CN116367063B - Bone conduction hearing aid equipment and system based on embedded - Google Patents
Bone conduction hearing aid equipment and system based on embedded Download PDFInfo
- Publication number
- CN116367063B CN116367063B CN202310442236.XA CN202310442236A CN116367063B CN 116367063 B CN116367063 B CN 116367063B CN 202310442236 A CN202310442236 A CN 202310442236A CN 116367063 B CN116367063 B CN 116367063B
- Authority
- CN
- China
- Prior art keywords
- spectrogram
- preset
- current
- noise reduction
- similarity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 210000000988 bone and bone Anatomy 0.000 title claims abstract description 34
- 230000009467 reduction Effects 0.000 claims abstract description 59
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000009499 grossing Methods 0.000 claims abstract description 8
- 238000011176 pooling Methods 0.000 claims abstract description 8
- 238000004364 calculation method Methods 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 238000001228 spectrum Methods 0.000 claims description 8
- 238000009432 framing Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 6
- 238000001914 filtration Methods 0.000 description 4
- 241000282414 Homo sapiens Species 0.000 description 3
- 230000009471 action Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 210000000056 organ Anatomy 0.000 description 3
- 208000032041 Hearing impaired Diseases 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 208000016354 hearing loss disease Diseases 0.000 description 2
- 230000005236 sound signal Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 210000003625 skull Anatomy 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 230000000638 stimulation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R25/00—Deaf-aid sets, i.e. electro-acoustic or electro-mechanical hearing aids; Electric tinnitus maskers providing an auditory perception
- H04R25/50—Customised settings for obtaining desired overall acoustical characteristics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R2225/00—Details of deaf aids covered by H04R25/00, not provided for in any of its subgroups
- H04R2225/43—Signal processing in hearing aids to enhance the speech intelligibility
Abstract
The invention provides an embedded bone conduction hearing aid device, which comprises a spectrogram generation module, a preprocessing module, a scene determination module and a noise reduction module, wherein a current spectrogram is generated, the current spectrogram is subjected to smoothing treatment, weights are obtained by calculation based on the utilization rate of a microprocessor and the utilization rate of a memory, whether downsampling is carried out or not is judged according to the weights, if so, the window size and the step length which are adopted under the weights are obtained, and average pooling downsampling is carried out on the spectrograms in a set of the current spectrogram and the preset spectrogram according to the window size and the step length; calculating the similarity between the preprocessed current spectrogram and each preprocessed preset spectrogram set according to a preset mode, and determining the current environment; and calling a noise reduction method and/or noise reduction parameters corresponding to the current scene to perform noise reduction processing on the voice signal. The invention can improve the effect of the bone conduction hearing aid device by utilizing limited embedded resources.
Description
Technical Field
The invention relates to the field of bone conduction, in particular to bone conduction hearing aid equipment and system based on embedded bone conduction.
Background
Auditory sense refers to the sense produced by the stimulation of human auditory organs by external sound, is one of five senses of human beings, and is an important way for people to communicate with each other. However, the hearing organs are also fragile organs, and once damaged, the hearing level is lowered with the increase of the age, so that the old and children are high-incidence people with hearing impairment. For people with hearing impairment, wearing hearing aids is an important and effective way to restore hearing.
The hearing aid mainly collects external sounds, and after the processing, a patient can hear the sounds. Hearing aids can be classified into behind-the-ear hearing aids, in-the-canal hearing aids, and the like according to the wearing position, and the principle of these hearing aids is to convert sound signals into electrical signals, process the electrical signals, and then output the electrical signals in a matched manner. In addition, there are bone conduction hearing aids, which, unlike conventional hearing aids, have the main principle of converting sound signals into vibrations, and allowing the patient to perceive the sound through skull vibrations, using the human skeleton as a transmission medium. Compared with the traditional hearing aid, the bone conduction hearing aid does not pass through the auditory canal, so that the hearing damage is not aggravated, the damage to the auditory canal is not caused, and the bone conduction hearing aid is particularly suitable for conductive hearing impaired people. But it is also the characteristics of bone conduction that determine its shortcoming, and it is not generally high to have traditional audiphones through bone transmission sound, tone quality and definition etc. moreover bone conduction earphone belongs to wearing equipment, and self resources are limited, how can carry out speech enhancement etc. under limited resources is the focus of bone conduction audiphones research.
