JP2010122966A - Character recognition processor and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently recognize a character desired by a user without deteriorating a learning effect by a handwritten input and even when character recognition performance is not high in a character recognition processor. <P>SOLUTION: When a user writes a desired character (a Chinese character composed of two components, e.g. "Go" or word) in handwritten character input areas 16L and 16R of a touch panel 16 (16t), a first recognition candidate character (a word having the same component on the left side, e.g. "Go" or error) in a recognition candidate character group verified and recognized on the basis of a recognition database is displayed in a recognized character display area 16N and the handwritten character input areas 16L and 16R. When a component (a), e.g. "Gon" or remark, which is also included in the desired character, of the word displayed in the handwritten character input areas 16L and 16R is selected and touched, only recognition candidate characters including the selected component (a) as a fixed recognition portion are selected and displayed as a correction candidate character (16TK) in the recognition candidate character group. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a character recognition processing device and a program for recognizing handwritten characters as printed characters.

  In recent years, electronic dictionaries in which a paper dictionary is stored in a storage device as electronic data and can be easily looked up by a user have been widely used. According to this electronic dictionary, it is possible to search for a headword that matches the keyed-in character, display explanation information corresponding to the headword, and check the meaning content of the word desired by the user.

  In recent electronic dictionaries, it is possible to search for a headword from the recognized characters by recognizing characters input by handwriting instead of key input.

  However, in a conventional character recognition apparatus that recognizes characters input by handwriting as described above, if a character input by handwriting is recognized as a character different from a character desired by the user, the recognized character is used. After deletion, it was necessary to input a desired character again by handwriting, which was very troublesome. The trouble is particularly noticeable when inputting complex characters with a large number of strokes such as kanji.

Therefore, a handwritten character input method that narrows down recognition candidate characters by handwriting input of only some components such as radicals of kanji to be input and displaying a list of kanji including the components recognized from now on. (For example, refer to Patent Document 1). According to this, since it is not necessary to input the entire character by handwriting, it is possible to improve the efficiency of the user work, and it is possible to input a desired character even if only a part of the kanji is known.
Japanese Patent Application Laid-Open No. 06-348687

  In the conventional handwritten character input method, since it is not necessary to input the entire character by handwriting, the efficiency of the user work can be improved, but the learning effect by inputting the entire character by handwriting is reduced. . In addition, when the character recognition performance of the apparatus is not high, there is a problem that even if only a part of the constituent elements of the kanji to be input are input by handwriting, they are not recognized as desired constituent elements.

  The present invention has been made in view of such problems, and can efficiently recognize a user-desired character without impairing the learning effect by handwriting input and when the character recognition performance is not high. An object of the present invention is to provide a character recognition processing apparatus and program.

  The character recognition processing device according to claim 1, a handwritten character input means for inputting handwritten characters in response to a user operation, a character recognition means for character recognition of handwritten characters input by the handwritten character input means, Recognized character display means for displaying recognition candidate characters obtained by character recognition by the character recognition means, and character portion designation for designating a recognition candidate character portion displayed by the recognized character display means in accordance with a user operation And a recognized character narrowing display means for narrowing and displaying characters including the character portion designated by the character portion designation means among the recognition candidate characters obtained by character recognition by the character recognition means. It is characterized by that.

  The character recognition processing device according to claim 2, a handwritten character input means for inputting a handwritten character in response to a user operation, a character recognition means for recognizing a handwritten character input by the handwritten character input means, Recognized character display means for displaying recognition candidate characters obtained by character recognition by the character recognition means, and character portion designation for designating a recognition candidate character portion displayed by the recognized character display means in accordance with a user operation A character designated by the character portion designation means among the recognition candidate characters obtained by the character recognition by the character recognition means, and a handwritten character re-input means for re-inputting handwritten characters in response to a user operation The character including the part is narrowed down, and the handwritten character re-input by the handwritten character re-input means is recognized from the narrowed-down character. It is characterized by comprising: a character recognition means.

  The character recognition processing device according to claim 3 is the character recognition processing device according to claim 1 or 2, wherein the character portion designating means includes a recognition candidate character displayed by the recognized character display means. It is characterized in that a plurality of parts are designated according to a user operation.

  The character recognition processing device according to claim 4 is the character recognition processing device according to claim 2, wherein the displayed character portion of the recognition candidate is designated by the character portion designation means in accordance with a user operation. And a positional relationship designating unit for designating a positional relationship between the designated character part and the additional part, and the re-character recognition unit is configured to recognize the character of the recognition candidate obtained by the character recognition by the character recognition unit. Among them, the character including the character portion specified by the character portion specifying means is narrowed down, and the character portion specified by the character portion specifying means is specified by the positional relationship specifying means from among the narrowed characters. Recognizing a character that is a combination of handwritten characters re-input by the handwritten character re-input means according to the positional relationship as the additional part. It is characterized.

