JP2006072730A - Character recognition device, character recognition method and character recognition program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a convenient handwritten character recognition device that can recognize a plurality of handwritten characters. <P>SOLUTION: The character recognition device comprises a handwritten character acquisition means for acquiring handwritten character information on a plurality of handwritten characters, written on a coordinate input device by an operator; a character segmentation means for dividing the plurality of handwritten characters into character units, according to the handwritten character information acquired by the handwritten character acquisition means; a character recognition means for recognizing each character segmented out by the character segmentation means; a misrecognized character selection means for selecting a misrecognized character misrecognized by the character recognition means by the operator's direction; and a character correction means for correcting the misrecognized character selected by the misrecognized character selection means to another character. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a handwritten character recognition device, a character recognition method, and a character recognition program for recognizing a handwritten character drawn by an operator on a coordinate input device.

  As a method for recognizing handwritten characters, a method is known in which a fixed frame is provided and a user is allowed to enter characters in the frame in units of one character (see, for example, “Patent Document 1”). This method is advantageous in that it does not require a process of cutting out one character unit and there is no erroneous recognition due to a cutting error. On the other hand, the user needs to perform input and recognition operations one character at a time, which is disadvantageous in that the burden on the operator is large and convenience is low.

  There is also known a method of freely writing a plurality of characters in a handwriting area without providing an input frame (see, for example, “Patent Document 2”). In this method, the operator can continuously input a plurality of handwritten characters in the handwriting area. Therefore, convenience is improved. However, in this method, each character must be automatically extracted from a plurality of handwritten characters. For this reason, a single character cut-out process is performed.

JP 2001-014103 A JP 2000-076380 A

  In the method of eliminating the input frame and freely writing a plurality of characters, it is necessary to perform the cut-out processing for each character as described above. In this cut-out process, there is a problem that recognition of one character may be wrong. On the other hand, there is a desire to further improve the convenience for the operator.

  The present invention has been made in view of the above, and provides a handwritten character recognition device, a character recognition method, and a character recognition program that can perform character recognition of a plurality of handwritten characters and are convenient. Objective.

  In order to solve the above-described problems and achieve the object, the invention according to claim 1 is a handwritten character recognition device for recognizing a handwritten character drawn by an operator on a coordinate input device, wherein the operator Handwritten character acquisition means for acquiring handwritten character information about a plurality of handwritten characters drawn on the input device, and character cutting for dividing the plurality of handwritten characters into character units based on the handwritten character information acquired by the handwritten character acquisition means A character recognition unit that performs character recognition on each character cut out by the character cutout unit, and a misrecognized character obtained by erroneous recognition by the character recognition unit according to an instruction from the operator. An error recognition character selection means for selecting, and a character correction means for correcting the error recognition character selected by the error recognition character selection means to another character are provided.

  The invention according to claim 2 is the character recognition device according to claim 1, wherein the misrecognized character selection means is the plurality of handwritten characters acquired by the handwritten character acquisition means in accordance with an instruction from the operator. A plurality of the erroneously recognized characters are selected from characters, and the character correcting means corrects each of the plurality of erroneously recognized characters selected by the erroneously recognized character selecting means to other characters.

  The invention according to claim 3 is the character recognition device according to claim 2, wherein the handwritten character display means provided in the coordinate input device displays the handwritten character drawn by the operator, The misrecognized character selection means, when the operator designates a predetermined handwritten character among the plurality of handwritten characters displayed on the handwritten character display means on the coordinate input device, the character designated by the operator The erroneously recognized character is selected.

  Moreover, the invention concerning Claim 4 is the character recognition apparatus as described in any one of Claim 1 to 3, Comprising: The said misrecognition character selection means is a some handwritten character displayed on the said handwritten character display means When the operator draws a surrounding line surrounding a predetermined handwritten character among the characters in the coordinate input device, the character surrounded by the surrounding line is selected as the erroneously recognized character.

  Moreover, the invention concerning Claim 5 is the character recognition apparatus as described in any one of Claim 1 to 4, Comprising: The said misrecognition character selection means is a some handwritten character displayed on the said handwritten character display means When the operator draws an underline for a predetermined character in the coordinate input device, the character corresponding to the underline is selected as the misrecognized character.

  The invention according to claim 6 is the character recognition device according to any one of claims 1 to 5, wherein the handwritten character display means is the plurality of handwritten characters acquired by the handwritten character acquisition means. Are displayed in units of lines, and the erroneously recognized character selection means allows the operator to use the coordinate input device in a line area provided corresponding to a predetermined handwritten character displayed on the handwritten character display means. When designated, a plurality of characters corresponding to the row area are selected as the erroneously recognized characters.

  The invention according to claim 7 is the character recognition device according to any one of claims 1 to 6, wherein the character recognition unit obtained by the character recognition unit and the character recognition unit by the character recognition unit Recognized character display means for displaying a recognized character candidate that has been determined to be possible, and a predetermined recognition from among the recognized character candidates displayed by the recognized character display means according to an instruction from the operator A corrected character selecting means for selecting a character candidate, wherein the character correcting means changes the erroneously recognized character selected by the erroneously recognized character selecting means to the recognized character candidate selected by the corrected character selecting means. It is characterized by doing.

  The invention according to claim 8 is the character recognition device according to claim 7, wherein the recognized character display means includes the recognized character for each of the plurality of handwritten characters and the each of the plurality of handwritten characters. A recognition character candidate, and the correction character selection means selects the recognition character candidate for the erroneous recognition character selected by the erroneous recognition character selection means according to an instruction from the operator, and the character correction means The misrecognized character selected by the misrecognized character selecting unit is changed to the recognized character candidate selected by the corrected character selecting unit.

  The invention according to claim 9 is the character recognition device according to claim 7 or 8, wherein the recognized character display means includes the error among the plurality of handwritten characters acquired by the handwritten character acquisition means. Recognized character for the erroneously recognized character selected by the recognized character selecting means and the recognized character candidate are displayed, and the corrected character selecting means selects the predetermined recognized character candidate in accordance with an instruction from the operator, The character correcting means changes the erroneously recognized character selected by the erroneously recognized character selecting means to the recognized character candidate selected by the corrected character selecting means.

  The invention according to claim 10 is the character recognition device according to any one of claims 1 to 9, wherein a corrected handwritten character for the erroneously recognized character selected by the erroneously recognized character selection unit is acquired. A corrected character acquiring means for performing character recognition on the corrected handwritten character acquired by the corrected character acquiring means, and the character correcting means is selected by the erroneously recognized character selecting means. A misrecognized character is corrected to a recognized character obtained for the corrected handwritten character.

