JP2005275652A - Apparatus and method for processing input trajectory - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enhance the performance of input trajectory processing by solving the problems such that operations to change the setting of a gesture are complicated, a user is required to learn the operations and it is impossible to invalidate one or more gestures and respond the needs for stop recognition processing. <P>SOLUTION: Position coordinate data input to a position coordinate input part is transferred to a hand writing gesture recognition means, the recognition means digitizes a shape characteristic of the data, then, outputs the most analogous gesture code as a recognition result to a gesture code processing means with reference to a recognition dictionary. The gesture code processing means sends commands in response to the kinds of gesture codes transmitted from the dictionary to a command processing means. A gesture adaptation history storage means is connected to the gesture code processing means and the gesture adaptation history storage means stores the number of times of execution of gesture processing and the number of times of gesture cancellation after recognition of each gesture for each kind of gestures. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an input locus processing apparatus and an input locus processing method.

  In recent years, character recognition devices that input handwritten characters and commands using coordinate input means such as tablets have become widespread. An input trajectory for command input is called a gesture, but it is necessary to clearly distinguish a gesture from other inputs, and a gesture is set according to a user's purpose and the like in order to improve operation performance.

  Under such circumstances, the following devices have been proposed for gesture discrimination and setting.

  In the apparatus of Patent Document 1, a gesture shape and a corresponding command are displayed by a predetermined key operation, and a gesture change program is operated.

  In the apparatus of Patent Document 2, a macro gesture operation registration program is activated by a predetermined operation, and a gesture is registered by pressing a registration start button and a registration end button before and after the gesture input.

  In the apparatus of Patent Literature 3, an input locus longer than a predetermined value is determined as a gesture, and an input locus shorter than the predetermined value is determined as a character, and recognition processing is performed.

  In the apparatus of Patent Document 4, it is determined that the input locus outside the icon is a gesture and the input locus within the icon is an icon designation, and processing such as drag and drop is performed when the icon is designated.

Japanese Patent No. 2989084 JP 7-44311 A Japanese Patent No. 2905013 Japanese Patent No. 2584782

  In the device of Patent Document 1, the operation for changing the gesture setting is complicated, and the user has to learn the operation. In addition, it has not been possible to meet the need for disabling one or more gestures and stopping the recognition process.

  In the apparatus of Patent Document 2, since it is necessary to activate a macro gesture operation registration program at the time of gesture registration, it is necessary to provide a program activation menu in every scene where gestures are used, and the operation screen becomes complicated. Further, since the registration target is only a combination of gesture operations, the usage is limited.

  In the apparatus of Patent Document 3, since the line length is used as a clue, it is necessary to input a long line length of the gesture, so that the lightness of input is lost. Further, since it is difficult to distinguish large characters from gestures, the character size is limited and the application is limited.

  In the device of Patent Document 4, since the starting point of the input trajectory is used as a clue, the user's intention may not be detected correctly due to a user's parallax, blurring of the input trajectory, erroneous detection based on the accuracy of the device itself, and the like. Was insufficient.

  Furthermore, in the conventional pen input device, since there is no difference in the display of the trajectory between the gesture and the character or the like, the gesture and the other input cannot be easily distinguished.

  In addition, there are pen input devices that make it easy to determine the input by limiting the recognition target for a predetermined period, but in order to process the recognition target before the limit period, it is necessary to wait for the end of the period, and operability is reduced. It sometimes gave the impression that it was bad.

  As described above, the conventional apparatus has insufficient operation performance and other performance of the input trajectory processing.

  According to the present invention, the performance of input trajectory processing can be improved.

  The input trajectory processing device according to the present invention includes means for inputting an input trajectory of gesture as position coordinate data, recognition means for extracting a gesture corresponding to the input trajectory by matching the input trajectory and the gesture dictionary, Gesture adaptation history storage means for storing the adaptation state of the gesture extracted by the recognition means, and means for correcting the processing of the recognition means based on the stored contents of the gesture adaptation history storage means. As a result, the performance of the input trajectory process can be improved.

  In the input trajectory processing device according to the present invention, the adaptation status stored by the gesture adaptation history storage means is, for example, a cancellation process of the extracted gesture.

  The input trajectory processing device according to the present invention may further include means for changing the adaptation status for a gesture written after the recognition means recognizes a specific gesture.

  The input trajectory processing device according to the present invention may further include means for changing the shape of the gesture when the adaptation state reaches a predetermined state.

  The input trajectory processing device according to the present invention includes means for inputting an input trajectory of gesture as position coordinate data, recognition means for extracting a gesture corresponding to the input trajectory by matching the input trajectory and the gesture dictionary, Means for stopping recognition of the gesture of the recognition means for a predetermined period after the recognition means recognizes a specific gesture.

  The input trajectory processing device according to the present invention includes means for inputting an input trajectory of gesture as position coordinate data, recognition means for extracting a gesture corresponding to the input trajectory by matching the input trajectory and the gesture dictionary, Means for causing the recognition means to recognize only a predetermined character type when the recognition means recognizes a specific gesture.

  In the input trajectory processing device according to the present invention, at least two specific gestures are defined, the character types recognized by the first gesture are a character code and an edit gesture, and the character types recognized by the second gesture are It may be a character code.

  In the input trajectory processing device according to the present invention, the means for recognizing only the predetermined character type returns to the recognition of all the character types when a predetermined time has elapsed since the start of the recognition of only the predetermined character type. Also good.

  In the input trajectory processing device according to the present invention, when the recognition unit recognizes a specific gesture, the restriction of the character type by the unit that causes the recognition unit to recognize only a predetermined character type may be stopped.

  In the input trajectory processing device according to the present invention, the means for recognizing only the predetermined character type recognizes all the character types corresponding to these gestures when a plurality of different predetermined gestures are input before and after. It may be.

  The input trajectory processing method according to the present invention includes a step of inputting an input trajectory of a gesture as position coordinate data, a recognition step of extracting a gesture corresponding to the input trajectory by matching the input trajectory and a gesture dictionary, A gesture adaptation history storage step for storing the adaptation state of the gesture extracted by the recognition means; and a step for correcting the processing of the recognition means based on the stored contents of the gesture adaptation history storage means.

  The input trajectory processing method according to the present invention includes a step of inputting an input trajectory of a gesture as position coordinate data, a recognition step of extracting a gesture corresponding to the input trajectory by matching the input trajectory and a gesture dictionary, And a step of stopping recognition of the recognition step gesture for a predetermined period after a specific gesture is recognized by the recognition step.

  The input trajectory processing method according to the present invention includes a step of inputting an input trajectory of a gesture as position coordinate data, a recognition step of extracting a gesture corresponding to the input trajectory by matching the input trajectory and a gesture dictionary, And a step of recognizing only a predetermined character type in the recognition step when a specific gesture is recognized in the recognition step.

  Next, preferred embodiments of the input locus processing apparatus and the input locus processing method according to the present invention will be described.

[Input locus processing device]
In FIG. 1, an input locus processing apparatus according to a first embodiment of the present invention is applied to a portable information device 1 for electronic mail transmission / reception.

  The portable information device 1 has a liquid crystal display unit 2 and a camera unit 4. A transparent resistive film digitizer for inputting coordinates is attached to the surface of the liquid crystal display unit 2, and the pen 3 is pressed on the display screen. Thus, a handwritten locus can be input.

  The camera unit 4 is a digital camera, and includes an optical system such as a lens and an image sensor such as a CCD.

  The pen 3 is used for inputting soft buttons, a soft keyboard, and the like on the display screen in addition to inputting handwritten trajectory information.

  The portable information device 1 recognizes the input handwriting trajectory data by a known handwritten character recognition means or captures it as an image memo as it is.

  The recognized character is applied to the text of the electronic mail, and the image memo is the attached data of the electronic mail.

  An image taken by the camera unit 4 is displayed on the liquid crystal display unit 2 and can be attached to an electronic mail.

  The electronic mail including the image memo is sent to the personal computer 5 through the USB interface or the like or recorded on the recording medium 6 inserted into the slot 6S according to a user's instruction operation. The electronic mail stored in the recording medium 6 is sent to a personal computer 5 or a printer (not shown). The electronic mail created by the portable information device 1 is sent to another portable information device or personal computer through a local network or a public line.

