Wearable device and signal processing method

Abstract

The invention discloses a wearable device and a signal processing method, and relates to the technical field of intelligent electronic products, the wearable device comprises a shell, a biological sound sensor and at least two auxiliary sensors, the shell is provided with a mounting cavity and a detection surface opposite to the mounting cavity, the detection surface is arranged to conduct biological sound, the biological sound sensor is arranged on the bottom wall of the mounting cavity, and the auxiliary sensors are arranged on the bottom wall of the mounting cavity. The signal acquisition end of the biological sound sensor is arranged towards the bottom wall of the mounting cavity, the biological sound sensor can acquire biological sound signals transmitted by the detection surface, the auxiliary sensors are arranged on the side wall of the mounting cavity and are electrically connected with the biological sound sensor, and the signal acquisition ends of the auxiliary sensors are arranged back to the biological sound sensor; the biological sound sensor collects biological sound signals transmitted by the detection face in the first direction, the direction, away from the biological sound sensor, of the signal collection end of the auxiliary sensor is the second direction, and an included angle is formed between the second direction and the first direction. The invention aims to perform multi-channel signal collection through the wearable device so as to improve the signal collection quality and accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent electronic products, and in particular to a wearable device and a signal processing method. Background Art

[0002] Bio-sound detection is an important test item in routine physical examinations. Bio-sound detection includes but is not limited to heart sound detection, lung sound detection, etc. Taking heart sound detection as an example, heart sound detection usually uses a bio-sound sensor to obtain the low-frequency vibration sound waves generated by the heart contraction, so as to analyze the heart health status of humans or animals through the waveform of the sound waves.

[0003] Existing bio-sound sensors are also gradually being widely used in smart wearable devices to help users perform daily bio-sound detection more quickly. For example, when a smart wearable watch is running, it collects bio-sound vibration signals in a direction perpendicular to the user's skin surface. However, due to other activities of the user and jitter at the wearing location, vibration signals unrelated to the bio-sound vibration signals and other noise interference will also be collected when collecting the bio-sound vibration signals, thereby affecting the accuracy and effect of the bio-sound detection. Summary of the invention

[0004] The main purpose of the present invention is to propose a wearable device and a signal processing method, aiming to collect multi-channel signals through the wearable device to improve the collection quality and accuracy of biological sound signals.

[0005] To achieve the above object, the present invention provides a wearable device, the wearable device comprising: A housing, the housing having a mounting cavity and a detection surface facing away from the mounting cavity, the detection surface being configured to conduct biological sound; A bio-sound sensor, wherein the bio-sound sensor is disposed on the bottom wall of the installation cavity, a signal collection end of the bio-sound sensor is disposed toward the bottom wall of the installation cavity, and the bio-sound sensor can collect the bio-sound signal conducted by the detection surface; and At least two auxiliary sensors, each of which is disposed on a side wall of the mounting cavity and is electrically connected to the bio-sound sensor, and a signal collection end of the auxiliary sensor is disposed facing away from the bio-sound sensor; The biological sound sensor collects the biological sound signal transmitted by the detection surface in a first direction, and the signal collection end of the auxiliary sensor is directed away from the biological sound sensor in a second direction, and the second direction is set at an angle to the first direction.

[0006] In one embodiment, the signal detection directions of at least two of the auxiliary sensors are arranged perpendicular to each other; And / or, the second direction and the first direction are arranged perpendicular to each other.

[0007] In one embodiment, the bio-sound sensor includes one of a bone conduction sensor, a microphone, an acceleration sensor, an inertial sensor, a piezoelectric ceramic, a pressure sensor, and a vibration detection sensor; And / or, the auxiliary sensor is an acceleration sensor; And / or, the auxiliary sensor is a vibration detection sensor.

[0008] In one embodiment, the housing is provided with a detection hole, the detection hole communicates with the installation cavity and passes through the detection surface; Part of the bio-sound sensors are disposed in the detection hole and protrude from the detection surface.

[0009] The present invention further proposes a signal processing method, which is implemented by the above-mentioned wearable device, and the steps of the signal processing method include: Acquire a first vibration signal collected by the bio-sound sensor within a preset time period; Acquire a second vibration signal collected by the auxiliary sensor within the preset time period; A filtering strategy is used to obtain the target biological sound signal; The steps of the filtering strategy include: determining whether there is a signal in the first vibration signal that is the same as the second vibration signal; If yes, filtering out the signal identical to the second vibration signal from the first vibration signal to obtain the target biological sound signal; If not, the collected first vibration signal is used as the target biological sound signal.

[0010] In one embodiment, the wearable device includes a plurality of bio-sound sensors, and the plurality of bio-sound sensors are arranged in an array on the bottom wall of the mounting cavity, and the signal processing method further includes: Acquire a plurality of the first vibration signals collected by the plurality of the bio-sound sensors within the preset time period; using the filtering strategy to compare the second vibration signal with each of the first vibration signals to obtain a plurality of the target biological sound signals; Within the preset time period, selecting a plurality of preset time points, and calculating the signal average of the target biological sound signal collected at the plurality of preset time points; Mapping each of the preset time points with a signal average value of each of the target biological sound signals; and determining the regional biological sound signal corresponding to the area to be monitored based on the signal average value.