  The invention according to claim 11 is the character recognition device according to claim 10, wherein the erroneously recognized character selection unit is configured to detect a plurality of erroneously recognized characters from the plurality of handwritten characters according to an instruction from the operator. The corrected character acquisition means acquires corrected handwritten characters corresponding to each of the plurality of erroneously recognized characters selected by the erroneously recognized character selection means, and the character recognition means acquires the corrected character acquisition means. Character recognition is performed on the plurality of corrected handwritten characters acquired, and the character correcting unit corrects the erroneously recognized character selected by the erroneously recognized character selecting unit to a recognized character obtained with respect to the corrected handwritten character. It is characterized by that.

  The invention according to claim 12 is a handwritten character recognition method for recognizing a handwritten character drawn by an operator on a coordinate input device, wherein the operator relates to a plurality of handwritten characters drawn on the coordinate input device. Based on the handwritten character information acquired in the handwritten character acquiring step, the character extracting step of dividing the plurality of handwritten characters into character units, and the character extracting step A character recognition step for recognizing each character, a misrecognized character selection step for selecting a misrecognized character obtained by misrecognition in the character recognition step according to an instruction from the operator, A character correcting step of correcting the misrecognized character selected in the character selecting step to another character. That.

  The invention according to claim 13 is a character recognition program, characterized by causing a computer to execute the character recognition method according to claim 12.

  According to the first aspect of the present invention, the handwritten character acquisition unit acquires handwritten character information regarding a plurality of handwritten characters drawn by the operator on the coordinate input device, and the character cutout unit acquires the handwritten character acquisition unit. Based on the handwritten character information, a plurality of handwritten characters are divided into character units, the character recognition means performs character recognition for each character cut out by the character cutout means, and the erroneously recognized character selection means performs character recognition. The erroneously recognized character obtained by the means being erroneously recognized is selected by the operator's instruction, and the character correcting means corrects the erroneously recognized character selected by the erroneously recognized character selecting means to another character. In a character recognition device that recognizes all at once, correction of erroneously recognized characters can be performed not only in units of one character but also in units of a plurality of characters, so that the convenience of the character recognition device can be improved. Achieve the cormorant effect.

  According to the invention of claim 2, the erroneously recognized character selecting means selects a plurality of the erroneously recognized characters from the plurality of handwritten characters acquired by the handwritten character acquiring means according to an instruction from the operator, and the character correcting means. Since each of the plurality of misrecognized characters selected by the misrecognized character selection means is corrected to other characters, the operator can freely select the misrecognized characters, and the convenience of the character recognition device is improved. There is an effect.

  According to the invention of claim 3, the handwritten character display means provided in the coordinate input device displays the handwritten character drawn by the operator, and the erroneously recognized character selection means is displayed on the handwritten character display means. When the operator specifies a predetermined handwritten character among the plurality of handwritten characters on the coordinate input device, the operator selects the character specified by the operator as the misrecognized character. A desired handwritten character can be selected while viewing the displayed handwritten character, and the convenience of the character recognition device can be improved.

  According to the invention of claim 4, the misrecognized character selection means is configured so that the operator inputs a surrounding line surrounding a predetermined handwritten character among the plurality of handwritten characters displayed on the handwritten character display means in the coordinate input device. When drawn, the character surrounded by the surrounding line is selected as a misrecognized character. As described above, by using the gesture processing, there is an effect that the operation by the operator can be simplified.

  According to the invention of claim 5, the erroneously recognized character selection unit is configured such that the operator draws an underline for a predetermined character among the plurality of handwritten characters displayed on the handwritten character display unit on the coordinate input device. Then, the character corresponding to the underline is selected as a misrecognized character. As described above, by using the gesture processing, there is an effect that the operation by the operator can be simplified.

  According to the invention of claim 6, the character display means displays a plurality of handwritten characters acquired by the handwritten character acquisition means in line units, and the erroneously recognized character selection means is displayed on the handwritten character display means. When an operator designates in a line area provided corresponding to a predetermined handwritten character on the coordinate input device, a plurality of characters corresponding to the line area are selected as misrecognized characters. As described above, by using the gesture processing, there is an effect that the operation by the operator can be simplified.

  Further, according to the invention of claim 7, the recognized character display means includes the recognized character obtained by the character recognizing means and the recognized character candidate determined by the character recognizing means to be a recognized character. The corrected character selection means selects a predetermined recognized character candidate from the recognized character candidates displayed by the recognized character display means according to an instruction from the operator, and the character correction means is selected by the erroneously recognized character selection means. Since the erroneously recognized character is changed to the recognized character candidate selected by the corrected character selection means, there is an effect that the operation by the operator can be simplified.

  According to the invention of claim 8, the recognized character display means displays the recognized characters for each of the plurality of handwritten characters and the recognized character candidates for each of the plurality of handwritten characters, and the corrected character selection means performs the erroneous recognition. A recognized character candidate for the erroneously recognized character selected by the character selecting means is selected by an instruction from the operator, and the character correcting means selects the erroneously recognized character selected by the erroneously recognized character selecting means by the corrected character selecting means. Therefore, there is an effect that the operation by the operator can be simplified.

  According to the invention of claim 9, the recognized character display means recognizes the recognized character for the erroneously recognized character selected by the erroneously recognized character selection means among the plurality of handwritten characters acquired by the handwritten character acquisition means, The character candidate is displayed, the corrected character selecting means selects a predetermined recognized character candidate according to an instruction from the operator, and the character correcting means selects the erroneously recognized character selected by the erroneously recognized character selecting means. Since it changes to the recognition character candidate selected by the means, there exists an effect that operation by an operator can be simplified.

  According to the invention of claim 10, the corrected character acquisition means acquires the corrected handwritten character for the erroneously recognized character selected by the erroneously recognized character selection means, and the character recognition means is acquired by the corrected character acquisition means. The corrected handwritten character is recognized, and the character correcting means corrects the erroneously recognized character selected by the erroneously recognized character selecting means to the recognized character obtained with respect to the corrected handwritten character. There is an effect that can be made.

  According to the invention of claim 11, the misrecognized character selection means selects a plurality of misrecognized characters from the plurality of handwritten characters according to an instruction from the operator, and the corrected character acquisition means selects the misrecognized character selection unit. The corrected handwritten character corresponding to each of the plurality of erroneously recognized characters selected by the means is acquired, the character recognizing means recognizes the plurality of corrected handwritten characters acquired by the corrected character acquiring means, and the character correcting means Since the misrecognized character selected by the recognized character selecting means is corrected to the recognized character obtained with respect to the corrected handwritten character, the operation by the operator can be simplified.