Disclosure of Invention
In order to improve experience of a hearing aid device wearer, in a first aspect, the invention provides an embedded bone conduction hearing aid device, wherein the hearing aid device at least comprises a microprocessor and a memory, and the hearing aid device adopts FreeRTOS as an operating system; the apparatus further comprises the following modules:
the spectrogram generation module is used for framing the voice signal when the preset condition is met, and generating a current spectrogram;
the preprocessing module is used for carrying out smoothing processing on the current spectrogram, calculating to obtain a weight based on the utilization rate of the microprocessor and the utilization rate of the memory, judging whether to carry out downsampling according to the weight, if so, obtaining window size and step length adopted under the corresponding weight, and carrying out average pooling downsampling on the spectrograms in the current spectrogram and the preset spectrogram set according to the window size and the step length;
the scene determining module is used for calculating the similarity between the preprocessed current spectrogram and each preprocessed preset spectrogram set according to a preset mode, and determining the current environment based on the similarity; the current environment is one of a plurality of preset use scenes, and each use scene corresponds to a preset language spectrum chart set;
the noise reduction module is used for calling a noise reduction method and/or noise reduction parameters corresponding to the current scene to perform noise reduction processing on the voice signal.
Preferably, the bone conduction hearing aid device further comprises a mode switching key for switching between a normal mode and a noise reduction mode and a scene switching key for switching between different scenes, the scene switching key being usable when the device is in the noise reduction mode.
Preferably, the calculating based on the utilization rate of the microprocessor and the utilization rate of the memory obtains a weight specifically as follows:
acquiring the occupied time of an idle task of the FreeRTOS in a preset time period, calculating the ratio r of the occupied time to the preset time period, and taking 1-r as the utilization rate of a microprocessor; acquiring the utilization rate of the memory;
obtaining the maximum value Max of the utilization rate of the microprocessor and the utilization rate of the memory, and performingAs the weight, wherein>Representing an upward rounding.
Preferably, the similarity between the current spectrogram and each preset spectrogram set is calculated according to a preset mode, specifically:
for each preset spectrogram in the preset spectrogram set, executing the following operations: overlapping the current spectrogram with a preset spectrogram, calculating the fitting degree of the overlapping part, then moving the current spectrogram by a distance of T/2N along the time direction, calculating the fitting degree of the overlapping part again until the current spectrogram moves for N times to obtain N+1 fitting degrees, and taking the maximum value in the N+1 fitting degrees as the similarity between the current spectrogram and the preset spectrogram; wherein N is a positive integer;
and taking an average value of the similarity of the current spectrogram and each preset spectrogram in the preset spectrogram set as the similarity of the current spectrogram and the preset spectrogram set.
Preferably, the calculation mode of the fitness is as follows:
according to the formulaCalculating the fit, wherein M is the number of pixel points of the overlapping part, < >>P is the value of the pixel point of the current spectrogram (x,y) And (x, y) represents the coordinates of the pixel points of the overlapped part for the value of the pixel points of the preset spectrogram.
The invention further provides a noise reduction method applied to bone conduction hearing aid equipment, which comprises the following steps:
step 1, framing a voice signal when a preset condition is met, and generating a current spectrogram;
step 2, carrying out smoothing treatment on the current spectrogram, calculating to obtain a weight based on the utilization rate of the microprocessor and the utilization rate of the memory, judging whether to carry out downsampling according to the weight, if so, obtaining window size and step length adopted under the corresponding weight, and carrying out average pooling downsampling on the spectrograms in the current spectrogram and the preset spectrogram set according to the window size and the step length;
step 3, calculating the similarity between the preprocessed current spectrogram and each preprocessed preset spectrogram set according to a preset mode, and determining the current environment based on the similarity; the current environment is one of a plurality of preset use scenes, and each use scene corresponds to a preset language spectrum chart set;
and 4, calling a noise reduction method and/or noise reduction parameters corresponding to the current scene to perform noise reduction processing on the voice signal.
Preferably, the bone conduction hearing aid device further comprises a mode switching key for switching between a normal mode and a noise reduction mode and a scene switching key for switching between different scenes, the scene switching key being usable when the device is in the noise reduction mode.