  The portable information device 1 stores an operation history of data creation, editing, display, reproduction, etc. in the liquid crystal display unit 2 in a RAM.

  As shown in the block diagram of FIG. 2, the portable information device 1 includes an image display unit 2-2 including a liquid crystal display unit 2 (including a liquid crystal display element, a liquid crystal control circuit, and a display memory), and a resistance of the surface of the liquid crystal display unit A position coordinate input unit 2-1 configured by a film digitizer is included, and the position coordinate input unit 2-1 and the image display unit 2-2 are connected to the CPU 2-6 via the system bus 11.

  When the pen 3 is pressed on the position coordinate input unit 2-1, the coordinate data of the pressed position is read by the CPU 2-6, and when the handwriting trajectory is written by the pen 3, the trajectory is a position coordinate data group. , Read by the CPU 2-6.

  The CPU 2-6 displays image data or handwritten trajectory data on the image display unit 2-2 based on the read position coordinate data, or performs recognition processing by handwriting recognition processing means.

  A keyboard 2-3 is connected to the bus 11, and a key code such as alphanumeric characters is input by pressing a key, and the CPU 2-6 reads the key code.

  The camera unit 4 includes a control circuit, and an image photographed by the CCD is sent to the CPU 2-6 via the system bus 11 by the control circuit. The CPU 2-6 stores the received image in the RAM 7. The RAM 7 is used as a work area for storing image data and a recognition processing program.

  The CPU 2-6 is connected to the ROM 8 via the system bus 11, and the CPU 2-6 performs a processing operation according to a program stored in the ROM 8.

  The ROM 8 further includes a handwriting gesture recognition processing procedure, dictionary data of the shape of the gesture locus to be recognized, a CCD control program, a liquid crystal display control program, other processing procedures, and other application processing procedures for the portable information device 1. Etc. are stored.

  An image media I / F 9 is connected to the system bus 11 so that data of a compact flash card such as a digital camera can be read and written, and image data and the like can be stored.

  The USB interface unit 10 is connected to the system bus 11 and communicates with the personal computer 5 while being controlled by the CPU 2-6 to exchange handwritten locus data files and image data files.

  The system bus 11 is connected to a communication card 2-7 with a built-in controller connected to the PHS line and an antenna, and connected to the public line by passing a prescribed command to the controller in the communication card 2-7. Electronic mail transmission / reception can be executed.

  If the communication card 2-7 is changed to a local area network card or the like, it is possible to support a local area network. Transmission / reception of electronic mail is based on general protocols such as RFC882, RFC2231, RFC2822, and RFC2045-2049.

  3, in the functional block diagram of the portable information device 1, the position coordinate data input to the position coordinate input unit 2-1 corresponds to the handwriting gesture recognition means 3-1 (CPU 2-6 and handwriting gesture recognition processing procedure). The handwriting gesture recognition means 3-1 digitizes the shape feature of the position coordinate data and then refers to the recognition dictionary 3-2 (corresponding to the dictionary data of the shape of the gesture trajectory). The gesture code having the highest similarity is output to the gesture code processing means 3-3 as a recognition result.

  The gesture code processing means 3-3 sends a command corresponding to the type of gesture code sent from the recognition dictionary 3-2 to the command processing means 3-5. For example, when the input locus is most similar to the delete gesture in the recognition dictionary 3-2, a delete command is output, and the command processing unit 3-5 displays the processing result on the display unit 2-2.

  In the recognition dictionary 3-2, a standard gesture shape is digitized and stored, and is generated by the recognition algorithm in the handwriting gesture recognition means 3-1.

  The display unit 2-2 receives an instruction from the CPU 2-6 via the system bus 11, and displays image data, file icons, and the like on the screen.

  A gesture adaptation history storage unit 3-4 is connected to the gesture code processing unit 3-3, and the gesture adaptation history storage unit 3-4 performs, for each type of gesture, the number of times the gesture processing has been executed after each gesture recognition, The number of times that the gesture has been canceled is stored.

  A key and menu processing means 3-6 are connected to the gesture code processing means 3-3 and command processing means 3-5, and the key and menu processing means 3-6 perform command processing on commands input from the keyboard 2-3. Send to means 3-5. For example, when the delete key on the keyboard 2-3 is pressed, a delete command is sent to the command processing means 3-5 to execute the delete process.

  FIG. 4 is a table showing a gesture dictionary, which includes a gesture adaptation history and a gesture recognition stop flag.

  The gesture dictionary stores gesture shape data, a recognition code, a gesture processing type, a gesture adaptation history, and a gesture stop flag.

  Gesture shape data is numerical data obtained by normalizing position coordinate data of a registered locus of a gesture.

  The recognition code is a character code as an output of the recognition result, and corresponds to a command such as a deletion process or a copy process.

  The gesture adaptation history is extracted as the recognition result of the input trajectory for each gesture, and the number of times the corresponding command is executed (referred to as adaptation count), and the number of times that the gesture command is canceled (referred to as cancellation count). ) Are recorded sequentially, separated by commas. For example, if the number of executions is 4 and the number of cancellations is 1, “4, 1” is stored in the gesture adaptation history storage area.

  The gesture stop flag is a flag for excluding each gesture from the recognition target. When the gesture stop flag is stopped, the gesture does not compare with the input locus. Therefore, even if a gesture whose gesture stop flag is stopped is input, the gesture is not regarded as a recognition result. The gesture stop flag of the recognition target gesture is set to “in operation”.

  In FIG. 4, the recognition code “FF01 (0xFF01)”, the gesture processing type “deletion”, the gesture adaptation history “4, 1”, and the gesture stop flag “in operation” are stored for the M-type gesture. The recognition code “FF02 (0xFF02)”, the gesture processing type “copy”, the gesture adaptation history “8,0”, the gesture stop flag “in operation” are stored, and the recognition code “FF03 (0xFF03), gesture for the P-type gesture is stored. The processing type “paste”, the gesture adaptation history “2,1”, the gesture stop flag “in operation” are stored, and the recognition code “FF04 (0xFF04)”, the gesture processing type “cancel”, and the gesture adaptation for gestures similar to the seal History "1,0" The gesture stop flag “stopped” is stored.

  5, 6, 7, and 8 are examples of operation display screens representing gesture command processing.

  In the screen of FIG. 5, “test character string” is selected and inverted, and a C-type gesture is entered on the inverted character string. Referring to the gesture dictionary of FIG. 4, a “copy” command is executed, and “test character string” is copied to the copy buffer (RAM 7).

  Following the screen of FIG. 5, the operation of the screen of FIG. 6 is executed. In FIG. 6, a P-type gesture is entered, and a “paste” command is executed based on the gesture dictionary of FIG.

  At this time, since “test character string” is stored in the copy buffer by the operation of FIG. 5, “test character string” is pasted at the position of the paste gesture.

Following the screen of FIG. 6, the operation of the screen of FIG. 7 is executed. In FIG. 7, immediately after recognizing the paste gesture of FIG. 6, a gesture similar to a seal is entered, and a “cancel” command is executed based on the gesture dictionary of FIG. 4. Here, it is assumed that the gesture stop flag is “active”. This cancels the paste gesture that is the previous gesture, cancels the paste of the “test character string” executed in FIG. Test string "is deleted.

As a result of executing the canceling process of FIG. 7, when the canceling frequency in the gesture adaptation history becomes higher than the threshold, the gesture setting screen of FIG. 8 is automatically displayed. The cancellation frequency evaluation will be described later with reference to FIG. 9, and the gesture setting screen will be described later with reference to FIG.
[Input locus processing method]
A first embodiment of the input trajectory processing method performed by the portable information device 1 of FIG. 1 will be described. A program for executing the processing of the flowchart is stored in the ROM 8.

  When the position coordinate data of the handwriting gesture trajectory is input from the position coordinate input unit 2-1 in FIG. 3, gesture recognition is performed by a predetermined algorithm, and any one of the recognition codes in FIG. 4 is extracted. The gesture recognition processing algorithm is the same as the algorithm described in Patent Document 4, for example.

  When the recognition code is determined, gesture code processing is executed by the following steps in the gesture code processing means 3-3.

  Step S9-1: First, a work area and the like for gesture code processing are initialized.

  Step S9-2: Following step S9-1, it is determined whether or not the recognition code is canceled (0xFF04). If it is cancel, the process proceeds to step S9-4. If it is a gesture other than cancel, the process proceeds to step S9-3.

