When a determination area with a “to” value is found, it is possible to search for an input trajectory pattern by determining whether or not there is a determination area with a “to” value around that determination area. Good.

  Also, for example, in the above embodiment, a plurality of determination areas are displayed on the screen. However, the plurality of determination areas are not necessarily displayed, and a plurality of determination areas may be defined internally in the character recognition device. Good.

  Further, a plurality of lines (ranges) may be provided in the input area for inputting shorthand characters, and shorthand characters may be input in units of lines. For example, an area in which a plurality of main determination areas M are arranged in the vertical direction is defined as the first line, and an area in which a plurality of main determination areas M are similarly arranged in the vertical direction is provided below the first line. Also good.

  In the above embodiment, a shorthand character is inputted and the inputted shorthand character is recognized. It may be recognized by this method. For example, an alphabetic script is written in the same manner as a shorthand character, but each character written in a script can be recognized by using the method described above. As described above, the handwritten character can be recognized by applying the present invention to the characters written concatenated (characters written with a single stroke).

  In another embodiment, the above-described processing may be performed by a system configured by a plurality (or a single) device. For example, the above processing may be executed by a plurality of devices connected to the network sharing the processing. The above processing may be performed by a distributed network system such as so-called cloud computing. Further, part or all of the above processing may be executed by a plurality of processors, and part or all of the above processing may be performed by a dedicated circuit.

DESCRIPTION OF SYMBOLS 1 Character recognition apparatus 2 Input means 8 Display means M Main determination area A Sub determination area B Sub determination area

Claims (12)

A character recognition device that recognizes characters from input made by a user,
Obtaining means for obtaining an input trajectory input by the user;
Area setting means for setting a plurality of determination areas;
Calculation means for calculating a determination value of the determination area for each determination area based on a positional relationship between the input locus and the determination area;
A character recognition device comprising: character determination means for determining a character based on a determination value for each determination area calculated by the calculation means.   The character recognition device according to claim 1, wherein the calculation unit calculates the determination value based on whether or not the input trajectory hits the determination area.   The calculation unit is configured to determine, as a determination value of the determination region, a first determination value based on whether or not the input locus has hit the determination region, and whether or not the input locus has passed a boundary of the determination region. The character recognition device according to claim 1 or 2, wherein the second determination value is calculated. Position information is set in each of the determination areas,
The character recognition device according to claim 1, wherein the character determination unit determines a character based on determination values of the plurality of determination areas and position information of the determination areas.   The character determination means determines a character based on determination values of the plurality of determination areas, position information of the determination areas, and an order in which the input trajectory hits the determination areas. The character recognition device described in 1.   When the character determination unit determines a character based on the determination values of the plurality of determination areas, the last determination area of the plurality of determination areas corresponding to the character is set as the determination area at the beginning of the next character. The character recognition device according to claim 4 or 5, wherein the next character is determined. A display control means for displaying the plurality of determination areas on a display means;
The character recognition device according to claim 1, wherein the acquisition unit acquires an input locus input on an input surface including a plurality of determination areas displayed on the display unit.   The character recognition device according to claim 1, wherein the plurality of determination areas include a first determination area and a second determination area smaller than the first determination area.   The character recognition device according to claim 1, wherein the input locus is a shorthand character. A character recognition program executed by a computer of a character recognition device for recognizing characters from input made by a user, the computer comprising:
Obtaining means for obtaining an input trajectory input by the user;
Area setting means for setting a plurality of determination areas;
Calculation means for calculating a determination value of the determination area for each determination area based on a positional relationship between the input locus and the determination area;
A character recognition program that functions as a character determination unit that determines a character based on a determination value for each determination region calculated by the calculation unit. A character recognition system for recognizing characters from input made by a user,
Obtaining means for obtaining an input trajectory input by the user;
Area setting means for setting a plurality of determination areas;
Calculation means for calculating a determination value of the determination area for each determination area based on a positional relationship between the input locus and the determination area;
A character recognition system comprising: character determination means for determining a character based on a determination value for each determination area calculated by the calculation means. A character recognition method for recognizing characters from input made by a user,
An acquisition means for acquiring an input trajectory input by a user;
A region setting means for setting a plurality of determination regions;
Calculating a determination value of the determination area for each determination area based on a positional relationship between the input locus and the determination area;
A character recognition method comprising: a character determination unit including a step of determining a character based on a determination value for each determination area calculated by the calculation unit.

Images