Preferably, the calculating based on the utilization rate of the microprocessor and the utilization rate of the memory obtains a weight specifically as follows:
acquiring the occupied time of an idle task of the FreeRTOS in a preset time period, calculating the ratio r of the occupied time to the preset time period, and taking 1-r as the utilization rate of a microprocessor; acquiring the utilization rate of the memory;
obtaining the maximum value Max of the utilization rate of the microprocessor and the utilization rate of the memory, and performingAs the weight, wherein>Representing an upward rounding.
Preferably, the similarity between the current spectrogram and each preset spectrogram set is calculated according to a preset mode, specifically:
for each preset spectrogram in the preset spectrogram set, executing the following operations: overlapping the current spectrogram with a preset spectrogram, calculating the fitting degree of the overlapping part, then moving the current spectrogram by a distance of T/2N along the time direction, calculating the fitting degree of the overlapping part again until the current spectrogram moves for N times to obtain N+1 fitting degrees, and taking the maximum value in the N+1 fitting degrees as the similarity between the current spectrogram and the preset spectrogram; wherein N is a positive integer;
and taking an average value of the similarity of the current spectrogram and each preset spectrogram in the preset spectrogram set as the similarity of the current spectrogram and the preset spectrogram set.
Preferably, the calculation mode of the fitness is as follows:
according to the formulaCalculating the fit, wherein M is the number of pixel points of the overlapping part, < >>P is the value of the pixel point of the current spectrogram (x,y) For the values of the pixels of the preset spectrogram, (x, y) represents the overlapping partial imagesCoordinates of the pixel.
Finally, the invention provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements a method as described above.
The traditional bone conduction hearing aid device brings great convenience to users, enables hearing-impaired people to hear sound, but has larger sound noise due to the bone conduction characteristic, and can be used in some scenes by adopting the same noise reduction method no matter where occasions are, but has good noise reduction effect in other scenes, and aiming at the problem, the invention provides novel bone conduction hearing aid device based on embedded type.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a structural diagram of a first embodiment;
FIG. 2 is a diagram illustrating a movement process of a current spectrogram relative to a preset spectrogram;
fig. 3 is a flowchart of a second embodiment.
Detailed Description
In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In a first aspect, the present invention provides an embedded bone conduction hearing aid device, where the hearing aid device at least includes a microprocessor and a memory, and the hearing aid device uses FreeRTOS as an operating system;
in a specific embodiment, the bone conduction hearing aid device further comprises a power source, bluetooth, or the like. Fig. 1 shows a simple structure diagram of an embedded bone conduction hearing aid device, and as shown in fig. 1, after a microphone 101 collects voice, an analog signal is converted into digital information through an ADC, then processed by a microprocessor 103, and finally output through a bone conduction vibrator 104.
The apparatus further comprises the following modules:
the spectrogram generation module is used for framing the voice signal when the preset condition is met, and generating a current spectrogram;
after the voice is collected by a collector such as a microphone, an analog signal is converted into a digital signal, as the voice is a continuous process, the voice signal is required to be subjected to frame division for analysis, feature extraction and the like of the voice signal, a window function is used for loading a time domain, preferably, the window function is a hamming window or a hanning window, then fourier transformation is performed on each voice frame to obtain a spectrogram, the spectrogram is obtained by the spectrogram, the abscissa of the spectrogram is time, the ordinate is frequency, the value of a pixel point represents the amplitude of the frequency, and the larger the amplitude is, the larger the value of the pixel point is.
In a specific embodiment, the preset condition is that the sum of absolute values of differences between each frequency of the current voice and each frequency of the voice at the previous time point is judged every preset time or every preset time, and if the sum is larger than a set value, the voice environment at the current moment and the voice environment at the previous time point are proved to have larger change, framing is triggered, a current spectrogram is generated, and execution of a subsequent preprocessing module, a scene determining module and a noise reducing module is performed; wherein the difference between the previous point in time and the current time is the predetermined time. That is, if the predetermined time is T1 and the current time is T, the previous time point is T-T1. The difference value of each frequency is the absolute value of the difference value of the amplitude of the same frequency of the voice at the current moment and the voice at the previous time point, and the sum of the absolute values of all the frequencies is calculated.
The preprocessing module is used for carrying out smoothing processing on the current spectrogram, calculating to obtain a weight based on the utilization rate of the microprocessor and the utilization rate of the memory, judging whether to carry out downsampling according to the weight, if so, obtaining window size and step length adopted under the corresponding weight, and carrying out average pooling downsampling on the spectrograms in the current spectrogram and the preset spectrogram set according to the window size and the step length;
the spectrogram is composed of a plurality of spectrograms, obvious cracks are formed in the spectrograms, and the subsequent analysis of the cracks can be reduced by carrying out balance treatment on the spectrograms. The smoothing is performed by, for example, mean filtering, median filtering, gaussian filtering, or the like.