  The program according to claim 5 is a program for controlling a computer including a handwriting input unit, wherein the computer inputs handwritten characters according to a user operation from the handwriting input unit. A character recognition unit for recognizing a handwritten character input by the handwritten character input unit; a recognized character display control unit for displaying a recognition candidate character obtained by character recognition by the character recognition unit on the handwriting input unit; A character part designating unit that designates a character part of a recognition candidate displayed by the recognized character display control unit according to a user operation, and the character part of the recognition candidate character obtained by character recognition by the character recognizing unit. Narrow down the character including the character part designated by the designation means It is characterized in that to function as a display control unit.

  According to the present invention, there is provided a character recognition processing device and a program capable of efficiently recognizing a user-desired character without impairing the learning effect by handwriting input and when the character recognition performance is not high. it can.

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

  FIG. 1 is a diagram showing an external configuration of an electronic dictionary device 10 having a handwriting input function according to an embodiment of a character recognition processing device of the present invention.

  The character recognition processing device is configured to be mounted on an electronic dictionary device 10 having a handwriting input function described below, or a PDA (personal digital assistants), a PC (personal computer), a mobile phone having a handwriting input function. It is configured to be installed in an electronic book, a portable game machine or the like.

  The electronic dictionary device 10 includes a foldable case in which a main body case 11 and a lid case 12 can be expanded / closed via a hinge portion 13. On the surface of the main body case 11 where the folding case is unfolded, a character input key 14a, a dictionary designation / function key 14b, a “translation / decision” key 14c, a cursor key 14d, a “recognition” key 14e, speakers 15L, 15R, etc. Are provided with an input unit 14 and a touch panel 16.

  The touch panel 16 has a structure in which a touch position detection device that detects a position touched by a user with a pen, a finger, and the like and a display device are integrated. A transparent tablet 16t is overlaid on the screen. The input area of the touch panel 16 is a correction to the handwritten character input areas 16L and 16R for inputting the handwritten character or the recognized character (recognized character) as necessary in accordance with the execution of the process for recognizing the handwritten character. It is switched to a correction candidate character area 16TK for displaying a list of candidate characters for selection (see FIG. 5).

  And the locus | trajectory accompanying the handwriting input in the state in which the said touch panel 16 was switched to the handwritten character input area 16L, 16R is displayed on the liquid crystal display screen.

  Further, a liquid crystal display unit 17 with a backlight of, for example, 480 × 320 dots is provided on the surface of the lid case 12 for almost the entire surface.

  FIG. 2 is a block diagram showing a configuration of an electronic circuit of the electronic dictionary device 10 having the handwriting input function.

  The electronic dictionary device 10 is configured by a computer that reads a program recorded in various storage media or a transmitted program and whose operation is controlled by the read program. The electronic circuit includes a CPU ( a central processing unit) 21 is provided.

  The CPU 21 is a device control program 22a stored in the main storage area (flash memory) 22 in advance, or an apparatus read into the main storage area 22 from the external storage medium 23 such as a ROM card via the storage medium read / write unit 24. A work area (RAM) 26 according to the control program 22a or the apparatus control program 22a read into the main storage area 22 from the Web server 30 (in this case, the program server) 30 on the Internet N via the communication control unit 25. The various work data are stored in and the operation of each part of the circuit is controlled.

  The device control program 22a stored in the main storage area 22 is an input signal corresponding to a user operation from the input unit 14, or each Web server 30 on the Internet N connected via the communication control unit 25. Or a connection communication signal with a memory card (storage medium) 23 such as an EEPROM, RAM, or ROM connected externally via the storage medium read / write unit 24.

  In addition to the main storage area 22, the storage medium read / write unit 24, the communication control unit 25, the work area 26, and the input unit 14, the CPU 21 is connected to the display unit 17 and the like.

  The device control program 22a includes a system program that controls the overall operation of the electronic dictionary device 10, Web servers 30 on the Internet N through the communication control unit 25, a user PC (Personal Computer) (not shown), and the like. In addition to storing a communication program for communication, a character recognition processing program for handwritten characters, a headword search processing program corresponding to the input search character string, and various explanatory information corresponding to the selected headword Various processing programs such as a reading / display processing program of meaning contents / example sentences) are stored.

  The entry word search processing program and various explanation information read / display processing programs stored as the device control program 22a are executed based on the dictionary database 22b stored in the main storage area 22, and the character recognition processing program is the same. It is executed based on the recognition database 22c stored in the main storage area 22.

  The dictionary database 22b has a plurality of types of dictionary contents 22b1, 22b2,. A plurality of dictionary data is stored in advance or downloaded.

  Each of the dictionary contents 22b1, 22b2,... Includes descriptive information such as a headword WO arranged in a predetermined order corresponding to each language, and a translation, meaning content, and example sentence of the headword associated with each headword WO. DI etc.

  In the recognition database 22c, for all the Chinese characters stored in the dictionary database 22b, data in which the character pattern of each Chinese character is associated with the component pattern constituting the Chinese character is stored.

  The work area 26 is a display data storage area corresponding to the display screen of the display unit 17, a dictionary search-processed entry word / description information storage area, a storage area necessary for handwritten character recognition processing, etc. It has various storage areas in which various data input / output to / from the CPU 21 according to the program 22a are stored as necessary.

