TWI401417B - Segmented voice recognition navigation system and its method - Google Patents

Description

Segmented speech recognition navigation system and method thereof

A voice navigation system and method thereof, in particular, a segmented voice recognition navigation system and method thereof.

The prior art navigation device generally uses a user to manually input a navigation destination as a main path planning module. However, if a user frequently navigates to a location, he or she can record or store the location, or if the navigation device has a folder like "My Favorites" and "Commonly Used Target", the location information can be recorded in the information. In the folder.

Moreover, in the current navigation device, most of the built-in systems have a calendar function, which can pre-enter the itinerary or activity that the user needs to perform in the future.

However, the prior art has problems that are difficult to overcome: First, if the "My Favorites" or "Commonly Used Targets" folder records too much location information, when the user wants to navigate, it needs to be from "My Favorites" or " It takes a certain amount of time to select and select the required location information for the commonly used target folder.

Secondly, although the calendar can be pre-entered into the travel required by the user, when the navigation is performed, the location of the trip is still required to be input into the navigation device, so that the route planning can be performed, and the inconvenience in use is improved.

Third, when the input interface of the navigation device and its functions are complicated, the user has a certain degree of difficulty in inputting, that is, it takes a longer time. The input of the location is performed, so that the inconvenience of use is increased, and the time is input at a wasteful place.

Fourth, the input method used in the navigation system or device by the prior art is mostly based on pinyin or stroke-based input method, and the keyboard input is inevitable, and the text input of the navigation is performed while the user is driving. It is a situation in which the user is distracted by driving, which inevitably increases the risk of driving.

In view of this, the problem to be solved by the present invention is to provide a navigation system and method thereof that do not affect the user's driving, avoid distraction of the user, and improve navigation accuracy.

In order to solve the above system problem, the technical means provided by the present invention discloses a segmented voice recognition navigation system, which comprises a regional database, a road database, a house database, a voice database, and a voice recognition module. A group, a microphone, a prompt module and a navigation module.

The regional database stores at least one regional data, and the road database stores at least one road data belonging to the regional data, and the house database records a plurality of house data corresponding to the road data. The voice database records a plurality of voice data corresponding to the geographical data, the road data and the house data.

The voice recognition module obtains the regional voice, the road voice and the house voice input by the user through the microphone in different stages, and separately determines whether the regional database, the road database and the house database store the matched regional data and the road data. And house information. And identifying the user through the microphone Enter one of the voices to select a destination, destination, and destination number from the matching geographic, road, and house data.

The prompting module is segmented to prompt the voice recognition module to determine the matching regional data, road data and house data, and the user can select the voice through the microphone input. The navigation module integrates the destination domain, the destination road and the destination door to form a navigation path.

In order to solve the above method problem, the technical means provided by the present invention discloses a segmented voice recognition navigation method, which first identifies a user to send a regional voice, analyzes a destination domain, and identifies a user to issue a road voice. Analyze a destination road; identify the user to issue a house number voice, analyze a destination number card; and plan a navigation path according to the destination domain, the destination road and the destination doorplate.

The invention has the effects that the prior art cannot achieve:

First, the voice input capability allows the user to perform voice-activated navigation operations while driving, thereby avoiding user distraction and reducing the risk of driving.

Secondly, the segmentation input navigation purpose and the prompt identification result are used to improve the correctness of the navigation location selection, so as to avoid the user having to spend more time to input the location due to the poor recognition rate, and further avoiding the user driving. Heart, reduce the risk of driving.

In order to further understand the purpose, structural features and functions of the present invention, the following detailed description is given in conjunction with the related embodiments and drawings:

Please refer to FIG. 1 , which is a navigation system architecture according to a first embodiment of the present invention. Schematic diagram. The navigation system includes a map database 110, a voice database 150, a voice recognition module 120, a prompt module 140, a microphone 130, and a navigation module 160. The map database 110 further includes a regional database 111, a road database 112, and a house database 113.

The regional database 111 stores at least one regional data, such as place names: Taipei City, Pingtung City, Taichung City, etc. The road database 112 stores at least one road material belonging to the geographical information, such as road names: Jianguo Road, Renai Road, Jiuru Road, and the like. The house number database 113 records a plurality of house number data corresponding to the road data, and includes a house number, such as a number of numbers 1 to 9 and a combination thereof, such as 100, 45, and 1. The voice database 150 records a plurality of voice data corresponding to the area data, the road data, and the house data.