Because the processor resources and the memory resources of the embedded system are limited, if the spectrogram is too large, the invention can generate a setback feeling, based on the result, the invention can also obtain the weight according to the utilization rate of the microprocessor and the utilization rate of the memory, judge whether to downsample the spectrogram according to the weight, and calculate the weight by adopting various methods, one example is to weight and sum the utilization rate of the microprocessor and the utilization rate of the memory, and other modes can also be adopted.
If the higher the utilization, the more heavily loaded the current embedded system, the larger the window and step size that needs to be downsampled, and vice versa. In a specific embodiment, a corresponding relationship between the weight and the window size, and the step size is preset, for example, the weight is 1, the window size is set to 2×2, the step size is 1, the weight is 2, the window size is set to 3×3, and the step size is 2, where the specific corresponding relationship is determined according to the size of the performance core memory of the microprocessor.
The current spectrogram and the preset spectrogram set are required to be matched later, the spectrogram sizes of the current spectrogram and the preset spectrogram set are required to be consistent, and similarly, the spectrograms in the preset spectrogram set are required to be subjected to average pooling downsampling in the same way. In the present invention, all the spectrograms are gray-scale images. In the invention, the current spectrogram and the preset spectrogram are only used for distinguishing two different types of spectrograms, and can also be expressed as a first spectrogram and a second spectrogram, wherein the first spectrogram is a spectrogram of a scene to be determined, and the second spectrogram is a preset spectrogram corresponding to the scene.
The scene determining module is used for calculating the similarity between the preprocessed current spectrogram and each preprocessed preset spectrogram set according to a preset mode, and determining the current environment based on the similarity; the current environment is one of a plurality of preset use scenes, and each use scene corresponds to a preset language spectrum chart set;
the wearer of the bone conduction hearing aid device is located at different places and the noise is different, and the scenes include but are not limited to subways, markets, amusement parks, offices and sports, different sounds in the scenes can generate different noise, and a preset language spectrum atlas is set for each scene as shown in the following table 1:
TABLE 1
Each preset spectrogram set comprises a plurality of preset spectrograms, and as shown in table 1, the preset spectrogram set spec_a corresponding to the subway comprises 4 preset spectrograms. And respectively calculating the similarity between the current spectrogram and 4 preset spectrograms in the spec_A to obtain the similarity between the current spectrogram and the spec_A, and if the similarity between the current spectrogram and the spectrogram set of the subway, the market, the amusement park, the office and the sports is respectively 1, 4, 5, 2 and 2, determining the current environment or scene as the amusement park, and then adopting a noise reduction method corresponding to the amusement park to carry out noise reduction treatment on the voice signals.
The noise reduction module is used for calling a noise reduction method and/or noise reduction parameters corresponding to the current scene to perform noise reduction processing on the voice signal.
In a specific embodiment, each scene corresponds to a noise reduction method, and the noise reduction method comprises at least one noise reduction task; in another embodiment, the same noise reduction method is adopted for all scenes, but the noise reduction parameters are different, that is, each scene corresponds to a group of noise reduction parameters, and the different noise reduction parameters have different noise suppression effects; in another embodiment, each scene corresponds to a noise reduction method including at least one noise reduction task and a set of noise reduction parameters including at least one noise reduction parameter. There are various noise reduction methods, such as spectral subtraction, wiener filtering, adaptive filters, or neural networks.
In a specific embodiment, the bone conduction hearing aid device further comprises a mode switching key and a scene switching key, wherein the mode switching key is used for switching between a normal mode and a noise reduction mode, the scene switching key is used for switching between different scenes, and the scene switching key is available when the device is in the noise reduction mode. When the user starts the noise reduction mode, the scene is judged, the noise reduction method and/or the noise reduction parameters corresponding to the scene are adopted to reduce the noise of the voice, namely, the noise reduction mode is started as a precondition for executing the noise reduction. Because the system cannot accurately determine the current scene, the embedded bone conduction hearing aid device is further provided with the scene switching key for manually selecting among a plurality of scenes, for example, the current scene is an office, but the system judges that the market is a market, and a user needs to manually switch through the scene switching key, so that the embedded bone conduction hearing aid device is more convenient for the user to use.
