JPH1152848A - Data base retrieval system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data base retrieval system shortening a time required for retrieval processing by reducing a load of processing required for retrieval. SOLUTION: A voice recognition part 52 retrieves recognitive information having a registered waveform most similar to a waveform of a voice signal inputted through a microphone 7 by retrieving a target information library with recognitive information read out from a CD-ROM 2 and stored in a retrieval memory 56 to extract storage positional information showing a storage position of target information. A data base retrieval part 54 specifies directly any target information from the target information library with recognitive information stored in the retrieval memory 56 based on the storage positional information extracted by the voice recognition part 52 to read out it. A target information picture generation part 58 displays the target information outputted from the data base retrieval part 54 on a screen of a display device 9.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database retrieval system for retrieving desired target information using voice recognition in an in-vehicle navigation device or the like.

[0002]

2. Description of the Related Art Many on-vehicle navigation systems include a database search system for searching for location data and facility information on various facilities from a location, a telephone number and the like. Recently, as a method of designating a facility to be searched, a method of designating a facility name or the like by voice instead of performing key input by a remote control (remote control) unit has been used.

FIG. 4 is a diagram showing a configuration of a database search system for inputting a destination or the like of a navigation operation by voice. FIG. 5 is a diagram showing the contents of a speech recognition dictionary used in the database search system shown in FIG. 4, and FIG. 6 is a diagram showing the contents of a target information dictionary.

[0004] The database search system shown in FIG.
Microphone 100, voice recognition unit 102, database search unit 104, voice recognition dictionary 106, target information dictionary 10
8 are included. The voice recognition unit 102 searches the voice recognition dictionary 106 to select recognition information having a registered waveform that is most similar to the waveform of the input voice, and extracts a corresponding key code. The database search unit 104 searches the target information dictionary 108 based on the key code extracted by the voice recognition unit 102, and extracts and outputs target information corresponding to the key code.

For example, when a search for “detailed information on amusement park B” as the purpose information is considered, first, a driver or a passenger utters “amusement park B” toward the microphone 100. The voice recognition unit 102 searches the voice recognition dictionary 106 shown in FIG. 5 based on the waveform of the voice signal input via the microphone 100, and finds “Y2” corresponding to the most similar recognition information (registered voice waveform). Is extracted as a key code, and the extracted key code “Y2” is output to the database search unit 104. Database search unit 10
4 is a key code “Y” output from the voice recognition unit 102.
Based on “2”, the target information dictionary 108 shown in FIG. 6 is searched, and “detailed information on amusement park B” is extracted as target information corresponding to the key code “Y2”.

[0006] The target information dictionary shown in FIG. 6 includes category information of facilities and the like in order to search for target information based on key input instead of voice input. For example, when the category “amusement park” is designated by a key input from the remote control unit, the database search unit 104
Extracts one or more target information candidates that match this category from the target information dictionary 108, and one of them is selected as final target information.

[0007]

By the way, in the conventional database search system utilizing the above-described speech recognition, a first search for extracting a key code based on an input speech is first performed in order to obtain target information. After that, it is necessary to perform a second search for extracting the target information based on the key code, and the search process is repeated twice, so that the load of the search process is large and the search takes time. For example, in a database search method in which key input is performed directly without performing voice recognition, the first search processing in which voice recognition is performed is not necessary. Since a process of searching for a key code is added, the processing time is correspondingly increased. In particular, in the in-vehicle navigation device, the speech recognition dictionary 106 and the target information dictionary
-A search process is performed after stored in a ROM or the like, read out as necessary and stored in a memory, and two read processes are performed on a CD-ROM or the like in order to perform two search processes. Done. Therefore, it takes time to read these dictionaries, and the time required to obtain the target information is further increased.

SUMMARY OF THE INVENTION The present invention has been made in view of the above points, and has as its object to reduce the processing load required for searching and thereby reduce the time required for searching processing. Is to provide.

[0009]

In order to solve the above-mentioned problems, according to the database search method of the present invention, target information is directly specified when a search dictionary is searched based on an input voice by a voice recognition unit. The target information reading means can directly read the target information based on the specified information. Therefore,
Since the target information can be obtained only by searching the search dictionary once, the load required for the search process can be reduced, and the search processing time can be shortened.

[0010] In particular, when recognition information necessary for speech recognition and the above-mentioned designation information and purpose information are included in one search dictionary, the search dictionary transfer means uses a CD-R.
Since the process of transferring a search dictionary from a storage device such as an OM or a DVD-ROM to the search dictionary storage means is required only once,
The search time including the time required for reading the dictionary can be significantly reduced.