  According to the twelfth aspect of the present invention, in the handwritten character acquisition step, handwritten character information relating to a plurality of handwritten characters drawn by the operator on the coordinate input device is acquired, and in the character cutting step, the handwritten character acquisition step is performed. Based on the acquired handwritten character information, a plurality of handwritten characters are divided into character units, and in the character recognition step, character recognition is performed on each character extracted in the character extraction step, and in the erroneous recognition character selection step, The misrecognized character obtained by misrecognizing in the character recognizing step is selected by the operator's instruction, and in the character correcting step, the misrecognized character selected in the misrecognized character selecting step is corrected to another character. In a character recognition device that recognizes handwritten characters in a batch, correction of misrecognized characters is not limited to single character units, but multiple character units. But it is possible to perform an effect that it is possible to improve the convenience of the character recognition device.

  According to the invention of claim 13, in the handwritten character acquisition step, handwritten character information related to a plurality of handwritten characters drawn by the operator on the coordinate input device is acquired, and in the character cutout step, the handwritten character acquisition step is performed. Based on the acquired handwritten character information, a plurality of handwritten characters are divided into character units, and in the character recognition step, character recognition is performed on each character extracted in the character extraction step, and in the erroneous recognition character selection step, The misrecognized character obtained by misrecognizing in the character recognizing step is selected by the operator's instruction, and in the character correcting step, the misrecognized character selected in the misrecognized character selecting step is corrected to another character. In a character recognition device that recognizes handwritten characters in a batch, correction of misrecognized characters is not limited to single character units, but multiple character units. But it is possible to perform an effect that it is possible to improve the convenience of the character recognition device.

  Exemplary embodiments of a character recognition device, a character recognition method, and a character recognition program according to the present invention will be explained below in detail with reference to the accompanying drawings.

(First embodiment)
FIG. 1 is a diagram illustrating an overall configuration of a character recognition system 1 including a character recognition device according to a first embodiment.

  The character recognition system 1 includes a character recognition device 100 and a handwritten character input device 200. The character recognition device 100 according to the present embodiment is a personal computer. Specifically, the handwritten character input device 200 is a large screen display device. The handwritten character input device 200 includes a handwritten character input screen 210 for inputting handwritten characters and the like, and a display unit 220 for displaying the handwritten characters input on the handwritten character input screen 210. Specifically, the handwritten character input screen 210 is a coordinate input device that acquires input from the user as coordinate values. The handwritten character input screen 210 and the display unit 220 are provided integrally.

  Note that a mouse or a tablet may be used as input means for the user to input characters or the like on the handwritten character input screen 210.

  The display unit 220 may be a CRT or a projector. Another example of the display unit 220 may be a device that is already on the market and that combines a coordinate input device and a projector.

  As shown in FIG. 1, the handwritten character input screen 210 of the handwritten character input device 200 detects a signal input by an operator using an input means such as a finger or a pen as a coordinate value. Then, it is sent to the character recognition device 100. The character recognition device 100 performs processing on a set of coordinate values (hereinafter referred to as a stroke) acquired from the handwritten character input screen 210. Then, the obtained processing content is displayed on the display unit 220.

  In the present embodiment, the action that the user performs on the handwritten character input screen 210 is associated with the left button of the mouse. For example, the action of the user touching the handwritten character input screen 210 is associated with the DOWN event of the left mouse button. The action of the operator moving while using the input means while touching the coordinate input unit is associated with the move event of the left mouse button. The action of the operator separating the input means from the coordinate input unit is associated with the UP event of the left mouse button.

  The handwritten character input device 200 is provided with a driver for controlling the handwritten character input screen 210. The control driver of the handwritten character input screen 210 outputs the coordinate value obtained by the operator's action and mouse event information to the character recognition device 100.

  FIG. 2 is a block diagram illustrating a functional configuration of the character recognition device 100. The character recognition device 100 includes a coordinate acquisition unit 102, a command processing unit 104, a handwritten character batch recognition unit 106, a character editing unit 108, a display processing unit 110, a coordinate value memory 120, a line unit memory 122, A result memory 124, a correction memory 126, and a rewrite memory 128 are included.

  The coordinate acquisition unit 102 acquires coordinate values input to the handwritten character input screen 210 from a driver that controls the handwritten character input screen 210. Then, the coordinate value is stored in the coordinate value memory 120 and the row unit memory 122.

  FIG. 3 shows the data structure of the coordinate value memory 120. The coordinate value memory 120 stores a plurality of coordinate values in units of stroke images. The stroke image is an image obtained by an event from the mouse DOWN to MOVE to UP.

  The row unit memory 122 has a plurality of pages. Each page stores a plurality of coordinate values corresponding to one stroke image. That is, one page stores all the coordinate values obtained during the stroke, that is, between the DOWN and UP of the mouse.

  Furthermore, the coordinate value memory 120 has a plurality of pages and stores coordinate values corresponding to a plurality of stroke images. Each page stores coordinate values corresponding to different strokes.

  FIG. 4 shows the data structure of the row unit memory 122. The line unit memory 122 stores handwritten characters written over a plurality of lines in units of lines. The row unit memory 122 has a plurality of pages. Each page stores information corresponding to one line of handwritten characters.

  Here, one line is a predetermined range in the vertical direction of the handwritten character input screen 210. This predetermined range, that is, height information indicating the height is held in advance. If even a part of a new stroke enters the vertical range in which a predetermined margin is provided in the height information, the stroke is recognized as a stroke for the same line.

  Each page stores information indicating a start page and an end page in the coordinate value memory 120 of the stroke image for one line included in each page. Each page also stores information indicating the bounding rectangle of the corresponding row. Specifically, the upper left coordinate and lower right coordinate of the row store information indicating a circumscribed rectangle.

  Each page further stores information indicating a start page and an end page of the result memory 124 that stores a recognition result for characters included in the corresponding line.

  Again, the description returns to FIG. The command processing unit 104 specifies a command from the user based on the coordinate value acquired by the coordinate acquisition unit 102. As commands from the user, there are a handwritten character batch recognition processing command and a character editing processing command. Here, the handwritten character batch recognition processing is processing for recognizing handwritten characters input on the handwritten character input screen 210 by the user. When a plurality of handwritten characters are input, a character recognition process is performed on all the plurality of handwritten characters. The character editing process is a process for changing a character already recognized and identified to another character. That is, when an erroneous character is recognized, a process for correcting it to a correct character is performed.

  The handwritten character batch recognition unit 106 performs character recognition processing on the handwritten characters input on the handwritten character input screen 210. When a plurality of characters are input on the handwritten character input screen 210, character recognition processing is performed on all the plurality of characters at once. Specifically, the handwritten character is first divided into character units. Then, character recognition processing is performed on each character obtained by the division. The recognized characters obtained by character recognition for each handwritten character are stored in the result memory 124.