The weight may be calculated in various ways, such as using weighted average as described above, but this cannot reflect the respective situations of the microprocessor and the memory, and in one embodiment, the weight is calculated based on the utilization of the microprocessor and the utilization of the memory, specifically:
acquiring the occupied time of an idle task of the FreeRTOS in a preset time period, calculating the ratio r of the occupied time to the preset time period, and taking 1-r as the utilization rate of a microprocessor; acquiring the utilization rate of the memory;
obtaining the maximum value Max of the utilization rate of the microprocessor and the utilization rate of the memory, and performingAs the weight, wherein>Representing an upward rounding.
The FreeRTOS has no direct function to call to get the utilization rate of the microprocessor, but the Idle Task (Idle Task) can reflect the load of the microprocessor, and the invention takes 1-r as the utilization rate of the microprocessor, then calculates the utilization rate of the microprocessor and the utilization rate of the memory, and the maximum value of the two is, for example, 40% of the utilization rate of the microprocessor, 30% of the utilization rate of the memory, max=40% of the maximum value, and the weight is 2.
In a specific embodiment, the calculating the similarity between the current spectrogram and each preset spectrogram set according to the preset mode specifically includes:
for each preset spectrogram in the preset spectrogram set, executing the following operations: overlapping the current spectrogram with a preset spectrogram, calculating the fitting degree of the overlapping part, then moving the current spectrogram by a distance of T/2N along the time direction, calculating the fitting degree of the overlapping part again until the current spectrogram moves for N times to obtain N+1 fitting degrees, and taking the maximum value in the N+1 fitting degrees as the similarity between the current spectrogram and the preset spectrogram; wherein N is a positive integer, T is the time coordinate size of the rightmost end when the leftmost side of the current spectrogram is positioned at the origin, namely T represents the time size of the current spectrogram or T represents the time width of the current spectrogram. Preferably, N is half the number of corresponding frames in the current spectrogram. For example, if the current spectrogram corresponds to a spectrogram of 4 frames of a speech signal, then n=2 is preferred. Fig. 2 shows a process of moving a current spectrogram over a preset spectrogram.
And taking an average value of the similarity of the current spectrogram and each preset spectrogram in the preset spectrogram set as the similarity of the current spectrogram and the preset spectrogram set.
The calculation mode of the fit degree is as follows:
according to the formulaCalculating the fit, wherein M is the number of pixel points of the overlapping part, < >>P is the value of the pixel point of the current spectrogram (x,y) And (x, y) represents the coordinates of the pixel points of the overlapped part for the value of the pixel points of the preset spectrogram.
The smaller the matching degree is, the more similar the two spectrograms are, but the probability that the two spectrograms are misplaced is that the two spectrograms are formed by the frequency spectrums of the frames.
In addition, the invention provides a noise reduction method applied to bone conduction hearing aid equipment, as shown in fig. 3, the method comprises the following steps:
step 1, framing a voice signal when a preset condition is met, and generating a current spectrogram;
step 2, carrying out smoothing treatment on the current spectrogram, calculating to obtain a weight based on the utilization rate of the microprocessor and the utilization rate of the memory, judging whether to carry out downsampling according to the weight, if so, obtaining window size and step length adopted under the corresponding weight, and carrying out average pooling downsampling on the spectrograms in the current spectrogram and the preset spectrogram set according to the window size and the step length;
step 3, calculating the similarity between the preprocessed current spectrogram and each preprocessed preset spectrogram set according to a preset mode, and determining the current environment based on the similarity; the current environment is one of a plurality of preset use scenes, and each use scene corresponds to a preset language spectrum chart set;
and 4, calling a noise reduction method and/or noise reduction parameters corresponding to the current scene to perform noise reduction processing on the voice signal.
Preferably, the bone conduction hearing aid device further comprises a mode switching key for switching between a normal mode and a noise reduction mode and a scene switching key for switching between different scenes, the scene switching key being usable when the device is in the noise reduction mode.
Preferably, the calculating based on the utilization rate of the microprocessor and the utilization rate of the memory obtains a weight specifically as follows:
acquiring the occupied time of an idle task of the FreeRTOS in a preset time period, calculating the ratio r of the occupied time to the preset time period, and taking 1-r as the utilization rate of a microprocessor; acquiring the utilization rate of the memory;
obtaining the maximum value Max of the utilization rate of the microprocessor and the utilization rate of the memory, and performingAs the weight, wherein>Representing an upward rounding.
