A voice control system for an identity - recognizable in - vehicle intelligent terminal

Abstract

The present invention relates to the field of voice control technology, and in particular, to a vehicle-mounted intelligent terminal voice control system capable of identifying identity. It comprises a voice receiving port, a noise recognition processing unit and a comparison and analysis processing unit. In the present invention, the comparison and analysis processing unit firstly judges whether there are voice commands similar to the voice quality of the car owner in the noise in combination with the car owner's usual voice quality, and removes the voice commands with large differences, and at the same time judges whether such verbal language appears in the voice commands similar to the car owner's usual voice quality in the noise in combination with the oral language storage information, and further removes irrelevant commands. After the comparison is completed, the effective commands are extracted by the voice command extraction unit, and implemented according to the actual commands. In the case of noise in the car, it can accurately analyze which voice command is issued by the car owner, improve the efficiency of voice command response, and avoid the car owner from repeating voice commands.

Description

Technical Field

[0001] The present invention relates to the technical field of voice control, and more specifically, to a voice control system for an in-vehicle intelligent terminal capable of identifying identity. Background Art

[0002] During the driving process of a vehicle, an in-vehicle intelligent terminal is generally equipped to receive the command information of the vehicle owner, helping the vehicle owner to concentrate on driving the vehicle. Most in-vehicle intelligent terminals receive commands through voice recognition, that is, the vehicle owner issues a voice command to the in-vehicle intelligent terminal, and the in-vehicle intelligent terminal recognizes the voice command and then executes the command content. However, when there are many people in the vehicle, the voice commands are miscellaneous. While the driver issues a command, other people may also talk, making a large amount of noise mixed in the command. At this time, it is difficult for the in-vehicle intelligent terminal to extract the correct command of the vehicle owner from the noise, resulting in a low response efficiency of the in-vehicle intelligent terminal. The vehicle owner needs to repeat the command repeatedly, and during this process, the vehicle owner will be distracted, affecting the vehicle owner's driving of the vehicle. Therefore, there is an urgent need for a voice control system for an in-vehicle intelligent terminal capable of identifying identity. Summary of the Invention

[0003] The purpose of the present invention is to provide a voice control system for an in-vehicle intelligent terminal capable of identifying identity to solve the problems raised in the above background art.

[0004] To achieve the above purpose, a voice control system for an in-vehicle intelligent terminal capable of identifying identity is provided, including a voice receiving port for receiving the voice inside the vehicle. The output end of the voice receiving port is connected to a noise recognition and processing unit for recognizing the impurity voice inside the vehicle and classifying each voice according to different timbres. The output end of the noise recognition and processing unit is connected to a comparison and analysis processing unit. The input end of the comparison and analysis processing unit is connected to a vehicle owner's voice quality storage unit for storing the voice quality of the vehicle owner in normal times. The input end of the vehicle owner's voice quality storage unit is further connected to a vehicle owner's voice habit storage unit for storing the vehicle owner's usual speaking manner and common catchphrases when issuing commands. The comparison and analysis processing unit combines the voice quality of the vehicle owner in normal times, the vehicle owner's usual speaking manner, and the information of common catchphrases when issuing commands, conducts a comparative analysis on the noise, compares each voice in the noise with the pre-stored voice pattern of the vehicle owner, and arranges each voice in the noise according to the similarity. The output end of the comparison and analysis processing unit is connected to a voice command extraction unit.

[0005] As a further improvement of the present technical solution, the noise recognition and processing unit includes a coherent instruction recognition module, which recognizes the instructions that can form coherence in the noise. The output end of the coherent instruction recognition module is connected to an instruction tone recognition module, which recognizes and processes the tones of different coherent instructions. The output end of the instruction tone recognition module is connected to an instruction classification and output module, which is used to classify each different coherent instruction according to the tones of different coherent instructions.

[0006] As a further improvement of the present technical solution, the output end of the voice receiving port is connected to a volume recognition unit, which is used to recognize the different instruction volumes in the noise.

[0007] As a further improvement of the present technical solution, the volume recognition unit includes a different instruction volume measurement module, which is used to measure the volumes of each different instruction in the noise. The output end of the different instruction volume measurement module is connected to a volume threshold preset module, which presets the volume threshold according to the measurement results of the volumes of each different instruction. The output end of the volume threshold preset module is connected to an irrelevant instruction elimination module.