  The handwritten character batch recognition unit 106 further stores, in the result memory 124, recognized character candidates that are highly likely to be recognized characters by the character recognition process. For example, when the recognized character is “u”, it is stored as a “ra” recognized character candidate similar to the recognized character “u”.

  FIG. 5 shows the data structure of the result memory 124. The result memory 124 stores the result of character recognition by the handwritten character batch recognition unit 106. The result memory 124 has a plurality of pages. Each page stores a recognized character and a recognized character candidate corresponding to one character. Each page further stores a start page and an end page of the coordinate value memory 120 corresponding to one stored character, and an upper left coordinate and a lower right coordinate of a line including the stored one character.

  The character editing unit 108 edits the recognized character obtained by the character recognition process by the handwritten character batch recognition unit 106 according to an instruction from the user. The correction memory 126 temporarily stores the recognized character and the recognized character candidate selected by the user at this time. The character editing unit 108 also corrects the recognized character based on the handwritten character acquired again from the user with respect to the recognized character obtained by the recognition processing by the handwritten character batch recognition unit 106. The recognized character and the recognized character candidate obtained in this rewriting process are stored in the rewriting memory 128.

  The display processing unit 110 displays information stored in the coordinate value memory 120, the row unit memory 122, the result memory 124, the correction memory 126, and the rewrite memory 128 on the display unit 220.

  The coordinate acquisition unit 102 according to the present embodiment corresponds to the handwritten character acquisition unit, the misrecognized character selection unit, the corrected character selection unit, and the corrected character acquisition unit described in the claims. Moreover, the handwritten character collective recognition unit 106 according to the present embodiment corresponds to a character cutout unit and a character recognition unit described in the claims. In addition, the character editing unit 108 according to the present embodiment corresponds to the character correction means described in the claims.

  FIG. 6 is a flowchart showing character recognition processing by the character recognition device 100. The coordinate acquisition unit 102 of the character recognition device 100 first acquires coordinate values from the handwritten character input screen 210 (step S102). Next, the acquired coordinate values are sequentially stored in the coordinate value memory 120 (step S104).

  Next, the command processing unit 104 performs command processing based on the coordinate values acquired by the coordinate acquisition unit 102. In the command processing, it is determined whether the command from the user is a handwritten character batch recognition processing command or a character editing processing command. Then, the determined command is issued to the handwritten character batch recognition unit 106 or the character editing unit 108.

  When the coordinate acquisition unit 102 acquires a coordinate value for one stroke, the coordinate acquisition unit 102 determines whether the stroke image is an input of a handwritten character or a command. If it is an instruction, command processing is performed. If none of the commands corresponds to input of handwritten characters or the like, the process returns to step S102 and the coordinate information acquisition process is repeated.

  In the command processing (step S106), when the command processing unit 104 determines that the command from the user is a handwritten character batch recognition processing command, the command processing unit 104 commands the handwritten character batch recognition unit 106 to perform processing. Then, the handwritten character batch recognition unit 106 performs handwritten character batch recognition on the handwritten characters input on the handwritten character input screen 210 (step S110).

  In the handwritten character batch recognition process, handwritten characters are recognized in batches in units of lines. The character recognition result for each character is stored in the result memory 124. Specifically, the recognized character and the recognized character candidate obtained as a result of character recognition and the result memory 124 are stored. Further, the display processing unit 110 causes the display unit 220 to display the recognized character and the recognized character candidate.

  On the other hand, in the command processing (step S106), when the command processing unit 104 determines that the command from the user is a character editing processing command, the command processing unit 104 commands the character editing unit 108 to perform processing. Then, the character editing unit 108 performs character editing processing on the handwritten character designated by the user (step S112). In the character editing process, the character recognition result obtained by the character recognition by the handwritten character batch recognition unit 106 is corrected. The character editing process (step S112) will be described later.

  In the display process (step S120), the result obtained in the handwritten character batch recognition process (step S110) or the character editing process (step S112) is acquired from each memory and displayed on the display unit 220. At the same time, the process is returned to the coordinate input process. Waiting for input from the user, the above processing is repeated.

  FIG. 7 is a flowchart showing detailed processing of the coordinate acquisition unit 102 in step S104 described in FIG.

  The coordinate acquisition unit 102 constantly monitors mouse events (step S140). If there are no mouse events, the loop continues to monitor for mouse events.

  When a mouse event is detected in the handwritten character input screen 210 (step S140, Yes), the coordinate acquisition unit 102 stores the coordinate value in the result memory 124 (step S142). Next, it is checked whether the event is an UP event (step S144). If there is no UP event (step S144, No), the process is returned to the mouse event monitoring. If it is an UP event (step S144, Yes), the page of the coordinate value memory 120 is changed to the next page (step S146). Specifically, the page number is increased by one. The subsequent coordinate values are stored in the next page.

  Through the above processing, the coordinate value group obtained between the DOWN and UP of the mouse can be stored in the coordinate value memory 120 as a stroke. FIG. 8 shows a stroke image for the handwritten character input screen 210.

  At the same time, in order to clearly indicate the stroke unit to the operator, the display processing unit 110 is instructed to display the stroke image in which the coordinate points are connected with straight lines as shown in FIG. The display processing unit 110 displays a line segment as a stroke image on the display unit 220.

  Next, the coordinate acquisition unit 102 determines a row (step S148). Specifically, height information is held in units of lines. And even if some new strokes enter in the range of the vertical direction which provided the predetermined margin in height information, it recognizes as the handwritten character in the same line. In this case, the page number of the coordinate value memory 120 is added to the line unit memory 122 (step S150).

  At the same time, a circumscribed rectangle (upper left coordinates and lower right coordinates of the line) of the stroke group for one line including the new line is obtained. Then, the obtained coordinate value is stored in the row unit memory 122. If there is no new stroke, it is recognized as a new line. Then, the page of the line unit memory 122 is turned. That is, the page number is increased by one. When the line determination is completed, the process proceeds to command processing.

  FIG. 9 is a flowchart showing detailed processing of the command processing unit 104 in the command processing (step S106). First, when there is a handwritten character batch recognition processing command from the user (step S162, Yes), the handwritten character batch recognition unit 106 performs character recognition.

  The handwritten character batch recognition processing command is issued by the following two methods. That is, first, a button for starting recognition is arranged at an arbitrary place on the handwritten character input screen 210. Then, when the user presses the button, a handwritten character batch recognition processing command is issued.

  The other is to investigate whether there is further input within a predetermined time after the mouse UP event. When there is no input, a handwritten character batch recognition processing command is issued.

  When a handwritten character batch recognition processing command is issued by any of the above methods, the handwritten character batch recognition unit 106 performs a handwritten character batch recognition process.

  When the handwritten character batch recognition processing command is not issued (step S162, No), the presence / absence of the character editing processing command is confirmed.