However, the above-mentioned search dictionary may be provided separately as a speech recognition dictionary required for speech recognition and a target information dictionary necessary for extracting target information. Desired target information can be directly read without searching the entirety, and the time required for the search processing can be reduced.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A navigation apparatus according to an embodiment to which a database search method of the present invention is applied stores a target information dictionary with recognition information in which a conventional speech recognition dictionary and a target information dictionary are integrated into a search memory. In addition, when performing a search by voice recognition, location information that can directly specify the target information is extracted, and thereafter, the target information is directly specified and extracted based on the position information. .
Hereinafter, a navigation device according to an embodiment of the present invention will be described with reference to the drawings.

(1) Overall Configuration of Navigation Apparatus FIG. 1 is a diagram showing the overall configuration of a navigation apparatus according to an embodiment to which the present invention is applied. The navigation device shown in FIG. 1 includes a navigation controller 1 that controls the entire navigation device, and a CD-ROM 2 that stores various map data and dictionaries necessary for database search.
And a disk reading device 3 for reading map data and the like recorded on the CD-ROM 2, a remote control unit 4 for a driver or a passenger to input various instructions by keys, and detection of the own vehicle position and own vehicle direction. GPS receiver 5, autonomous navigation sensor 6, and microphone 7 that collects voices of drivers and the like
And a display device 9 for displaying a map image and purpose information
And

The above-mentioned disk reading device 3 can be loaded with one or a plurality of CD-ROMs 2 and any one of the CD-ROMs 2 is controlled by the navigation controller 1.
-Read map data and the like from the ROM 2. The remote control unit 4 has various operation keys, and an infrared signal corresponding to the operation state of the keys is transmitted to the navigation controller 1. The GPS receiver 5 has a plurality of G
The radio wave transmitted from the PS satellite is received, and the absolute position and the azimuth of the vehicle are calculated by performing the three-dimensional positioning process or the two-dimensional positioning process. These are calculated based on the own vehicle position) and output together with the positioning time. Also,
The autonomous navigation sensor 6 includes an angle sensor such as a vibrating gyroscope that detects the vehicle rotation angle as a relative bearing, and a distance sensor that outputs one pulse every predetermined traveling distance.
Detect the relative position and direction of the vehicle. The display device 9 displays, based on image data output from the navigation controller 1, map information around the own vehicle together with a vehicle position mark, a departure point mark, a destination mark, and the like, and the target information extracted by database search. Is displayed.

(2) Detailed Configuration and Operation of Navigation Controller Next, a detailed configuration of the navigation controller 1 shown in FIG. 1 will be described. The navigation controller 1 includes a map buffer 16 for displaying a predetermined map on the display device 9 based on the map data read from the CD-ROM 2, a map reading control unit 18, a map drawing unit 20, a VRAM 22, a reading control unit 24, Image synthesis unit 26
And a vehicle position calculation unit 30, a route search processing unit 36, a mark image generation unit 44, and a guidance route drawing for performing the calculation of the vehicle position, the map matching process, the route search process, and the route guidance process and displaying the results. And a part 42. Further, the navigation controller 1 uses a remote control unit 60, an operation screen generation unit 64 for displaying various operation screens to a driver or the like and transmitting operation instructions from the remote control unit 4 to each unit, and a voice recognition. A voice recognition unit 52, a database search unit 54, a search memory 56, and a target information screen generation unit 58 for searching for desired target information and displaying the target information extracted by the search are provided.

The map buffer 16 is provided in the disk reading device 3.
Is for temporarily storing the map data read from the CD-ROM 2 by the above method. When the screen center position is calculated by the reading control unit 24, an instruction to read map data in a predetermined range including the screen center position is sent from the map reading control unit 18 to the disk reading device 3, and the map necessary for map display is displayed. Data is read from the CD-ROM 2 and stored in the map buffer 16. When searching for the target information using voice recognition, a dictionary reading instruction required for the search is sent from the map reading control unit 18 to the disk reading device 3 prior to this search processing, and the recognition information is added. The target information dictionary is read from the CD-ROM 2 and stored in the search memory 56.

The map drawing section 20 creates a map image required for display based on a drawing unit included in the map data stored in the map buffer 16. The created map image data is stored in the VRAM 22, and the read control unit 24
Thereby, map image data for one screen is read. The image synthesizing unit 26 adds the read map image data to
A mark image generating unit 44, an operation screen generating unit 64, which will be described later,
Each image data output from each of the target information screen generation units 58 is superimposed to perform image synthesis, and the synthesized image is displayed on the screen of the display device 9.

The vehicle position calculating section 30 calculates the position of the own vehicle based on the data detected by the GPS receiver 5 and the autonomous navigation sensor 6, and when the calculated own position is not on the road of the map data. A map matching process for correcting the own vehicle position is performed.