During the navigation operation, the prompting module 140 promptly prompts the user to input the region, the road and the house number to be reached at the destination point. The user inputs the regional voice, the road voice, and the house voice through the microphone 130 in a segmented manner. The voice recognition module 120 first identifies the regional voice to determine whether the regional database 111 stores the matched at least one regional data, and then identifies the road voice to determine whether the road database 112 stores the matched at least one road data, and finally the door voice. It is determined whether the house database 113 stores at least one of the matching house materials.

In the meantime, the prompt module 140 prompts the user to match all the geographical information, the road data and the house data. The user then inputs the selected voice through the microphone 130 for the voice recognition module 120 to identify and select the user. Destination domain, destination road and destination house number. Finally, the navigation module 160 integrates the destination domain, the destination road, and the destination door to plan a navigation path.

The prompt module 140 can be a display, which displays a plurality of prompt texts corresponding to all matching regional data, road data and house data output by the voice recognition module 120. However, to improve safety, the cueing module 140 can be designed as a heads up display. Moreover, the prompting module 140 can be a voice player, which plays a plurality of prompt voices corresponding to all matching geographical data, road data and house data. Moreover, the system further includes a button module 170 for the user to select the regional data, the road data and the house data matched by the voice recognition module 120.

In addition, in order to improve the accuracy of road navigation, the road database 112 further includes a plurality of intersecting road materials, such as the roads intersected by Jianguo Road: Youth Road, Boai Road, Jiuru Road, Renai Road, etc., and record roads form Jianguo Road/ Youth Road, Jianguo Road/Boai Road, Jianguo Road/Jiuru Road, Jianguo Road/Renai Road...etc. The user can input a crossroad voice through the microphone 130, such as the intersection of Jianguo Road and Youth Road. The voice recognition module 120 is configured to determine whether the road database 112 stores the matched at least one intersection road data, and then the prompt module 140 sends a related prompt message or prompt voice for the user to use all the intersecting road data currently matched. In, choose the desired intersection path.

Please refer to FIG. 2 , which is a flowchart of a navigation method according to an embodiment of the present invention. Please also refer to Figure 1 for your understanding. This navigation method includes the following steps:

Identify one of the regional voices sent by the user and analyze a destination domain (step Step S110). Referring to FIG. 3 simultaneously, the local voice sent by the user is obtained by using the microphone 130 (step S111), and the voice recognition module 120 reads the voice data of the regional database 111 to identify the area by using a plurality of voice data. Voice (step S112).

Based on the identification result, it is determined whether the regional database 111 stores at least one matching geographical information (step S113). As described above, if the area data stored in the area database 111 is a place name, the voice database 150 must store voice data corresponding to the place name. The speech recognition module 120 compares the regional speech and the speech data one by one, and obtains the speech data with the highest similarity, and then finds the regional data corresponding to the speech data from the regional database 111.

If the result of the determination is YES, the prompt module 140 is used to prompt the user to match the geographical information (step S114). For example, the voice recognition module 120 may identify a plurality of regional data due to a user's accent problem. For example, Taichung City and Taitung City have similar sounds, and the voice recognition module 120 may simultaneously identify two regions with similar sounds. data. If the determination is no, the prompting module 140 prompts the user to re-enter the regional voice (step S116), and returns to the microphone 130 to acquire the regional voice sent by the user (step S111).

Thereafter, the microphone 130 is used to obtain one of the user input voices, and the selected voice is recognized to select the destination domain from all the matched region data (step S115).

Identifying one of the road voices sent by the user and analyzing a destination road (step S120), please refer to FIG. 4 and FIG. 3 simultaneously, and the detailed flow shown in FIG. The process is similar. First, the road voice sent by the user is acquired by the microphone 130 (step S121), and the voice recognition module 120 reads the voice data of the road database 112 to recognize the road voice by using a plurality of voice data (step S122).

Based on the identification result, it is determined whether the road database 112 stores at least one matching road data (step S123). As described above, if the road data stored in the road database 112 is a road name, the voice database 150 must store voice data corresponding to the road name. The speech recognition module 120 compares the road voice and the voice data one by one, and obtains the voice data with the highest similarity, and then finds the road data corresponding to the voice data from the regional database 111.

If the result of the determination is YES, the prompt module 140 is used to prompt the user to match the road data (step S124). For example, the voice recognition module 120 may recognize a plurality of road materials due to a user's accent problem, such as Jianguo Road and Jiangong Road, and the sound recognition module 120 may recognize two sounds at the same time. Road information. Thereafter, the microphone 130 is used to obtain one of the user input voices to select the destination road from all the matched area data (step S125).

If the determination is no, the prompting module 140 prompts the user to re-enter the road voice (step S126), and returns to use the microphone 130 to acquire the road voice issued by the user (step S121).

The identification user sends out a house number voice, and analyzes a destination card (step S130). Referring to FIG. 5, the user 130 is first used to obtain the user. The house number voice (step S131), the voice recognition module 120 reads the voice data of the house database 113 to identify the house voice with a plurality of voice data (step S132).

Based on the identification result, it is determined whether the house number library 113 stores at least one matching house number data (step S133). As described above, if the house data stored in the house database 113 includes various house numbers, that is, a plurality of digital combinations of numbers 1 to 9, the voice database 150 must store voice data corresponding to the house number or combination of numbers. The voice recognition module 120 compares the voice and voice data of the house card one by one, and obtains the voice data with the highest similarity, and then finds the house data corresponding to the voice data from the house database 113.

If the result of the determination is YES, the prompt module 140 is used to prompt the user to match the house number data (step S134). Thereafter, the microphone 130 is used to obtain one of the user input voices, and the selected voice is recognized to select the destination house number from all the matched house materials (step S135).

If the determination is no, the voice data of one of the highest recognition rates is analyzed, and the house data corresponding to the voice data of the highest recognition rate is used as the destination card (step S136).

In addition, the user can use the microphone 130 to input a crossroad voice. The voice recognition module 120 determines whether the road database 112 stores related crossroad data according to the voice data, and then obtains related prompts by the prompt module 140. The microphone 130 inputs the selected voice and analyzes the desired one of the intersecting roads to increase the reference condition of the navigation module 160 to plan the navigation path.

A navigation path is planned according to the destination domain, the destination road, and the destination house number (step S140). Referring to FIG. 6, the navigation module 160 determines a destination according to the destination domain, the destination road, and the destination door (step S141), and uses the prompt module 140 to inquire whether the destination is correct (step S142). If the user inputs the selected voice by using the button module 170 or the microphone 130 to confirm that the destination is correct, the navigation module 160 plans a navigation path according to the current location and the destination (step S143).

When the confirmation destination is not correct, the prompt module 140 prompts the user to re-enter the regional voice (step S144), and returns to the microphone 130 to acquire the regional voice sent by the user (step S110).

Please refer to FIG. 7 , which is a schematic structural diagram of a second embodiment of the present invention. The difference from the first embodiment lies in that the voice data corresponding to the regional data, the road data and the house data are stored in the regional database 111 and the road data. Library 112 and house library 113. Therefore, when the voice recognition module 120 obtains the regional voice, the road voice, and the house voice, the voice data of the local database 111, the road database 112 and the house database 113 are compared to identify the matching region. Information, road data and house information, for users to select information such as destination domain, destination road and destination house number.

While the present invention has been described above in terms of the preferred embodiments thereof, it is not intended to limit the invention, and the equivalent of the modification and retouching of the present invention is still within the spirit and scope of the present invention. Within the scope of patent protection of the present invention.

110‧‧‧Map database

111‧‧‧Regional database

112‧‧‧Road database

113‧‧‧Household database

120‧‧‧Voice recognition module

130‧‧‧Microphone

140‧‧‧ prompt module

150‧‧‧Voice Database

160‧‧‧Navigation module

170‧‧‧Key Module

1 is a schematic structural diagram of a first embodiment of the present invention; FIG. 2 is a flowchart of a navigation method according to an embodiment of the present invention; FIG. 3 is a flowchart of matching geographical regions according to an embodiment of the present invention; FIG. 5 is a flow chart showing a matching of the door data according to the embodiment of the present invention; FIG. 6 is a flow chart of the navigation confirmation according to the embodiment of the present invention; and FIG. 7 is a schematic structural view of the second embodiment of the present invention.

110‧‧‧Map database

111‧‧‧Regional database

112‧‧‧Road database

113‧‧‧Household database

120‧‧‧Voice recognition module

130‧‧‧Microphone

140‧‧‧ prompt module

150‧‧‧Voice Database

160‧‧‧Navigation module

170‧‧‧Key Module

Claims (16)