  The character editing processing command is specified by the following method. That is, an ON / OFF button for editing is arranged at an arbitrary location on the handwritten character input screen 210. When the user presses the button, ON / OFF is switched. If the button state is ON, it is determined that there has been a character editing process command (step S164, Yes), and the process proceeds to character editing process. If it is OFF, it is determined that there is no character editing processing command (step S164, No), and the process returns to step S102.

  FIG. 10 is a flowchart showing detailed processing of the handwritten character batch recognition unit 106 in the handwritten character batch recognition processing (step S110) shown in FIG. The handwritten character batch recognition unit 106 first refers to the coordinate value memory 120 and the line unit memory 122 and calls a stroke group for one line (step S200). Next, the stroke group is divided into strokes in units of one character by cutting out one character (step S202). By the process of dividing into strokes, it is possible to specify how many pages to how many pages in the coordinate memory are one character.

  Next, character recognition is performed using the recognition engine for each divided stroke group of one character unit (step S204, Yes, step S206). Then, the recognized character and the recognized character candidate are stored in the result memory 124. At the same time, the page number of the result memory is added to the result memory page of the row unit memory 122.

  Various methods have already been proposed as a method for cutting out one character, and any of these methods may be used. Similarly, the recognition method for the recognition engine is not limited.

  Next, a character area in units of one character, that is, a circumscribed rectangle is calculated (step S208). Since the stroke for one character is known by cutting out one character, the upper left coordinate and the lower right coordinate of the region where the stroke group for one character exists are calculated with reference to the coordinate memory, and stored in the result memory.

  When the recognition has been completed for all characters (step S204, No), it is checked whether or not the next line exists (step S210). If the next line exists (step S210, Yes), the process returns to step S200.

  On the other hand, when the next line does not exist (step S210, No), the process proceeds to the display process (step S120). Thus, the handwritten character batch recognition process in (Step S110) is completed. As a result, the recognition result can be stored in the result memory 124.

  FIG. 11 is a flowchart showing detailed processing of the character editing unit 108 in the character editing processing (step S112) shown in FIG. First, referring to the line unit memory 122 or the result memory 124, a target character to be subjected to character editing processing is determined (step S220).

  Gesture processing is used to select the target character. The gesture process is a process for selecting a character in a predetermined unit. There are three types of gesture processing: line-by-line selection, multi-character unit selection, and single-character selection. The gesture processing will be described in detail with reference to FIGS.

  FIG. 12 shows the result of the handwritten character batch recognition process displayed on the display unit 220. When the user inputs a handwritten character and the handwritten character batch recognition process is executed, a recognized character corresponding to each handwritten character is displayed.

  As shown in FIG. 12, handwritten characters are displayed, and recognized characters obtained by character recognition by the handwritten character collective recognition unit 106 are displayed under each handwritten character.

  Here, a rectangular area including one line of handwritten characters is the line area 300. Moreover, the rectangular area | region containing each handwritten character is character area 310a-e.

  When the display area shown in FIG. 12 is clicked on the line area 300 as shown in FIG. 13 and does not correspond to the character areas 310a to 310e for any handwritten character, it is recognized as a selection in line units. The That is, the determination condition is “clicked in a line area not included in any character area”.

  In this case, the character editing unit 108 refers to the line unit memory 122 and specifies a line area including the click coordinates. Then, the target character is determined from the result memory start page and the result memory end page in the result memory 124 corresponding to the specified row area.

  Further, when the user draws a surrounding line surrounding a predetermined character in the row area 300 as shown in FIG. 14 with respect to the display content shown in FIG. 12, it is recognized as a selection of a plurality of characters. That is, it corresponds to the discrimination condition “enclosing line including any character area”.

  In this case, the character editing unit 108 first determines a target character. Specifically, a circumscribed rectangle of the encircling line is calculated. Then, based on the upper left coordinates and lower right coordinates of the result memory 124, the page of the result memory 124 existing inside the calculated circumscribed rectangle is determined as the target character.

  Also, as shown in FIG. 15, when the user draws an underline corresponding to a predetermined character in the row area 300, it is recognized as a selection of a plurality of characters as in the case of the encircled line shown in FIG. The This case also corresponds to the discrimination condition “line segment existing near the character area”.

  In this case, the character editing unit 108 determines whether or not the character regions 310a to 310e exist within a predetermined distance in the y coordinate direction from the coordinate group constituting the line segment with reference to the result memory 124. If the character area 310 exists, the page of the memory 124 is determined as the target character as a result.

  In addition, when any of the character areas 310a to 310e is clicked on the display content shown in FIG. 12 as shown in FIG. 16, it is recognized as a single character selection. That is, the determination condition is “clicked in any character area”.

  In this case, the character editing unit 108 refers to the result memory 124 to specify in which character region 310a to 310e the click coordinate exists. As a result, the memory page is determined as the target character.

  In the character editing process (step S112), first, a recognized character and a recognized character candidate for the target character determined by the above method are acquired from the result memory 124. The acquired recognized characters and recognition candidates are stored in the correction memory 126. During the character string editing process, the process is reflected on the recognized character and the recognition candidate stored in the correction memory 126.

  The description returns to FIG. 11 again. If the target character determined in step S220 is a single character (step S222, Yes), a single character recognized character candidate dialog is displayed (step S224). On the other hand, if the character is not a single character (step S222, No), a recognized character candidate dialog for a plurality of characters is displayed (step S226).

  In this manner, different recognition character candidate dialogs are displayed depending on whether the character is a single character or a plurality of characters. Thereby, a user's visibility can be improved.

  As another example, a recognition character candidate dialog for a plurality of characters may be displayed even when the target character is a single character.

  FIG. 17 shows an example of a recognized character candidate dialog 400 for a plurality of characters. The recognized character candidate dialog 400 for multiple characters includes a confirmation button 402, a recognized character display area 404, a character candidate display area 406, and a rewrite button 408.

  The recognized character candidate dialog 400 is displayed as a result when the recognition result of the handwritten character “Aiueo” shown in FIG. 12 is selected in units of lines or a plurality of characters. In the recognized character display area 404 of the recognized character candidate dialog 400, a recognized character “Aira no O” corresponding to a plurality of handwritten characters drawn in one line is displayed. In the character candidate display area 406, recognized character candidates corresponding to each handwritten character are displayed in the vertical direction in the order of recognition candidates. Each recognized character candidate is displayed as a button, and the user can select a correct character from these.

  Each recognition candidate button is associated with a recognized character candidate ID for identifying each recognized character candidate. When the user selects a predetermined recognized character candidate button in the character candidate display area 406, the character editing unit 108 can determine a recognized character candidate corresponding to the selected recognized character candidate button as a recognized character.