  Step S9-3: Command processing corresponding to the gesture type is executed based on the recognition code. That is, if it is a “delete” gesture, the selected character string is deleted, and if it is a “copy” gesture, the selected character string is stored in the copy buffer. For example, the character string stored in the copy buffer is pasted.

  Step S9-4: Cancel processing is performed on the immediately preceding gesture processing data. The immediately preceding gesture processing data is stored in step S9-11, which will be described later. For example, if paste processing has been executed immediately before, the pasted character string is deleted.

  Step S9-5: For the canceled gesture, “1” is added to the number of cancellations in the gesture adaptation history. For example, when the paste process is canceled, the gesture adaptation history storage area of the paste process is specified with reference to the recognition code (0xFF03) of the immediately preceding process. In FIG. 4, since the gesture adaptation history is “2, 1”, “1” is added to the number of cancellations to obtain “2, 2”.

  Step S9-6: “1” is added to the number of adaptations in the gesture adaptation history storage area for the gesture to be processed. For example, when the copy process is executed, the gesture adaptation history is “8, 0” in FIG. 4, and “1” is added to the adaptation count to obtain “9, 0”.

  Step S9-7: As a result of counting the number of cancellations, it is determined whether or not the cancellation frequency exceeds a threshold value. When the threshold value is exceeded, the process proceeds to step S9-8, and when the threshold value is not exceeded, the process proceeds to step S9-10.

  The cancellation frequency threshold is, for example, “number of adaptations”, and it is determined that the frequency is high when the number of adaptations is less than the number of cancellations in the gesture adaptation history. For example, if the adaptation number of paste processes = 2 and the number of cancellations = 3, it is determined that the cancellation frequency is high.

  The frequency evaluation method and the threshold setting method should be selected optimally according to the number of gesture types, the application that implements the gesture, and the like.

  Step S9-8: The gesture setting screen process of FIG. 8 is displayed.

  The gesture setting screen will be described in detail with reference to FIG.

  The gesture setting screen includes a gesture type 10-1, a recognition stop radio button 10-2, 10-3, a current gesture shape display box 10-4, a new gesture shape input box 10-5, a setting end button 10-6, A clear button 10-7 and a cancel button 10-8 are provided.

  The current gesture shape display box 10-4 displays the currently set gesture shape, and the user can confirm the current gesture shape together with the gesture type 10-1.

  The recognition stop radio buttons 10-2 and 10-3 indicate that the radio button 10-2 is ON and the radio button 10-3 is OFF. Radio buttons 10-2, 10-3 are displayed as black circles or white circles, and activated radio buttons are black circles. When the ON radio button 10-2 becomes a black circle, the displayed gesture is excluded from the recognition target, and when the OFF radio button 10-3 becomes a black circle, the displayed gesture becomes the recognition target.

  A white circle radio button 10-2 or 10-3 can be inverted to a black circle by pointing with the pen 3 or the like, and the other radio buttons 10-3 or 10-2 are inverted to a white circle.

  In the new gesture shape input box 10-5, the shape of a new gesture to be registered in place of the current gesture can be input with the pen 3 or the like.

  The setting end button 10-6 is a button for validating the setting performed in the recognition stop radio buttons 10-2 and 10-3 and the new gesture shape input box 10-5 and ending the gesture setting.

  When the clear button 10-7 is pressed, the locus input in the new gesture shape input box 10-5 is deleted.

  When the cancel button 10-8 is pressed, the settings made in the recognition stop radio buttons 10-2 and 10-3 and the new gesture shape input box 10-5 are invalidated, and the gesture setting is ended.

  Step S9-9: Following step S9-8, the gesture adaptation history is initialized. When a new gesture shape is registered in step S9-8, the adaptation history of the gesture is initialized, and the number of adaptations = the number of cancellations = 0.

  Step S9-10: The gesture code process is terminated and the work area used for the process is released.

  Step S9-11: The content of the gesture process processed this time is stored. That is, the type of gesture processed immediately before is stored and can be referred to in the next processing. For example, if the processed gesture is a “paste” gesture, the recognition code = 0xFF03 and the gesture processing type “paste” are stored.

  In FIG. 11, the process of the gesture setting screen of FIG. 10 is executed by the following steps.

  Step S11-1: The processing of the gesture setting screen is started, the work area and the like used by the setting screen processing are initialized, and the buttons and frames of the setting screen are displayed.

  Step S11-2: The gesture type display 10-1 is displayed, and the process proceeds to Step S11-3.

  Step S11-3: Refer to the gesture recognition stop state in the gesture dictionary and display the recognition stop radio buttons 10-2 and 10-3.

  Step S11-4: Refer to the gesture shape in the gesture dictionary, and display the gesture shape in the current gesture shape region 10-4. For example, in the case of a “paste” gesture, the “paste” gesture shape is read and displayed as shown in FIG.

  Step S11-5: It is determined whether or not the recognition stop radio button 10-2 or 10-3 is pressed by the pen 3. When the radio button 10-2 or 10-3 is pressed, the process proceeds to step S11-6. When the radio button 10-2 or 10-3 is not pressed, the process jumps to step S11-7.

  Step S11-6: The content of rewriting the gesture stop flag is recorded in the work area in response to pressing of the radio button 10-2 or 10-3.

  Step S11-7: It is determined whether or not there is a pen input in the new gesture input area 10-5. If there is an input, the process proceeds to step S11-8. If there is no input, the process jumps to step S11-10.

  When pen down is a position coordinate within 10-5, it is handled as a new input of gesture, and the position coordinate data until pen up is handled as registered locus data of gesture.

  Step S11-8: The input locus is drawn in the new gesture input area 10-5, the position coordinate data of the locus is stored in the buffer, and the process proceeds to Step S11-9.

  Step S11-9: It is determined whether or not the pen is up. When the pen is up, the process proceeds to step S11-10, and when the pen-down continues, the process returns to step S11-8.

  Step S11-10: It is determined whether or not the setting end button 10-6 has been pressed. When the setting end button 10-6 is pressed, the process proceeds to step S11-11. When the setting end button 10-6 is not pressed, the process proceeds to step S11-13.

  Step S11-11: A setting storing process to the gesture dictionary is performed. That is, the state of the gesture stop flag stored in the buffer in step S11-6 is stored in the gesture stop flag storage area of the gesture dictionary. In addition, the input trajectory position coordinates in the new gesture input area 10-5 stored in the buffer in step S11-8 are normalized, and feature points are extracted by a recognition algorithm, thereby generating dictionary data and storing the gesture shape. Store in the area.

  Step S11-12: It is determined whether or not the clear button 10-7 has been pressed. When the clear button 10-7 is pressed, the process proceeds to step S11-13. When the clear button 10-7 is not pressed, the process proceeds to step S11-14.

  Step S11-13: The input area is cleared. That is, the drawing of the locus input in the input area 10-5 is deleted, and the position coordinate data stored in the buffer is deleted.

  Step S11-14: It is determined whether or not the cancel button 10-8 is pressed. When the cancel button 10-8 is pressed, the process proceeds to step S11-15 to end the gesture setting screen process. When the cancel button 10-8 is not pressed, the process waits for input and returns to step S11-5.

  Step S11-15: The gesture setting screen process is terminated, and the secured work area and the like are released. Further, the setting screen closing process of FIG. 10 is performed.

  In the gesture setting screen process, since a gesture with a high cancellation frequency is automatically displayed, it is possible for the user to easily determine whether the gesture is necessary and to easily set the recognition stop. It is also easy to register the new shape by determining the suitability of the shape of the gesture.

  Thereby, the gesture shape and the recognition target gesture can be customized to be easy to use for the user.

Next, a second embodiment of the input locus processing apparatus and the input locus processing method according to the present invention will be described with reference to the drawings.
[Input locus processing device]
In the first embodiment, since the gesture setting screen is displayed when the frequency of canceling the gesture is high, it is impossible to positively exclude the gesture from the recognition target by the user's judgment.

  In the second embodiment, a gesture prohibiting gesture for prohibiting recognition of a gesture is provided, and a gesture input after inputting the gesture prohibiting gesture is excluded from recognition targets.