[0008] As a further improvement of the present technical solution, the volume recognition unit adopts a threshold comparison algorithm, and its algorithm formula is as follows:

[0009] A: {a 1 , a 2 , …, a n};

[0010]

[0011]

[0012] Among them, A is the set of instructions with different volumes in the noise, a 1 to a n are the instructions with different volumes in the noise, is the volume threshold, n is the sum of all instructions in the noise, f(a) is the judgment function, a is the judgment volume. When the judgment volume a of the instruction is less than the volume threshold at this time, the judgment function f(a) outputs 0, and it is judged that the instruction is an irrelevant instruction. When the judgment volume a of the instruction is not less than the volume threshold at this time, the judgment function f(a) outputs 1, and it is judged that the instruction is a relevant instruction.

[0013] As a further improvement of the present technical solution, the comparison and analysis processing unit includes a sound quality comparison module, the output end of the sound quality comparison module is connected to a result integration module, the input end of the result integration module is connected to a voice keyword comparison module, and the voice keyword comparison module is used to extract keywords from the noise and compare them with the owner's habitual spoken language.

[0014] As a further improvement of the present technical solution, the output end of the voice instruction extraction unit is connected to a repeating voice instruction unit, and the repeating voice instruction unit is used to repeat the instruction after identifying the vehicle owner.

[0015] As a further improvement of the present technical solution, the repeated voice command unit includes a car owner command receiving module, which is used to receive the car owner commands obtained through analysis, and the output end of the car owner command receiving module is connected to a car owner command copying module, which is used to copy the car owner commands, and the output end of the car owner command copying module is connected to a reread command issuing module.

[0016] As a further improvement of the technical solution, the output end of the repeated voice instruction unit is connected to the input end of the voice receiving port.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] 1. In the identity-recognizable in-vehicle intelligent terminal voice control system, the comparison and analysis processing unit first determines whether there are voice commands with similar voice quality to the car owner's usual voice in the noise based on the car owner's usual voice quality, and eliminates the voice commands with large differences. At the same time, it determines whether such verbal language appears in the voice commands with similar voice quality to the car owner's usual voice in the noise based on the oral language storage information, and further eliminates irrelevant commands. After the comparison is completed, the voice command extraction unit extracts the valid commands and implements them according to the actual commands. In the case of noise in the car, it can accurately analyze which voice command is issued by the car owner, improve the voice command response efficiency, and avoid the car owner from repeating voice commands.

[0019] 2. In the identity-recognizable in-vehicle intelligent terminal voice control system, the volume of each different command is monitored by the volume recognition unit, and each different command is sorted according to the volume, and a volume threshold is set in advance. The command volume in the noise that is lower than the volume threshold will be eliminated. These command volumes below the volume threshold are voice commands issued away from the driving seat, which are definitely not commands issued by the car owner. The command volumes below the volume threshold are eliminated in advance, and irrelevant commands are filtered out of the noise in advance to improve the efficiency of later comparison.

[0020] 3. In the voice control system of the vehicle-mounted intelligent terminal capable of identifying identity, the repeated voice instruction unit repeats and prompts the instruction after the vehicle owner is identified, reminding the vehicle owner whether the identified vehicle owner instruction is correct, performing secondary verification of the voice instruction, and improving the correct rate of instruction feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is the overall flowchart of the present invention;

[0022] Figure 2 is the flowchart of the noise recognition and processing unit of the present invention;

[0023] Figure 3 is the flowchart of the volume recognition unit of the present invention;

[0024] Figure 4 is the flowchart of the comparative analysis and processing unit of the present invention;

[0025] Figure 5 is the flowchart of the repeated voice instruction unit of the present invention

[0026] The meanings of the various reference numerals in the figure are as follows:

[0027] 10. Voice receiving port;

[0028] 20. Noise recognition and processing unit; 210. Coherent instruction recognition module; 220. Instruction tone recognition module; 230. Instruction classification and output module;

[0029] 30. Volume recognition unit; 310. Different instruction volume measurement module; 320. Volume threshold preset module; 330. Irrelevant instruction elimination module;

[0030] 40. Comparative analysis and processing unit; 410. Sound quality comparison module; 420. Result integration module; 430. Voice keyword comparison module;

[0031] 50. Vehicle owner sound quality storage unit;

[0032] 60. Voice habit storage unit;

[0033] 70. Voice instruction extraction unit;

[0034] 80. Repeated voice instruction unit; 810. Vehicle owner instruction receiving module; 820. Vehicle owner instruction copying module; 830. Repeated instruction sending module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0036] Please refer to Figures 1 - 5 As shown, a voice control system for an in-vehicle intelligent terminal capable of identifying identity is provided, including a voice receiving port 10 for receiving voices inside the vehicle. The output end of the voice receiving port 10 is connected to a noise recognition and processing unit 20 for recognizing the impurity voices inside the vehicle and classifying each voice according to different timbres. The output end of the noise recognition and processing unit 20 is connected to a comparison and analysis processing unit 40. The input end of the comparison and analysis processing unit 40 is connected to a vehicle owner's voice quality storage unit 50 for storing the voice quality of the vehicle owner in normal times. The input end of the vehicle owner's voice quality storage unit 50 is further connected to a vehicle owner's voice habit storage unit 60 for storing the vehicle owner's usual speaking manner and common catchphrases when giving instructions. The comparison and analysis processing unit 40 combines the voice quality of the vehicle owner in normal times, the vehicle owner's usual speaking manner, and the information of common catchphrases when giving instructions, conducts a comparative analysis on the noise, compares each voice in the noise with the voice pattern of the pre-stored vehicle owner, and arranges each voice in the noise according to the similarity. The output end of the comparison and analysis processing unit 40 is connected to a voice command extraction unit 70.

[0037] In specific use, when there are more than one vehicle owner in the vehicle, the voice commands issued in the vehicle are more complicated, the voice receiving port 10 receives the voice in the vehicle, converts the voice signal issued in the vehicle into an electrical signal that can be recognized by the on-board intelligent terminal, generates voice reception information, and transmits the voice reception information to the noise recognition processing unit 20, the noise recognition processing unit 20 recognizes the impure voice in the vehicle, and classifies each voice according to different sound qualities, generates voice classification information, and transmits the voice classification information to the comparison analysis processing unit 40, the owner's voice quality storage unit 50 stores the owner's usual voice The voice quality is stored. When there is only the owner in the car, the voice command issued by the owner can be judged as issued by the owner himself through the owner voice quality storage unit 50, and the vehicle-mounted intelligent terminal directly responds to the voice command. When there are more than one passengers in the vehicle, the owner voice quality storage unit 50 generates the owner's usual voice quality and transmits the owner's usual voice quality to the comparison and analysis processing unit 40. The comparison and analysis processing unit 40 first determines whether there is a voice command with a similar voice quality to the owner's usual voice quality in the noise based on the owner's usual voice quality, and removes the voice commands with large differences (for example: the owner is a male , the sound quality is thick, while there is a female voice quality in the noise, the sound quality is crisp), at this time, this type of sound quality will be marked as irrelevant sound quality and eliminated. When two or more voice commands with similar voice quality to the owner's usual voice appear in the noise, the owner's voice habit storage unit 60 stores the owner's usual speaking style and the common spoken language when issuing commands (for example, when the owner usually issues commands, he will say "hello" or specific pronouns in advance), and systematically stores such spoken language, generates spoken language storage information, and transmits the spoken language storage information to the comparison analysis processing unit 40. The comparison analysis processing unit 40 concludes The spoken language storage information is combined to determine whether such spoken language appears in the voice instructions with similar voice quality to the owner's usual voice in the noise, and irrelevant instructions are further eliminated. After the comparison is completed, the comparison and analysis processing unit 40 transmits the irrelevant instructions and the final recognized valid instructions (determined to be the instructions issued by the owner himself) to the voice instruction extraction unit 70. The voice instruction extraction unit 70 extracts the valid instructions and implements them according to the actual instructions. In the event of noise in the car, it can accurately analyze which voice instruction is issued by the owner, thereby improving the voice instruction response efficiency and avoiding the owner from repeating voice instructions.

[0038] In addition, the noise recognition processing unit 20 includes a coherent instruction recognition module 210 that recognizes instructions that can form a coherent sequence in the noise. The output end of the coherent instruction recognition module 210 is connected to an instruction tone recognition module 220 that recognizes and processes the tones of different coherent instructions. The output end of the instruction tone recognition module 220 is connected to an instruction classification output module 230 that classifies each different coherent instruction according to the tones of different coherent instructions. In specific use, the coherent instruction recognition module 210 recognizes instructions in the noise that can form a coherent and grammatically correct instruction sentence and can convey direct information, generates coherent instruction recognition information, and transmits the coherent instruction recognition information to the instruction tone recognition module 220. The instruction tone recognition module 220 recognizes and processes the tones of different coherent instructions, generates instruction tone recognition information, and transmits the instruction tone recognition information to the instruction classification output module 230. The instruction classification output module 230 classifies each different coherent instruction according to the tones of different coherent instructions to distinguish the coherent instructions with different tones in the noise.

[0039] Furthermore, the output end of the voice receiving port 10 is connected to a volume recognition unit 30 that recognizes the volumes of different instructions in the noise. In specific use, the voice receiving port 10 transmits the noise information to the volume recognition unit 30. The volume recognition unit 30 monitors the volumes of each different instruction, sorts each different instruction according to the volume size, and sets a volume threshold in advance. Instructions in the noise with a volume lower than the volume threshold will be excluded. These instructions with a volume lower than the volume threshold are voice instructions issued far from the driver's seat and are definitely not instructions issued by the vehicle owner. Excluding these instructions with a volume lower than the volume threshold in advance filters out irrelevant instructions in the noise and improves the later comparison efficiency.

[0040] Furthermore, the volume recognition unit 30 includes a different instruction volume measurement module 310 for measuring each different instruction volume in the background noise. The output end of the different instruction volume measurement module 310 is connected to a volume threshold presetting module 320, which presets the volume threshold according to the measurement results of each different instruction volume. The output end of the volume threshold presetting module 320 is connected to an irrelevant instruction elimination module 330. During specific use, the different instruction volume measurement module 310 measures each different instruction volume in the background noise, generates volume measurement information, and transmits the volume measurement information to the volume threshold presetting module 320. The volume threshold presetting module 320 presets the volume threshold according to the measurement results of each different instruction volume, generates volume threshold information, and transmits the volume threshold information to the irrelevant instruction elimination module 330. The irrelevant instruction elimination module 330 classifies the instructions in the background noise that are lower than the volume threshold as irrelevant instructions and eliminates these irrelevant instructions.

[0041] Specifically, the volume recognition unit 30 adopts a threshold comparison algorithm, and its algorithm formula is as follows:

[0042] A: {a 1 , a 2 , …, a n};

[0043]

[0044]

[0045] Among them, A is the set of instructions with different volumes in the background noise, a 1 to a n are the instructions with different volumes in the background noise, is the volume threshold, n is the sum of all instructions in the background noise, f(a) is the judgment function, a is the judgment volume. When the judgment volume a of the instruction is less than the volume threshold at this time, the judgment function f(a) outputs 0, and it is judged that the instruction is an irrelevant instruction. When the judgment volume a of the instruction is not less than the volume threshold at this time, the judgment function f(a) outputs 1, and it is judged that the instruction is a relevant instruction.

[0046] In addition, the comparison and analysis processing unit 40 includes a sound quality comparison module 410. The output end of the sound quality comparison module 410 is connected to a result integration module 420. The input end of the result integration module 420 is connected to a voice keyword comparison module 430. The voice keyword comparison module 430 is used to extract keywords in the background noise and compare them with the owner's habitual spoken language. Specifically, when in use, the sound quality comparison module 410 compares different instruction sound qualities in the background noise with the pre-stored sound quality of the owner, pre-stores the different instruction sound qualities similar to the owner's sound quality, generates sound quality comparison information. At the same time, the voice keyword comparison module 430 extracts keywords in the background noise, compares them with the owner's habitual spoken language, generates keyword comparison information, and transmits the keyword comparison information to the result integration module 420, which conducts comparison and integration.

[0047] Furthermore, the output end of the voice command extraction unit 70 is connected to a repeated voice command unit 80. The repeated voice command unit 80 is used to repeat and prompt the command after the owner is recognized. When the vehicle-mounted intelligent terminal recognizes the command issued by the owner from the background noise, the repeated voice command unit 80 repeats and prompts the command after the owner is recognized, reminding the owner whether the recognized owner command is correct, conducting a secondary verification of the voice command, and improving the correct rate of command feedback.

[0048] Still further, the repeated voice command unit 80 includes an owner command receiving module 810. The owner command receiving module 810 is used to receive the analyzed owner command. The output end of the owner command receiving module 810 is connected to an owner command copying module 820. The owner command copying module 820 is used to copy the owner command. The output end of the owner command copying module 820 is connected to a repeated reading command issuing module 830. Specifically, when in use, the owner command receiving module 810 receives the analyzed owner command, generates owner command receiving information, and transmits the command receiving information to the owner command copying module 820. The owner command copying module 820 copies the owner command, generates an owner command copy, and transmits the owner command copy to the repeated reading command issuing module 830. The repeated reading command issuing module 830 conducts a repeated reading of the owner's voice according to the owner command copy, reminding the owner whether the recognized voice is the owner's own intention and verifying the correctness of the recognized owner voice.

[0049] In addition, the output end of the repeated voice command unit 80 is connected to the input end of the voice receiving port 10. The repeated voice command unit 80 repeats and prompts the command after the owner is recognized, and transmits the repeated prompt of the command after the owner is recognized to the voice receiving port 10. When the owner receives the repeated prompt, the owner determines the command through the voice receiving port 10, thereby further improving the correct rate of command response.

[0050] The basic principles, main features and advantages of the present invention have been shown and described above. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification are only preferred examples of the present invention and are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. An identity - recognizable in - vehicle intelligent terminal voice control system, including a voice receiving port (10), and the voice receiving port (10) is used for receiving voices inside the vehicle. Characterized in that: The output end of the voice receiving port (10) is connected to a noise recognition and processing unit (20). The noise recognition and processing unit (20) is used for recognizing impurity voices inside the vehicle and classifying each voice according to different timbres. The output end of the noise recognition and processing unit (20) is connected to a comparison and analysis processing unit (40). The input end of the comparison and analysis processing unit (40) is connected to an owner voice quality storage unit (50). The owner voice quality storage unit (50) stores the voice quality of the owner in normal times. The input end of the owner voice quality storage unit (50) is further connected to an owner voice habit storage unit (60). The owner voice habit storage unit (60) is used for storing the owner's usual speaking manner and common catchphrases when giving instructions. The comparison and analysis processing unit (40) combines the voice quality of the owner in normal times, the owner's usual speaking manner, and the information of common catchphrases when giving instructions, conducts a comparative analysis on the noise, compares each voice in the noise with the voice pattern of the pre - stored owner, and arranges each voice in the noise according to the similarity. The output end of the comparison and analysis processing unit (40) is connected to a voice command extraction unit (70); The noise recognition and processing unit (20) includes a coherent command recognition module (210). The coherent command recognition module (210) recognizes the commands that can form coherence in the noise. The output end of the coherent command recognition module (210) is connected to a command timbre recognition module (220). The command timbre recognition module (220) recognizes and processes different coherent command timbres. The output end of the command timbre recognition module (220) is connected to a command classification output module (230). The command classification output module (230) is used for classifying each different coherent command according to different coherent command timbres.

2. The identity - recognizable in - vehicle intelligent terminal voice control system according to claim 1, Characterized in that: The output end of the voice receiving port (10) is connected to a volume recognition unit (30). The volume recognition unit (30) is used for recognizing different command volumes in the noise.

3. The identity - recognizable in - vehicle intelligent terminal voice control system according to claim 2, Characterized in that: The volume recognition unit (30) includes a different command volume measurement module (310). The different command volume measurement module (310) is used for measuring each different command volume in the noise. The output end of the different command volume measurement module (310) is connected to a volume threshold presetting module (320). The volume threshold presetting module (320) presets a volume threshold according to the measurement results of each different command volume. The output end of the volume threshold presetting module (320) is connected to an irrelevant command elimination module (330).

4. The voice control system of the vehicle-mounted intelligent terminal capable of identifying identity according to claim 3, wherein: The volume recognition unit (30) adopts a threshold comparison algorithm, and its algorithm formula is as follows: ； ； ； Among them, A is a set of instructions with different volumes in the noise, to are instructions with different volumes in the noise, is the volume threshold, is the sum of all instructions in the noise, is the judgment function, is the judgment volume. When the judgment volume of the instruction is less than the volume threshold , at this time the judgment function outputs 0, and it is judged that the instruction is an irrelevant instruction. When the judgment volume of the instruction is not less than the volume threshold , at this time the judgment function outputs 1, and it is judged that the instruction is a relevant instruction.

5. The voice control system of the vehicle-mounted intelligent terminal capable of identifying identity according to claim 1, wherein: The comparison and analysis processing unit (40) includes a sound quality comparison module (410). The output end of the sound quality comparison module (410) is connected to a result integration module (420). The input end of the result integration module (420) is connected to a voice keyword comparison module (430). The voice keyword comparison module (430) is used to extract keywords in the background noise and compare them with the owner's habitual catchphrases.

6. The voice control system of the vehicle-mounted intelligent terminal capable of identifying identity according to claim 1, wherein: The output end of the voice command extraction unit (70) is connected to a repeated voice command unit (80). The repeated voice command unit (80) is used to repeat and prompt the command after identifying the owner.

7. The voice control system of the vehicle-mounted intelligent terminal capable of identifying identity according to claim 6, wherein: The repeated voice command unit (80) includes an owner command receiving module (810). The owner command receiving module (810) is used to receive the owner command obtained through analysis. The output end of the owner command receiving module (810) is connected to an owner command copying module (820). The owner command copying module (820) is used to copy the owner command. The output end of the owner command copying module (820) is connected to a repeated command issuing module (830).

8. The voice control system of the vehicle-mounted intelligent terminal capable of identifying identity according to claim 7, wherein: The output end of the repeated voice command unit (80) is connected to the input end of the voice receiving port (10).

Images