A noise filtering system and method

Abstract

The present invention discloses a noise screening system, which includes an audio acquisition device. The output end of the audio acquisition device is bidirectionally electrically connected to a terminal processor. The output end of the terminal processor is bidirectionally electrically connected to a noise direction twin model establishment module. The output end of the noise direction twin model establishment module is bidirectionally electrically connected to a convolutional learning system. In the present invention, a plurality of audio acquisition devices simultaneously collect surrounding audio and transmit the data to the terminal processor. The terminal processor receives the data and calculates based on the noise intensity collected by the audio acquisition devices in different directions. The terminal processor confirms the direction and displays the noise orientation through a human-computer interaction module, solving the problem that the existing noise screening system does not have the effect of collecting and judging the noise direction, cannot effectively intercept the source direction according to the noise range, seriously affecting the accuracy of audio acquisition and increasing the difficulty of audio noise elimination.

Description

Technical Field

[0001] The present invention relates to the technical field of noise filtering, and specifically to a noise filtering system and method. Background Art

[0002] In order to reduce the influence of surrounding noise on audio acquisition, a noise filtering system is installed at the output end of some audio acquisition devices. However, the existing noise filtering system does not have the effect of collecting and judging the direction of noise, and cannot effectively intercept the source direction according to the noise range, seriously affecting the accuracy of audio acquisition and increasing the difficulty of audio noise elimination. Summary of the Invention

[0003] To solve the problems raised in the above background art, the purpose of the present invention is to provide a noise filtering system and method, which has the advantage of judging the direction of noise, and solves the problems that the existing noise filtering system does not have the effect of collecting and judging the direction of noise, cannot effectively intercept the source direction according to the noise range, seriously affects the accuracy of audio acquisition, and increases the difficulty of audio noise elimination.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: A noise filtering system and method, including an audio acquisition device;

[0005] The output end of the audio acquisition device is bidirectionally electrically connected to a terminal processor, the output end of the terminal processor is bidirectionally electrically connected to a noise direction twin model establishment module, the output end of the noise direction twin model establishment module is bidirectionally electrically connected to a convolutional learning system, the output end of the terminal processor is bidirectionally electrically connected to a signal processing system, the output end of the convolutional learning system is bidirectionally electrically connected to the input end of the signal processing system, the output end of the signal processing system is bidirectionally electrically connected to a wireless transceiver module, and the output end of the wireless transceiver module is bidirectionally electrically connected to a cloud server.

[0006] Preferably, the output end of the audio acquisition device is bidirectionally electrically connected to a band screening unit, the output end of the band screening unit is electrically connected to the input end of the terminal processor, and the band screening unit is composed of a filter and a control circuit.

[0007] Preferably, the terminal processor is composed of an azimuth calculation system and an audio refresh system.

[0008] Preferably, the input end of the terminal processor is bidirectionally electrically connected to a human-computer interaction module, and the human-computer interaction module is composed of a display circuit and a key circuit.

[0009] Preferably, the noise direction twin model establishment module is composed of an audio direction logic model, an audio direction concept model, and a data integration model.

[0010] Preferably, the convolutional learning system of the present invention is composed of a convolutional neural network, an autoencoder neural network, and a deep belief neural network.

[0011] Preferably, the signal processing system of the present invention is composed of a signal amplifier, an interference cancellation module, and a signal noise reduction unit, and the cloud server is composed of long short-term network data and cloud storage data.

[0012] A noise screening system and method, comprising the following steps:

[0013] S1: Install a number of audio collection devices on the same horizontal plane. The audio collection devices simultaneously collect the surrounding audio and transmit the data to the terminal processor. The terminal processor receives the data and calculates according to the noise intensity collected by the audio collection devices in different directions;

[0014] S2: When the noise distances collected by each audio collection device are independently calculated and confirmed, the terminal processor transmits the data to the noise direction twin model establishment module. The noise direction twin model establishment module imports the noise direction into the audio direction logic model and the audio direction concept model through the data integration model. After the noise direction twin model is established, the terminal processor confirms the direction and displays the noise orientation through the human-computer interaction module;

[0015] S3: During the operation of the noise direction twin model establishment module, the convolutional learning system can collect data and use the signal processing system to extract and verify the long short-term network data and cloud storage data stored in the cloud server. At the same time, the convolutional learning system connected to the cloud server can store information for a long time and involve the gradient descent process during backpropagation through time, increasing the operation efficiency of the noise direction twin model establishment module. The user can also control the acquisition range of the filter through the key circuit, so that the noise screening system can independently judge the directions of different band noises.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0017] 1. In the present invention, a number of audio collection devices simultaneously collect the surrounding audio and transmit the data to the terminal processor. The terminal processor receives the data and calculates according to the noise intensity collected by the audio collection devices in different directions. The terminal processor confirms the direction and displays the noise orientation through the human-computer interaction module, solving the problem that the existing noise screening system does not have the effect of collecting and judging the noise direction, cannot effectively intercept the source direction according to the noise range, seriously affecting the audio collection accuracy and increasing the difficulty of audio noise cancellation.