The route search processing section 36 searches for a traveling route connecting a preset destination and departure place under predetermined conditions. The guide route drawing unit 42 draws the guide route set by the route search processing unit 36 so as to overlap the map image. The mark image generating section 44 generates a vehicle position mark at the position of the own vehicle or generates a cursor mark having a predetermined shape.

The voice recognition unit 52 searches the target information dictionary with recognition information stored in the search memory 56 to find out the feature (registered waveform) most similar to the waveform of the voice signal input via the microphone 7. Retrieval information is retrieved, and corresponding storage location information is extracted. This storage position information indicates the storage position of the target information corresponding to the recognition information. The database search unit 54 includes the search memory 5
6, the target information is directly specified and read out from the target information dictionary with the recognition information stored in 6. The target information screen generation unit 58 generates image data for displaying the search result (target information) output from the database search unit 54 on the screen of the display device 9. It is preferable that the above-described speech recognition unit 52 be implemented using a DSP suitable for parallel processing, and the database search unit 54 be implemented using a general-purpose processor suitable for logical operation.

FIG. 2 is a diagram showing a specific example of the target information dictionary with recognition information stored in the search memory 56. As shown in the figure, the target information dictionary with recognition information of the present embodiment stored in the search memory 56 is used as a registered voice waveform required for the voice recognition unit 52 to compare features with an input voice waveform. Search information, target information stored corresponding to the recognition information, storage position information indicating a storage position of each target information in the dictionary, and target information searched by a conventionally used method. The necessary categories are included.

The above-described search memory 56 is used as a search dictionary storage unit, the voice recognition unit 52 is used as a voice recognition unit, the database search unit 54 is used as a target information reading unit, and the CD-ROM 2 is used.
The disk reading device 3 corresponds to a storage device, and the map reading control unit 18 corresponds to a search dictionary transfer unit. The target information dictionary with recognition information stored in the search memory 56 corresponds to the search dictionary, and the storage position information in the target information dictionary with recognition information corresponds to the designated information in the search dictionary.

(3) Operation of Navigation Apparatus The navigation apparatus to which the present invention is applied has the above-described configuration. Next, the operation of the navigation apparatus when extracting desired target information using voice recognition will be described. explain.

FIG. 3 is a flowchart showing an operation procedure of the navigation device according to the embodiment, and shows an operation in a case of displaying any target information specified by voice. For example, it is assumed that the name of the amusement park is specified by voice and detailed information of the specified amusement park is searched.

The voice recognition unit 52 constantly determines whether or not to enable the voice recognition processing (step 100). For example, when the remote control unit 4 is operated by a driver or the like, a database search using voice recognition is performed. When an instruction to perform is given, a database search with the voice recognition process enabled is started. First, an instruction is sent from the map reading control unit 18 to the disk reading device 3, and the target information dictionary with recognition information read from the CD-ROM 2 is stored in the search memory 5
6 (step 101).

Next, the voice recognition unit 52
It is determined whether there is a voice input by the user (step 10).
2) If any voice input is made, the search memory 56
(Step 103), it is determined whether or not there is recognition information having a speech waveform similar to the signal waveform of the input speech (step 1).
04). If there is no recognition information having a similar voice waveform, the process returns to step 102 to wait for voice input. If there is recognition information having a similar voice waveform, the voice recognition unit 52 extracts storage position information corresponding to the recognition information and notifies the database search unit 54 (step 105). By using the target information dictionary with recognition information shown in FIG. 2, for example, when a driver or the like utters “amusement park B”, storage position information “2” indicating the storage position of the corresponding target information is extracted. Is notified to the database search unit 54.

Upon receiving the storage position information notified from the voice recognition unit 52, the database search unit 54 retrieves the target information specified by the storage position information from the target information dictionary with recognition information stored in the search memory 56. Read (step 106). This purpose information is read out by designating the storage position in the dictionary, so that it is not necessary to search the entire dictionary. For example, “2” is notified as storage position information, and “detailed information on amusement park B” is read as corresponding purpose information.

Next, the target information screen generating section 58 generates image data based on the target information read by the database search section 54, and displays it on the display device 9 (step 107). For example, as the detailed information of the amusement park B, various information such as a transportation method for going to the amusement park B and a description of a typical amusement facility of the amusement park B are displayed.

As described above, the navigation device according to the present embodiment includes the recognition information necessary for recognizing the input voice and the storage position information necessary for directly specifying the target information corresponding to the recognition information. The dictionary is read from the CD-ROM 2 and stored in the search memory 56 prior to the database search. The speech recognition unit 52 searches the dictionary for recognition information similar to the signal waveform of the input speech, extracts corresponding storage location information, and the database search unit 54 generates target information based on the storage location information. Is read directly.