[0011] In one embodiment, the step of using the filtering strategy to compare the second vibration signal with each of the first vibration signals to obtain a plurality of the target biological sound signals includes: Acquire the second vibration signal of each of the auxiliary sensors; comparing each of the second vibration signals with the first vibration signal in sequence; determining whether the first vibration signal contains a signal identical to each of the second vibration signals; If yes, removing the signal identical to the second vibration signal from the first vibration signal to obtain the target biological sound signal; If not, the collected first vibration signal is used as the target biological sound signal.

[0012] In one embodiment, after acquiring the second vibration signal collected by the auxiliary sensor within the preset time period, the signal processing method further includes: Determining whether the bio-sound sensor is in a stable state within the preset time period; If yes, acquiring the first vibration signal through the bio-sound sensor and acquiring the second vibration signal through the auxiliary sensor, and increasing the sampling frequency of the bio-sound sensor in the stable state; If not, stop collecting the first vibration signal through the bio-sound sensor, and stop collecting the second vibration signal through the auxiliary sensor.

[0013] In one embodiment, the step of determining whether the bio-sound sensor is in a stable state within the preset time period includes: storing the collected second vibration signal to form noise vibration data; Finding a maximum value in the noise and vibration data as an abnormal vibration value; The time when the abnormal vibration value appears within the preset time length is regarded as the abnormal point; Determine whether the number of abnormal points is less than or equal to the preset number of abnormal points within the preset time length; If yes, the bio-sound sensor is in a stable state within the preset time period; If not, the bio-sound sensor is not in a stable state within the preset time period.

[0014] In one embodiment, the step of filtering out the signal identical to the second vibration signal in the first vibration signal comprises: generating corresponding noise reduction control parameters according to the signal characteristic index of the second vibration signal; Establishing a noise generation adversarial network according to the noise reduction control parameters; When the first vibration signal contains a signal identical to the second vibration signal, the noise generation adversarial network generates and outputs a corresponding reverse second vibration signal; The frequency and amplitude of the reverse second vibration signal are equal to those of the second vibration signal, and the phase of the reverse second vibration signal is opposite to that of the second vibration signal; The reverse second vibration signal is input to the bio-sound sensor to remove the second vibration signal.

[0015] The wearable device of the technical solution of the present invention includes a shell, a bio-sound sensor and at least two auxiliary sensors. The shell has an installation cavity and a detection surface facing away from the installation cavity. The detection surface is configured to conduct bio-sound. The bio-sound sensor is arranged on the bottom wall of the installation cavity. The signal acquisition end of the bio-sound sensor is arranged toward the bottom wall of the installation cavity. The bio-sound sensor can collect bio-sound signals conducted by the detection surface. Each auxiliary sensor is arranged on the side wall of the installation cavity and is electrically connected to the bio-sound sensor. The signal acquisition end of the auxiliary sensor is arranged facing away from the bio-sound sensor. The bio-sound sensor collects bio-sound signals conducted by the detection surface in a first direction, and the auxiliary sensors collect bio-sound signals conducted by the detection surface in a second direction. The direction in which the signal acquisition end of the auxiliary sensor is away from the biological sound sensor is the second direction, and the second direction is set at an angle to the first direction. When the biological sound sensor collects the biological sound signal transmitted by the detection surface along the first direction, the auxiliary sensor will also collect other vibration signals in the second direction, thereby realizing multi-channel collection and acquisition of vibration signals, so that the acquired vibration signal can be processed by means of signal fusion or signal filtering, so as to obtain the vibration signal of the biological sound more completely, restore the vibration characteristics of the biological sound to the greatest extent, effectively improve the signal-to-noise ratio of signal acquisition, and improve the acquisition quality and accuracy of the biological sound signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0017] Figure 1 This is a schematic diagram of the structure of a wearable device in one embodiment of the present invention; Figure 2 A schematic diagram of the use of a wearable device in one embodiment of the present invention; Figure 3 A schematic diagram of the steps of a signal processing method in one embodiment of the present invention; Figure 4 A schematic diagram of steps of a signal processing method in another embodiment of the present invention; Figure 5 A schematic diagram of the steps of using a filtering strategy to compare the second vibration signal with each first vibration signal to obtain multiple target biological sound signals in one embodiment of the present invention; Figure 6 A schematic diagram of steps of a signal processing method in another embodiment of the present invention; Figure 7 A schematic diagram of the steps of determining whether the bio-sound sensor is in a stable state within a preset time period in one embodiment of the present invention; Figure 8 The figure is a schematic diagram of the steps of filtering out the signal identical to the second vibration signal in the first vibration signal in one embodiment of the present invention.

[0018] Description of Figure Numbers: 100. Wearable device; 1. Shell; 1a. Surface shell; 1b. Bottom shell; 11. Installation cavity; 12. Detection surface; 2. Bio-sound sensor; 3. Auxiliary sensor; 4. Control board.