  12, the portable information device 1 to which the second embodiment is applied includes a position coordinate input unit 2-1, a display unit 2-2, a handwritten gesture recognition unit 3-1, a recognition dictionary 3-2, similar to FIG. 3. Command processing means 3-5, key and menu processing means 3-6 are provided.

  In the second embodiment, a gesture code processing means 12-2 is connected to the handwritten gesture recognition means 3-1 and the key / menu processing means 3-6, and a gesture prohibition setting storage means 12- is connected to the gesture code processing means 12-2. 1 is connected.

  When the gesture code processing means 12-2 recognizes the prohibited gesture, it sets a stop flag for the next input gesture. For this reason, the gesture code processing unit 12-2 sends the gesture recognition code input after the prohibited gesture to the gesture prohibition setting storage unit 12-1.

The gesture prohibition setting storage unit 12-1 sets the gesture stop flag of the gesture dictionary to “stopped” for the gesture of the recognition code sent from the gesture code processing unit 12-2.
The description of the same configuration in FIG. 12 as in FIG. 3 is omitted.
[Input locus processing method]
In FIG. 13, the input trajectory processing method implemented by the input trajectory processing apparatus of FIG. 12 is executed by the following steps. A program for executing the processing of the flowchart is stored in the ROM 8.

  A gesture is input, the gesture is recognized in the same manner as in the first embodiment, and a corresponding recognition code is output when the gesture stop flag of the recognized gesture is in operation. Thereafter, the processes after step S13-1 are executed.

  Step S13-1: The gesture code process is started, and a work area and the like for the gesture code process are initialized.

  Step S13-2: It is determined whether or not the input gesture is a prohibited gesture. If it is a prohibited gesture, the process proceeds to step S13-3. If it is not a prohibited gesture, the process proceeds to step S13-4.

  Step S13-3: Wait for input of the next input event, and determine whether or not the next input is a gesture. If the next input is a gesture, the process proceeds to step S13-5. If it is not a gesture, that is, if it is a keyboard input or the like, the process proceeds to step S13-6 and the process ends.

  Step S13-4: Similar to step S9-3 of the first embodiment, command processing corresponding to the gesture type is executed. Since this process is the same as that of the first embodiment, the description thereof is omitted.

  Step S13-5: Perform recognition setting prohibition processing for the input gesture. “Stopped” is stored in the gesture stop flag storage area of the dictionary of gestures input next to the prohibited gesture. Thus, the gesture is excluded from recognition targets in subsequent gesture recognition.

  Step S13-6: The gesture code process is terminated, and the work area and the like are released.

  As described above, by providing the gesture prohibited gesture, the user can easily set a specific gesture to be excluded from the recognition target.

  Similarly, if a gesture activation gesture is provided, the gesture stop flag of the gesture can be rewritten during operation by inputting a gesture that is stopped after inputting the gesture activation gesture.

Next, a third embodiment of the input locus processing apparatus and the input locus processing method according to the present invention will be described with reference to the drawings.
[Input locus processing device]
The third embodiment enables macro gesture setting and processing.

  14, the portable information device 1 to which the fourth embodiment is applied includes a position coordinate input unit 2-1, a display unit 2-2, a handwritten gesture recognition unit 3-1, a recognition dictionary 3-2, Command processing means 3-5, key and menu processing means 3-6 are provided.

  In the third embodiment, a gesture code processing means 14-2 is connected to the handwritten gesture recognition means 3-1 and the key / menu processing means 3-6, and a macro gesture macro storage means 14- is connected to the gesture code processing means 14-2. 1 is connected.

  When the gesture code processing unit 14-2 recognizes the macro setting gesture, the gesture code processing unit 14-2 performs a process of passing and storing a series of subsequent gestures and menu operations to the macro gesture macro storage unit 14-1.

  15, in the macro gesture macro storage structure, for example, for macro 1, all character string selection operations are stored in storage area 1, copy gestures are stored in storage area 2, and macro 2 is stored in storage area 1. An image file selection operation is stored, an image resolution conversion operation is stored in the storage area 2, and a mail transfer operation is stored in the storage area 3.

In FIG. 16, the macro gesture dictionary is composed of a gesture shape, a recognition code, a gesture processing type, and a macro storage area.
As the gesture shape, for example, a round 1 type gesture, a round 2 type gesture, a round a type gesture, a round B type gesture, and a round E type gesture are registered.

For example, the recognition code FF05, the gesture processing type macro gesture 1 and the macro storage area 1 are registered for the round type 1 gesture. For example, the recognition code FF06, the gesture processing type macro gesture 2 is registered for the round type 2 gesture. , And the macro storage area 2 are registered, for example, the recognition code FF07, the gesture processing type macro start gesture 1, and the macro storage area 1 are registered for the round a-type gesture, and for the round B-type gesture, For example, the recognition code FF08, the gesture processing type macro start gesture 2, and the macro storage area 2 are registered, and for example, the recognition code FF09 and the gesture processing type macro end gesture are registered for the round E-type gesture.
[Input locus processing method]
In FIG. 17, the input trajectory processing method implemented by the input trajectory processing apparatus of FIG. 14 is executed by the following steps. A program for executing the processing of the flowchart is stored in the ROM 8.

  When the gesture is input and the gesture is recognized in the same manner as in the first embodiment, the recognition code of the gesture dictionary is identified for each gesture, and the recognition code of the macro gesture in the macro gesture dictionary is identified, and step S17-1 The following processing is executed.

  Step S17-1: Macro gesture code processing is started. First, the work area for the gesture code processing is initialized.

  Step S17-2: It is determined whether or not the input gesture is a macro start gesture (recognition code 0xFF07 or 0xFF08). If it is a macro start gesture, the process proceeds to step S17-3. If it is a gesture other than the macro start gesture, the process proceeds to step S17-4.

  Step S17-3: First, the type of the macro start gesture is identified, and a storage location corresponding to the type is selected. For example, when the macro start gesture is the macro start gesture 1, the macro storage area 1 is used, and when the macro start gesture 2 is used, the macro storage area 2 is used.

  Step S17-4: Similar to step S9-3 of the first embodiment, processing of gestures other than the macro start gesture is executed.

  Step S17-5: Waits for the next input event to be input, and determines whether or not the next input is a macro end gesture (recognition code FF09). When the next input is a macro end gesture, the process proceeds to step S17-10 to end the process, and when it is not a macro end gesture, the process proceeds to step S17-6.

  Step S17-6: It is determined whether or not the input is a gesture. If the input is a gesture, the process proceeds to step S17-7. If the input is other than a gesture, the process proceeds to step S17-8.

  Step S17-7: The recognition code of the input gesture is stored in the storage area set in step S17-3. For example, if the macro start gesture is the macro start gesture 1, the gesture recognition codes are stored in order from the storage area 1 of the storage location of the macro 1.

  Step S17-8: It is determined whether or not the input is a menu selection instruction. If the input is a menu selection instruction, the process proceeds to step S17-9. If the input is not a menu selection instruction, the process returns to step S17-3 and waits for input.

  Step S17-9: Store the menu selection instruction in the macro storage area. For example, if the first macro start gesture is the macro start gesture 2, commands executed by menu selection are sequentially stored from the storage area 1 of the storage location of the macro 2.

  Step S17-10: The macro gesture code process is terminated, and the work area, dictionary file, etc. are released.

  When generating the macro 2 of FIG. 15, for example, in the image file display application in which the thumbnail of the image file is displayed, the macro start gesture 2 is first input in the image file thumbnail display area.

  Next, when an image file is selected with a selection gesture, the selection command is stored in the storage area 1 of the macro 2.

  Next, when image resolution conversion is selected from the menu, the image resolution conversion command is stored in the storage area 2 of the macro 2, and when mail transfer is selected from the menu, the mail transfer command is stored in the storage area 3 of the macro 2. Is remembered.

  Finally, when the macro end gesture is input, the storage process is ended.

  When the macro gesture 2 (recognition code 0xFF06) is written on the image thumbnail after the macro command of macro 2 in FIG. 15 is set, the command is recognized as the macro gesture 2 and stored in the macro storage area 2 Are executed sequentially. As a result, the image file under the macro gesture 2 is selected, and image resolution conversion and mail transfer are sequentially executed.

  In the third embodiment, the macro function of the gesture is realized by providing the macro gesture, the macro start gesture, and the macro end gesture, and more complicated and advanced input trajectory processing is possible.