Preferably, the similarity between the current spectrogram and each preset spectrogram set is calculated according to a preset mode, specifically:
for each preset spectrogram in the preset spectrogram set, executing the following operations: overlapping the current spectrogram with a preset spectrogram, calculating the fitting degree of the overlapping part, then moving the current spectrogram by a distance of T/2N along the time direction, calculating the fitting degree of the overlapping part again until the current spectrogram moves for N times to obtain N+1 fitting degrees, and taking the maximum value in the N+1 fitting degrees as the similarity between the current spectrogram and the preset spectrogram; wherein N is a positive integer.
And taking an average value of the similarity of the current spectrogram and each preset spectrogram in the preset spectrogram set as the similarity of the current spectrogram and the preset spectrogram set.
Preferably, the calculation mode of the fitness is as follows:
according to the formulaCalculating the fit, wherein M is the number of pixel points of the overlapping part, < >>P is the value of the pixel point of the current spectrogram (x,y) And (x, y) represents the coordinates of the pixel points of the overlapped part for the value of the pixel points of the preset spectrogram.
Finally, the invention provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements a method as described above.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by adding necessary general purpose hardware platforms, or may be implemented by a combination of hardware and software. Based on such understanding, the foregoing aspects, in essence and portions contributing to the art, may be embodied in the form of a computer program product, which may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (6)
1. An embedded bone conduction hearing aid device at least comprises a microprocessor and a memory, wherein the hearing aid device adopts FreeRTOS as an operating system; characterized in that the device further comprises the following modules:
the spectrogram generation module is used for framing the voice signal when the preset condition is met, and generating a current spectrogram;
the preprocessing module is used for carrying out smoothing processing on the current spectrogram, calculating to obtain a weight based on the utilization rate of the microprocessor and the utilization rate of the memory, judging whether to carry out downsampling according to the weight, if so, obtaining window size and step length adopted under the corresponding weight, and carrying out average pooling downsampling on the spectrograms in the current spectrogram and the preset spectrogram set according to the window size and the step length;
the scene determining module is used for calculating the similarity between the preprocessed current spectrogram and each preprocessed preset spectrogram set according to a preset mode, and determining the current environment based on the similarity; the current environment is one of a plurality of preset use scenes, and each use scene corresponds to a preset language spectrum chart set;
the noise reduction module is used for calling a noise reduction method and/or noise reduction parameters corresponding to the current scene to perform noise reduction processing on the voice signal;
the similarity between the current spectrogram and each preset spectrogram set is calculated according to a preset mode, and the similarity is specifically as follows:
for each preset spectrogram in the preset spectrogram set, executing the following operations: overlapping the current spectrogram with a preset spectrogram, calculating the fitting degree of the overlapping part, then moving the current spectrogram by a distance of T/2N along the time direction, calculating the fitting degree of the overlapping part again until the current spectrogram moves for N times to obtain N+1 fitting degrees, and taking the maximum value in the N+1 fitting degrees as the similarity between the current spectrogram and the preset spectrogram; wherein T represents the time width of the current spectrogram, and N is a positive integer;
taking an average value of the similarity of the current spectrogram and each preset spectrogram in the preset spectrogram set as the similarity of the current spectrogram and the preset spectrogram set;
the calculation mode of the fit degree is as follows:
according to the formulaCalculating the fit, wherein M is the number of pixel points of the overlapping part, < >>For the value of the pixel point of the current spectrogram,/->And (x, y) represents the coordinates of the pixel points of the overlapped part for the value of the pixel points of the preset spectrogram.
2. The device of claim 1, wherein the bone conduction hearing assistance device further comprises a mode switching key for switching between a normal mode and a noise reduction mode and a scene switching key for switching between different scenes, the scene switching key being available when the device is in the noise reduction mode.
3. The apparatus of claim 1, wherein the weight is calculated based on the utilization of the microprocessor and the utilization of the memory, specifically:
acquiring the occupied time of an idle task of the FreeRTOS in a preset time period, calculating the ratio r of the occupied time to the preset time period, and taking 1-r as the utilization rate of a microprocessor; acquiring the utilization rate of the memory;
obtaining the maximum value Max of the utilization rate of the microprocessor and the utilization rate of the memory, and performingAs the weight, wherein>Representing an upward rounding.