[0019] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components in a certain specific posture. If the specific posture changes, the directional indication will also change accordingly.

[0022] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features limited to "first" and "second" may explicitly or implicitly include at least one of the features. In addition, if "and / or" or "and / or" appears in the full text, its meaning includes three parallel solutions. Taking "A and / or B" as an example, it includes solution A, solution B, or solutions that satisfy both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of ordinary technicians in the field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0023] To achieve the above purpose, please refer to Figure 1 and Figure 2 As shown, the present invention proposes a wearable device 100, which includes a shell 1, a bio-sound sensor 2 and at least two auxiliary sensors 3. The shell 1 has an installation cavity 11 and a detection surface 12 facing away from the installation cavity 11, the detection surface 12 is configured to conduct bio-sound, the bio-sound sensor 2 is arranged on the bottom wall of the installation cavity 11, and the signal acquisition end of the bio-sound sensor 2 is arranged toward the bottom wall of the installation cavity 11. The bio-sound sensor 2 can collect the bio-sound signal conducted by the detection surface 12. Each auxiliary sensor 3 is arranged on the side wall of the installation cavity 11 and is electrically connected to the bio-sound sensor 2. The signal acquisition end of the auxiliary sensor 3 is arranged away from the bio-sound sensor 2; the bio-sound sensor 2 collects the bio-sound signal conducted by the detection surface 12 in a first direction, the signal acquisition end of the auxiliary sensor 3 is away from the bio-sound sensor 2 in a second direction, and the second direction is arranged at an angle to the first direction.

[0024] In this embodiment, the wearable device 100 can be a smart watch, a smart bracelet, a smart necklace, a smart ring and other devices. Bio-sound refers to the relevant sounds generated during the heart beating, lung breathing, intestinal digestion, muscle contraction, joint movement and other processes, such as heart sounds generated by heart beating, such as the sound of atrial contraction and ventricular contraction, such as breath sounds, wheezing sounds, dry rales during lung breathing, and bowel sounds during intestinal digestion. The bio-sound sensor 2 is placed on the chest near the heart or lung area, or on the abdomen to obtain vibration signals. The sound is transmitted to the outside world in the form of mechanical waves and can be received by relevant devices or sensors.

[0025] In this embodiment, the bio-sound sensor 2 includes at least one of a bone conduction sensor, a microphone, an acceleration sensor, an inertial sensor, a piezoelectric ceramic, a pressure sensor, and a vibration detection sensor, and the auxiliary sensor 3 is at least one of a vibration detection sensor or an acceleration sensor.

[0026] It can be understood that the bio-sound sensor 2 can be one or more of the above-mentioned bone conduction sensor, microphone, acceleration sensor, inertial sensor, piezoelectric ceramic, pressure sensor, and vibration detection sensor, and when multiple of the above-mentioned sensors are set, they can be the same sensor, or different sensors can be used together to adapt to different detection areas and detection environments.

[0027] The wearable device 100 in the present application includes at least one bio-sound sensor 2. The at least one bio-sound sensor 2 can be one of the above-mentioned sensors, or multiple types of sensors mentioned above can be used together to acquire signals. There is no limitation on this. The bio-sound sensor 2 receives pressure and sound wave vibrations, and makes the vibrations form analog signals, and then converts the analog signals into digital signals. After signal amplification and signal filtering, the background noise is removed to obtain the final bio-sounds such as heart sound characteristics and breathing characteristics.

[0028] In this embodiment, the shell 1 is the main supporting structure of the wearable device 100, which is a thin shell structure. The shell 1 includes a face shell 1a and a bottom shell 1b. The face shell 1a and the bottom shell 1b are detachably connected, such as by snaps or bolts. The face shell 1a and the bottom shell 1b together enclose an installation cavity 11, and the bio-sound sensor 2 and the auxiliary sensors 3 are arranged in the installation cavity 11.

[0029] The shell 1 is provided with a detection surface 12 on the side of the bottom shell 1b facing away from the surface shell 1a, that is, on the side facing away from the bottom wall of the installation cavity 11. When the user uses the wearable device 100, the detection surface 12 is close to the user's skin or the outside of the clothes. The bio-sound sensor 2 is arranged on the bottom wall of the installation cavity 11, and the signal collection end of the bio-sound sensor 2 is arranged toward the bottom wall of the installation cavity 11, so as to transmit the user's bio-sound vibration sound waves to the bio-sound sensor 2 through the detection surface 12 and the bottom shell 1b. At the same time, the auxiliary sensor 3 is arranged on the bottom wall of the installation cavity 11. On the side wall, the wearable device 100 also includes a control board 4, which includes but is not limited to an MCU microcontroller, a communication module such as Bluetooth or Wi-Fi, a memory, a display driver module, an input and output interface, a power management module, etc. The bio-sound sensor 2 and the auxiliary sensor 3 are both electrically connected or communicatively connected to the control board 4, such as directly electrically connected to the MCU controller of the control board 4, or communicatively connected to the control board 4 via Bluetooth or Wi-Fi, etc., so as to realize the electrical connection or communication connection between the bio-sound sensor 2 and the auxiliary sensor 3.