Next, a fourth embodiment of the input locus processing apparatus and the input locus processing method according to the present invention will be described with reference to the drawings.
[Input locus processing device]
In the fourth embodiment, the gesture recognition process stop period is set by the recognition process change gesture, and the edit gesture or the like is not recognized during the gesture recognition process stop period. This facilitates normal drag and drop processing and character input operations.

  In the functional block diagram of FIG. 18, the portable information device 1 to which the fourth embodiment is applied has a position coordinate input unit 2-1 and a display unit 2-2 similar to those in FIG. 3, and the position coordinate input unit 2-1 Is connected with handwritten character recognition means 3-1. The handwritten character recognition means 3-1 includes a recognition dictionary 3-2, a recognized character type change instruction means 3-3, an input character editing processing means 3-4, a handwriting gesture processing means 3-5, and a drag drop processing means 3-6. Are connected, and the input character editing processing means 3-4, the handwriting gesture processing means 3-5, and the drag drop processing means 3-6 are connected to the display unit 2-2.

  A handwritten character or a handwritten locus of a gesture is input to the position coordinate input unit 2-1, and the input handwritten locus data is sent to the handwritten character recognizing unit 3-1.

  The handwritten character recognition unit 3-1 recognizes the input XY position coordinate data string of the handwriting trajectory, and outputs a character code, a gesture recognition code, an XY coordinate of the code input position, and the like. The recognition process of the handwritten character recognition unit 3-1 is processed by a known recognition algorithm, for example, an algorithm described in Japanese Patent No. 3155777. In this algorithm, the distance between the position coordinates of the start point, intermediate point, and end point of the input stroke and the position coordinate of the dictionary stroke is calculated to determine the similarity of all input strokes.

  The output character code and recognition code are recognized character type change instruction means 3-3, input character edit processing means 3-4, handwritten gesture processing means 3-5 and drag drop processing means 3 according to the type of the code. Passed to -6.

  In the recognition dictionary 3-2, an average value of input locus sample data of a plurality of users is registered as dictionary data by the handwritten character recognition unit 3-1. The dictionary data associates standardized input trajectory feature data with character codes and gesture recognition codes.

  When the recognition process change gesture is sent from the handwritten character recognition unit 3-1, the recognized character type change instruction unit 3-3 sends a control command to stop the gesture recognition process to the handwritten character recognition unit 3-1. A process stop timer is started, and when a specified time has elapsed, a gesture recognition process start control message is sent to the handwritten character recognition unit 3-1.

  When the character code is sent from the handwritten character recognizing unit 3-1, the input character editing processing means 3-4 inputs the character code in the character input area, displays the character font, and starts the character editing process. The processing of the input character editing processing means 3-4 is the same as that of a known handwritten character input device.

  The handwriting gesture processing means 3-5 starts the editing process at the input position when the gesture recognition code is sent from the handwritten character recognition unit 3-1.

  The drag drop processing means 3-6 is an operation by a pointing device, and performs a known drag drop process. For example, the file icon can be moved to another position by pressing the pen on the file icon in the display unit and moving the pen while the button is pressed. Move the file icon over the file folder icon and release the pen to move the selected file into the file folder.

  FIG. 19 shows the configuration of the recognition dictionary, which stores, for example, the correspondence between types of handwritten trajectories, output character codes, and processing examples after recognition.

  In FIG. 19, handwritten trajectories are classified into three types: handwritten characters, gesture shapes, and recognition process change gestures. As the handwritten character, the number 2 is given as an example, but English characters and kanji characters can naturally be processed.

  Gestures are used for editing input characters and the like, and gestures such as “insert” and “delete” are defined.

  The recognition processing change gesture has a shape that can be clearly distinguished from the writing method used when inputting normal characters, such as a shape in which a straight line is drawn from the bottom to the top.

  When the recognition process change gesture is input, the handwritten character recognition unit 3-1 outputs the recognition code “0xFFFF” to the recognized character type change instruction unit 3-3 as a result of the recognition process. Accordingly, the recognized character type change instruction unit 3-3 sends a control command for stopping the gesture recognition process to the handwritten character recognition unit 3-1, and stops the gesture recognition process for a certain period.

  During the gesture recognition process stop period, even if an edit gesture or the like is input, it is not recognized, so that a normal drag drop process or character input operation is facilitated. When the gesture recognition process stop period ends, the gesture recognition process is resumed.

  In the recognition dictionary of FIG. 19, a handwritten character “2”, a mountain-shaped similar editing gesture, and a vertical line-like recognition processing change gesture are defined as the handwriting locus type, and the recognition code “0xFF12”, “0xFFF0”, “0xFFFF”, processing example “2 is input in the character input area”, “what is stored in the copy buffer where the gesture is written is inserted”, “a certain period of time The gesture recognition process is forbidden during this period. Even if the editing gesture is written immediately after recognizing this locus, no editing operation is performed.

[Input locus processing method]
Next, a fourth embodiment of the input trajectory processing method implemented by the input trajectory processing apparatus of FIG. 18 will be described with reference to the drawings.

  20, FIG. 21, FIG. 22 and FIG. 23 are examples of display screens of the display unit 2-2 at the time of gesture input according to the fourth embodiment, and FIG. 24 is a flowchart showing an input trajectory processing method.

  In FIG. 24, the input trajectory processing method is executed by the following steps.

  A program for executing the processing of the flowchart is stored in the ROM 8.

  Step S24-1: The input position coordinate storage buffer and the like are initialized, pen input processing is started, and the process proceeds to Step S24-2.

  Step S24-2: The starting point coordinate position is checked. That is, using the first position pressed with the pen as the starting point, the coordinates of the starting point are acquired, and the process proceeds to step 24-3.

  Step S24-3: It is determined whether or not an icon representing a file or the like as shown in FIG. 20 exists at the start point position, and whether or not the gesture recognition process is stopped.

  FIG. 20 shows a display screen of a file browser that displays a file as an icon, and a known operation such as displaying the contents of the file by pressing the pen on the icon and moving it without moving is possible.

  If the start position icon does not exist and the gesture recognition process is not stopped, the process proceeds to step S24-5. If the start position icon exists or the gesture recognition process is not stopped, step S24-5 is performed. move on.

  That is, when there is an icon at the start point position, a drag drop process is executed. The fourth embodiment is adapted to input processing in a file browser, but other processing is adopted when adapting to a word processor or other applications.

  Step S24-4: A normal drag drop process or the like is executed. For example, the electronic mail 1. If you press the pen on the ETX file icon, move the pen over the trash can icon while holding down the pen, and release the pen, the electronic mail The ETX file is deleted.

  Step S24-5: An input locus recognition process is performed, and the process proceeds to step S24-6. The position coordinate data string from the input start point to the end point (a point separated from the point where the pen is pressed) is normalized, and feature points are extracted. The feature points of the input trajectory are compared with the feature points of the dictionary pattern in the recognition dictionary, and the recognition code of the dictionary pattern with the highest degree of matching is used as the recognition result.

  As the recognition processing algorithm, a known algorithm described in Japanese Patent No. 3155777 is used.

  For example, when a vertical line-like gesture from the bottom to the top is input, the matching degree of the recognition processing change gesture of FIG. 19 is the highest, and the dictionary pattern of the number “2” has a low matching degree. The recognition code having the highest degree of coincidence detected in this way is stored.

  Step S24-6: It is determined whether or not the input locus recognized in step S24-5 is a recognition process change gesture (recognition code 0xFFFF). If the input locus is a recognition process change gesture, the process proceeds to step S24-7, and if it is not a recognition process change gesture, the process proceeds to step S24-9.

  Step S24-7: A gesture prohibition flag is set, and a gesture prohibition process is performed. By setting the gesture prohibition flag, the process branches from step S24-3 to step S24-4 in the next input trajectory process, so that a drag drop process is always performed. Next, the process proceeds to step S24-8.

  Step S24-8: A gesture prohibition timer process is started and a gesture recognition process stop period is set. A gesture recognition processing stop period, for example, 1 minute is set in the timer, and when the set time has elapsed, processing is performed to reset the gesture prohibition flag. Next, the process proceeds to step S24-9.

  Step S24-9: A known gesture editing process or character input process is executed, and the process proceeds to Step S24-10. For example, if the number “2” is recognized, the number 2 is input in the character input area at the input position, and the font “2” is displayed on the screen. When a gesture is recognized, processing corresponding to the function of the gesture is executed. For example, when the “delete” gesture is recognized, the character or file icon at the input position is deleted.