A segmented speech recognition navigation system, comprising: a regional database for recording at least one regional data; a road database for recording at least one road data of the regional data; a house database for recording a plurality of the road data a phonetic data library, which records a plurality of voice data corresponding to the geographical information, the road data and the house data; a voice recognition module, which uses the voice data to identify a regional voice, and first determines the regional data. Whether the library stores at least one regional data, identifies a road voice, determines whether the road database stores at least one road data, and then identifies a house voice, and determines whether the house database stores at least one matching card data. And identifying a selected voice to select a destination domain, a destination road and a destination door from all the matched geographical data, the matched road data and the matched house data; a prompting module prompting the voice sequentially The area data, the road data and the house information determined by the identification module to be matched; a wind, obtaining the local voice, the road voice and the house voice, and the user inputting the selected voice; and a navigation module integrating the destination domain, the destination road and the destination door to form a Navigation path. The segmented voice recognition navigation system according to claim 1, wherein the prompt module is a display, and the voice recognition mode is displayed. All the regional data, the road data and the plurality of prompt texts corresponding to the house data are outputted by the group. The segmented voice recognition navigation system of claim 1, wherein the prompt module is a head-up display, which displays all the regional data output by the voice recognition module, and the road data corresponds to the house data. Multiple prompt texts. The segmented voice recognition navigation system of claim 1, wherein the prompt module is a voice player, which plays all the regional data, the road data and the house number output by the voice recognition module. The data corresponds to a plurality of prompt voices. The segmented voice recognition navigation system of claim 1, further comprising a button module, wherein the user selects all the regional data, the road data and the house data output by the voice recognition module. . The segmented voice recognition navigation system of claim 1, wherein the road database further comprises a plurality of intersecting road data, wherein the microphone obtains one of the user input cross road voices for the voice recognition module Identifying, determining whether the road database stores at least one intersecting road data. A segmented speech recognition navigation system, comprising: a regional database, recording at least one regional data and corresponding voice data; a road database, recording at least one road data of the regional data and corresponding voice data; a card database for recording a plurality of house data of the road data and corresponding voice data; a voice recognition module, first using the voice data of the regional database to identify a regional voice, and determining whether the regional database stores and matches At least one regional data, and then using the voice data of the road database to identify a road voice, determining whether the road database stores at least one road data, and then using the voice data of the house database to identify a house card Voice, determining whether the house database stores at least one of the matching house materials, and identifying a selected voice to select a destination domain, a destination from each of the matched regional data, the matched road data, and the matched house data. a road and a destination door; a reminder module prompting all the regional data, the road data and the house data output by the voice recognition module; a microphone, obtaining the local voice, the road voice and the house number issued by the user Voice, and for the user to input the selected voice; and a navigation module to integrate the Geographical, this object is the road to form a house object navigation route. The segmented voice recognition navigation system of claim 7, wherein the prompt module is a display, which displays all the regional data output by the voice recognition module, and the road data corresponds to the house data. Multiple prompt texts. For example, the segmented voice recognition navigation system described in claim 7 is a head-up display, which displays the voice recognition mode. All the regional data, the road data and the plurality of prompt texts corresponding to the house data are outputted by the group. The segmented voice recognition navigation system of claim 7, wherein the prompting module is a voice player, and all the regional data, the road data and the house number output by the voice recognition module are played. The data corresponds to a plurality of prompt voices. The segmented voice recognition navigation system of claim 7, further comprising a button module, wherein the user selects all the regional data, the road data and the house data output by the voice recognition module. . The segmented voice recognition navigation system of claim 7, wherein the road database further comprises a plurality of intersecting road data, wherein the microphone obtains one of the user input cross road voices for the voice recognition module Identifying, determining whether the road database stores the matched at least one cross road data, and prompting the matching cross road data through the prompt module. A segmented speech recognition navigation method, comprising at least the following steps: identifying a user to send a regional voice, and using a microphone to obtain the regional voice sent by the user; using a plurality of voice data to identify the regional voice; Whether the regional database stores at least one matching geographical data to determine whether to prompt the user to match the geographical data; use the microphone to obtain one of the user input selection voices, and Identifying the selected voice to select a destination domain from all the matched regional data; identifying a user to issue a road voice, analyzing a destination road; identifying a user to issue a door voice, analyzing a destination number; and The destination domain, the destination road, and the destination doorplate plan a navigation path. The segmented speech recognition navigation method according to claim 13 , wherein the identifying user sends out a road voice, and analyzing the target road step comprises the steps of: obtaining the road issued by the user by using a microphone Voice; identifying the road voice by using a plurality of voice data; determining whether a road database stores the matched at least one road data to determine whether to output the matched road data; and using the microphone to obtain one of the user input selection voices, The selected voice is identified to select the destination road from all of the matched road data. The segmented speech recognition navigation method according to claim 13, wherein the identifying user sends out a house number voice, and analyzing the target door step comprises the steps of: obtaining the road issued by the user by using a microphone Voice; using a plurality of voice data to identify the house voice; Determining whether a house database stores at least one of the matching house materials to determine whether to output the matched house data; and using the microphone to obtain one of the user input selection voices, and identifying the selected voice to select from all the selected house cards The data is selected for the purpose of the house number. The segmented speech recognition navigation method of claim 13, further comprising the step of recognizing a cross-road voice of the user input to analyze a destination crossroad.