[0030] The bio-sound sensor 2 in the present application collects the bio-sound signal transmitted from the detection surface 12 along the first direction. However, during the signal collection process, changes in the external environment, or instability of the wearable device 100, movement or shaking of the user, and bio-sounds generated by other parts of the body, etc., may cause the bio-sound sensor 2 to collect other vibration signals in addition to the target bio-sound signal, thereby causing the bio-sound sensor 2 to obtain vibration sound wave signals in other directions. The vibration sound wave signals in multiple directions are intertwined, and the target bio-sound vibration sound wave signal transmitted along the first direction cannot be highlighted.

[0031] Based on this, the present application sets at least two auxiliary sensors 3 on the basis of setting the biological sound sensor 2. The at least two auxiliary sensors 3 detect vibration signals along a direction away from the biological sound sensor 2, that is, vibration signal detection along a second direction set at an angle to the first direction, so that the wearable device 100 can perform multi-channel vibration signal detection, and remove other vibration signals along the second direction in the biological sound sensor 2 by means of signal fusion or signal filtering, so as to accurately and completely obtain the target biological sound vibration signal, restore the vibration characteristics of the biological sound to the greatest extent, effectively improve the signal-to-noise ratio of signal acquisition, and improve the acquisition quality and accuracy of the biological sound signal.

[0032] The wearable device 100 of the present technical solution includes a shell 1, a bio-sound sensor 2 and at least two auxiliary sensors 3. The shell 1 has an installation cavity 11 and a detection surface 12 facing away from the installation cavity 11. The detection surface 12 is configured to conduct bio-sounds. The bio-sound sensor 2 is arranged on the bottom wall of the installation cavity 11. The signal collection end of the bio-sound sensor 2 is arranged toward the bottom wall of the installation cavity 11. The bio-sound sensor 2 can collect bio-sound signals conducted by the detection surface 12. Each auxiliary sensor 3 is arranged on the side wall of the installation cavity 11 and is electrically connected to the bio-sound sensor 2. The signal collection end of the auxiliary sensor 3 is arranged away from the bio-sound sensor 2. The bio-sound sensor 2 collects the bio-sound signals conducted by the detection surface 12. The biological sound signal is in the first direction, and the direction in which the signal collection end of the auxiliary sensor 3 is away from the biological sound sensor 2 is the second direction. The second direction is set at an angle to the first direction. When the biological sound sensor 2 collects the biological sound signal transmitted by the detection surface 12 along the first direction, the auxiliary sensor 3 will also collect other vibration signals in the second direction, thereby realizing multi-channel collection and acquisition of vibration signals, and the obtained vibration signals can be processed by means of signal fusion or signal filtering, so as to obtain the vibration signals of the biological sound more completely, restore the vibration characteristics of the biological sound to the greatest extent, effectively improve the signal-to-noise ratio of signal acquisition, and improve the acquisition quality and accuracy of the biological sound signal.

[0033] In one embodiment, if Figure 1 As shown, the direction in which the signal collection end of the auxiliary sensor 3 is away from the bio-sound sensor 2 is the detection direction, and the detection directions of at least two auxiliary sensors 3 are arranged perpendicular to each other; optionally, the second direction is arranged perpendicular to the first direction.

[0034] It can be understood that the signal acquisition end of the auxiliary sensor 3 is set away from the biological sound sensor 2, and the direction in which the signal acquisition end of each auxiliary sensor 3 is away from the biological sound sensor 2 is the detection direction. The detection directions of at least two auxiliary sensors 3 together form the above-mentioned second direction that is at an angle to the first direction, wherein the detection directions of at least two auxiliary sensors 3 are set perpendicular to each other, so that at least two auxiliary sensors 3 can detect two mutually orthogonal directions to obtain a larger and more complete detection range, and can be effectively fused or filtered in the biological sound sensor 2 to obtain a more accurate and complete target biological sound vibration signal.

[0035] At the same time, the second direction and the first direction can also be set perpendicular to each other, so that the first direction and the detection direction of at least two auxiliary sensors 3 are set perpendicular to each other, so that the biological sound sensor 2 can effectively remove or fuse the noise vibration signal in the plane perpendicular to the first direction, so as to obtain the vibration signal of the biological sound more completely, restore the vibration characteristics of the biological sound to the greatest extent, effectively improve the signal-to-noise ratio of signal acquisition, and improve the acquisition quality and accuracy of the biological sound signal.

[0036] In one embodiment, the bio-sound sensor 2 includes one of a bone conduction sensor, a microphone, an acceleration sensor, an inertial sensor, a piezoelectric ceramic, a pressure sensor, and a vibration detection sensor; optionally, the auxiliary sensor 3 is an acceleration sensor; optionally, the auxiliary sensor 3 is a vibration detection sensor.