  Step S24-10: The process of one input locus is terminated. Here, the work storage area or the like is released.

  With such a configuration, a gesture prohibiting gesture can be realized as one of the recognition processing change gestures by performing the pen input process.

  In this way, the gesture prohibiting gesture can be applied to a file browser application as shown in FIG. When the edit gesture is valid, for example, if the W-type “delete” gesture of FIG. 20 is written on the file icon “Sea Photo.JPG”, it is recognized as a “delete” gesture in step S24-5. Then, the “sea photo.JPG” file is deleted, and a display screen as shown in FIG. 21 is obtained.

  On the other hand, if the “delete” gesture is input as shown in FIG. 22 after inputting the gesture prohibition gesture, the “delete” gesture is not recognized, and the input trajectory is passed to the drag drop process in step S24-4. As a result, for example, it is processed as multiple selection of files.

  When a plurality of files are selected, for example, as shown in FIG. 23, “Sea Photo.JPG”, “Athletic Meet.MQV”, “Rule of Lord.JPG”, “take2.MP3”, “take55.MP3” , “Take3.MP3” is selected and displayed in a frame or the like.

  By adopting the gesture prohibition gesture in this way, the operability of input other than the gesture is remarkably improved. Further, the gesture prohibiting process itself can be executed without complicated operations such as a setting menu, and excellent operability can be realized even in a small device such as a PDA.

Next, a fifth embodiment of the input locus processing apparatus and the input locus processing method according to the present invention will be described with reference to the drawings.
[Input locus processing device]
In the fourth embodiment, gesture recognition is prohibited by implementing a gesture prohibition gesture. However, in the fifth embodiment, recognition of a specific character is prohibited when handwritten characters are input.

  Regarding the prohibition of recognition of a specific character, in the conventional apparatus, a recognition character type setting button or the like is installed around the character input area, and the recognition target character type is changed by pressing the button. In such a configuration, an installation area for the character type setting button is required, and application layout becomes difficult in a small-screen portable information device or the like. In addition, when a character type setting button is provided, the character input area and the character type setting button may not be placed close to each other due to layout limitations, and in this case, the operability deteriorates.

  In the fifth embodiment, the recognized character type restriction changing gesture and the recognized character type restriction releasing gesture are provided to enable the recognition character type restriction and release.

  In the functional block diagram of FIG. 25, the portable information device 1 to which the fifth embodiment is applied has a position coordinate input unit 2-1 and a display unit 2-2 similar to those in FIG. Is connected to a character cutting means 10-1 for handwritten character handwriting data. Connected to the character cutting means 10-1 for handwritten character handwriting data are a recognized character type restriction changing instruction means 10-2, a gesture recognition means 10-3, an alphanumeric recognition processing means 10-4, and a hiragana / kanji recognition processing means 10-5. An input character editing processing means 18-4 is connected to the gesture recognition means 10-3, alphanumeric recognition processing means 10-4 and hiragana / kanji recognition processing means 10-5, and further to the character editing processing means 18-4. A display unit 2-2 is connected.

  The character cutting means 10-1 for handwritten character handwriting data cuts the handwriting data input to the position coordinate input means 2-1 into handwriting data for one character according to the pen-up time and the character frame. The process of separating the handwriting data for one character is processed by a known algorithm. The handwriting data divided into one character is the gesture recognition means 10-3, alphanumeric recognition processing means 10-4, hiragana / kanji recognition processing means 10-5 according to the instruction of the recognized character type restriction change instruction means 10-2. Is passed on. The recognition processing of the alphanumeric recognition processing means 10-4 and the hiragana / kanji recognition processing means 10-5 is also based on a known recognition algorithm.

  The gesture recognizing means 10-3 normalizes the handwritten data for one character passed, extracts the feature points, matches the feature points with the feature points of the gesture handwritten data in the dictionary, and has the highest degree of coincidence. The high gesture recognition code is output to the input character editing processing means 18-4.

  When the gesture recognizing unit 10-3 recognizes the character type restriction changing gesture or the character type restriction releasing gesture, the recognition code is passed to the recognized character type restriction changing instruction unit 10-2 to change the restriction on the recognition target character type.

  The alphanumeric character recognition processing means 10-4 normalizes the handwritten data for the passed characters and extracts the feature points. The extracted feature points are matched with the feature points of alphanumeric handwriting data in a recognition dictionary (not shown), and the character code having the highest degree of matching is output to the input character editing processing means 18-4.

  The hiragana / kanji recognition processing means 10-5 normalizes the handwritten data for one character passed and extracts the feature points. The extracted feature points are matched with the feature points of the hiragana / kanji handwriting data in the recognition dictionary (not shown), and the character code having the highest matching degree is output to the input character editing processing means 3-4.

  The input character editing means 10-4 is the best match based on the degree of coincidence between the character code passed from the gesture recognition means 10-3, the alphanumeric recognition processing means 10-4, and the hiragana / kanji recognition processing means 10-5 and the recognition dictionary. A character code or recognition code having a high degree is used as a recognition result.

If the recognition result is a character code of alphanumeric characters and hiragana / kanji characters, display data of characters is sent to the display unit 2-2 in order to display a character font in the character input area.
If the recognition result is a gesture, editing processing corresponding to the gesture is performed, character deletion or character string insertion is performed, and the result is displayed on the display unit 2-2.

  FIG. 26 shows the configuration of a recognition dictionary, which stores, for example, the correspondence between types of input trajectories, output character codes, and processing examples after recognition.

  In FIG. 26, the recognition dictionary defines handwritten alphanumeric characters “at mark”, handwritten kanji & hiragana for “letter”, vertical wave type editing gesture, vertical line type character type restriction change gesture, and gamma type character type restriction release gesture. The identification code “0xFF20” is associated with the number “at mark”, and an example of processing after recognition “@ is input into the character input area”. The recognition code “0x5B57” is associated with the handwritten kanji and hiragana of “letter”. , The processing example after recognition “Character is input in character input area” is associated, the recognition code “0xFFF1” is used for the longitudinal wave type editing gesture, and the processing example after recognition “ When a gesture is written, the character is erased after recognition. ”Is associated with the recognition code“ 0xFFFF ”for the vertical line type restriction change gesture. An example of processing after recognition “Restrict recognition character type to alphanumeric characters” is associated, recognition code “0xFFFE” is used for a gamma type character type restriction release gesture, and an example of processing “Recognition restriction on recognized character type is released.” Are associated.

[Input locus processing method]
Next, a fifth embodiment of the input trajectory processing method performed by the input trajectory processing apparatus of FIG. 25 will be described with reference to the drawings.

  27 and 28 are screens for editing a new message for electronic mail. The input trajectory before recognition is displayed, and FIG. 28 displays the font after recognition. FIG. 29 is a flowchart showing the input trajectory processing method.

  In FIG. 29, the input trajectory processing method for one character is executed by the following steps.

  A program for executing the processing of the flowchart is stored in the ROM 8.

  Step S29-1: The input trajectory data processing for one character is started. That is, an XY position coordinate data string input to the position coordinate input unit 2-1 with the pen 3 is acquired, and data for one character is separated from the XY position coordinate data string by a known algorithm. As shown in FIG. 27, when handwritten characters are written in one square input frame, the XY position coordinate data string of each character frame is regarded as handwriting data for one character. Then, when the input to the character frame is started, the input of the previous character frame is confirmed and the recognition process is executed. Here, the input position coordinate storage buffer for one character is initialized.

  Step S29-2: The process of reading the position coordinate data of the input locus for one character is performed.

  The position coordinate data of the input handwriting for one character generated by the character cutting means 10-1 for the handwritten handwriting data is stored in the primary storage area of the RAM 7.

  When “W” is input by handwriting into the character frame as indicated by reference symbol A in FIG. 27, the handwriting of “W” is read as position coordinate data.

  Step S29-3: The state of restriction of the recognized character type is checked. When the recognition target character type is not limited, for example, alphanumeric characters, hiragana, katakana, and kanji are recognized character types. When the recognition target character type is limited to alphanumeric characters, the alphanumeric character is the recognition character type. .