  The storage area necessary for the handwritten character recognition process in the work area 26 includes a recognition candidate character group memory 26a, a fixed recognition partial memory 26b, an additional input partial installation position memory 26c, and the like.

  In the recognition candidate character group memory 26a, characters that have been character-recognized by collating the handwritten character data input by the touch panel 16 (16t) with the character patterns of each Chinese character stored in the recognition database 22c are collated. For example, up to 30 characters are stored as a recognition candidate character group in descending order of degree (fitness).

  In the fixed recognition partial memory 26b, when a portion (part) that matches the user-desired character is designated for the correction candidate character selected and displayed by the user from the recognition candidate character group, the correction candidate character The designated part (part) is stored as a fixed recognition part for character recognition correction.

  In the additional input portion setting position memory 26c, when the fixed recognition portion partially matching the user-desired character is displayed on the touch panel 16 (16t), the remaining portion is additionally input by handwriting and the character is recognized again. Information (additional input portion installation position) indicating whether the remaining portion is additionally installed at the upper, lower, left or right position of the fixed recognition portion is displayed on the touch panel 16 (16t) together with the display of the fixed recognition portion. It is stored in response to a user operation on the installation position designation buttons u (up), d (down), l (left), r (right) (see FIG. 8).

  Next, a handwritten character recognition process in the electronic dictionary device 10 having the handwriting input function having the above-described configuration will be described.

  In the handwritten character recognition process in the electronic dictionary device 10, based on a recognition candidate character group (26a) obtained by character recognition of a handwritten character, a matching portion with a desired character is defined as a fixed recognition portion (26b). The processing mode <recognized character re-input mode (OFF)> for selectively narrowing down and inputting the desired character while designating, and the matching portion with the desired character are designated as the fixed recognition portion (26b), and the desired There is a processing mode <recognized character re-input mode (ON)> in which the entire character or the remaining portion other than the fixed recognition portion (26a) is additionally input by handwriting again, character recognition is performed and the desired character is narrowed down and input. . The <recognized character re-input mode> (OFF / ON) is set in advance according to the user operation prior to the handwritten character recognition process.

  FIG. 3 is a flowchart showing handwritten character recognition processing (part 1) in the electronic dictionary device 10 having the handwriting input function.

  FIG. 4 is a flowchart showing handwritten character recognition processing (part 2) in the electronic dictionary device 10 having the handwriting input function.

  FIG. 5 is a diagram showing an operation display state of the touch panel 16 in <recognized character re-input mode (OFF)> associated with handwritten character recognition processing in the electronic dictionary device 10 having the handwriting input function.

  When dictionary content 22bn desired by the user stored in the dictionary database 22b is designated and the dictionary search mode is activated (step S1), a search keyword input screen (not shown) is displayed on the display unit 17. The handwritten character input areas 16L and 16R are displayed on the touch panel 16 (16t), and a state of waiting for input of handwritten characters is entered (step S2).

  At this time, as shown in FIG. 5A, in the touch panel 16 (16t), a central area and a right area which are horizontally divided into three are a left input area 16L and a right input area 16R for handwritten characters. The upper part of the left area is the recognized character display area 16N. In the recognized character display area 16N, the first recognition candidate in the recognition candidate character group (26a) obtained from the recognition database 22c by character recognition of the handwritten character input to the left input area 16L or the right input area 16R. Characters (characters with the highest matching (matching). The same applies hereinafter) are displayed for three characters.

  In the lower part of the recognized character display area 16N, when the first recognized candidate characters corresponding to each of handwritten characters exceeding three characters are arrayed and displayed in the recognized character display area 16N, the hidden recognized characters are scrolled. “←” button CL, “→” button CR for displaying, and “rewrite” button E for erasing handwritten characters displayed by handwriting input in the handwritten character input areas 16L, 16R and instructing rewriting. Is displayed.

  For example, in order to input a user-desired character “word”, as shown in FIG. 5A, the handwritten character “word” is input and displayed in the left input area 16L of the touch panel 16 (16t) (step S2). When the “recognition” key 14e is operated or handwriting input of the next character is started (step S3 (YES)), the handwritten input character “word” is stored in the recognition database 22c. Character recognition is performed by collating with the kanji character pattern (step S4).

  Then, for example, recognition candidate characters for 30 characters are stored in the recognition candidate character group memory 26a in descending order of matching (accuracy) acquired from the recognition database 22c, and the first recognition candidate character (for example, “false”) is stored. ) Is displayed in the recognized character display area 16N (step S5).

  If the recognized character “false” displayed in the recognized character display area 16N is an incorrect recognized character, and the recognized character “false” is touched with the pen P (YES in step S6), the recognition candidate Among the 30 recognition candidate characters stored in the character group memory 26a, 10 recognition candidate characters are extracted in descending order of matching (fitness). As shown in FIG. 5B, correction candidate characters are extracted. A list is displayed as the area 16TK (step S7).

  In the display state of the correction candidate character area 16TK, for example, if “correction” is selected and pen-touched as the correction candidate character closest to the user-desired character “word”, it is identified and displayed by the reverse display H ( In step S8 (YES), as shown in FIG. 5C, the selected correction candidate character “wrong” is displayed by being overwritten on the recognized character display area 16N, and is displayed in the left input area 16L. The image is enlarged and displayed (step S9).

  At this time, the correction candidate character “wrong” enlargedly displayed in the left input area 16L is the constituent elements a, b, c constituting the correction candidate character “wrong” stored in association with each other in the recognition database 22c. Are displayed with a dotted line.

  Here, regarding the correction candidate character “wrong” displayed in the left input area 16L in an enlarged manner, when the biased component “a” that matches the user-desired character “word” is pen-touched as a fixed recognition portion (step) S10), the biased component a is identified and displayed by the reverse display H and stored in the fixed recognition partial memory 26b (step S11).

  Then, it is currently determined whether the recognized character re-input mode is “ON” or “OFF”, and if it is determined that it is <recognized character re-input mode (OFF)> (step S12 (NO)). Among the 30 recognition candidate characters stored in the recognition candidate character group memory 26a, the fixed recognition parts (in this case, the words in the fixed recognition part memory 26b are stored in descending order of matching (matching degree). Ten recognition candidate characters including partial component a) are narrowed down and extracted, and are displayed as a list of correction candidate character regions 16TK as shown in FIG. 5D (step S13).

  In the display state of the correction candidate character area 16TK including the fixed recognition portion, when the user-desired character “word” is selected and pen-touched and identified and displayed by the reverse display H (step S8 (YES)), FIG. As shown in (E), the selected correction candidate character “word” is displayed by being overwritten in the recognized character display area 16N, and is enlarged and displayed in the left input area 16L (step S9). .

  Also in this case, the correction candidate character “word” enlarged and displayed in the left input area 16L is the component a, constituting the correction candidate character “word” stored in association in the recognition database 22c. b and c are displayed separated by a dotted line.

  In this way, with the user-desired character “word” being displayed in the recognized character display area 16N and the left input area 16L, handwriting input of the next character is performed in the right input area 16R as shown in FIG. When the operation is started or the “recognition” key 14e is operated, the desired character “word” displayed in the recognized character display area 16N is fixed and the search keyword input screen (FIG. 5) displayed on the display unit 17 is displayed. (Step S10 (NO) → S6 (NO) → S14).

  FIG. 6 is a diagram showing an operation display state (part 1) of the touch panel 16 in <recognized character re-input mode (ON)> associated with handwritten character recognition processing in the electronic dictionary device 10 having the handwriting input function.

  As in the case shown in FIGS. 5A to 5C, as shown in FIGS. 6A to 6C, first, a user-desired character “word” input by handwriting is erroneously recognized. When the character “false” is displayed in the recognized character display area 16N (steps S1 to S5) and the wrong recognized character “false” is touched with the pen P (step S6 (YES)), the recognition candidate Of the 30 recognition candidate characters stored in the character group memory 26a, 10 recognition candidate characters are extracted in descending order of matching (fitness) and displayed as a list of correction candidate character areas 16TK (step S7). ).

  In the display state of the correction candidate character area 16TK, for example, if “correction” is selected and pen-touched as the correction candidate character closest to the user-desired character “word”, it is identified and displayed by the reverse display H ( In step S8 (YES), as shown in FIG. 6C, the selected correction candidate character “wrong” is displayed by being overwritten on the recognized character display area 16N, and is displayed in the left input area 16L. The image is enlarged and displayed (step S9).

  Here, regarding the correction candidate character “wrong” displayed in the left input area 16L in an enlarged manner, when the biased component “a” that matches the user-desired character “word” is pen-touched as a fixed recognition portion (step) S10), the biased component a is identified and displayed by the reverse display H and stored in the fixed recognition partial memory 26b (step S11).

  Then, it is currently determined whether the recognized character re-input mode is “ON” or “OFF”, and if it is determined to be <recognized character re-input mode (ON)> (step S12 (YES)). As shown in FIG. 6D, only the fixed recognition part (in this case, the biased component a) touched in the left input area 16L is left and displayed, and the right input area 16R is blank. Then, a state of waiting for acceptance of handwritten characters is entered (step S15).

  Here, as shown in FIG. 6D, in the right input area 16R of the touch panel 16 (16t), the user-desired character “word” is input again and displayed (step S16), and the “recognition” key is displayed. When 14e is operated (step S17 (YES)), it is determined that the character input is for the right input area 16R (step S18 (YES)).

  Then, among the kanji characters stored in the recognition database 22c, handwriting input is again performed in the right input area 16R for only the character group including the fixed recognition portion (in this case, the prejudice component a). The user-desired character “word” is collated and recognized (step S19).

  Then, for example, recognition candidate characters for 30 characters are stored in the recognition candidate character group memory 26a in descending order of the degree of matching (relevance) of the character recognition, and as shown in FIG. A candidate character (for example, “word”) is displayed in the recognized character display area 16N (step S5).

  By re-inputting the user-desired character “word”, the correct recognized character “word” recognized only from the recognition character group including the fixed recognition portion (speaking component a) is recognized character display area 16N. When the handwriting input of the next character is started in the left input area 16L or the “recognition” key 14e is operated, the desired character “word” displayed in the recognized character display area 16N is displayed. Is determined and displayed on the search keyword input screen (not shown) displayed on the display unit 17 (step S6 (NO) → S14).

  FIG. 7 is a diagram showing an operation display state (part 2) of the touch panel 16 in <recognized character re-input mode (ON)> associated with handwritten character recognition processing in the electronic dictionary device 10 having the handwriting input function.

  For example, as in the case shown in FIG. 6D, only the fixed recognition part (word component a) is displayed in the left input area 16L of the touch panel 16 (16t) (step S15), and the right input area 16R is displayed. When a user-desired character “word” is input again by handwriting (step S16) and the character is recognized (step S17 → S18 → S19), FIG. The state displayed on the display area 16N is shown (step S5).

  Here, when the erroneous recognized character “review” displayed in the recognized character display area 16N is touched with the pen P (step S6 (YES)), the collated character is recognized in the step S19 and the recognition candidate character group. Of the recognition candidate character groups stored in the memory 26a, 10 recognition candidate characters are extracted in descending order of matching (matching) and displayed as a list as the correction candidate character region 16TK (step S7).

  If, for example, “consultation” is selected and pen-touched as a correction candidate character closest to the user-desired character “word” in the display state of the correction candidate character region 16TK (step S8 (YES)), FIG. As shown in FIG. 7 (B), the selected correction candidate character “consultation” is displayed by being overwritten on the recognized character display area 16N and enlarged and displayed on the left input area 16L (step). S9).

  Here, the correction candidate character “consultation” enlarged and displayed in the left input area 16L is added to the prejudice component a that matches the user desired character “word” and is already stored as a fixed recognition part. When the “mouth” -shaped component c is touched with a pen as a fixed recognition part (step S10), the “mouth” -shaped component c is also identified and displayed by the reverse display H in addition to the biased component a. And additionally stored in the fixed recognition partial memory 26b (step S11).

  If it is currently determined that it is <recognized character re-input mode (ON)> (step S12 (YES)), as shown in FIG. 7C, the fixed recognition portion selected in the left input area 16L. (In this case, only the biased component a and the “mouth” -shaped component c) are left and displayed, and the right input area 16R is left blank and is in a state of waiting for acceptance of handwritten characters (step S15).

  Here, as shown in FIG. 7C, in the right input area 16R of the touch panel 16 (16t), the user-desired character “word” is again inputted by hand (step S16), and the “recognition” key is displayed. When 14e is operated (step S17 (YES)), it is determined that the character input is for the right input area 16R (step S18 (YES)).

  Then, among the kanji characters stored in the recognition database 22c, only the character group including the fixed recognition portion (in this case, the word component “a” and the “mouth” component c)) is used as the target. The user-desired character “word” rewritten by handwriting in the input area 16R is collated and recognized (step S19).

  Then, for example, recognition candidate characters for 30 characters are stored in the recognition candidate character group memory 26a in descending order of the recognition degree (matching degree) of the character recognition, and as shown in FIG. A recognition candidate character (for example, “word”) is displayed in the recognized character display area 16N (step S5).

  By re-inputting the user-desired character "word", the correct recognized character "recognized only from the recognition character group including the fixed recognition portion (the word component" a "and the" mouth "component c)). When the word is displayed in the recognized character display area 16N, when the handwriting input of the next character is started in the left input area 16L or the “recognition” key 14e is operated, the recognized character display area 16N The desired character “word” displayed on the screen is confirmed and displayed on the search keyword input screen (not shown) displayed on the display unit 17 (step S6 (NO) → S14).

  FIG. 8 is a diagram showing an operation display state (part 3) of the touch panel 16 in <recognized character re-input mode (ON)> associated with handwritten character recognition processing in the electronic dictionary device 10 having the handwriting input function.

  As shown in FIG. 8A, only the fixed recognition part (in this case, the biased component a) is displayed in the left input area 16L, and the handwritten character input areas 16L and 16R are displayed with the right input area 16R blank. In the state of waiting for input (step S15), the remaining portion “た” obtained by removing the fixed recognition portion (speaking component a) from the desired character “word” is input by handwriting in the left input area 16L (step S15). S16) When the “recognition” key 14e is operated (step S17 (YES)), it is determined that the input is an additional handwriting input to the left input area 16L (step S20 (YES)).

  Then, among the kanji stored in the recognition database 22c, only the character group including the fixed recognition part (in this case, the word radical a), the fixed recognition part (the word component a) ) And the additional handwritten input portion “吾” are collated and the character “word” is collated (step S19).

  Then, for example, recognition candidate characters for 30 characters are stored in the recognition candidate character group memory 26a in descending order of the degree of matching (relevance) of the recognized characters, and as shown in FIG. The recognition candidate character “word” is overwritten and displayed in the recognized character display area 16N (step S5).

  With respect to the user-desired character “word”, the correct recognized character “word” which is recognized again by combining the portion “吾” additionally input by handwriting with the fixed recognition portion (word-constant component “a”) is the recognized character. When the handwriting input of the next character is started in the right input region 16R or the “recognition” key 14e is operated in the state displayed in the display region 16N, the desired character displayed in the recognized character display region 16N is displayed. The character “word” is confirmed and displayed on the search keyword input screen (not shown) displayed on the display unit 17 (step S6 (NO) → S14).

  On the other hand, similarly to the case shown in FIGS. 6A to 6C, first, an erroneously recognized character “false” is displayed in the recognized character display area 16N for the user-desired character “word” input by handwriting. In this state (steps S1 to S5), when the erroneous recognition character “false” is touched with the pen P (step S6 (YES)), the 30 characters stored in the recognition candidate character group memory 26a. Among the recognition candidate characters, 10 recognition candidate characters are extracted in descending order of matching degree (matching degree) and displayed as a list as the correction candidate character area 16TK (step S7).

  In the display state of the correction candidate character area 16TK, for example, if “correction” is selected and pen-touched as the correction candidate character closest to the user-desired character “word”, it is identified and displayed by the reverse display H ( Step S8 (YES)), as shown in FIG. 6C, the selected correction candidate character “wrong” is displayed by being overwritten on the recognized character display area 16N, and is displayed in the left input area 16L. The image is enlarged and displayed (step S9).

  Here, regarding the correction candidate character “wrong” displayed in the left input area 16L in an enlarged manner, when the biased component “a” that matches the user-desired character “word” is pen-touched as a fixed recognition portion (step) S10), the biased component a is identified and displayed by the reverse display H and stored in the fixed recognition partial memory 26b (step S11).

  If it is currently determined that it is <recognized character re-input mode (ON)> (step S12 (YES)), as shown in FIG. 8B, the fixed recognition portion that is pen-touched in the left input area 16L. (In this case, only the prejudice component a) is left and displayed, and the right input area 16R is left blank, waiting for acceptance of handwritten characters (step S15).

  At this time, each of the upper recognition unit (word component a) displayed on the left input region 16L is desired by the user at the upper end, lower end, left end, and right end of the left input region 16L. Additional input part installation position designation buttons u, d, l, and r are displayed for designating in which of the upper, lower, left and right directions the remaining part “吾” of the character “word” is installed.

  Here, for the character “word” desired by the user, the position where the remaining part “吾” is to be added to the fixed recognition part (word component a) currently displayed in the left input area 16L is on the right side. If the additional input portion installation position designation button r is identified and displayed (H) by a touch operation (steps S16 and S22), information on the installation position “right” of the additional input portion that has been identified and displayed (H) is added. It is stored and set in the input part installation position memory 26c (step S23).

  Then, as shown in FIG. 8C, in the right input area 16R, the part “吾” to be added to the fixed recognition part (word component a) displayed in the left input area 16L is handwritten. When input is performed (step S16) and the "recognition" key 14e is operated (step S17 (YES)), it is determined that the input is a character input to the right input area 16R (step S18 (YES)), and an additional input portion It is determined that the information on the additional input part installation position (in this case, “right”) is set in the installation position memory 26c (step S24 (YES)).

  Then, among the kanji characters stored in the recognition database 22c, only the character group including the fixed recognition part (in this case, the word component a) is added by handwriting input to the right input area 16R. Is added to the “right” of the fixed recognition part (word-constant component a) set in the additional input part installation position memory 26c, and the character “word” is collated and character recognition is performed. (Step S25).

  Then, for example, recognition candidate characters for 30 characters are stored in the recognition candidate character group memory 26a in descending order of the degree of matching (relevance) of the recognized characters, and as shown in FIG. A recognition candidate character (for example, “word”) is displayed in the recognized character display area 16N (step S5).

  In order to be additionally installed and matched with such a fixed recognition part (word bias component a), the installation position “right” is designated and the additional part “吾” is input, whereby the fixed recognition part (word bias) is specified. In the state where the correct recognized character “word” recognized only from the recognition character group including the component a) is displayed in the recognized character display area 16N, handwriting input of the next character is performed in the left input area 16L. When the operation is started or the “recognition” key 14e is operated, the desired character “word” displayed in the recognized character display area 16N is fixed and the search keyword input screen (FIG. 5) displayed on the display unit 17 is displayed. (Step S6 (NO) → S14).

  In the present embodiment, the additional input portion installation position designation buttons u, d, l, and r are displayed and selected after the fixed recognition portion is designated. However, the present invention is not limited to such an embodiment. For example, before the fixed recognition part is designated, the additional input part installation position designation buttons may be able to select u, d, l, and r.

  In the present embodiment, the installation position of the part additionally input by the additional input part installation position designation buttons u, d, l, and r is designated. However, the present invention is not limited to such an embodiment, and the fixed recognition part. For example, a button for designating the installation position of the fixed recognition part may be provided to designate the installation position of the fixed recognition part.

  Therefore, according to the handwritten character recognition function in the electronic dictionary device 10 having the above configuration, when a user-desired character (for example, “word”) is input by handwriting in the handwritten character input areas 16L and 16R of the touch panel 16 (16t), the recognition is performed. A first recognition candidate character (for example, “wrong”) in the recognition candidate character group (26a) that is collated and recognized based on the database 22c is displayed in the recognized character display region 16N and the handwritten character input regions 16L and 16R. When the component “word” a that matches the desired character “word” is selected and touched with respect to the recognized character “false” displayed in the handwritten character input areas 16L and 16R, the recognition candidate character group Of these, only the recognition candidate characters including the selected component “word” a as a fixed recognition portion are narrowed down and displayed as correction candidate characters (16TK). For this reason, it becomes possible to recognize a user desired character efficiently, without impairing the learning effect by handwriting input, and also when character recognition performance is not high.

  Further, according to the handwritten character recognition function in the electronic dictionary device 10 having the above-described configuration, when a user-desired character (for example, “word”) is input by handwriting in the handwritten character input areas 16L and 16R of the touch panel 16 (16t), the recognition is performed. A first recognition candidate character (for example, “wrong”) in the recognition candidate character group (26a) that is collated and recognized based on the database 22c is displayed in the recognized character display region 16N and the handwritten character input regions 16L and 16R. For the recognized character “false” displayed in the handwritten character input areas 16L and 16R, after selecting and touching the component “word” a that matches the desired character “word”, the desired character When “word” is input again by handwriting, the re-input handwritten character “word” is collated again from only the narrowed character group including the component “word” a as a fixed recognition portion in the recognition candidate character group. The first recognition candidate character is recognized and overwritten in the recognized character display area 16N. For this reason, it becomes possible to recognize a user desired character efficiently, without impairing the learning effect by handwriting input, and also when character recognition performance is not high.

  Further, according to the handwritten character recognition function in the electronic dictionary device 10 having the above configuration, the first handwritten character input areas 16L and 16R are recognized and displayed in accordance with the handwritten input of a user desired character (for example, “word”). For one recognition candidate character (for example, “consultation”), when there are a plurality of components that match the desired character “word” (“word” a and “mouth” c), the plurality of components are Select and touch as a fixed recognition part. Then, only the further narrowed-down character group including the plurality of constituent elements “word” and “mouth” as the fixed recognition portion among the recognition candidate character group is displayed as the correction candidate character (16TK). Alternatively, the re-input handwritten character “word” can be collated and recognized again only from the further narrowed-down character group. For this reason, it becomes possible to recognize a user desired character more efficiently.

  According to the handwritten character recognition function in the electronic dictionary device 10 having the above configuration, the recognized character “false” displayed in the handwritten character input areas 16L and 16R matches the desired character “word”. After selecting and touching the component “word” a, the installation position r “right” of the remaining portion “吾” of the desired character “word” for the selected component “word” is designated, and the remaining portion Enter “吾” in addition by handwriting. Then, the remaining part of the recognition candidate character group that is additionally written by hand with respect to the fixed recognition portion “word” from only the narrowed character group including the selected component “word” a as the fixed recognition portion. The character “word” in which “吾” is matched at the designated installation position r “right” is again collated and recognized, and the first recognition candidate character is overwritten and displayed in the recognized character display area 16N. For this reason, it becomes possible to recognize a user desired character efficiently, without impairing the learning effect by handwriting input, and also when character recognition performance is not high.

  Note that each processing method by the electronic dictionary device 10 having the handwriting input function described in the embodiment, that is, the handwritten character recognition process (part 1) shown in the flowchart of FIG. 3, and the handwritten character recognition shown in the flowchart of FIG. Each method such as processing (2) is a program that can be executed by a computer, such as a memory card (ROM card, RAM card, etc.), a magnetic disk (floppy disk, hard disk, etc.), an optical disk (CD-ROM). , DVD, etc.), and can be stored and distributed in an external storage medium 23 such as a semiconductor memory. The computer of the electronic device (10) having the touch panel 16 reads the program and the character recognition database (22c) stored in the external storage medium 23 into the storage device (22), and the operation is controlled by the read program. By this, the recognition function of the handwritten character demonstrated in the said embodiment is implement | achieved, and the same process by the method mentioned above can be performed.

  Further, the program and the character recognition database (22c) for realizing each of the above methods can be transmitted on the network N in the form of program codes, and the program and the character recognition database (22c) are connected to the network N. The above-described handwritten character recognition function can also be realized by taking in the computer of the electronic device (10) having the touch panel 16 by the communication control unit 25.

  Note that the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the invention at the stage of implementation. Further, each of the embodiments includes inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some constituent elements are deleted from all the constituent elements shown in each embodiment or some constituent elements are combined in different forms, the problems described in the column of the problem to be solved by the invention If the effects described in the column “Effects of the Invention” can be obtained, a configuration in which these constituent requirements are deleted or combined can be extracted as an invention.