[0018] 2. By setting a band screening unit, the present invention enables users to screen the frequency bands of the collected noise, meeting the requirements for azimuth discrimination of noises with different frequencies.

[0019] 3. By setting an azimuth calculation system and an audio refresh system, the present invention can improve the efficiency of noise azimuth calculation and increase the noise collection frequency at the same time.

[0020] 4. By setting a human-computer interaction module, the present invention enables users to operate the noise screening system, reducing the operation difficulty of the noise screening system.

[0021] 5. By setting an audio direction logic model, an audio direction concept model, and a data integration model, the present invention can integrally display the direction intensity of surrounding noises and can accurately confirm the range by establishing a twin model.

[0022] 6. By setting a convolutional neural network, an autoencoder neural network, and a deep belief neural network, the present invention can improve the operation efficiency of the terminal processor and increase the judgment accuracy of noise signals.

[0023] 7. By setting a signal amplifier, an interference cancellation module, and a signal noise reduction unit, the present invention can stably transmit remote signals. By setting long short-term network data and cloud storage data, the present invention can record operation data, facilitating the long-term preservation of information by the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of the system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with 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 the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] As Figure 1 shown, a noise screening system provided by the present invention includes an audio collection device;

[0027] The output end of the audio acquisition device is bidirectionally electrically connected to a terminal processor. The output end of the terminal processor is bidirectionally electrically connected to a noise direction twin model establishment module. The output end of the noise direction twin model establishment module is bidirectionally electrically connected to a convolutional learning system. The output end of the terminal processor is bidirectionally electrically connected to a signal processing system. The output end of the convolutional learning system is bidirectionally electrically connected to the input end of the signal processing system. The output end of the signal processing system is bidirectionally electrically connected to a wireless transceiver module. The output end of the wireless transceiver module is bidirectionally electrically connected to a cloud server.

[0028] Reference Figure 1 , the output end of the audio acquisition device is bidirectionally electrically connected to a band screening unit. The output end of the band screening unit is bidirectionally electrically connected to the input end of the terminal processor. The band screening unit consists of a filter and a control circuit.

[0029] As a technical optimization scheme of the present invention, by setting the band screening unit, it is convenient for the user to screen the frequency bands of the collected noise, meeting the azimuth discrimination requirements of different frequency noises.

[0030] Reference Figure 1 , the terminal processor consists of an azimuth calculation system and an audio refresh system.

[0031] As a technical optimization scheme of the present invention, by setting the azimuth calculation system and the audio refresh system, the efficiency of noise azimuth calculation can be improved, and at the same time, the noise acquisition frequency can be increased.

[0032] Reference Figure 1 , the input end of the terminal processor is bidirectionally electrically connected to a human-computer interaction module. The human-computer interaction module consists of a display circuit and a key circuit.

[0033] As a technical optimization scheme of the present invention, by setting the human-computer interaction module, it is convenient for the user to operate the noise screening system, reducing the operation difficulty of the noise screening system.

[0034] Reference Figure 1 , the noise direction twin model establishment module consists of an audio direction logic model, an audio direction concept model, and a data integration model.

[0035] As a technical optimization scheme of the present invention, by setting the audio direction logic model, the audio direction concept model, and the data integration model, the intensity of the surrounding noise direction can be integrally displayed, and at the same time, the range can be accurately confirmed by establishing a twin model.

[0036] Reference Figure 1 , the convolutional learning system consists of a convolutional neural network, an autoencoder neural network, and a deep belief neural network.

[0037] As a technical optimization solution of the present invention, by setting up a convolutional neural network, an autoencoder neural network, and a deep belief neural network, the operating efficiency of the terminal processor can be improved, and the judgment accuracy of noise signals can be increased.

[0038] Reference Figure 1 , the signal processing system is composed of a signal amplifier, an interference cancellation module, and a signal noise reduction unit, and the cloud server is composed of long short-term network data and cloud storage data.

[0039] As a technical optimization solution of the present invention, by setting up a signal amplifier, an interference cancellation module, and a signal noise reduction unit, stable transmission of remote signals can be achieved, and by setting up long short-term network data and cloud storage data, operating data can be recorded to facilitate the system to store information for a long time.

[0040] Reference Figure 1 , a noise screening method includes the following steps:

[0041] S1: Install a number of audio collection devices on the same horizontal plane. The audio collection devices simultaneously collect the surrounding audio and transmit the data to the terminal processor. The terminal processor receives the data and calculates according to the noise intensity collected by the audio collection devices in different directions;

[0042] S2: When the noise distances collected by each audio collection device are independently calculated and confirmed, the terminal processor transmits the data to the noise direction twin model establishment module. The noise direction twin model establishment module imports the noise direction into the audio direction logic model and the audio direction concept model through the data integration model. After the noise direction twin model is established, the terminal processor confirms the direction and displays the noise azimuth through the human-computer interaction module;

[0043] S3: During the operation of the noise direction twin model establishment module, the convolutional learning system can collect data and use the signal processing system to extract and verify the long short-term network data and cloud storage data stored in the cloud server. At the same time, the convolutional learning system connected to the cloud server can store information for a long time and the gradient descent process involved in backpropagation through time, increasing the operation efficiency of the noise direction twin model establishment module. The user can also control the collection range of the filter through the key circuit, so that the noise screening system can independently judge the directions of noises in different frequency bands.

[0044] Working principle and usage process of the present invention: When in use, a certain number of audio collection devices are installed on the platform surface. The audio collection devices simultaneously collect the surrounding audio and transmit the data to the terminal processor. The terminal processor receives the data and calculates based on the noise intensity collected by the audio collection devices in different directions. When the noise distances collected by each audio collection device are independently calculated and confirmed, the terminal processor transmits the data to the noise direction twin model establishment module. The noise direction twin model establishment module imports the noise direction into the audio direction logic model and the audio direction concept model through the data integration model. After the noise direction twin model is established, the terminal processor confirms the direction and displays the noise azimuth through the human-computer interaction module. During the operation of the noise direction twin model establishment module, the convolutional learning system can collect data and use the signal processing system to extract and verify the long-term and short-term network data and cloud storage data stored in the cloud server. At the same time, the convolutional learning system connected to the cloud server can store information for a long time and involve the gradient descent process during backpropagation through time, increasing the operation efficiency of the noise direction twin model establishment module. The user can also control the collection range of the filter through the key circuit, enabling the noise screening system to independently judge the directions of different band noises.

[0045] In summary, for the noise screening system and method, a certain number of audio collection devices simultaneously collect the surrounding audio and transmit the data to the terminal processor. The terminal processor receives the data and calculates based on the noise intensity collected by the audio collection devices in different directions. The terminal processor confirms the direction and displays the noise azimuth through the human-computer interaction module, solving the problem that the existing noise screening system does not have the effect of collecting and judging the noise direction, cannot effectively intercept the source direction according to the noise range, seriously affecting the audio collection accuracy and increasing the difficulty of audio noise elimination.

[0046] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0047] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A noise filtering system, comprising an audio acquisition device; It is characterized in that: The output end of the audio acquisition device is bi-directionally electrically connected to a terminal processor. The output end of the terminal processor is bi-directionally electrically connected to a noise direction twin model establishment module. The output end of the noise direction twin model establishment module is bi-directionally electrically connected to a convolutional learning system. The output end of the terminal processor is bi-directionally electrically connected to a signal processing system. The output end of the convolutional learning system is bi-directionally electrically connected to the input end of the signal processing system. The output end of the signal processing system is bi-directionally electrically connected to a wireless transceiver module. The output end of the wireless transceiver module is bi-directionally electrically connected to a cloud server. The output end of the audio acquisition device is bi-directionally electrically connected to a band screening unit. The output end of the band screening unit is bi-directionally electrically connected to the input end of the terminal processor. The band screening unit consists of a filter and a control circuit. The terminal processor consists of an azimuth calculation system and an audio refresh system. The input end of the terminal processor is bi-directionally electrically connected to a human-computer interaction module. The human-computer interaction module consists of a display circuit and a key circuit. The noise direction twin model establishment module consists of an audio direction logic model, an audio direction concept model, and a data integration model. The convolutional learning system consists of a convolutional neural network, an autoencoder neural network, and a deep belief neural network.

2. The noise filtering system according to claim 1, wherein: The signal processing system consists of a signal amplifier, an interference cancellation module, and a signal noise reduction unit. The cloud server consists of long short-term network data and cloud storage data.

3. The usage method of a noise filtering system according to claim 1, characterized in that: Including the following steps: S1: Install a number of audio acquisition devices on the same horizontal plane. The audio acquisition devices simultaneously collect the surrounding audio and transmit the data to the terminal processor. The terminal processor receives the data and calculates based on the noise intensity collected by the audio acquisition devices in different directions; S2: When the noise distances collected by each audio acquisition device are independently calculated and confirmed, the terminal processor transmits the data to the noise direction twin model establishment module. The noise direction twin model establishment module imports the noise direction into the audio direction logic model and the audio direction concept model through the data integration model. After the noise direction twin model is established, the terminal processor confirms the direction and displays the noise azimuth through the human-computer interaction module; S3: During the operation of the noise direction twin model establishment module, the convolutional learning system can collect data and use the signal processing system to extract and verify the long short-term network data and cloud storage data stored in the cloud server. At the same time, the convolutional learning system connected to the cloud server can store information for a long time and involve the gradient descent process during backpropagation through time, increasing the operation efficiency of the noise direction twin model establishment module. The user can also control the acquisition range of the filter through the key circuit, enabling the noise filtering system to independently judge the directions of noises in different bands.

Images