  Step S29-4: A process for calling a recognition algorithm for the character type to be recognized is performed. If the limit value of the character type to be recognized is “0” (no limit), the gesture recognition processing function program in the gesture recognition means 10-3, alphanumeric recognition processing means 10-4, hiragana / kanji recognition processing means 10-5, The alphanumeric recognition processing function program and the hiragana / kanji recognition processing function program are called, and the input locus position coordinate data for one character is passed to each function program. If the limit value of the character type to be recognized is “1” (limited to alphanumeric characters), the gesture recognition processing function program and the alphanumeric recognition processing function program are called, and the input locus position coordinate data for one character is stored in each function program. To pass.

  When the recognition target character type is limited to alphanumeric characters, dictionary data matching of alphanumeric handwriting is performed using the alphanumeric character as the recognition character type, and the alphanumeric character code having the highest degree of matching is output.

  Step S29-5: A process of reading the degree of coincidence between the character code of the recognition result of each function program called in step S29-4 and the dictionary pattern is performed.

  From each recognition processing function program, the character code and the matching degree of the dictionary pattern having the highest matching degree with the input trajectory data are read, and the character code having the highest matching degree among them is output as the recognition result.

  Step S29-6: It is determined whether or not the character code of the recognition result is a character type restriction changing gesture. If the character code of the recognition result determined in step S29-5 is the character code of the character type restriction change gesture of FIG. 26 (OxFFFF), the process proceeds to step S29-7. Proceed to S29-8.

  Step S29-7: A recognition target character type restriction process is performed. The value of the character type restriction state storage area in the RAM is rewritten to “1”. Thereby, the input locus to be recognized next is limited to alphanumeric characters or gestures.

  For example, in the text input / edit application for handwritten electronic mail shown in FIG. 27, since only an alphanumeric page is required to input an Internet page address, “http://WWW.ca” or the like is input after inputting a character type restriction changing gesture. To do. Thereby, a desired address can be accurately input.

  In the examples of FIGS. 27 and 28, whether or not the character type restriction process is being executed is not displayed, but when the user should be informed that character type restriction is being implemented, the color of the input locus is different from the unrestricted state. Or the like. For example, the input locus is displayed in black in the unlimited state, and the input locus is displayed in red in the character type restriction processing state.

  Step S29-8: It is checked whether or not the character code of the recognition result is a character type restriction release gesture. If the character code of the recognition result is the recognition code 0xFFFF of the character type restriction release gesture, the process proceeds to step S29-9, and if it is a code other than the character type restriction release gesture, the process proceeds to step S29-10.

  Step S29-9: The value of the character type restriction state storage area in the RAM 7M is set to “0” in order to release the restriction on the recognition target character type. As a result, the input trajectory to be recognized next includes all alphanumeric characters, editing gestures, hiragana, katakana, and kanji.

  Step S29-10: If the character code is alphanumeric, kanji, hiragana, or katakana, the character code and the position coordinates of the input character are sent to the display unit 2-2, and the character font is displayed at that location. If it is an edit gesture, an edit process corresponding to the edit gesture is performed. For example, in the case of an “erase” gesture, the character font at the position where the gesture is written is erased.

  Step S29-11: The processing of the trajectory data for one character is terminated, and the data such as the work area is deleted. If the next character has been input, the process from step S29-1 is resumed. If no character has been input, the process waits for input.

  In the fifth embodiment, an information processing apparatus that implements a recognized character type change gesture can be realized, and a recognition rate can be improved by restricting character types, and an environment in which a desired character string can be easily input can be obtained.

  It is also possible to adopt a configuration in which the effective period of the character type restriction is set in advance and the restriction is automatically released when a predetermined time has elapsed.

  In the fifth embodiment, an example in which the character type to be recognized is limited to alphanumeric characters has been described. However, there are two-byte and one-byte alphanumeric character codes, which may be handled differently by a computer and have limited applications. is there. For alphanumeric characters, half-width codes are often used. Therefore, by adding a single-byte full-width code switching function to the recognized character type change gesture, more effective character type restriction is possible.

  For example, the character type restriction change gesture is changed to the character type restriction change + half-width gesture. If this gesture is recognized, the recognition target character type is the alphanumeric editing gesture, and the alphanumeric code to be output as the recognition result is changed. Restrict to single-byte alphanumeric codes.

  On the other hand, if this gesture is recognized as a character type restriction release gesture as a character type restriction release + full-width gesture, the recognition target character types are all types of alphanumeric characters, hiragana katakana kanji, and edit gestures. Numerals, hiragana, katakana, and kanji codes are double-byte codes.

  During the process of step S29-10, a value of “0” or “1” is stored in the character type restriction state storage area in the RAM 7, and if it is “0”, the alphanumeric code is a full-width code, and if “1”, it is English. A single-byte code of a numeric code and a rule may be set.

  When a full-width character code is stored in the character code of the recognition dictionary of the fifth embodiment, it is not necessary to store the half-width code if the half-width character code is appropriately generated by a character code conversion table or the like.

  In the fifth embodiment, the description has been made with the configuration in which the recognition algorithm is prepared for each character type. However, the recognition dictionary is prepared for the character type, and one recognition algorithm is used, according to the limit value of the recognition target character type, A recognition dictionary to be read may be selected.

Next, a sixth embodiment of the input trajectory processing method implemented in the portable information device 1 of FIG. 25 will be described with reference to the drawings.
[Input locus processing method]

  In the fifth embodiment, the mode for recognizing all character types to be recognized and the mode for recognizing alphanumeric characters and editing gestures are changed by gestures. In the sixth embodiment, the number of modes is further increased. For example, gesture 1 for limiting the character type to be recognized only to alphanumeric characters, gesture 2 for limiting the character type to be recognized to katakana, gesture 3 for limiting to hiragana and kanji, and gesture 4 for returning to the mode for recognizing all character types And provide.

  In FIG. 30, the sixth embodiment is executed by the following steps following steps S30-1 to S30-5 corresponding to steps S29-1 to S29-5 of the fifth embodiment.

  Step S30-6: It is checked whether the character code of the recognition result is any character type restriction changing gesture. When the character code is gesture 1, the process proceeds to step S30-7. When the character code is gesture 2, the process proceeds to step S30-8. When the character code is gesture 3, the process proceeds to step S30-9. . If the character code is not a character type restriction changing gesture, the process proceeds to step S30-13.

  Step S30-7: It is determined whether or not the previous character to be recognized is gesture 2 or gesture 3. When it is gesture 2 or gesture 3, the process proceeds to step S30-10, and when it is neither gesture 2 nor gesture 3, the process proceeds to step S30-11.

  Step S30-8: It is determined whether or not the previous character to be recognized is gesture 1 or gesture 3. When it is gesture 1 or gesture 3, the process proceeds to step S30-10, and when it is neither gesture 1 nor gesture 3, the process proceeds to step S30-11.

  Step S30-9: It is determined whether or not the previous character to be recognized is gesture 1 or gesture 2. When it is gesture 1 or gesture 2, the process proceeds to step S30-10, and when it is neither gesture 1 nor gesture 2, the process proceeds to step S30-11.

  Step S30-10: Two types of character types are set as recognition targets when two different types of character type restriction change gestures are input before and after. By enabling such OR processing, various recognition targets can be set, and the recognition rate can be increased while maintaining processing efficiency.

  Further, by finely setting the restrictions on the recognized character type, the degree of freedom of the gesture shape is increased, and a gesture shape that is easy to write, such as a relatively simple shape, can be employed. This can prevent gesture input errors and misrecognition.

  Step S30-11: Similar to step S29-7 in the fifth embodiment, a process is performed in which only one character type is recognized.

  Step S30-12: Similarly to step S29-9 of the fifth embodiment, the character type restriction is released.

  Step S30-13: Character fonts are displayed as in step S29-10 of the fifth embodiment.

  Step S30-14: The process ends in the same manner as in step S29-11 of the fifth embodiment.