The figure which shows the external appearance structure of the electronic dictionary apparatus 10 provided with the handwriting input function which concerns on embodiment of the character recognition processing apparatus of this invention. The block diagram which shows the structure of the electronic circuit of the electronic dictionary apparatus 10 provided with the said handwriting input function. The flowchart which shows the handwritten character recognition process (the 1) in the electronic dictionary apparatus 10 provided with the said handwriting input function. The flowchart which shows the handwritten character recognition process (the 2) in the electronic dictionary apparatus 10 provided with the said handwriting input function. The figure which shows the operation display state of the touch panel 16 in <recognized character re-input mode (OFF)> accompanying the handwritten character recognition process in the electronic dictionary apparatus 10 provided with the said handwriting input function. The figure which shows the operation display state (the 1) of the touch panel 16 in <recognized character re-input mode (ON)> accompanying the handwritten character recognition process in the electronic dictionary apparatus 10 provided with the said handwriting input function. The figure which shows the operation display state (the 2) of the touch panel 16 in <recognized character re-input mode (ON)> accompanying the handwritten character recognition process in the electronic dictionary apparatus 10 provided with the said handwriting input function. The figure which shows the operation display state (the 3) of the touch panel 16 in <recognized character re-input mode (ON)> accompanying the handwritten character recognition process in the electronic dictionary apparatus 10 provided with the said handwriting input function.

Explanation of symbols

10: Electronic dictionary device (character recognition processing device) with handwriting input function
DESCRIPTION OF SYMBOLS 11 ... Main body case 12 ... Lid case 13 ... Hinge part 14 ... Input part 14a ... Character input key 14b ... Dictionary designation / function key 14c ... "Translation / decision" key 14d ... Cursor key 14e ... "Recognition" key 15L, 15R ... Speaker 16 ... Touch panel 16t ... Transparent tablet 16L, 16R ... Handwritten character input area 16N ... Recognition character display area 16TK ... Correction candidate character area 17 ... Liquid crystal display unit 21 ... CPU
DESCRIPTION OF SYMBOLS 22 ... Main memory area 22a ... Apparatus control program 22b ... Dictionary content 22c ... Recognition database 23 ... External storage medium 24 ... Storage medium reading / writing part 25 ... Communication control part 26 ... Work area 26a ... Recognition candidate character group memory 26b ... Fixed recognition part Memory 26c ... Additional input part installation position memory 30 ... WEB server N ... Communication network P ... Pen H ... Inverted display a, b, c, ... Components u, d, l, r ... Additional part installation position designation buttons

Claims (5)

Handwritten character input means for inputting handwritten characters in response to a user operation;
Character recognition means for recognizing a handwritten character input by the handwritten character input means;
Recognized character display means for displaying characters of recognition candidates obtained by character recognition by the character recognition means;
A character portion specifying means for specifying a character portion of the recognition candidate displayed by the recognized character display means in accordance with a user operation;
Of the recognition candidate characters obtained by the character recognition by the character recognition means, a recognized character narrowing display means for narrowing and displaying characters including the character portion designated by the character portion designation means;
A character recognition processing device comprising: Handwritten character input means for inputting handwritten characters in response to a user operation;
A character recognition means for recognizing a handwritten character input by the handwritten character input means;
Recognized character display means for displaying characters of recognition candidates obtained by character recognition by the character recognition means;
A character portion specifying means for specifying a character portion of the recognition candidate displayed by the recognized character display means in accordance with a user operation;
Handwritten character re-input means for re-inputting handwritten characters in response to a user operation;
Of the recognition candidate characters obtained by the character recognition by the character recognition means, narrow down the characters including the character portion designated by the character portion designation means, and the handwritten character re-input means from among the narrowed-down characters Re-character recognition means for recognizing handwritten characters re-input by
A character recognition processing device comprising: The character portion designation means designates a plurality of portions of the recognition candidate characters displayed by the recognized character display means according to a user operation;
The character recognition processing device according to claim 1, wherein the character recognition processing device is a character recognition processing device. Positional designation means for designating the positional relation between the designated character portion and the additional portion, and the character portion designated by the character portion designation means is designated in accordance with a user operation. Prepared,
The re-character recognizing unit narrows down the characters including the character portion designated by the character part designating unit from among the recognition candidate characters obtained by the character recognition by the character recognizing unit, and among the narrowed-down characters The handwritten character re-input by the handwritten character re-input means according to the positional relationship specified by the positional relationship specifying means is added to the character portion specified by the character portion specifying means as the additional portion. Character recognition for the combined characters,
The character recognition processing device according to claim 2. A program for controlling a computer having a handwriting input unit,
The computer,
Handwritten character input means for inputting handwritten characters according to user operations from the handwriting input unit;
Character recognition means for recognizing handwritten characters input by the handwritten character input means;
Recognized character display control means for displaying characters of recognition candidates obtained by character recognition by the character recognition means on the handwriting input unit,
A character part designating unit for designating a character part of a recognition candidate displayed by the recognized character display control unit according to a user operation;
Recognized character narrowing display control means for narrowing down characters including the character portion designated by the character portion designation means from among the recognition candidate characters obtained by the character recognition by the character recognition means, and displaying them on the handwriting input unit,
A computer-readable program designed to function as a computer.

Images