4. A method of noise reduction for use in a bone conduction hearing aid device, the method comprising the steps of:
step 1, framing a voice signal when a preset condition is met, and generating a current spectrogram;
step 2, carrying out smoothing treatment on the current spectrogram, calculating to obtain a weight based on the utilization rate of a microprocessor and the utilization rate of a memory, judging whether to carry out downsampling according to the weight, if so, obtaining window size and step length adopted under the corresponding weight, and carrying out average pooling downsampling on the spectrograms in the current spectrogram and the preset spectrogram set according to the window size and the step length;
step 3, calculating the similarity between the preprocessed current spectrogram and each preprocessed preset spectrogram set according to a preset mode, and determining the current environment based on the similarity; the current environment is one of a plurality of preset use scenes, and each use scene corresponds to a preset language spectrum chart set;
step 4, invoking a noise reduction method and/or noise reduction parameters corresponding to the current scene to perform noise reduction processing on the voice signal;
the similarity between the current spectrogram and each preset spectrogram set is calculated according to a preset mode, and the similarity is specifically as follows:
for each preset spectrogram in the preset spectrogram set, executing the following operations: overlapping the current spectrogram with a preset spectrogram, calculating the fitting degree of the overlapping part, then moving the current spectrogram by a distance of T/2N along the time direction, calculating the fitting degree of the overlapping part again until the current spectrogram moves for N times to obtain N+1 fitting degrees, and taking the maximum value in the N+1 fitting degrees as the similarity between the current spectrogram and the preset spectrogram; wherein T represents the time width of the current spectrogram, and N is a positive integer;
taking an average value of the similarity of the current spectrogram and each preset spectrogram in the preset spectrogram set as the similarity of the current spectrogram and the preset spectrogram set;
the calculation mode of the fit degree is as follows:
according to the formulaCalculating the fit, wherein M is the number of pixel points of the overlapping part, < >>For the value of the pixel point of the current spectrogram,/->And (x, y) represents the coordinates of the pixel points of the overlapped part for the value of the pixel points of the preset spectrogram.
5. The method of claim 4, wherein the calculating the weight based on the utilization of the microprocessor and the utilization of the memory is specifically:
acquiring the occupied time of an idle task of the FreeRTOS in a preset time period, calculating the ratio r of the occupied time to the preset time period, and taking 1-r as the utilization rate of a microprocessor; acquiring the utilization rate of the memory;
obtaining the maximum value Max of the utilization rate of the microprocessor and the utilization rate of the memory, and performingAs the weight, wherein>Representing an upward rounding.
6. A computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the method according to any of claims 4-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310442236.XA CN116367063B (en) | 2023-04-23 | 2023-04-23 | Bone conduction hearing aid equipment and system based on embedded |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310442236.XA CN116367063B (en) | 2023-04-23 | 2023-04-23 | Bone conduction hearing aid equipment and system based on embedded |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116367063A CN116367063A (en) | 2023-06-30 |
| CN116367063B true CN116367063B (en) | 2023-11-14 |
Family
ID=86921713
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310442236.XA Active CN116367063B (en) | 2023-04-23 | 2023-04-23 | Bone conduction hearing aid equipment and system based on embedded |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116367063B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101593522A (en) * | 2009-07-08 | 2009-12-02 | 清华大学 | A kind of full frequency domain digital hearing aid method and apparatus |
| CN107533839A (en) * | 2015-12-17 | 2018-01-02 | 华为技术有限公司 | A kind of processing method and equipment to surrounding environment sound |
| CN109859767A (en) * | 2019-03-06 | 2019-06-07 | 哈尔滨工业大学（深圳） | A kind of environment self-adaption neural network noise-reduction method, system and storage medium for digital deaf-aid |
| CN113016189A (en) * | 2018-11-16 | 2021-06-22 | 三星电子株式会社 | Electronic device and method for recognizing audio scene |
