KR100424215B1 - Method and apparatus for outputting traffic message digitally encoded by synthetic voice - Google Patents

Abstract

A method and apparatus for outputting a digitally encoded traffic message including a place code and an event code by a synthesized voice, wherein one carrier record is selected from among a plurality of carrier records stored according to each received traffic message, Place-related and event-related words are inserted into the open position of the carrier record, which are derived from place codes and event codes.

Description

Method and apparatus for outputting traffic message digitally encoded by synthetic voice

The invention relates to a method according to the preamble of the independent claims.

A radio-data-system (RDS) makes it possible to additionally and non-audibly transmit digital data in parallel with a radio broadcast program in a data channel. The specification of the radio-data-system for microwave-radio broadcasting is in particular the publication of the European Radio Broadcasting Union (EBU) in March 1984 Tech. 3244-E. ≪ / RTI > A wireless broadcast receiver having an appropriate RDS-decoder can receive and decode data transmitted in addition to wireless broadcast to the same receiver. This data transmission is made up of 32 groups for each 108 bits, and a certain service is assigned to each group. A group 8a is provided for the transmission of a digitally encoded traffic message,

The construction and coding of this traffic message is specified in detail in ALERT C, a tentative standard issued by the RDS ATT ALERT Consortium in November 1990. The essential elements of a traffic message are Location and Event. This information is cataloged. That is, a clear code is assigned to each event related to each location and traffic related to the traffic. If you connect places in the location table along the existing roads, the course is given. Devices for decoding, for storage, for subsequent processing, and for outputting traffic messages are needed for use of the Traffic Message Channel (TMC) in addition to the general devices of the receiving device with the RDS-decoder .

For voice output, a record should be formed that can be understood by the user directly from the information received with the traffic message and the information stored in the receiver, and sounded naturally.

It is an object of the present invention to provide a method for outputting a traffic message that is digitally encoded by synthesized voice with a low memory cost as little as possible. In addition, the method according to the present invention should be able to take into account features of various languages.

This object is achieved according to the invention by means of which a carrier record is selected from among several carrier records stored according to a respective received traffic message and within this open position of the carrier record, the place and related words derived from the place code and the event code and / Records are inserted. Due to the many different place codes and a large number of different voice formats for describing the contents of each of the event code and the traffic message, the number of voice outputs output as records is very large. With the method according to the invention, a sufficient number of carrier records, respective place codes and event codes and associated words need only be stored.

In a preferred embodiment of the method according to the invention, a carrier record, a place related word and an event related word and / or record are stored in the form of a phonetic transcription, The notes for the place and event related words and / or records are read out, the notes forming each one word are transformed, then the carrier record with the inserted words and / or records is transformed between the words Filtered, and supplied to the sound reproducing means.

The embodiment has the advantage that it is not necessary to convert, for example, place and event related words represented in a general script into phonetic representations.

Place-related words are typically stored in the language spoken in each locale. However, if these words are inserted into the carrier record with event related words and / or records, the intonation profile in the language of the event related words and carrier records can be considered. Thus, in a further refinement of the method according to the invention, corresponding incentive profiles are stored with carrier records and event related words or records, respectively, which are considered when reading notes from the sound catalog and are accented with place related words Crystals are made after conversion to sound.

This improvement example is for reasons of semantic reasoning that usually store place-related information in an exchangeable memory, for example a chip card, while also storing information dependent on each language, such as carrier record and event related words And / or records are preferably stored in a different memory.

Matching the place-related words given in other languages to the sound of the language of the carrier record means that according to another embodiment of the present invention, if the place-related words are stored in a different language than the carrier record, .

This rhyme-conversion can be done, for example, by a motor vehicle operator in France who outputs a traffic message in German, listening to the nominal paris (pronounced s at the end). However, names in other languages are also common in a series of places, for example Aachen is Aix la Chapelle in French. In order to be able to take account of these specificities also in speech output, in accordance with another embodiment, at least at least one place-related common words in one language, as well as miso-related words in the local language, are also stored under the place code, Place-related words stored in the language are inserted into the carrier record and place-related words are inserted into each local language for place codes where the place-related words are not stored in the language of the carrier record.