By enabling OR setting for each recognition target character type restriction in steps 30-7 to S30-10, the following various character type restriction settings are possible.
When a restricted gesture 1 is recognized and then a character is entered: it is recognized restricted to alphanumeric characters.
When the restriction gesture 2 is recognized and a character is input next: It is recognized as limited to katakana.
When the restricted gesture 3 is recognized and a character is input next: it is recognized limited to Hiragana / Kanji.
When the restricted gesture 4 is recognized and then a character is input: All character types to be recognized are recognized.
When a restriction gesture 1 is recognized, then a restriction gesture 2 is recognized, and then a character is input: recognition is limited to alphanumeric characters and katakana.
When a restricted gesture 2 is recognized, then a restricted gesture 3 is recognized, and then a character is input: it is recognized limited to katakana and hiragana kanji.
When a restriction gesture 1 is recognized, then a restriction gesture 3 is recognized, and then a character is input: recognition is restricted to alphanumeric characters and hiragana kanji.

  In the fourth and fifth embodiments, the recognition target is limited in units of general character types such as hiragana katakana. In the seventh embodiment, so-called user customization of the recognition target is realized in which the recognition target is limited with a higher degree of freedom.

  For example, one or a plurality of user customized restricted character type gestures are defined, and each user customized restricted character type gesture is associated with an arbitrary character or character string. Then, by inputting a character or the like to be recognized after the user customization restricted character type gesture, the character or the like can be reliably recognized.

  Correspondence between the user customization restricted character type gesture and the character or the like is executed by application software or the like.

  For example, when one user-customized restricted character type gesture is associated with “123., high” and “123., high” is to be recognized, “123., high” follows the user-customized restricted character type gesture. "Is easy to recognize" 123., high ".

  This allows customization to an easy-to-use recognition device, and is particularly effective for users who have a low recognition rate in the standard state.

It is a figure which shows the portable information device to which Example 1 of the character recognition apparatus which concerns on this invention is applied. (Example 1) It is a block diagram of the portable information device of FIG. (Example 1) It is a functional block diagram of the portable information device of FIG. (Example 1) It is a block diagram of the gesture dictionary with which the portable information device of FIG. 1 was equipped. (Example 1) It is a figure which shows the screen of the character input state in the portable information device of FIG. (Example 1) It is a figure which shows the screen of the other character input state in the portable information device of FIG. (Example 1) It is a figure which shows the screen of character input cancellation operation in the portable information device of FIG. (Example 1) It is a figure which shows the screen displayed when character input cancellation operation is performed frequently in the portable information device of FIG. (Example 1) It is a flowchart which shows Example 1 of the character recognition method implemented in the portable information device of FIG. (Example 1) It is a figure which shows the gesture setting screen in the portable information device of FIG. (Example 1) It is a flowchart which shows the process of the gesture setting of FIG. (Example 1) It is a functional block diagram which shows the portable information device to which Example 2 of the character recognition apparatus which concerns on this invention is applied. (Example 2) It is a flowchart which shows Example 2 of the character recognition method implemented in the portable information device of FIG. (Example 2) It is a functional block diagram which shows the portable information device to which Example 3 of the character recognition apparatus which concerns on this invention is applied. Example 3 It is a block diagram of the macro storage area in the character recognition apparatus of FIG. Example 3 It is a block diagram of the macro gesture dictionary with which the character recognition apparatus of FIG. 14 was equipped. Example 3 It is a flowchart which shows Example 3 of the character recognition method implemented in the portable information device of FIG. Example 3 It is a functional block diagram which shows the portable information device to which Example 4 of the character recognition apparatus which concerns on this invention is applied. (Example 4) It is a table | surface which shows the example according to the shape for every kind of handwritten locus | trajectory in the character recognition apparatus of FIG. 18, an output code, and recognition. (Example 4) It is a figure which shows the display screen in the character recognition apparatus of FIG. (Example 4) It is a figure which shows the other display screen in the character recognition apparatus of FIG. (Example 4) It is a figure which shows the further another display screen in the character recognition apparatus of FIG. (Example 4) It is a figure which shows the further another display screen in the character recognition apparatus of FIG. (Example 4) It is a flowchart which shows Example 4 of the character recognition method implemented in the portable information device of FIG. (Example 4) It is a functional block diagram which shows the portable information device to which Example 5 of the character recognition apparatus which concerns on this invention is applied. (Example 5) It is a table | surface which shows the process example after the shape for every kind of handwritten locus | trajectory in the character recognition apparatus of FIG. 25, an output code, and recognition. (Example 5) It is a figure which shows the display screen in the character recognition apparatus of FIG. (Example 5) It is a figure which shows the other display screen in the character recognition apparatus of FIG. (Example 5) It is a flowchart which shows Example 5 of the character recognition method implemented in the character recognition apparatus of FIG. (Example 5) It is a flowchart which shows Example 6 of the character recognition method implemented in the character recognition apparatus of FIG. (Example 6)

Explanation of symbols

1 Mobile Information Terminal 2-1 Position Coordinate Input Unit 2-2 Image Display Unit 2-3 Keyboard 2-6 CPU
2-7 Communication Card 3 Pen 3-1 Handwritten Gesture Recognition Unit 3-2 Recognition Dictionary 3-3 Gesture Code Processing Unit 3-4 Gesture Adaptation History Storage Unit 3-5 Command Processing Unit 4 Camera Unit 5 Personal Computer 6 Transfer File 7 RAM
8 ROM
9 Image media interface 10 USB interface 11 System bus

Claims (15)

Means for inputting a gesture input locus as position coordinate data;
Recognizing means for extracting a gesture corresponding to the input locus by matching the input locus with a gesture dictionary;
Gesture adaptation history storage means for storing adaptation states of gestures extracted by the recognition means;
Means for correcting the processing of the recognition means based on the stored contents of the gesture adaptation history storage means;
An input trajectory processing device. 2. The input trajectory processing apparatus according to claim 1, wherein the adaptation status stored in the gesture adaptation history storage means is cancellation processing of the extracted gesture. 3. The input trajectory processing apparatus according to claim 1, further comprising: a unit that changes the adaptation state of a gesture written after the recognition unit recognizes a specific gesture. 4. 4. The input trajectory processing apparatus according to claim 1, further comprising means for changing a shape of the gesture when the adaptation state reaches a predetermined state. Means for inputting a gesture input locus as position coordinate data;
Recognizing means for extracting a gesture corresponding to the input locus by matching the input locus with a gesture dictionary;
Means for stopping recognition of the gesture of the recognition means for a predetermined period after the recognition means recognizes a specific gesture;
An input trajectory processing device. Means for inputting a gesture input locus as position coordinate data;
Recognizing means for extracting a gesture corresponding to the input locus by matching the input locus with a gesture dictionary;
Means for causing the recognition means to recognize only a predetermined character type when the recognition means recognizes a specific gesture;
An input trajectory processing device. The at least two specific gestures are defined, the character type recognized by the first gesture is a character code and an edit gesture, and the character type recognized by the second gesture is a character code. Item 7. The input trajectory processing device according to Item 6. 8. The input locus according to claim 6, wherein the means for recognizing only the predetermined character type returns to the recognition of all the character types when a predetermined time elapses after starting the recognition of only the predetermined character type. Processing equipment. 9. The input trajectory process according to claim 6, wherein when the recognition unit recognizes a specific gesture, the restriction of the character type by the unit that causes the recognition unit to recognize only a predetermined character type is stopped. apparatus. The means for recognizing only the predetermined character type recognizes all character types corresponding to these gestures when a plurality of different predetermined gestures are input before and after each other. The input locus processing device according to any one of the above. Inputting a gesture input locus as position coordinate data;
A recognition step of extracting a gesture corresponding to the input trajectory by matching the input trajectory with a gesture dictionary;
A gesture adaptation history storage step for storing an adaptation state of the gesture extracted by the recognition means;
Modifying the processing of the recognition means based on the stored contents of the gesture adaptation history storage means;
An input trajectory processing method comprising: Inputting a gesture input locus as position coordinate data;
A recognition step of extracting a gesture corresponding to the input trajectory by matching the input trajectory with a gesture dictionary;
A step of stopping recognition of the recognition step gesture for a predetermined period after a specific gesture is recognized by the recognition step;
An input trajectory processing method comprising: Inputting a gesture input locus as position coordinate data;
A recognition step of extracting a gesture corresponding to the input trajectory by matching the input trajectory with a gesture dictionary;
A step of recognizing only a predetermined character type in the recognition step when a specific gesture is recognized in the recognition step;
An input trajectory processing method comprising: A computer-executable program including program code for causing a computer to execute the input trajectory processing method according to claim 11. A computer-readable storage medium storing the computer-executable program according to claim 14.

Images