[0037] In one embodiment, the bio-sound sensor 2 includes at least one of a bone conduction sensor, a microphone, an acceleration sensor, an inertial sensor, a piezoelectric ceramic, a pressure sensor, and a vibration detection sensor.

[0038] It can be understood that the bio-sound sensor 2 can be one or more of the above-mentioned bone conduction sensor, microphone, acceleration sensor, inertial sensor, piezoelectric ceramic, pressure sensor, and vibration detection sensor, and when multiple of the above-mentioned sensors are set, they can be the same sensor, or different sensors can be used together to adapt to different detection areas and detection environments.

[0039] Among them, if the bio-sound sensor 2 is a bone conduction sensor, it can directly contact the bones and detect the propagation of bio-sound in the body by detecting the vibration of the body bones; if the bio-sound sensor 2 is a microphone, it can detect the bio-sound generated in the body or on the body surface through the vibration sound waves in the air; if the bio-sound sensor 2 is an accelerometer sensor, when the heart beats or the lungs breathe, the body will produce tiny movements and vibrations, and the accelerometer can obtain the changes in the above movements; if the bio-sound sensor 2 is a pressure sensor, the pressure sensor can be placed on the surface of the body to capture the pressure changes caused by the heart beating or lung activity. The above types of sensors can convert tiny vibrations into electrical signals and effectively extract the characteristics of bio-sounds to realize the detection of bio-sounds.

[0040] In one embodiment, the housing 1 is provided with a detection hole, which communicates with the mounting cavity 11 and passes through the detection surface 12 ; a portion of the bio-sound sensor 2 is disposed in the detection hole and protrudes from the detection surface 12 .

[0041] In this embodiment, a detection hole is opened in the bottom shell 1b of the shell 1, and the detection hole connects the installation cavity 11 and the external environment. The bio-sound sensor 2 includes a substrate, an elastic detection protrusion and one or more of the above-mentioned conductive sensors, microphones, acceleration sensors, inertial sensors, piezoelectric ceramics, pressure sensors, and vibration detection sensors. The substrate is connected to the bottom wall of the installation cavity 11, and the elastic detection protrusion is movably connected to the substrate. The above-mentioned sensor is arranged in the elastic detection protrusion and is located between the elastic detection protrusion and the substrate. The elastic detection protrusion and the substrate jointly press against the sensor, and the elastic detection protrusion can extend out of the detection hole. When the wearable device 100 presses against the user for bio-sound detection, the elastic detection protrusion will produce elastic deformation and press against the sensor, so that the elastic detection protrusion is always in close contact with the skin surface of the user, so that the internal sensor can detect vibration sound waves.

[0042] It can be understood that the present application sets the sensor in an elastic detection protrusion made of elastic material or elastic structure, so that the sensor is isolated by the elastic detection protrusion and the elastic contact with the substrate to avoid contact between the sensor and the housing or other structures of the wearable device 100, and the detection end of the elastic detection protrusion contacts the user's skin surface, and can also transmit vibration to the sensor through the elastic deformation of the elastic detection protrusion itself to realize sensor detection, so that it is no longer affected by the housing of the wearable device 100, and can obtain vibration signals more completely, restore vibration characteristics to the greatest extent, effectively improve the signal-to-noise ratio of signal acquisition, and improve the accuracy and effect of biological sound detection.

[0043] The present invention also proposes a signal processing method, such as Figure 3 As shown, the signal processing method is implemented by the wearable device 100, and the steps of the signal processing method include: S11, obtaining a first vibration signal collected by the bio-sound sensor 2 within a preset time period; S21, obtaining a second vibration signal collected by the auxiliary sensor 3 within a preset time period; S31, using a filtering strategy to obtain a target biological sound signal; S41, the step of filtering strategy includes: determining whether there is a signal in the first vibration signal that is the same as the second vibration signal; S41a, if yes, filter out the signal identical to the second vibration signal in the first vibration signal to obtain the target biological sound signal; S41b, if not, use the collected first vibration signal as the target biological sound signal.

[0044] In this embodiment, the signal processing method is implemented based on the wearable device 100. The signal processing method first obtains the first vibration signal collected by the bio-sound sensor 2 within a preset time length, and the first vibration signal is a vibration signal in multiple directions including the user's bio-sound vibration signal. Then, the second vibration signal collected by the auxiliary sensor 3 within a preset time length is obtained. The number of auxiliary sensors 3 is at least two, and therefore at least two second vibration signals will be generated accordingly. A filtering strategy is adopted to obtain the user's bio-sound signal. The filtering strategy is to determine whether there is a signal identical to the second vibration signal in the first vibration signal. If so, the signal identical to the second vibration signal is filtered out in the first vibration signal, and then the signal after filtering in the first vibration signal is the target bio-sound signal. If not, the first vibration signal can be directly used as the target bio-sound signal.

[0045] It can be understood that the preset duration is the time when the wearable device 100 abuts against the user, and the bio-sound sensor 2 and the auxiliary sensor 3 start to detect the vibration signal at the same time, that is, within the preset duration, the bio-sound sensor 2 and the auxiliary sensor 3 start and end the detection of the vibration signal at the same time, so as to obtain the target bio-sound signal in the same time dimension, and when there is a signal identical to the second vibration signal in the first vibration signal, the signal identical to the second vibration signal can be filtered out in the same time dimension to obtain the target bio-sound signal.

