JP2007086309A - Voice synthesizer, voice synthesizing method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that although PIN(personal identification number) and ID numbers are prohibited to read in a prohibit list when reading a text and outputing it in voice, the PINs are always changed for safety and their registration is a troublesome load each time on the user or the system, the read control on XML tag reading is not simple, and only the text maker can make the control information without allowing the user to make any control to meet the situations. <P>SOLUTION: This voice synthesizer which converts the input text into a voice signal and outputs it in voice creates a pattern morpheme of the information that is not desired to read depending on the use situation by the user, such as PINs, telephone numbers, and card numbers, by using a pattern morpheme creating means, and skips over the information to which a pattern morpheme is created in the read content converter, or changes it into other contents, such as a beep tone or muting. It makes it possible to decide whether to change the text content to read in the read control signal input or to read it without changing the content depending on the situations of the use by the user. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an information device such as a mobile phone, a PDA (Personal Digital Assistant), a personal computer, an in-vehicle device such as a car navigation system, an ETC (Electronic Toll Collection System), an ATM (automatic cash dispenser), a CD (cash dispenser). The present invention relates to a text analysis technology applied to office equipment such as a machine, and relates to a text-to-speech synthesizer that reads a text analysis result, a method thereof, and a program for causing a computer to implement the method.

  Synthesizing speech signals artificially from arbitrary sentences is called text-to-speech synthesis. Text-to-speech synthesis is generally performed in three stages: a language processing unit (text analysis), a phoneme processing unit (prosodic setting), and a speech synthesis unit. The input text is first subjected to morphological analysis and syntactic analysis in the language processing unit, and then subjected to accent and intonation processing in the phonological processing unit, so that the phoneme environment such as phonological symbol, pitch length, duration length, etc. Information is output. Then, based on the phoneme environment information, a speech unit registered in the speech unit dictionary is selected. Finally, the speech is synthesized from the speech unit selected by the speech synthesizer and information such as the phoneme symbol, the pitch length, and the duration time.

In the conventional speech synthesizer, when the input sentence contains a term whose pronunciation output is not valid (discriminatory term, etc.), the pronunciation output is prohibited in the reading prohibition term table in order to prohibit the pronunciation of this type of term. The words that are prohibited from being read out are stored in advance. When the text that is the input sentence is input, the reading prohibition term judging means cuts out the input text in units of words, searches the reading prohibition table, and the words included in the input text are read prohibition terms. Determine whether or not. Then, the pronunciation prohibiting means prohibits pronunciation of a word corresponding to the reading prohibited term based on the determination result of the reading prohibited term determining means. For example, the pronunciation of a prohibited word is prohibited by generating a beep sound or silence in place of the prohibited word. Moreover, it is converted into a substitution expression and pronounced. (Disclosed in Patent Document 1)
Further, a method is disclosed in which XML (Extended Markup Language) tags are described to perform control such as reading out different from the original text content or not reading out anything (Non-Patent Document 1). In this method, a text creator embeds a control tag for designating reading and skipping in a sentence so that the intended creator reads out in advance.

JP-A-5-165486 (first page to fifth page, FIG. 1) Microsoft “SpeechSDK” Version 5.1

The conventional speech synthesizer disclosed in Patent Document 1 is configured as described above, but has the following problems. For example, consider a case where a sentence such as “Your PIN is 1234” or “Your ID is abcdefg” is entered as text. Even if a word such as “password” or “ID number” and “1234” of “password itself” or “abcdefg” of “ID number itself” are read out independently without being related, “password” or “ID” “No.” is a general noun that has no particular problem in reading, and “1234” and “abcdefg” are merely a list of numbers and alphabets, and do not become a big problem. However, when “PIN is 1234” is read out, it can be a big problem because it is associated with “PIN = (equal) 1234”.
In the conventional speech synthesizer, only words registered in the reading prohibition list can be prohibited from reading out. Therefore, it is necessary to register the above “1234” and “abcdefg” in the reading prohibition list. In addition to enormous numbers of combinations, it is necessary to always change the password for security, and registration each time places a burden on the user or system, and cannot cope with the above problems.