  The description returns to FIG. 11 again. When the recognized character candidate dialog 400 is displayed (step S226), the contents input by the user in the recognized character candidate dialog 400 for a plurality of characters are determined (step S230). There are three types of input from the user: “press one of the recognition candidate buttons”, “press the confirmation button”, and “press the rewrite button”. Each case is described below.

  If the input is “press any of the recognition candidate buttons” (step S230, press the recognition candidate button), the ID of each recognition candidate button is checked. Then, the recognized character candidate selected by the user from among the plurality of recognized character candidates in the correction memory 126 is specified. The recognized character candidate selected by the user is stored as a recognized character in the correction memory 126 (step S232).

  Next, the recognized character candidate is reflected in the recognized character display area 404 of the recognized character candidate dialog 400. That is, the selected recognized character candidate is displayed as a recognized character in the recognized character display area 404 (step S234). And it returns to step S230 again and waits for the input from a user.

  The process from step S230 to step S234 will be described more specifically with reference to FIGS. 18-1 to 18-3. In the recognized character display area 404 shown in FIG. 18A, the recognized characters “La” and “No” are misrecognized characters obtained by misrecognition. Therefore, the user presses the “U” button from the recognition candidates of the recognized character “La” in the character candidate display area 406 shown in FIG. Thereby, in the correction memory 126, “U” is stored as the recognized character instead of the recognized character “La”. Further, as shown in FIG. 18B, the recognized character “La” in the recognized character display area 404 is changed to “U”.

  Next, the recognized character “NO” in the recognized character display area 404 shown in FIG. In other words, the “e” button is pressed from the recognition candidates for the recognized character “NO” in the character candidate display area 406 shown in FIG. Thereby, in the correction memory 126, “e” is stored as the recognized character instead of the recognized character “no”. Further, as shown in FIG. 18C, the recognized character “NO” in the recognized character display area 404 is changed to “E”.

  The description returns to FIG. 11 again. In step S230, when the input from the user is “pressing the confirmation button” (step S230, pressing the confirmation button), the data in the correction memory 126 is stored in the page of the target character in the result memory 124 (step S240). Then, the recognized character candidate dialog 400 is hidden (step S242). Thus, the character editing process (step S112) is completed, and the process proceeds to the display process (step S120) described with reference to FIG.

  When the user presses the confirm button 402 after the correction process as described with reference to FIGS. 18A to 18C is performed, the recognized characters for the handwritten characters shown in FIG. As shown in Fig. 5, the character is changed to a corrected recognition character.

  In step S230, when the input from the user is “pressed the rewrite button” (step S230, press the rewrite button), the rewrite process is executed (step S236). Next, the contents of the correction memory 126 are reflected in the recognized character candidate dialog 400 for a plurality of characters (step S234). And it returns to step S230 again.

  Here, the rewriting process will be described with reference to FIGS. 20-1 and 20-2. The rewriting process is a process when there is no correct recognized character in the recognized character candidate dialog 400 for a plurality of characters. As shown in FIG. 20A, the recognized character “NO” is displayed in the character candidate display area 406 as a correct recognized character as a recognized character candidate. Therefore, when the user presses the button for the recognized character candidate “E”, the recognized character “NO” can be corrected to “E” as shown in FIG.

  On the other hand, in the recognized character “La”, as shown in FIG. 20A, there is no “U” button as a recognized character candidate. Therefore, when the correct recognized character is “U”, it cannot be corrected to the correct recognized character by the method of selecting the recognized character candidate button.

  As described above, when the correct recognized character does not exist in the character candidate display area 406, the rewriting process is performed. Specifically, the rewriting process starts when the user presses the rewriting button 408 shown in FIG.

  FIG. 21 is a flowchart showing detailed processing in the rewriting processing (step S236) shown in FIG. First, the rewriting dialog is displayed with reference to the correction memory 126 (step S250).

  22-1 to 22-3 show an example of the rewrite dialog 420. FIG. As illustrated in FIGS. 22-1 to 22-3, the rewrite dialog 420 includes a recognized character 422, a rewrite area 424 corresponding to the recognized character, and a confirm button 426. Rewrite areas 424 are provided as many as the number of recognized characters 422. In FIG. 22-1, the recognized characters “La” and “E” are displayed, and two rewrite areas 424 corresponding to these are displayed.

  As shown in FIG. 22-2, handwritten characters are input again by the user in the rewriting area 424 of the rewriting dialog 420 shown in FIG. 22-1. When the handwritten character input to the rewriting area 424 is correctly recognized, the correct recognized character “U” is displayed as the recognized character 422 as shown in FIG.

  The description returns to FIG. 21 again. After displaying the rewrite dialog 420, the input is determined (step S252). There are five inputs: “mouse DOWN in rewrite area”, “mouse move in rewrite area”, “mouse UP in rewrite area”, “timer time reached”, and “determine button pressed”. Each case will be described below.

  If the input is “mouse DOWN in rewrite area”, the recognition timer is stopped (step S254). Then, the process returns to the input determination process (step S252). Here, the recognition timer is a timer that starts after a mouse UP event. When the recognition timer starts, an input from the user is waited for a predetermined time. If there is no input, a handwritten character batch recognition command is issued.

  If there is a mouse DOWN in the rewriting area, the user has started to write handwritten characters. Therefore, the recognition timer is not started until the user finishes input.

  If the input from the user in step S252 is “mouse move in rewrite area”, the coordinate values in the mouse move are stored in the rewrite memory 128 temporarily prepared (step S260). The rewrite memory 128 is a memory equivalent to the coordinate value memory 120. When the storage of the coordinate values is completed, the contents of the correction memory 126 and the rewrite memory 128 are reflected in the rewrite dialog 420 (step S262). That is, in order to display the progress of the user's input of handwritten characters, the coordinate values in the rewriting memory are connected and displayed by line segments as shown in FIG. Then, the process returns to the input determination process (step S252).

  If the input is “mouse UP in rewrite area”, the recognition timer is started (step S256). Then, the process returns to the input determination process (step S252). The fact that there was a mouse UP in the rewriting area means that the user may have finished writing handwritten characters. In this case, there is a possibility that there is an additional input from the user. Therefore, the recognition timer is started. Then, after waiting for a predetermined time, recognition is started (step S270). In this case, the waiting time is preferably 1 to 2 seconds.

  Thus, when the input is “timer time reached”, it is determined that a predetermined time has elapsed since the user's previous input completion (mouse UP in the rewriting area), and recognition is started. At this time, the recognition timer is stopped.