[0046] Among them, the first vibration signal obtained by the biosound sensor 2 and the second vibration signal obtained by the auxiliary sensor 3 are both continuous analog signals within a preset time length, and then the continuous analog signals are converted into digital signals through an analog-to-digital converter, and the similarity between the first vibration signal and the second vibration signal is analyzed through signal processing methods such as spectrum analysis or time domain analysis to determine whether there is a signal in the first vibration signal that is the same as the second vibration signal.

[0047] It can be understood that the first vibration signal not only includes the bio-sound signal transmitted from the detection surface 12 collected by the bio-sound sensor 2 along the first direction, but also includes vibration signals generated in other directions at an angle to the first direction. That is to say, the first vibration signal collected by the bio-sound sensor 2 is a variety of vibration signals including target bio-sound signals and environmental vibration noise signals. The target bio-sound signal specifically refers to the user's bio-sound vibration signal collected along the direction perpendicular to the detection surface 12 (the first direction), and the vibration signal collected by the auxiliary sensor 3 along the second direction set at an angle to the signal collection direction of the bio-sound sensor 2 is the collected environmental vibration noise. After the judgment and filtering steps, the environmental noise in the first vibration signal can be filtered out to retain the bio-sound vibration signal, that is, the target bio-sound signal to be obtained.

[0048] It can be understood that the biological sound sensor 2 collects the biological sound signal transmitted by the detection surface 12 in the first direction, and the signal collection end of the auxiliary sensor 3 is away from the biological sound sensor 2 in the second direction. The second direction is set at an angle to the first direction. When the biological sound sensor 2 collects the biological sound signal transmitted by the detection surface 12 along the first direction, the auxiliary sensor 3 will also collect other vibration signals in the second direction, thereby realizing multi-channel collection and acquisition of vibration signals, so that the acquired vibration signals can be processed by means of signal fusion or signal filtering, so as to obtain the vibration signals of the biological sound more completely, restore the vibration characteristics of the biological sound to the greatest extent, effectively improve the signal-to-noise ratio of signal acquisition, and improve the acquisition quality and accuracy of the biological sound signal.

[0049] In one embodiment, if Figure 4 As shown, the wearable device 100 includes a plurality of bio-sound sensors 2, and the plurality of bio-sound sensors 2 are arranged in an array on the bottom wall of the mounting cavity 11. The signal processing method further includes: S12, obtaining a plurality of first vibration signals collected by a plurality of bio-sound sensors 2 within a preset time period; S22, using a filtering strategy to compare the second vibration signal with each of the first vibration signals to obtain a plurality of target biological sound signals; S32, selecting a plurality of preset time points within a preset time period, and calculating the signal average of the target biological sound signal collected at the plurality of preset time points; S42, mapping each preset time point with the signal average value of each target biological sound signal; and determining the regional biological sound signal corresponding to the area to be monitored according to the signal average value.

[0050] In this embodiment, the wearable device 100 includes a plurality of bio-sound sensors 2, and the plurality of bio-sound sensors 2 can be arranged in parallel along a straight line or in a rectangular array, and the plurality of bio-sound sensors 2 are all arranged on the bottom wall of the mounting cavity 11, and the signal acquisition ends of the plurality of bio-sound sensors 2 are all arranged toward the detection surface 12, so that the plurality of bio-sound sensors 2 can occupy the entire bottom wall of the mounting cavity 11 as much as possible, so as to maximize the detection surface 12 that corresponds to the back of the bottom wall of the mounting cavity 11, so as to maximize the acquisition of the bio-sound signal transmitted from the detection surface 12, that is, to accurately and completely acquire the complete bio-sound vibration characteristics of the detection area covered by the detection surface 12.

[0051] At the same time, multiple bio-sound sensors 2 are combined to form a bio-sound sensor 2 array structure, which is arranged on the bottom wall of the mounting cavity 11 as a complete sensor structure, so that at least two auxiliary sensors 3 are arranged at the periphery of the bio-sound sensor 2 array structure, and at least two auxiliary sensors 3 are arranged to surround the bio-sound sensor 2 array structure, so as to obtain a second vibration signal relative to the entire bio-sound sensor 2 array structure through the auxiliary sensors 3.

[0052] In this embodiment, based on the fact that multiple bio-sound sensors 2 collect multiple first vibration signals within a preset time period, the above-mentioned filtering strategy is used to compare the second vibration signal with each first vibration signal. After the comparison, the wearable device 100 obtains multiple target bio-sound signals, and selects multiple preset time points within the entire preset time period, and calculates the average value of the target bio-sound signals collected at the multiple preset time points.

[0053] It can be understood that the wearable device 100 calculates the signal average of the target biological sound signal collected at the preset time point, and can mathematically average multiple target biological sound signals collected at the preset time point. Signal averaging can effectively reduce the impact of random noise and improve the stability and reliability of the signal.