Therefore, in order to obtain the final target information after performing the voice input, only one search processing of the dictionary is required, and the load of the processing required for the search is reduced, and the processing time is shortened. In the related art, the two dictionaries shown in FIGS. 5 and 6 are separately read from the CD-ROM and stored in the search memory. However, in the present embodiment, the dictionaries need only be read and stored once. Also, the time required to read the dictionary can be reduced. When the two dictionaries shown in FIGS. 5 and 6 are considered, index information is provided for each of the category and the key code. In the dictionary of the present embodiment shown in FIG. Since it is only necessary to provide the index information, the size of the entire dictionary required for the search is reduced. When the entire dictionary is read at a time and stored in the search memory 56, the search memory 5
6 can be made smaller.

The present invention is not limited to the above-described embodiment, and various modifications can be made within the scope of the present invention. In the above-described embodiment, the detailed information of the amusement park, which is a leisure facility, is searched for as the purpose information, and the search result is displayed on the screen of the display device 9. However, the search result does not always need to be displayed. For example, the position data of the destination facility for the route search may be stored as the target information, and the position data as the search result may be transmitted from the database search unit 54 to the route search processing unit 36.

In the present embodiment, the "category" item is included in the target information dictionary with recognition information in order to use the conventional search method, but only the search using voice recognition is performed. In this case, the "category" item may be deleted. Further, in the above-described embodiment, the voice recognition unit 52 and the database search unit 54 are separately provided, but each processing may be performed by one processing unit.

In this embodiment, the dictionary is read from the CD-ROM 2 only once by including the recognition information and the target information in one dictionary as shown in FIG. The speech recognition dictionary including the recognition information and the storage location information may be separated from the target information dictionary including the target information and the category. Even if the dictionaries are divided in this way, the storage location information in the speech recognition dictionary can be extracted by one search process, and the target information in the target information dictionary is directly extracted based on the storage location information. Since the reading can be performed, the processing load can be reduced while the processing load can be reduced.

[0034]

As described above, according to the present invention, when a search dictionary is searched based on an input voice by a voice recognition unit, designation information for directly specifying target information is extracted. Since the target information can be read directly by the reading means, the target information can be obtained only by searching the search dictionary once, and the load of the search processing can be reduced and the processing time can be reduced. In particular, when the recognition information necessary for speech recognition and the above-described designation information and purpose information are included in one search dictionary, the search dictionary transfer means transfers the search dictionary from a storage device using a CD-ROM or the like. Since the process of transferring the search dictionary to the storage means only needs to be performed once, the search time including the time required for reading the dictionary can be significantly reduced.

[Brief description of the drawings]

FIG. 1 is a diagram showing an overall configuration of a navigation device according to an embodiment to which the present invention is applied.

FIG. 2 is a diagram showing the contents of a target information dictionary with recognition information stored in a search memory.

FIG. 3 is a flowchart illustrating an operation procedure of the navigation device according to the embodiment;

FIG. 4 is a diagram showing a configuration of a conventional database search system for inputting a destination or the like of a navigation operation by voice.

5 is a diagram showing the contents of a speech recognition dictionary used in the database search system shown in FIG.

6 is a diagram showing the contents of a target information dictionary used in the database search system shown in FIG.

[Explanation of symbols]

 Reference Signs List 1 navigation controller 2 CD-ROM 3 disk reader 4 remote control unit 7 microphone 9 display device 52 voice recognition unit 54 database search unit 56 search memory 58 purpose information screen generation unit

Continued on the front page (51) Int.Cl. ⁶ Identification code FI G10L 3/00 551 G06F 15/40 370E 370C 15/403 310Z

Claims (3)

[Claims] A search dictionary storage means for storing a search dictionary including recognition information to be compared with an input voice, target information corresponding to the recognition information, and designation information necessary for directly specifying the target information. When a voice is input, a voice recognition unit that searches the search dictionary to select the recognition information having similar characteristics to the input voice and extracts the corresponding specification information; and the voice recognition unit. And a target information reading means for reading the corresponding target information from the search dictionary based on the specified information extracted by the method. 2. The storage device according to claim 1, wherein the storage device stores various data including the search dictionary, and reads the search dictionary stored in the storage device prior to searching for target information. A database search method, further comprising: a search dictionary transfer unit that stores in a storage unit. 3. The search dictionary according to claim 1, wherein the search dictionary includes a speech recognition dictionary having the recognition information and the designation information, and a target information dictionary having the target information. Database search method.