  The recognition engine executes recognition with the stroke group stored in the rewriting memory 128 as one character (step S272). The recognition result is stored in the correction memory 126. Specifically, the recognized character and the recognized character candidate are stored (step S274). Which recognition character and recognition character candidate in the correction memory 126 corresponds to the recognition result is specified from the ID of the rewriting area based on which rewriting area is input. Further, the data stored in the rewrite memory 128 is initialized (erased) (step S276) and prepared for the next input. Next, the contents of the correction memory 126 and the rewrite memory 128 are reflected in the rewrite dialog 420 (step S262), and the process returns to the input determination process (step S252).

  If the input is “depress the confirmation button”, it is determined that the rewriting process is completed, and the rewriting dialog 420 is hidden (step S280). Then, the process proceeds to step S236 described with reference to FIG.

  FIG. 23 shows a recognized character candidate dialog 400 displayed after the rewriting process is completed. When the confirm button 426 shown in FIGS. 22-1 to 22-3 is pressed, the recognized character candidate dialog 400 is displayed again.

  In the recognized character candidate dialog 400 for a plurality of characters displayed before the rewriting process is performed, “ra” is displayed as a recognized character in the recognized character display area 404 as shown in FIG. In the recognized character candidate dialog 400 for a plurality of characters shown in FIG. 23, the recognized character “La” is changed to “U” in the recognized character display area 404. Further, in FIG. 23, a recognized character candidate that is a result of re-recognition by the rewriting process is displayed in the character candidate display area 406. When the confirm button 402 is pressed here, the result of the character recognition process is displayed.

  FIGS. 24-1 and 24-2 illustrate a recognized character candidate dialog 430 for a single character. As described with reference to FIG. 16, when a single character is selected in step S222 of FIG. 11, a recognized character candidate dialog 430 for a single character shown in FIG. 24-1 is displayed.

  The recognized character candidate dialog 400 for multiple characters has a large display area and is not suitable for a single character. Therefore, as shown in FIGS. 24-1 and 24-2, by displaying a recognized character candidate dialog 430 for a single character having a smaller display area than the recognized character candidate dialog 400 for a plurality of characters, convenience for the user is displayed. Can be improved.

  Note that the recognized character candidate dialog for a single character is different only in the display method, and the processing content is the same as the processing after step S230 described with reference to FIGS.

  In the recognized character candidate dialog 430 for a single character shown in FIG. 24A, when the recognized character “La” recognition candidate “U” button is pressed, the recognized character “La” in the editing memory is replaced with “U”. Then, “U” is displayed as a recognized character as reflected in the recognized character candidate dialog 430 for a single character, as shown in FIG. When the confirm button is pressed, the contents of the correction memory 126 are stored again in the result memory 124, and “ra” is changed to “u” as a recognized character and displayed as shown in FIG.

  FIG. 26 is a flowchart showing detailed processing of the display processing unit 110 in the display processing (step S120). The display processing receives processing from coordinate input processing, handwritten character batch recognition processing, and character editing processing.

  When the process is received from the coordinate input process, the coordinate points in the coordinate value memory 120 are connected by line segments and displayed on the display device (step S320).

  When the processing is received from the handwritten character recognition processing and the character editing processing, the recognized characters in the result memory 124 are displayed as text on the display device (step S300), and the upper left coordinates and lower right coordinates are obtained from the result memory 124. The character area is read (step S310), the upper left and lower right coordinates are read from the line unit memory 122, and the line area is displayed (step S312), and the coordinate points in the coordinate value memory 120 are connected by line segments. Is displayed (step S320).

  Each process described above is performed for all strokes, all recognized characters, and all lines. The Text display method is performed using a TextBox such as Visual Basic. Then, the process returns to the coordinate input process.

  FIG. 27 is a diagram illustrating a hardware configuration of the character recognition device 100 according to the first embodiment. The character recognition device 100 has, as a hardware configuration, a ROM 52 that stores a character recognition program that executes character recognition processing in the character recognition device 100, and a CPU 51 that controls each unit of the character recognition device 100 according to a program in the ROM 52. The RAM 53 stores various data necessary for controlling the character recognition device 100, the communication I / F 57 is connected to the network and communicates, and the bus 62 connects each unit.

  The character recognition program in the character recognition device 100 described above is a file in an installable or executable format on a computer-readable recording medium such as a CD-ROM, floppy (R) disk (FD), or DVD. It may be recorded and provided.

  In this case, the character recognition program is loaded onto the main storage device by being read from the recording medium and executed by the character recognition device 100, and each unit described in the software configuration is generated on the main storage device. It has become.

  Further, the character recognition program of the present embodiment may be provided by being stored on a computer connected to a network such as the Internet and downloaded via the network.

  As described above, the present invention has been described using the embodiment, but various changes or improvements can be added to the above embodiment.

It is a figure which shows the whole structure of the character recognition system 1 containing the character recognition apparatus concerning Embodiment 1. FIG. 2 is a block diagram showing a functional configuration of a character recognition device 100. FIG. 3 is a diagram illustrating a data configuration of a coordinate value memory 120. FIG. 3 is a diagram showing a data configuration of a row unit memory 122. FIG. It is a figure which shows the data structure of the result memory. 5 is a flowchart illustrating character recognition processing by the character recognition device 100. It is a flowchart which shows the detailed process of the coordinate acquisition part 102 in step S104 demonstrated in FIG. It is a figure which shows the stroke image with respect to the handwritten character input screen. It is a flowchart which shows the detailed process of the command process part 104 in a command process (step S106). It is a flowchart which shows the detailed process of the handwritten character batch recognition part 106 in the handwritten character batch recognition process (step S110) shown in FIG. It is a flowchart which shows the detailed process of the character edit part 108 in the character edit process (step S112) shown in FIG. It is a figure which shows the result of the handwritten character batch recognition process displayed on the display part. It is a figure for demonstrating the discrimination conditions that it clicked in the line area | region which is not contained in any character area. It is a figure for demonstrating the discrimination conditions called the surrounding line which includes any one of character areas. It is a figure for demonstrating the discrimination conditions called the line segment which exists in the character area vicinity. It is a figure for demonstrating the discrimination conditions that it clicked in one of the character areas. It is a figure which shows an example of the recognition character candidate dialog 400 for multiple characters. It is a figure for demonstrating more concretely about processing of Step S230 to Step S234. It is a figure for demonstrating more concretely about processing of Step S230 to Step S234. It is a figure for demonstrating more concretely about processing of Step S230 to Step S234. It is a figure which shows the recognition character displayed when the confirmation button 402 is pressed after correction processing is performed. It is a figure for demonstrating a rewriting process. It is a figure for demonstrating a rewriting process. It is a flowchart which shows the detailed process in the rewriting process (step S236) shown in FIG. 5 is a diagram illustrating an example of a rewrite dialog 420. FIG. 5 is a diagram illustrating an example of a rewrite dialog 420. FIG. 5 is a diagram illustrating an example of a rewrite dialog 420. FIG. It is a figure which shows the recognition character candidate dialog 400 displayed after the rewriting process is completed. It is a figure which shows an example of the recognition character candidate dialog 430 for single characters. It is a figure which shows an example of the recognition character candidate dialog 430 for single characters. It is a figure which shows the recognition character after a change. It is a flowchart which shows the detailed process of the display process part 110 in a display process (step S120). It is a figure which shows the hardware constitutions of the character recognition apparatus 100 which concerns on Embodiment 1. FIG.