| WO2021218377A1 (en) * | 2020-04-29 | 2021-11-04 | Oppo广东移动通信有限公司 | Noise reduction mode control method and apparatus, electronic device and storage medium |
| CN114125639A (en) * | 2021-12-06 | 2022-03-01 | 维沃移动通信有限公司 | Audio signal processing method and device and electronic equipment |
| CN114245280A (en) * | 2021-12-20 | 2022-03-25 | 清华大学深圳国际研究生院 | Scene self-adaptive hearing aid audio enhancement system based on neural network |
| CN114820869A (en) * | 2022-05-11 | 2022-07-29 | 中南大学 | Incomplete scatter diagram overlap removing method |
-
2023
- 2023-04-23 CN CN202310442236.XA patent/CN116367063B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101593522A (en) * | 2009-07-08 | 2009-12-02 | 清华大学 | A kind of full frequency domain digital hearing aid method and apparatus |
| CN107533839A (en) * | 2015-12-17 | 2018-01-02 | 华为技术有限公司 | A kind of processing method and equipment to surrounding environment sound |
| CN113016189A (en) * | 2018-11-16 | 2021-06-22 | 三星电子株式会社 | Electronic device and method for recognizing audio scene |
| CN109859767A (en) * | 2019-03-06 | 2019-06-07 | 哈尔滨工业大学（深圳） | A kind of environment self-adaption neural network noise-reduction method, system and storage medium for digital deaf-aid |
| WO2021218377A1 (en) * | 2020-04-29 | 2021-11-04 | Oppo广东移动通信有限公司 | Noise reduction mode control method and apparatus, electronic device and storage medium |
| CN114125639A (en) * | 2021-12-06 | 2022-03-01 | 维沃移动通信有限公司 | Audio signal processing method and device and electronic equipment |
| CN114245280A (en) * | 2021-12-20 | 2022-03-25 | 清华大学深圳国际研究生院 | Scene self-adaptive hearing aid audio enhancement system based on neural network |
| CN114820869A (en) * | 2022-05-11 | 2022-07-29 | 中南大学 | Incomplete scatter diagram overlap removing method |
Non-Patent Citations (1)
| Title |
|---|
| 基于卷积神经网络的交通声音事件识别方法;张文涛;韩莹莹;黎恒;;现代电子技术(14);正文第1章-第4章 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116367063A (en) | 2023-06-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109326302B (en) | Voice enhancement method based on voiceprint comparison and generation of confrontation network | |
| US11812223B2 (en) | Electronic device using a compound metric for sound enhancement | |
| CN109410976B (en) | Speech enhancement method based on binaural sound source localization and deep learning in binaural hearing aid | |
| ES2775799T3 (en) | Method and apparatus for multisensory speech enhancement on a mobile device | |
| CN109036460B (en) | Voice processing method and device based on multi-model neural network | |
| EP3866165B1 (en) | Method for enhancing telephone speech signals based on convolutional neural networks | |
| CN111916101B (en) | Deep learning noise reduction method and system fusing bone vibration sensor and double-microphone signals | |
| CN105741849A (en) | Voice enhancement method for fusing phase estimation and human ear hearing characteristics in digital hearing aid | |
| JP2003520469A (en) | Noise reduction apparatus and method | |
| TWI581254B (en) | Environmental noise elimination system and application method thereof | |
| US9640193B2 (en) | Systems and methods for enhancing place-of-articulation features in frequency-lowered speech | |
| JP2008065090A (en) | Noise suppressing apparatus | |
| EP4044181A1 (en) | Deep learning speech extraction and noise reduction method fusing signals of bone vibration sensor and microphone | |
| US20090257609A1 (en) | Method for Noise Reduction and Associated Hearing Device | |
| CN214226506U (en) | Sound processing circuit, electroacoustic device, and sound processing system | |
| CN116367063B (en) | Bone conduction hearing aid equipment and system based on embedded | |
| CN113782044A (en) | Voice enhancement method and device | |
| CN116132875B (en) | Multi-mode intelligent control method, system and storage medium for hearing-aid earphone | |
| WO2005029463A1 (en) | A method for recovering target speech based on speech segment detection under a stationary noise | |
| CN114822566A (en) | Audio signal generation method and system, and non-transitory computer readable medium | |
| CN111968627B (en) | Bone conduction voice enhancement method based on joint dictionary learning and sparse representation | |
| CN113411715B (en) | Prompting method for speaking sound volume, earphone and readable storage medium | |
| CN113611319A (en) | Wind noise suppression method, device, equipment and system based on voice component | |
| KR101022457B1 (en) | Method to combine CASA and soft mask for single-channel speech separation | |
| CN112118511A (en) | Earphone noise reduction method and device, earphone and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |