A method for pressing a sound reduction method of a blasting ball
By setting up multiple sound pickup devices in the popping bead pressing device and combining algorithms and models to process the popping bead breaking sound, the problems of echo and noise in existing devices are solved, achieving a highly efficient noise reduction effect.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHONGQING CHINA TOBACCO IND CO LTD
- Filing Date
- 2022-01-26
- Publication Date
- 2026-05-19
AI Technical Summary
Existing press-to-burst bead devices lack noise reduction and damping functions when picking up the sound of the bursting beads, resulting in severe echo and noise interference.
The sound of the popping bead being broken is acquired using a first and a second sound pickup device. The signal is then processed using a linear echo cancellation algorithm and a Butterworth filter. Combined with an initial recurrent neural network model, the cancellation signal is determined and echoes and noise are removed to obtain a noise-reduced sound signal.
It effectively removes echoes and noise from the popping sound of the beads, improving the clarity and quality of the sound signal.
Smart Images
Figure CN114627845B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of sound noise reduction technology, and in particular to a method for sound noise reduction using a popping bead. Background Technology
[0002] With the increasing use of flavor capsules in cigarettes, the control and evaluation of their quality has become increasingly important. The quality of flavor capsules is mainly evaluated by testing their chemical and physical properties. Regarding chemical properties, there are reports on the detection of contents within the flavor capsule, the detection of formaldehyde and acetaldehyde in the flavor capsule, and the detection of pigments in the wall material. However, regarding physical property testing, there are only reports on devices and methods for testing the bursting strength and deformation of flavor capsules. Since the process of bursting the flavor capsule also brings consumers a richer and more personalized experience, the study of the sound signal of the flavor capsule bursting is also a key focus of flavor capsule physical property research. However, existing devices for pressing the flavor capsule and picking up the bursting sound are usually small devices without sound damping or noise reduction functions, and the picked-up bursting sound will have echoes and noise. Summary of the Invention
[0003] The main objective of this invention is to propose a method for noise reduction of the sound of a popping bead being pressed, which aims to reduce the noise of the popping bead breaking sound and remove the echo and noise in the sound of the popping bead breaking.
[0004] To achieve the above objectives, the present invention provides a method for noise reduction of a popping bead by pressing, wherein the popping bead is placed in a popping bead pressing device, and a first sound pickup device and a second sound pickup device are provided within the popping bead pressing device. The method for noise reduction of a popping bead by pressing includes the following steps:
[0005] The pickup position information of the first and second pickup devices is obtained, and the first and second popping bead breaking sounds are obtained based on the first and second pickup devices. The pickup position information includes the distance between the first and second pickup devices and the popping bead, and the angle between the first and second pickup devices. The first popping bead breaking sound and the second popping bead breaking sound are obtained by the first and second pickup devices respectively.
[0006] Using the sounds of the first and second popping beads breaking as samples, the microphone signal is processed using a linear echo cancellation algorithm to obtain an echo signal.
[0007] The first and second popping bead breaking sounds are subjected to high-frequency filtering based on the Butterworth filtering program to obtain the first filtered sound signal and the second filtered sound signal.
[0008] The first cancellation signal and the second cancellation signal are determined based on the pickup location information and the echo signal;
[0009] The first filtered sound signal is processed by the first cancellation signal to obtain the first noise-reduced sound signal, and the second filtered sound signal is processed by the second cancellation signal to obtain the second noise-reduced sound signal.
[0010] When the waveform overlap between the first noise-reduced sound signal and the second noise-reduced sound signal exceeds a preset threshold, the first noise-reduced sound signal is determined to be the noise-reduced popping bead sound.
[0011] Optionally, the step of performing high-frequency filtering on the first and second burst bead sounds based on a Butterworth filtering procedure to obtain a first filtered sound signal and a second filtered sound signal includes:
[0012] The first sampling frequency, the first fluctuation frequency, and the first filter order of the first pickup device are obtained;
[0013] The second sampling frequency, the second fluctuation frequency, and the second filter order of the second pickup device are obtained;
[0014] The first filter amplitude is calculated based on the first sampling frequency, the first fluctuation frequency, and the first filter order.
[0015] The second filter amplitude is calculated based on the second sampling frequency, the second fluctuation frequency, and the second filter order.
[0016] The signal amplitude in the first popping bead sound is replaced with the first filtered amplitude to obtain the first filtered sound signal;
[0017] The signal amplitude in the second popping bead breaking sound is replaced with the second filtered amplitude to obtain the second filtered sound signal.
[0018] Optionally, the step of calculating the first filter amplitude based on the first sampling frequency, the first fluctuation frequency, and the first filter order includes:
[0019] Calculate the first cutoff frequency based on the first sampling frequency and the first fluctuation frequency;
[0020] The signal amplitude in the sound of the first popping bead breaking is subjected to high-frequency filtering based on the first cutoff frequency and the filtering order to obtain the first filtered amplitude.
[0021] Optionally, the first cutoff frequency Among them, f c f is the first fluctuation frequency. s This is the first sampling frequency;
[0022] The square of the first filter amplitude Where N is the first filter order.
[0023] Optionally, the step of determining the first cancellation signal and the second cancellation signal based on the pickup location information and the echo signal includes:
[0024] Obtain an initial recurrent neural network model, which includes multiple original matrices as parameters;
[0025] Decompose at least one of the original matrices to obtain the matrix parameters formed by the multiplication formula of the two component matrices;
[0026] The initial recurrent neural network model is trained to obtain the recurrent neural network model;
[0027] The pickup location information and echo signal are input into the recurrent neural network model to obtain the first cancellation signal and the second cancellation signal.
[0028] Optionally, training the initial recurrent neural network model to obtain the recurrent neural network model includes:
[0029] Acquire training data, which includes near-sound signals of the sample, ambient noise signals of the sample, residual echo signals of the sample, and output audio signals of the sample.
[0030] Obtain an ideal ratio mask based on the training data;
[0031] The initial recurrent neural network model is trained based on the ideal ratio mask and the training data to obtain the recurrent neural network model, and the output of the recurrent neural network model is fitted with the ideal ratio mask.
[0032] This invention processes the acquired first and second popping bead breakage sounds to obtain echo information, a first filtered sound signal, and a second filtered sound signal. Then, based on the pickup location information and the echo signal, a first cancellation signal and a second cancellation signal are determined. The first filtered sound signal is then processed using the first cancellation signal, and the second filtered sound signal is processed using the second cancellation signal. When the similarity between the processed first and second filtered sound signals is higher than a preset threshold, a noise-reduced popping bead sound is obtained. This achieves noise reduction of the acquired popping bead breakage sound, removing echoes and noise from the popping bead breakage sound. Attached Figure Description
[0033] Figure 1 This is a flowchart illustrating an embodiment of the method for reducing noise from the popping bead sound according to the present invention.
[0034] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0035] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.
[0036] Reference Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the method for reducing the noise of a popping bead by pressing according to the present invention.
[0037] In this embodiment of the invention, the method for reducing the noise of a press-activated menthol capsule is applied to a device for reducing the noise of a press-activated menthol capsule, and the method includes:
[0038] Step S10: Obtain the pickup position information of the first and second pickup devices, and obtain the first and second popping bead breaking sounds based on the first and second pickup devices. The pickup position information includes the distance between the first and second pickup devices and the popping bead, and the angle between the first and second pickup devices. The first popping bead breaking sound and the second popping bead breaking sound are respectively obtained by the first and second pickup devices.
[0039] In this embodiment, to achieve noise reduction of the acquired popping bead sound, echoes and noise are removed from the sound. The popping bead sound noise reduction device acquires the pickup position information of the first and second pickup devices, and acquires the first and second popping bead sounds based on the first and second pickup devices. The pickup position information includes the distance between the first and second pickup devices and the popping bead, and the angle between the first and second pickup devices. The first and second popping bead sounds are acquired by the first and second pickup devices respectively. The popping bead pressing device can be a high-precision popping bead sound analyzer, such as the one in patent CN109341850A. The first and second pickup devices are respectively housed within the casing of the high-precision popping bead sound analyzer.