[0054] At the same time, the wearable device 100 maps each preset time point with the signal average value of the corresponding target bio-sound signal. Within a preset period of time, such as one second, several preset time points are selected, such as the 10th millisecond, the 50th millisecond and the 90th millisecond of each second, and all target bio-sound signals at these three time points are collected respectively. For each preset time point, the values ​​of all target bio-sound signals are added, and then divided by the number of target bio-sound signals, and the signal average value at the time point is obtained, thereby reducing the influence of random noise and improving the stability and accuracy of the signal. Finally, through these signal average values, the bio-sound state of the area to be monitored, such as heart rate, respiratory rate, etc., can be inferred. According to the signal average value, the wearable device 100 can determine the regional bio-sound signal corresponding to the area to be monitored to analyze the changes in the bio-sound signals in different areas at different time points. This process helps to analyze the health status or physiological state of the organism.

[0055] In one embodiment, if Figure 5 As shown, the steps of using a filtering strategy to compare the second vibration signal with each first vibration signal to obtain multiple target biological sound signals include: S221, obtaining a second vibration signal of each auxiliary sensor 3; S222, comparing each second vibration signal with the first vibration signal in sequence; S223, determining whether there is a signal in the first vibration signal that is identical to each of the second vibration signals; S223a, if yes, remove the signal identical to the second vibration signal from the first vibration signal to obtain the target biological sound signal; S223b. If not, the collected first vibration signal is used as the target biological sound signal.

[0056] In this embodiment, each auxiliary sensor 3 collects vibration signals along a second direction that is angled with the first direction, and each auxiliary sensor 3 obtains at least two second vibration signals, which are compared with the first vibration signal respectively. The comparison method is a similarity measurement, such as cross-correlation, cosine similarity, Euclidean distance, etc. to determine whether there is a signal in the first vibration signal that is identical to each second vibration signal. If cross-correlation is used for similarity judgment, a maximum threshold can be set. According to the set maximum threshold, if the similarity exceeds the maximum threshold, it can be determined that each second vibration signal is identical to a part of the signal in the first vibration signal, and then the identical signal is taken out from the first vibration signal, such as by direct truncation, signal replacement or interpolation for smoothness processing to obtain the final target biological sound signal.

[0057] At the same time, the above steps can be repeated until all the second vibration signals are compared with the first vibration signals, or all the second vibration signals are compared with each first vibration signal to obtain one or more target biological sound signals.

[0058] It can be understood that by comparing each second vibration signal with the first vibration signal in turn, multiple vibration noises in the second direction that is angled with the first direction can be gradually removed, that is, the vibration noise of the bio-sound sensor 2 in the second direction can be removed, so as to retain the bio-sound vibration signal collected by the bio-sound sensor 2 through the detection surface 12 in the first direction, thereby realizing multi-channel collection and acquisition of vibration signals, and processing the acquired vibration signals through signal filtering and other means, so as to more completely acquire the vibration signal of the bio-sound, restore the vibration characteristics of the bio-sound to the greatest extent, effectively improve the signal-to-noise ratio of signal acquisition, and improve the acquisition quality and accuracy of the bio-sound signal.

[0059] In one embodiment, if Figure 6 As shown, after obtaining the second vibration signal collected by the auxiliary sensor 3 within a preset time period, the signal processing method further includes: S51, determining whether the bio-sound sensor 2 is in a stable state within a preset time period; S51a, if yes, acquiring the first vibration signal through the bio-sound sensor 2 and the second vibration signal through the auxiliary sensor 3, and increasing the sampling frequency of the bio-sound sensor 2 in a stable state; S51b, if not, stop collecting the first vibration signal through the bio-sound sensor 2, and stop collecting the second vibration signal through the auxiliary sensor 3.

[0060] In this embodiment, during the signal collection process of the bio-sound sensor 2 and the auxiliary sensor 3 in the wearable device 100, since the user may be in a state of walking, moving, shaking, etc., the wearable device 100 is unstable as a whole. Compared with the vibration signal collected in a stable state, the vibration signal collected by the bio-sound sensor 2 and the auxiliary sensor 3 has a large deviation and low accuracy. Based on this, the signal processing method also needs to determine whether the bio-sound sensor 2 is in a stable state within a preset time length. If it is in a stable state, the first vibration signal is obtained through the bio-sound sensor 2 and the second vibration signal is collected through the auxiliary sensor 3.

[0061] In this embodiment, to determine whether the bio-sound sensor 2 is in a stable state within a preset time, the stability of the signal can be monitored in real time by the bio-sound sensor 2 and the auxiliary sensor 3, and the environmental vibration signal can also be collected by other auxiliary sensors 3, such as accelerometers, gyroscopes, etc., and then a statistical method (such as mean, standard deviation, peak value, etc.) can be used to determine whether the signal is in a stable state. For example, a signal threshold can be set, and when the standard deviation of the signal is less than the signal threshold, the signal is considered to be in a stable state.