Explanation of symbols

1 Character recognition system 51 CPU
52 ROM
53 RAM
57 Communication I / F
62 Bus 100 Character recognition device 102 Coordinate acquisition unit 104 Command processing unit 106 Handwritten character batch recognition unit 108 Character editing unit 110 Display processing unit 120 Coordinate value memory 122 Line unit memory 124 Result memory 126 Correction memory 128 Rewriting memory 200 Handwritten character input device 210 handwritten character input screen 220 display unit 300 line area 310a-e character area 400 recognized character candidate dialog for multiple characters 402 confirmation button 404 recognized character display area 406 character candidate display area 408 rewrite button 420 rewrite dialog 422 recognized character 424 rewrite area 426 Confirm button 430 Recognized character candidate dialog for single character

Claims (13)

A handwritten character recognition device that recognizes a handwritten character drawn by an operator on a coordinate input device,
Handwritten character acquisition means for acquiring handwritten character information about a plurality of handwritten characters drawn by the operator on the coordinate input device;
Based on the handwritten character information acquired by the handwritten character acquisition means, character cutting means for dividing the plurality of handwritten characters into character units;
Character recognition means for performing character recognition on each character cut out by the character cutting means;
A misrecognized character selection unit that selects a misrecognized character obtained by misrecognition by the character recognizing unit according to an instruction of the operator;
A character recognition apparatus comprising: a character correcting unit that corrects the erroneously recognized character selected by the erroneously recognized character selecting unit to another character. The misrecognized character selection means selects a plurality of the misrecognized characters from the plurality of handwritten characters acquired by the handwritten character acquisition means according to an instruction from the operator,
The character recognition device according to claim 1, wherein the character correcting unit corrects each of the plurality of erroneously recognized characters selected by the erroneously recognized character selecting unit to other characters. The handwritten character display means provided in the coordinate input device displays the handwritten character drawn by the operator,
The misrecognized character selection means is a character designated by the operator when the operator designates a predetermined handwritten character among the plurality of handwritten characters displayed on the handwritten character display means on the coordinate input device. The character recognition device according to claim 2, wherein the character recognition character is selected as the erroneous recognition character.   When the operator draws a surrounding line surrounding a predetermined handwritten character among the plurality of handwritten characters displayed on the handwritten character display means, the surrounding line is drawn by the operator in the coordinate input device. 4. The character recognition device according to claim 1, wherein a character surrounded by is selected as the misrecognized character. 5.   When the operator draws an underline for a predetermined character among the plurality of handwritten characters displayed on the handwritten character display unit, the character corresponding to the underline is selected by the operator. The character recognition apparatus according to claim 1, wherein the character recognition device is selected as the erroneously recognized character. The handwritten character display means displays the plurality of handwritten characters acquired by the handwritten character acquisition means in line units,
The misrecognized character selection means, when the operator designates the line area provided corresponding to the predetermined handwritten character displayed on the handwritten character display means on the coordinate input device, The character recognition device according to claim 1, wherein a plurality of characters corresponding to a region are selected as the erroneously recognized characters. Recognized character display means for displaying a recognized character obtained by the character recognizing means and a recognized character candidate determined to be a recognized character by the character recognizing means,
A correction character selection means for selecting a predetermined recognition character candidate from the recognition character candidates displayed by the recognition character display means according to an instruction from the operator;
The said character correction means changes the said misrecognized character selected by the said misrecognized character selection means to the said recognition character candidate selected by the said corrected character selection means, The any one of Claim 1 to 6 characterized by the above-mentioned. The character recognition device according to one item. The recognized character display means displays the recognized characters for each of the plurality of handwritten characters and the recognized character candidates for each of the plurality of handwritten characters,
The corrected character selecting means selects the recognized character candidate for the erroneously recognized character selected by the erroneously recognized character selecting means according to an instruction from the operator,
8. The character according to claim 7, wherein the character correcting means changes the erroneously recognized character selected by the erroneously recognized character selecting means to the recognized character candidate selected by the corrected character selecting means. Recognition device. The recognized character display means displays a recognized character for the erroneously recognized character selected by the erroneously recognized character selection means and the recognized character candidate among the plurality of handwritten characters acquired by the handwritten character acquisition means. ,
The modified character selection means selects the predetermined recognized character candidate according to an instruction from the operator,
The said character correction means changes the said misrecognized character selected by the said misrecognized character selection means to the said recognized character candidate selected by the said corrected character selection means, The Claim 7 or 8 characterized by the above-mentioned. Character recognition device. A correction character acquisition means for acquiring a corrected handwritten character for the erroneous recognition character selected by the erroneous recognition character selection means;
The character recognition means recognizes the corrected handwritten character acquired by the corrected character acquisition means,
The said character correction means corrects the said misrecognized character selected by the said misrecognized character selection means to the recognized character obtained with respect to the said corrected handwritten character, The any one of Claim 1 to 9 characterized by the above-mentioned. The character recognition device according to one item. The misrecognized character selection means selects a plurality of misrecognized characters from the plurality of handwritten characters according to an instruction from the operator,
The corrected character acquisition means acquires corrected handwritten characters corresponding to each of the plurality of erroneously recognized characters selected by the erroneously recognized character selection means,
The character recognition means recognizes a plurality of corrected handwritten characters acquired by the corrected character acquisition means,
11. The character recognition according to claim 10, wherein the character correcting unit corrects the erroneously recognized character selected by the erroneously recognized character selecting unit to a recognized character obtained with respect to the corrected handwritten character. apparatus. A handwritten character recognition method for recognizing a handwritten character drawn by an operator on a coordinate input device,
A handwritten character acquisition step of acquiring handwritten character information regarding a plurality of handwritten characters drawn by the operator on the coordinate input device;
Based on the handwritten character information acquired in the handwritten character acquisition step, a character cutting step for dividing the plurality of handwritten characters into character units;
A character recognition step of performing character recognition for each character cut out in the character cutout step;
A misrecognized character selection step of selecting a misrecognized character obtained by misrecognition in the character recognizing step according to an instruction of the operator;
And a character correcting step of correcting the erroneously recognized character selected in the erroneously recognized character selecting step to another character.   A character recognition program for causing a computer to execute the character recognition method according to claim 12.

Images