[0040] Step S20: Using the sounds of the first and second popping beads breaking as samples, the microphone signal is processed using a linear echo cancellation algorithm to obtain an echo signal;
[0041] In this embodiment, after acquiring the sound pickup location information, the sound of the first popping bead breaking, and the sound of the second popping bead breaking, the noise reduction device for the popping bead sound uses the first popping bead sound and the second popping bead sound as samples to process the microphone signal through a linear echo cancellation algorithm to obtain an echo signal.
[0042] Step S30: Perform high-frequency filtering on the first popping bead breaking sound and the second popping bead breaking sound based on the Butterworth filtering program to obtain the first filtered sound signal and the second filtered sound signal;
[0043] In this embodiment, after acquiring the first popping bead breaking sound and the second popping bead breaking sound, the popping bead sound noise reduction device performs high-frequency filtering on the first popping bead breaking sound and the second popping bead breaking sound based on the Butterworth filtering program to obtain the first filtered sound signal and the second filtered sound signal.
[0044] Step S30 performs high-frequency filtering on the first and second burst bead sounds based on a Butterworth filtering procedure to obtain a first filtered sound signal and a second filtered sound signal, which may include:
[0045] Step S31: Obtain the first sampling frequency, the first fluctuation frequency, and the first filter order of the first pickup device;
[0046] In this embodiment, after acquiring the sound of the first popping bead breaking, the noise reduction device acquires the first sampling frequency, the first fluctuation frequency, and the first filtering order of the first sound pickup device.
[0047] Step S32: Obtain the second sampling frequency, the second fluctuation frequency, and the second filter order of the second pickup device;
[0048] In this embodiment, after acquiring the sound of the first popping bead breaking, the noise reduction device for the popping bead acquires the second sampling frequency, the second fluctuation frequency, and the second filtering order of the second sound pickup device.
[0049] Step S33: Calculate the first filter amplitude based on the first sampling frequency, the first fluctuation frequency, and the first filter order;
[0050] In this embodiment, after acquiring the first sampling frequency, the first fluctuation frequency, and the first filtering order, the sound noise reduction device for pressing the popping bead calculates the first filtering amplitude based on the first sampling frequency, the first fluctuation frequency, and the first filtering order.
[0051] Step S33, which calculates the first filter amplitude based on the first sampling frequency, the first fluctuation frequency, and the first filter order, may include:
[0052] Calculate the first cutoff frequency based on the first sampling frequency and the first fluctuation frequency;
[0053] Based on the first cutoff frequency and the filter order, the signal amplitude in the sound of the first popping jelly is subjected to high-frequency filtering to obtain the first filter amplitude. First cutoff frequency Among them, f c f is the first fluctuation frequency. s This is the first sampling frequency;
[0054] The square of the first filter amplitude Where N is the first filter order.
[0055] Step S34: Calculate the second filter amplitude based on the second sampling frequency, the second fluctuation frequency, and the second filter order;
[0056] In this embodiment, after acquiring the second sampling frequency, the second fluctuation frequency, and the second filtering order, the sound noise reduction device for pressing the popping bead calculates the second filtering amplitude based on the second sampling frequency, the second fluctuation frequency, and the second filtering order.
[0057] Step S35: Replace the signal amplitude in the first popping bead breaking sound with the first filtered amplitude to obtain the first filtered sound signal;
[0058] In this embodiment, after obtaining the first filtered amplitude, the noise reduction device for the popping sound replaces the signal amplitude in the sound of the first popping bead breaking with the first filtered amplitude to obtain the first filtered sound signal.
[0059] Step S36: Replace the signal amplitude in the second popping bead breaking sound with the second filtered amplitude to obtain the second filtered sound signal.
[0060] In this embodiment, after obtaining the second filter amplitude, the noise reduction device for the popping sound replaces the signal amplitude in the sound of the second popping bead breaking with the second filter amplitude to obtain the second filtered sound signal.
[0061] Step S40: Determine the first cancellation signal and the second cancellation signal based on the pickup location information and the echo signal.
[0062] In this embodiment, after receiving the echo signal, the sound noise reduction device for pressing the popping bead determines the first cancellation signal and the second cancellation signal based on the sound pickup location information and the echo signal.