In addition, in the conventional speech synthesizer disclosed in Non-Patent Document 1, speech reading control is performed using XML tags, but information to be sent out other than the text body is necessary for the reading control, which is not simple. . Furthermore, the control information can be set only by the text creator, and can be controlled only by the text creator's intention, and cannot be controlled by the user of the speech synthesizer.
For example, a text writer may specify that the phone number range should be “highlighted” as control information because he / she wants to hear the phone number without fail. Depending on the usage situation, it may be personal information and you may not want it to be read out.

  The present invention has been made to solve the above-described problems. Depending on the use situation of the user, information that is not desired to be read out such as a personal identification number, a telephone number, and a card number is skipped, or a beep or The purpose is to make it possible to read out after changing to another content such as silence.

A speech synthesizer according to the present invention uses a text dictionary as input, finds, reads, and uses a language dictionary having accent-type information, and divides it into analysis results including reading information and accent information;
Prosody control means for generating prosodic information for controlling intonation and rhythm based on reading information and accent information obtained by text analysis means,
Based on the reading information and accent information obtained by the text analysis means, and the prosody information obtained by the prosody control means, a segment selection means for selecting a phoneme from an acoustic dictionary storing phonemes;
In a speech synthesizer having a speech synthesizer that creates a synthesized sound in accordance with the prosodic information obtained by the prosodic control means, the phoneme selected by the segment selection means
A pattern morpheme generation unit that generates a morpheme that matches a pattern defined in the reading control rule based on a reading control rule for extracting a specific pattern during the processing of the text analysis unit;
Reading contents changing means for changing the reading contents of the input text by changing the reading contents of the morpheme obtained by the pattern morpheme generation means,
There is a reading control signal input means for inputting a reading control signal as to whether or not to execute the reading contents change of the reading contents changing means.

  According to the speech synthesizer according to the present invention, information that is not desired to be read out depending on the use situation of the user, such as a personal identification number, a telephone number, and a card number, is generated by the pattern morpheme generating unit, and the reading content changing unit Read the morpheme-generated information or change it to another content such as a beep or silence, and change the content of the input text according to the user's usage status with the reading control signal input means, or do not change the content Since it is possible to select whether to read out, there is an effect that it can be used in accordance with the usage status of the user.

Embodiment 1 FIG.
The best mode for carrying out the present invention will be described below with reference to the drawings.
FIG. 1 is a block diagram showing a configuration of a speech synthesizer for realizing a speech synthesis method according to an embodiment of the present invention. In FIG. 1, 1 is a text input terminal for inputting text. A text analysis unit 2 analyzes text input from the text input terminal 1 and generates language information necessary for reading, accent information, and prosody control. Here, the processing in the text analysis unit 2 is performed using a known morphological analysis algorithm such as the longest match, the minimum number of clauses method, and the minimum cost method (for example, Kimiaki Shudo, Kenji Yoshimura “Japan Word structure and its analysis ", Information Processing Vol.27, No.8, pp.947 ～ 954, 1986).

  Reference numeral 3 denotes a language dictionary used for analyzing text by the text analysis unit 2 and generating language information necessary for reading, accent information, and prosodic control. Reference numeral 4 denotes a prosody control unit that generates prosody information such as intonation and rhythm based on the reading, accent information, and language information necessary for prosody control generated by the text analysis unit 2. Here, the prosody control unit generates prosody information such as intonation and rhythm based on a known technique represented by the Fujisaki model.

  Reference numeral 5 denotes an acoustic dictionary storing phonemes used when creating a synthesized sound. Reference numeral 6 denotes a phoneme used for creating a synthesized sound based on readings output from the prosody control unit 4 and prosodic information such as intonation and rhythm. A segment selection unit 7 for selecting a segment from the acoustic dictionary 5 generates a synthesized speech by matching the phoneme segment selected by the segment selection unit 6 with prosody information such as intonation and rhythm generated by the prosody control unit 4. It is a speech synthesizer. Here, as a method of synthesizing speech by changing the pitch period and the phoneme duration, for example, a residual drive LSP method for synthesizing on an LSP (Line Spectral Pair) parameter, an MBE (Multi Band) for synthesizing on a spectral parameter, etc. Excitation), a pitch waveform superposition method for superposing and synthesizing two pitch long waveforms, and a waveform editing method for connecting and synthesizing signal waveforms such as phoneme units can be used. Reference numeral 8 denotes an output terminal for outputting synthesized speech generated by the speech synthesis unit 7.

  Reference numeral 9 is a reading control signal input terminal, and the synthesized speech output from the output terminal 8 is read out in accordance with the contents of the text input at the input terminal 1, or important keywords are not read out and are silent, beep sounds, etc. A reading control signal for controlling whether to change to another sound and read out is input. 10 is a reading control rule that extracts a character string of a specific pattern in the text and defines the contents to be read out for the extracted character string pattern. 11 is a call control rule that is called during the processing of the text analysis unit 2. A pattern morpheme generator 12 that adds morpheme information based on each rule described in (1), is called during the processing of the text analyzer 2, and the morpheme in the morpheme analysis result is a morpheme generated by the pattern morpheme generator 11, This is a reading content changing unit that changes the reading contents in response to a request to change the reading contents from the reading control signal input terminal 9.

FIG. 2 is a flowchart showing the processing contents of the text analysis unit 2. In particular, S14 represents processing in the pattern morpheme generation unit 11, and S17 represents processing in the reading content change unit 12.
FIG. 3 shows details of the processing contents of S14 in FIG. 2, and is a flowchart showing the processing contents in the pattern morpheme generation unit 11. In the following description, N is 3 in the process of S25.
FIG. 4 shows in detail the processing contents of S17 in FIG. 2, and is a flowchart showing the processing contents in the reading content changing unit 12.

FIG. 5 is an example of the reading control rule 10, and 21 to 24 are reading control rule examples.
FIG. 6 is an example of a specific matching pattern of the rule name designated by the conversion target character pattern rule of the reading control rule 10, and 31 to 34 are examples of matching patterns, respectively.
FIG. 7 is an example of processing data in the text analysis unit 2, 51 is an example of input text, 52 is an example of a matching character string, 53 is an example of an optimal morpheme, 54 is an example of a reading control rule generation morpheme, and 55 to 56 is a text analysis. It is an example of a result.
FIG. 8 is an example of morpheme candidate generation in the text analysis unit 2, and 61 to 63 are actual morpheme examples.
FIG. 9 is another example of processing data in the text analysis unit 2, 71 is an example of input text, 72 is an example of a preceding word string, 73 is an example of a matching character string, 74 is an example of an optimal morpheme, and 75 is a generation of a reading control rule. Morphological examples 76 to 77 are examples of text analysis results.

Next, the operation will be described.
The operation when the input text example 51 shown in FIG. 7 is input to the text input terminal 1 of FIG. 1 will be described. Further, it is assumed that several levels of levels can be input to the reading control signal input terminal 9, and in the following embodiments, three levels of levels 0 to 2 are input. Here, the level designation is set by the user in advance.
The input text example 51 input to the text input terminal 1 is passed to the text analysis unit 2. The text analysis unit 2 extracts sentences one by one from the text and generates the most likely analysis result. Processing in the text analysis unit 2 for each sentence will be described with reference to FIG.

  Since the input text example 51 consists of one sentence, the entire input text example 51 is transferred to S11, the current position is set at the beginning of the sentence, and the process proceeds to process S12. In S12, since the current processing position is the beginning of the sentence, the process proceeds to S13. In S13, a word starting from the current position is searched from the language dictionary 3 and registered as a morpheme candidate. When the current position is the sentence head, the morphemes of the morpheme example 61 and the morpheme example 62 shown in FIG. 8 are generated.

When generation of the morpheme by dictionary lookup is completed, the process moves to S14 to check whether there is a pattern that matches the reading control rule of FIG. 5 and the matching pattern of FIG. sign up. Since there is no pattern matching the reading control rule and the matching pattern at the current position (= the beginning of the sentence), the process proceeds to S15, and the current position is moved toward the end of one character sentence.
When the processing of S15 is completed, the processing is shifted to S12, and the processing from S12 to S15 is repeatedly executed.

  Here, the specific processing of S14 when the current position of the input text example 51 becomes “(045) 930-0010, contact ...” will be described. The process is moved to S21 in FIG. 3, and in S21, the first reading control rule is set. Therefore, the rule example 21 shown in FIG. 5 is selected, and the process moves to S22. In S22, since all the reading control rules have not been processed, the process proceeds to S23. In S23, in the rule example 21, since the preceding word string “FAX” is defined, the process proceeds to S25. Since there is no preceding word string “FAX” within the front N characters (N = 3), the process proceeds to S27, the next rule example 22 shown in FIG. 5 is selected, and the process returns to S22.

  Similar to rule example 21, rule example 22 has a preceding word string, and “TEL” does not exist within the first N characters of input text example 51. Therefore, the determination process from S23 to S25 is performed, and the process proceeds to S27. , The next rule example 23 is selected, and the process returns to S22.

  In S22, since the full reading control rule has not been processed, the process proceeds to S23. In S23, since the preceding word string is not defined in the rule example 23, the process proceeds to S24. In S24, whether the conversion target character pattern rule of rule example 23 matches the partial character string of the text “(045) 930-0010, contact ...” from the current position of input text example 51 shown in FIG. Determine if. In rule example 23, since the conversion target character pattern rule is “rule (phone2)”, it is determined whether or not the matching pattern description in the matching pattern examples 31 to 33 in the matching pattern in FIG.

  Here, in the matching pattern description of FIG. 6, it is possible to describe a method of describing the character string directly and that a specific character type continues for a specified number of times. As a method of describing the character string directly, it is expressed by enclosing it with “” ”and“ ””, and as a method of describing that a specific character type continues for a specified number, <character type> (< (Minimum number of characters>, <Maximum number of characters>).

Further, as a matching condition, matching of a specific character type matched by the matching pattern description is associated with a variable in order from the top, and “len (<variable>)” is the length of the character string, “[val (<variable >> ), <Minimum>, <maximum>] ”defines the range of numerical values.
For example, in the matching pattern description of the matching pattern example 31, numbers “1” to “10” are consecutive in the description “NUM (1, 10)”, and then the character “-” appears by ““-””. 1 to 10 continuous, the character “-” appears, indicating that the numbers are 1 to 10 continuous. Each NUM (*, *) is assigned to variables $ 1 to $ 3 in order from the left. On the other hand, according to the matching condition “[len ($ 1) + len ($ 2) + len ($ 3), 10, 11]”, the condition is that the sum of consecutive numbers is 10 or 11.

  Accordingly, the matching pattern example 31 and the matching pattern example 33 do not match the text “(045) 930-0010, contact ...” from the current position of the input text example 51, and the matching pattern example 32 is If they match, the process proceeds to S26. As a result, in S26, the morpheme 63 of FIG. 8 is generated and registered, and the process proceeds to S27. At this time, in addition to general information such as reading, the level of the reading control rule is set in the morpheme. In S27, the next reading control rule is selected, and the processes in S22 to S27 are repeated. Finally, it is determined in S22 that all reading control rules have been processed, and the processing in the pattern morpheme generation unit 11 ends.

  As described above, when the generation of the morpheme candidate is completed, the text analysis processing unit 2 selects the most likely optimal morpheme in S16 according to a known morpheme analysis method such as the minimum cost method or the longest phrase matching method. For the input text example 51, an optimal morpheme example 53 shown in FIG. 7 is generated. When the optimum morpheme is selected in S16, the process proceeds to S17, and the reading content changing unit 12 changes the reading content.

The reading content changing process in S17 by the reading content changing unit 12 will be described in accordance with the processing flow of FIG. First, in S31, a synthetic sound creation level is determined from an input value to the reading control signal input terminal 9. The reading control signal input terminal 9 is a value that is input according to the usage environment of the device. Here, the reading control signal input terminal 9 accepts input in three stages, 0: reading all, 1: not reading the most important keyword, 2 : It is possible to input a level such as non-important keyword reading. In the following description, it is assumed that level 1 is input from the reading control signal input terminal 9.
In S32, the first morpheme “goyo” is selected from the optimal morpheme example 53 in FIG. 7, and the process proceeds to S33. In S33, since the processing of all morphemes has not been completed, the process proceeds to S34. In S34, the first morpheme “goyo [goyo ']” is a processing target, and since this morpheme is not a morpheme generated by the reading control rule, the process proceeds to S37.

  In S37, the next morpheme “no” is selected, and the process returns to S33. The above processing is repeated until the morpheme example 54 shown in FIG. 7 is selected in S37. When the morpheme example 54 is selected in S37 and the processing is shifted to S33, since all the morphemes have not been processed, the processing is shifted to S34. Here, since the morpheme example 54 is a morpheme generated by the reading control rule, the processing is shifted to S35. In S35, since the synthesized sound creation level (= 1) ≧ rule level (= 1), the process proceeds to S36. In S36, the reading is rewritten according to a predetermined rule. Here, it shall be rewritten to silence. The process of S36 is completed, and the next morpheme is selected in S37.

Thereafter, the processes from S32 to S37 are repeated, and the process of S17 is terminated. When the process of S17 ends, the process moves to S18, and the accent position as an accent phrase is controlled. In the processing of the accent position, it is assumed that the accent due to the connection of particles, auxiliary verbs, etc. is deformed by a known accent rule such as “NHK Japanese pronunciation accent dictionary”. Then, a text analysis result example 56 shown in FIG. 7 is generated, and the processing of the text analysis unit 2 is finished.
If the synthesized sound creation level is 0, the morpheme example 54 does not satisfy the condition in S35, and the process proceeds to S37 as it is, so that the text analysis result example 55 shown in FIG. 7 is generated.

  When the text analysis process ends, the process moves to the prosody control unit 4 to generate prosody information such as intonation and rhythm. Further, the segment selection unit 6 selects a segment in accordance with the utterance content from the acoustic dictionary 5. Finally, the speech synthesizer 7 generates a synthesized speech by transforming the segment selected by the segment selection unit 6 to match the intonation and rhythm generated by the prosody control unit 4 and outputs the synthesized speech to the output terminal 8. . Note that the processing in the prosody control unit 4, the segment selection unit 6, and the speech synthesis unit 7 can be realized by a known speech synthesis method, and thus the details are omitted.

Next, an example of an analysis result in another input text example will be briefly described with reference to FIG.
When the input text example 71 in FIG. 9 is input, the preceding word string in the reading control rule example 24 in FIG. 5 matches the preceding word string example 72, so that the matching pattern in FIG. The matching pattern of Example 34 is searched.
As a result, a matching character string example 73 is found, a morpheme is generated, and an optimal morpheme 74 is generated. Since there is a reading control rule generation morpheme example 75 in the optimum morpheme 74, the text analysis result example 77 is displayed when the level of the reading control signal input terminal 9 is 1, and the text analysis is performed when the level is 0. Result example 76 is generated.

In the above-described embodiment, an example of silence is shown as the processing in the reading content changing unit 12, but the content of reading is mapped to a meaningless word other than the silence as the changed content. Also good.
It is also possible to replace with a specific sound effect (animal bark or signal sound such as “pea”).

  In the above embodiment, the user designates the level from the reading control signal input terminal 9 in advance. However, the user confirms the reading control signal at the time of the synthesized sound output and confirms the result. It is also possible to read out at a specified level input from the input terminal 9.

  Furthermore, in the said embodiment, although illustrated with respect to a Japanese text, it is applicable also to foreign languages, such as English. For example, in “Your password number is 1234”, “password” may be replaced with “password”.

  In the above embodiment, the reading control is performed on the audio output, but the present invention can also be applied to a display device.

  Furthermore, in the above embodiment, the pattern matching in FIG. 6 is a pattern description at the character string level. However, it can be easily extended to be described at the morpheme level, thereby controlling the reading of personal names. It is also possible to control address reading.

The present invention relates to an information device such as a mobile phone, a PDA (Personal Digital Assistant), a personal computer, an in-vehicle device such as a car navigation system, an ETC (Electronic Toll Collection System), an ATM (automatic cash dispenser), a CD (cash dispenser). ) Applicable to office equipment such as machines.

It is a block diagram which shows the structure of the speech synthesizer which concerns on Embodiment 1 of this invention. It is a flowchart which shows the processing content of a text analysis process part. It is a flowchart which shows the processing content in a pattern morpheme production | generation part. It is a flowchart which shows the processing content in the reading content change part. It is explanatory drawing of the example of a reading control rule. It is explanatory drawing of the example of a specific matching pattern in a reading control rule. It is explanatory drawing of the example of processing data in a text analysis part. It is explanatory drawing of the example of morpheme candidate generation in a text analysis part. It is explanatory drawing of the example of another process data in a text analysis part.

Explanation of symbols

1. 1. Text input terminal 2. text analysis unit; Language dictionary, 4. Prosody control part,
5. 5. Acoustic dictionary 6. Segment selection unit, Speech synthesis unit, 8. 8. Audio output terminal 9. Reading control signal input terminal 10. Reading control rules 11. pattern morpheme generator, Reading content changing section, 21-24. Reading control rule examples, 31-34. 51. matching pattern example 52. Input text example Matching string example 53. 54. Optimal morpheme example Reading control rule generation morpheme example, 55-56. Text analysis result example 61-63. 71. morpheme examples 72. Input text example 73. preceding word string example 74. matching character string example 75. Optimal morpheme example Reading control rule generation morpheme examples 76-77. Text analysis result example.

Claims (7)

A text analysis means for taking text as input, using a language dictionary with finding, reading, and accent type information, and dividing into analysis results including reading information and accent information;
Prosody control means for generating prosodic information for controlling intonation and rhythm based on reading information and accent information obtained by text analysis means,
Based on the reading information and accent information obtained by the text analysis means, and the prosody information obtained by the prosody control means, a segment selection means for selecting a phoneme from an acoustic dictionary storing phonemes;
In a speech synthesizer having a speech synthesizer that creates a synthesized sound in accordance with the prosodic information obtained by the prosodic control means, the phoneme selected by the segment selection means
A pattern morpheme generation unit that generates a morpheme that matches a pattern defined in the reading control rule based on a reading control rule for extracting a specific pattern during the processing of the text analysis unit;
Reading contents changing means for changing the reading contents of the input text by changing the reading contents of the morpheme obtained by the pattern morpheme generation means,
A speech synthesizer characterized by comprising reading control signal input means for inputting a reading control signal indicating whether or not to execute reading contents change of the reading contents changing means.   The reading control signal input from the reading control signal input means is input after the speech synthesizer user confirms whether or not to change the reading content when the reading content changes. The speech synthesizer according to claim 1, wherein   3. The speech synthesizer according to claim 1, wherein the reading control signal input from the reading control signal input means is configured to be able to set a change in reading contents in several stages.   4. The speech synthesizer according to claim 1, wherein the reading content changing means is configured to silence the changed voice.   4. The speech synthesizer according to claim 1, wherein the reading content changing means is configured to make the changed sound a specific sound effect or a continuous sound effect. A text analysis step that takes a text as input, divides it into an analysis result including reading information and accent information, using a language dictionary with finding, reading and accent type information;
Prosody control step for generating prosody information for controlling intonation and rhythm based on reading information and accent information obtained by the text analysis step,
Based on the reading information and accent information obtained by the text analysis step, and the prosodic information obtained by the prosody control step, a segment selection step for selecting a phoneme from an acoustic dictionary storing phonemes;
In a speech synthesis method having a speech synthesis step of creating a synthesized sound in accordance with the prosodic information obtained in the prosody control step by the phoneme segment selected in the segment selection step,
During the processing of the text analysis step, a pattern morpheme generation step for generating a morpheme that matches a pattern defined in the reading control rule based on a reading control rule for extracting a specific pattern;
A reading content change step for changing the reading content of the input text by changing the reading content of the morpheme obtained by the pattern morpheme generation step;
A speech synthesis method, comprising: a reading control signal input step for inputting a reading control signal indicating whether or not to execute reading content change in the reading content changing step. A text analysis process that takes text as input, divides it into analysis results that include reading information and accent information, using a language dictionary with finding, reading, and accent type information,
Prosody control processing for generating prosody information for controlling intonation and rhythm based on reading information and accent information obtained by text analysis processing,
Based on the reading information and accent information obtained by the text analysis processing, and the prosody information obtained by the prosody control processing, a segment selection process for selecting a phoneme from an acoustic dictionary storing the phonemes;
In a speech synthesis program for causing a computer to implement speech synthesis means for creating synthesized speech in accordance with the prosodic information obtained by prosody control processing for the phoneme segment selected in the segment selection process,
A pattern morpheme generation process that generates a morpheme that matches a pattern defined in the reading control rule based on the reading control rule for extracting a specific pattern during the text analysis process;
Reading contents change processing for changing the reading contents of the input text by changing the reading contents of the morpheme obtained by the pattern morpheme generation process,
A speech synthesis program for causing a computer to further realize a reading control signal input process for inputting a reading control signal for determining whether or not to read a reading contents change in the reading contents changing process.

Images