In addition, once the optical display of the traffic message is made, the method according to the present invention can store place and event related words and carrier records as script characters for display on an optical playback device. In most applications, it is sufficient to store in local language, because the name can be seen on the traffic sign, which makes it easy to compare with the visual display.

Despite steady improvements in storage density, the storage capacity of the exchangeable memory is limited to a predictable time or is relatively expensive. In order to keep the required storage capacity of the interchangeable data carriers and possibly the internal memory as small as possible, other embodiments of the method according to the present invention allow the often appearing components of place and / And the components are stored as allophones under the additional phonetic transcription in the notes catalog.

In an advantageous device for the implementation of the method according to the invention for a radio receiver having a radio-data signal-decoder and a device for decoding a digitally encoded traffic message, the speech synthesis device comprises carrier records, event words or events A memory for records and place related words, a speech message decoded in the speech synthesizer is provided, and an output of the speech synthesizer is connected to an input of an audio-amplifier, and the amplifier controls the speaker device. With this device, a very large number of traffic messages can be output, and the memory space required is relatively small.

The advantageous use of a place list with at least partly place-related words in several languages is achieved by means of a device according to the invention in which a selector is arranged between the memory and the speech synthesizer for the place-related words, In the language when the records are retrieved from the memory, each place related word is stored, and each place related word of the language is delivered to the speech synthesizer.

Matching a foreign language word to the sound of a carrier record means that in an embodiment of the apparatus according to the invention the place related words provided in the speech synthesizer are not given in the language when the carrier records are taken out of memory, Lt; / RTI > At this time, this rhythm-conversion can be done according to a given rule or by catalog.

In another embodiment of the device according to the invention, the carrier records and event words or records are stored in several languages. As a result, a radio broadcast receiver which is a multilingual voice output is realized.

By arranging the memory for the place-related words in the exchangeable data carrier, a particularly good system is obtained in which a variety of sound quality device classes using the same chip card are possible. This is therefore particularly desirable because a sales channel that is independent of the wireless broadcast receiver for the chip card can be selected.

Embodiments of the present invention are illustrated in the drawings and described in further detail in the following description.

The same parts have the same reference numerals in the drawings. Embodiments and portions thereof are shown as a block diagram. However, this does not mean that the circuit arrangement according to the present invention is limited to implementations using individual circuits corresponding to the blocks. The circuit arrangement according to the invention is advantageously realized as a highly integrated circuit. In this case, a digital signal processor may be used which executes the processing steps shown in the block diagram by appropriate programming. These embodiments form important parts of a wireless broadcast receiver in an integrated circuit with other circuit devices.

The traffic message output according to the method according to the invention has, for example, the following form:

A7, in the direction of Kassel, Hannover, between Derneburg and Hildesheim, 2km.

The next carrier record is selected from the plurality of stored carrier records as a function of the various codes included in the encoded traffic message to be transmitted before the record is made for voice output:

First, the open positions are filled with a place holder, where word type 1 is the highway number, word type 2 is the place description, and word type 3 is the event description.

The event description is also carried out by the carrier record - hereafter referred to as the event record. E.g:

Congestion <4> km. At this time, word type 4 is a number. For convenience, the entire event is treated like a carrier record, and events without variable parameters, for example, "Attention, runaway driver comes to you" are treated like carrier records. This carrier record is stored in the radio broadcast receiver and numbered. Depending on the ALERT-Specification and parameters described above for record formation, one or more carrier records are selected and combined and supplemented with a specific word instead of a placeholder.

There is a place list 1, an event record list 2 and a carrier record list 3, respectively, in the embodiment schematically shown in FIG. 1, in which the carrier records and words to be spoken . The proper phonetic alphabet is known as SAMPA and is suitable for a variety of European languages. A word is written by SAMPA in its usual rhyme, for example in German or French. No human speaker is needed to encode the words into SAMPA, which is independent of the speaker.

All segments required for the words to be output are stored as segments in the segment library 4, also referred to as the sound catalog, and are called as functions of phonetic notations read from the lists 1, 2, 3, , Phonem, diphone, and allophone are substituted for each phonetic notation. These segments are spoken, recorded, numbered and encoded in a known digital form, quantized, stored and addressed by a human speaker for the formation of a segment library. The coding method is known as RELP, CELP and LPC.

The acoustic quality of this segment depends on the encoding and the manner of quantization. It can be divided into parameterized method and non-parameterized method. In a parameterized method, modification may be performed at a playback point in a limited frame. For example, the pitch height, segment duration can be modified. The intonation profile stored in addition to each event or each carrier record in lists (2 and 3) in (6) and (7) is considered.

In the embodiment according to FIG. 1, the place list 1 is stored in an exchangeable data carrier valid for one area, and the data carrier can be used irrespective of the language in which the traffic message is output. Therefore, the intonation profile is not stored in the list (1). Decisions of nomination are made later in step (8) for each language.

A coded speech signal representing a place, an event record and a carrier record is transformed into words in the sentences (9, 10, 11), and (12) a transformation is made between the words after the combination of the components of the traffic message. (13 ') after negative decoding, which is stored as a parameter value in the decoder 13, the sound formation is effected by appropriate control of the digital filter, for example simulating a human throat. Analog-to-digital conversion is performed in the digital-to-analog converter 14, which is then amplified by the audio-amplifier 15 and provided to the speaker 16.

In the second, third and fourth diagrams, only the members and functions necessary for the explanation of the audio output according to the present invention among the radio receivers are shown. Signals of the group 8a from the RDS decoder not shown here are provided to the traffic message processing unit 21 in the base station 22. According to the ALERT-Specification described above, this data includes information on the location, type, predictable duration, scale and direction of the event. This information is preferably decoded and selected in a device 10 that can be realized in the form of a microcomputer or a digital signal processor. Since the selection is made according to the selection criterion provided through the input unit 23, for example, only the traffic message related to the road segment input in the input direction is reproduced.

The data stored in the portion 25 of the place list 1 is already required for the decryption of the traffic message, for example, the location of the place mentioned after the place mentioned in the traffic message in the road section and the belonging to the one place. These and other information are stored in the chip card 24 in one place list and read by the reading device in the receiver as other information described later. In the example shown, location number 4711 belongs to area 2009 and neighbor locations 4710 and 4712.

The decoded traffic message is provided to an apparatus 26 for speech synthesis and the output signals thereof are provided to a speaker 16 via a digital / analog-to-digital converter 14 and an audio amplifier 15. The device 26 can also be realized by a digital signal processor, and a program 27 is provided to the processor. Data necessary for language synthesis is provided from the place list 1, the event record list 2, the carrier record list 3 and the segment library 4 stored in the chip card 24 as shown in FIG. This synthesis is performed in the same language in which event record list 2, carrier record list 3 and segment library 4 are created, according to the method described in connection with the first aspect. Locations are used as a common language in each locality.

In the place list (1), there are also numbers related to the connection point (15), for example, in addition to the place names of shoes. Since it is important to hear the number of connection points in the language in which the remainder of the output is made, number 15 is stored as a number in place list 1. In the case of speech synthesis, a particular part of the segment library 4, in which the pronunciation of the number is stored in each language of the receiver, is used. In order to enable optical audio output by the chip card 24, the place name is indicated by one or a plurality of foreign words in common letters and phonetic letters in the list (1). However, this foreign language is not evaluated by the receiver according to FIG. 2.

The receiver according to FIG. 3 expands so that the nomenclature inserted in the carrier record is linguistically consistent with the language of the receiver, as compared to the receiver according to FIG. To this end, two other measures are provided in the embodiment according to FIG. On the other hand, if the place name in the other language (Exonyme) is used in addition to the place name in the local language (Endonyme) in the chip card, it is stored. (Cologne), for example Cologne in French. In this case, the receiver according to FIG. 3 can select the place name which exists in the language of the receiver by means of the selection device 28. In the case shown, a choice of, for example, English, Dutch, German and French is made. If many place names are not stored in the language of the receiver, many places in one country are such cases), the place names in the place list 1 are used for voice synthesis.

Another measure for improving speech quality is to provide a name to the rhyme-converter 29 that is provided in a different language than the language of the receiver. This kind of conversion is well known and changes the voice melody in the direction of the receiver language.

After the receiver is switched on by special adjustment or at the exchange of the place list (for example, replacement of the chip card), the composition of the speech synthesis is made. The German receiver includes, for example, a chip card with a list of places for France. After switching on, this receiver recognizes that it is a chip card with a list of places for France, including German designation. The receiver then selects the conversion function of the French rhyme to the German rhyme in the rhyme-converter. Also check in the place list whether or not there is a German name for one place at each access. Without the German name, the "source prosody" (in this case French) changes to the "goal prosody" (here German).

In the receiver according to FIG. 4, the event record list and the carrier record list are stored in the event record list (2de, 2en, 2fr) and the carrier record lists (3de, 3en and 3fr) in multiple languages, ). In addition, the segment library 4 'is expanded by a special segment of the additional language, and all languages are spoken by the same speaker. This traffic message can therefore be output in multiple languages. The place names included in the place list by the selection device 28 and the rhyme-converter 29 are matched to each selected language.

Figure 1 is a schematic diagram showing the generation of a voice signal comprising a complete record consisting of received individual or receiver generated codes;

Figure 2 is a block diagram of a receiver designed to output a traffic message in one language.

Figure 3 is a block diagram of a receiver that is designed to output a language traffic message but whose foreign language designation matches the language of the receiver in relation to the melody.

4 is a block diagram of a receiver in which various languages can be selected for output.

DESCRIPTION OF THE REFERENCE NUMERALS

1, 2, 3: Listing 4: Segment Library

15: Audio amplifier 16: Speaker

23: Input unit 24: Data carrier

26: speech synthesizer 27: program

28: selection device 29: rhyme converter

Claims (11)

A method of outputting a digitally encoded traffic message using synthetically generated speech, the traffic message comprising location codes and event codes, the method comprising: And wherein the location and event related words derived from the location codes and event codes are inserted into the open position of the carrier record, Wherein said carrier records, place related words and event related words are stored in the form of phonetic transcriptions and the sounds are stored in the form of carrier records, place related words and event related And each of the words forming the word is transformed and subsequently the carrier record and inserted words are transformed between the words and digitally filtered and supplied to the sound reproduction means . The method according to claim 1, Each intonation profile is stored with the carrier records and event related words, and the intonation profile is taken from the note catalog at the time of reading the notes, and the intonation is defined for the place related words after conversion to notes &Lt; / RTI &gt; 3. The method according to claim 1 or 2, Location related words are stored in a different language than the carrier records, the location related words are proscribed before being inserted into the carrier records. 3. The method according to claim 1 or 2, Related words stored in the language of the carrier records are inserted into the carrier records, and the location-related words stored in the language of the carrier records are stored in at least one other language, Location-related words are inserted into each local language, for place codes in which the words are not stored in the language of the carrier records. 3. The method according to claim 1 or 2, Location-related and event-related words and carrier records are stored as a script character for display on a visual reproduction device. 3. The method according to claim 1 or 2, Characterized in that the often occurring components of said place and / or event related words are stored as additional phonetic transcription characters and said components are stored as allophones between additional phonetic transcription characters in the phoneme catalog How to. An apparatus for implementing a method according to claim 1 or 2 for a wireless broadcast receiver comprising a wireless data signal decoder and an apparatus for decoding a digitally encoded traffic message, The speech synthesizer 26 is connected to the memories 3, 2, 1 for the carrier records, event words or event records and place related words, and the encoded traffic message is sent to the speech synthesizer 26 Characterized in that the output of the speech synthesizer (26) is connected to the input of an audio amplifier (15), which controls the speaker device (16). 8. The method of claim 7, A selection device 28 is arranged between the memory 1 for the place related words and the speech synthesizer 26 and the selection device 28 is arranged in a language in which the carrier records are taken out of the memory 3 Related words of the language to the speech synthesizer (26) if the respective place-related words are stored. 8. The method of claim 7, Characterized in that the place related words supplied to the speech synthesizer (26) are provided to the rhythm converter (29) if the words are not given in the language when the carrier records are retrieved from the memory (3) Device. 9. The method of claim 8, Characterized in that the carrier records and the event words or records are stored in a plurality of languages. 8. The method of claim 7, Characterized in that the memory (1) for the place related words is arranged in a replaceable data carrier (24).