[0063] Step S40, which determines the first cancellation signal and the second cancellation signal based on the pickup location information and the echo signal, may include:
[0064] Step S41: Obtain an initial recurrent neural network model, wherein the initial recurrent neural network model includes multiple original matrices as parameters;
[0065] In this embodiment, an initial recurrent neural network model is obtained, which includes multiple original matrices as parameters.
[0066] Step S42: Decompose at least one of the original matrices to obtain the matrix parameters formed by the multiplication formula of the two component matrices;
[0067] In this embodiment, at least one of the original matrices is decomposed to obtain the matrix parameters formed by the multiplication formulas of the two component matrices.
[0068] Step S43: Train the initial recurrent neural network model to obtain the recurrent neural network model;
[0069] In this embodiment, the initial recurrent neural network model is trained to obtain the recurrent neural network model.
[0070] Step S43, training the initial recurrent neural network model to obtain the recurrent neural network model, may include:
[0071] Step S431: Obtain training data, which includes near-end sound signal of the sample, ambient noise signal of the sample, residual echo signal of the sample, and output audio signal of the sample.
[0072] In this embodiment, training data is acquired, which includes near-field sound signals of the samples, ambient noise signals of the samples, residual echo signals of the samples, and output audio signals of the samples.
[0073] Step S432: Obtain the ideal ratio mask based on the training data;
[0074] In this embodiment, an ideal ratio mask is obtained based on the training data.
[0075] Step S433: Train the initial recurrent neural network model according to the ideal ratio mask and the training data to obtain the recurrent neural network model, and fit the output of the recurrent neural network model with the ideal ratio mask.
[0076] In this embodiment, the initial recurrent neural network model is trained based on the ideal ratio mask and the training data to obtain the recurrent neural network model, and the output of the recurrent neural network model is fitted with the ideal ratio mask.
[0077] Step S44: Input the pickup location information and echo signal into the recurrent neural network model to obtain the first cancellation signal and the second cancellation signal.
[0078] In this embodiment, the pickup location information and echo signal are input into the recurrent neural network model to obtain the first cancellation signal and the second cancellation signal.
[0079] Step S50: Process the first filtered sound signal with the first cancellation signal to obtain a first noise-reduced sound signal, and process the second filtered sound signal with the second cancellation signal to obtain a second noise-reduced sound signal;
[0080] In this embodiment, after obtaining the first cancellation signal and the second cancellation signal, the noise reduction device for the popping bead sound processes the first filtered sound signal through the first cancellation signal to obtain the first noise-reduced sound signal, and processes the second filtered sound signal through the second cancellation signal to obtain the second noise-reduced sound signal.
[0081] Step S60: When the waveform overlap between the first noise-reduced sound signal and the second noise-reduced sound signal exceeds a preset threshold, the first noise-reduced sound signal is determined to be the noise-reduced popping bead sound.
[0082] In this embodiment, after obtaining the first noise-reduced sound signal and the second noise-reduced sound signal, the noise reduction device for the popping sound of the pressed bead will determine the first noise-reduced sound signal as the noise-reduced popping sound when the waveform overlap between the first noise-reduced sound signal and the second noise-reduced sound signal exceeds a preset threshold.
[0083] This embodiment obtains the pickup position information of the first and second pickup devices through the above scheme, and acquires the first and second burst sounds of the popping bead based on the pickup devices. The pickup position information includes the distance between the first and second pickup devices and the popping bead, and the angle between the first and second pickup devices. The first and second burst sounds are acquired by the two pickup devices respectively. Using the first and second burst sounds as samples, the microphone signal is processed by a linear echo cancellation algorithm to obtain the echo signal. The system performs high-frequency filtering on the first and second popping bead sounds using a Butterworth filter to obtain a first filtered sound signal and a second filtered sound signal. A first cancellation signal and a second cancellation signal are determined based on the pickup location information and echo signal. The first filtered sound signal is processed using the first cancellation signal to obtain a first noise-reduced sound signal, and the second filtered sound signal is processed using the second cancellation signal to obtain a second noise-reduced sound signal. When the waveform overlap between the first and second noise-reduced sound signals exceeds a preset threshold, the first noise-reduced sound signal is identified as the noise-reduced popping bead sound. This achieves noise reduction of the acquired popping bead sound, removing echoes and noise from the popping bead sound.
[0084] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.
[0085] The sequence numbers of the above embodiments of the present invention are for description only and do not represent the superiority or inferiority of the embodiments.
[0086] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.
[0087] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.
Claims
1. A method for noise reduction of the sound from a pressed popping bead, characterized in that, The popping bead is placed in a popping bead pressing device, and a first sound pickup device and a second sound pickup device are installed inside the popping bead pressing device. The method for reducing noise from the popping bead sound includes the following steps: The pickup position information of the first and second pickup devices is obtained, and the first and second popping bead breaking sounds are obtained based on the first and second pickup devices. The pickup position information includes the distance between the first and second pickup devices and the popping bead, and the angle between the first and second pickup devices. The first popping bead breaking sound and the second popping bead breaking sound are obtained by the first and second pickup devices respectively. Using the sound of the first popping bead breaking and the sound of the second popping bead breaking as samples, the microphone signal is processed by a linear echo cancellation algorithm to obtain an echo signal; The first and second popping bead breaking sounds are subjected to high-frequency filtering based on the Butterworth filtering program to obtain the first filtered sound signal and the second filtered sound signal. The first cancellation signal and the second cancellation signal are determined based on the pickup location information and the echo signal; The first filtered sound signal is processed by the first cancellation signal to obtain the first noise-reduced sound signal, and the second filtered sound signal is processed by the second cancellation signal to obtain the second noise-reduced sound signal. When the waveform overlap between the first noise-reduced sound signal and the second noise-reduced sound signal exceeds a preset threshold, the first noise-reduced sound signal is determined to be the noise-reduced popping sound. The step of determining the first cancellation signal and the second cancellation signal based on the pickup location information and the echo signal includes: Obtain an initial recurrent neural network model, which includes multiple original matrices as parameters; Decompose at least one of the original matrices to obtain matrix parameters composed of the multiplication formulas of the two component matrices; The initial recurrent neural network model is trained to obtain the recurrent neural network model; The pickup location information and echo signal are input into the recurrent neural network model to obtain the first cancellation signal and the second cancellation signal.
2. The method for reducing noise from a press-activated popping bead as described in claim 1, characterized in that, The step of performing high-frequency filtering on the first and second burst bead sounds based on the Butterworth filtering procedure to obtain a first filtered sound signal and a second filtered sound signal includes: The first sampling frequency, the first fluctuation frequency, and the first filter order of the first pickup device are obtained; The second sampling frequency, the second fluctuation frequency, and the second filter order of the second pickup device are obtained; The first filter amplitude is calculated based on the first sampling frequency, the first fluctuation frequency, and the first filter order. The second filter amplitude is calculated based on the second sampling frequency, the second fluctuation frequency, and the second filter order. The signal amplitude in the first popping bead sound is replaced with the first filtered amplitude to obtain the first filtered sound signal; The signal amplitude in the second popping bead breaking sound is replaced with the second filtered amplitude to obtain the second filtered sound signal.
3. The method for reducing noise from a press-activated popping bead according to claim 2, characterized in that, The step of calculating the first filter amplitude based on the first sampling frequency, the first fluctuation frequency, and the first filter order includes: Calculate the first cutoff frequency based on the first sampling frequency and the first fluctuation frequency; The signal amplitude in the sound of the first popping bead breaking is subjected to high-frequency filtering based on the first cutoff frequency and the filtering order to obtain the first filtered amplitude.
4. The method for reducing noise from pressed popping beads according to claim 3, characterized in that, First cutoff frequency ,in, The first fluctuation frequency, This is the first sampling frequency; The square of the first filter amplitude , where N is the first filter order.
5. The method for reducing noise from a press-activated popping bead according to claim 4, characterized in that, The step of training the initial recurrent neural network model to obtain the recurrent neural network model includes: Acquire training data, which includes near-sound signals of the sample, ambient noise signals of the sample, residual echo signals of the sample, and output audio signals of the sample. Obtain an ideal ratio mask based on the training data; The initial recurrent neural network model is trained based on the ideal ratio mask and the training data to obtain the recurrent neural network model, and the output of the recurrent neural network model is fitted with the ideal ratio mask.