[0062] Furthermore, after confirming that the signal is in a stable state, the sampling frequency of the bio-sound sensor 2 can be dynamically adjusted as needed. This can be achieved by increasing the number of sampling points or reducing the sampling time interval, or setting a suitable sampling frequency according to the characteristics of the collected signal (such as frequency range, change rate, etc.). For example, if the frequency of the signal is high, the sampling frequency can be appropriately increased to avoid aliasing. Alternatively, a real-time feedback mechanism can be used to dynamically adjust the sampling frequency according to the current signal changes. For example, when the rate of change of the signal is detected to increase, the sampling frequency can be temporarily increased.

[0063] At the same time, the vibration signal characteristic parameters of the bio-sound sensor 2 are monitored in real time to ensure that the bio-sound sensor 2, the auxiliary sensor 3 and the entire wearable device 100 are still in a stable state. If the characteristic parameters fluctuate or change suddenly, the sampling frequency needs to be adjusted in time or other measures need to be taken to restore the stable state. At the new sampling frequency, the first vibration signal and the second vibration signal are re-collected, and corresponding signal processing and feature extraction are performed.

[0064] It is understandable that determining the stable state can ensure that the wearable device 100 collects the first vibration signal and the second vibration signal in the best state, thereby obtaining the biological sound signal more completely and accurately, restoring the vibration characteristics of the biological sound to the greatest extent, effectively improving the signal-to-noise ratio of signal collection, and improving the collection quality and accuracy of the biological sound signal. At the same time, increasing the sampling frequency can capture more subtle signal changes and improve the resolution of the data.

[0065] In one embodiment, if Figure 7 As shown, the step of determining whether the bio-sound sensor 2 is in a stable state within a preset time period includes: S511, storing the collected second vibration signal to form noise vibration data; S512, searching for a maximum value in the noise vibration data as an abnormal vibration value; S513, taking the time when the abnormal vibration value appears within the preset time as the abnormal point; S514, determining whether the number of abnormal points is less than or equal to the preset number of abnormal points within a preset time length; S514a, if yes, the bio-sound sensor 2 is in a stable state within a preset time period; S514b: If not, the bio-sound sensor 2 is not in a stable state within the preset time period.

[0066] In this embodiment, the collected second vibration signal is stored to form noise vibration data, so as to facilitate the subsequent processing and analysis of the noise vibration data, and the maximum value is found in the noise vibration data, which determines whether the system considers that the bio-sound sensor 2 is in a stable state. In the noise vibration data, the maximum value is found to identify possible abnormal vibrations. These abnormal vibrations may be caused by external interference, equipment failure or other abnormal conditions. A threshold can be set, and only when the vibration signal exceeds the threshold, it is regarded as an abnormal vibration value. In addition, it can be considered to use sliding window technology to analyze the data step by step to better capture instantaneous abnormal changes. A preset threshold value for the number of abnormal points is set. If the number of abnormal points found within a preset time period is less than or equal to this threshold, then the working state of the sensor can be considered to be stable. This is because under normal circumstances, the vibration of the sensor should fluctuate within a relatively stable range, and only when encountering abnormal conditions will a larger vibration occur, that is, an abnormal point appears. If the number of abnormal points is too large, it means that there may be a continuous unstable factor affecting the performance of the sensor. If yes, the bio-sound sensor 2 is in a stable state within the preset time; if no, the bio-sound sensor 2 is not in a stable state within the preset time: According to the result of the previous step, it can be concluded that if the number of abnormal points meets the preset conditions, it means that the working state of the sensor is stable during the time period, otherwise it is considered that its working state is not stable and needs further inspection or adjustment.

[0067] It can be understood that through the above steps, the collected second vibration signal can be used to determine whether the bio-sound sensor 2 is in a stable state within a preset time length, wherein the number of abnormal points is used as an indicator to measure the frequency of abnormal situations, which helps to promptly discover possible problems or fluctuations, that is, it can determine whether the bio-sound sensor 2 is in a stable state.

[0068] In one embodiment, if Figure 8 As shown, the step of filtering out the signal identical to the second vibration signal in the first vibration signal includes: S41a1, generating corresponding noise reduction control parameters according to the signal characteristic index of the second vibration signal; S41a2, establishing a noise generation adversarial network according to the noise reduction control parameters; S41a3, when the first vibration signal contains a signal identical to the second vibration signal, a noise adversarial network is generated to generate and output a corresponding reverse second vibration signal; S41a4, the frequency and amplitude of the reverse second vibration signal are equal to those of the second vibration signal, and the phase of the reverse second vibration signal is opposite to that of the second vibration signal; S41a5, input the reverse second vibration signal to the bio-sound sensor 2 to remove the second vibration signal.

[0069] In this embodiment, the signal characteristic indicators of the second vibration signal include but are not limited to frequency characteristics, such as the center frequency and bandwidth of the second vibration signal; amplitude characteristics, such as the maximum amplitude and root mean square amplitude of the second vibration signal; phase characteristics and time characteristics, etc. The above-mentioned signal characteristic indicators are extracted and converted into noise reduction control parameters, and the noise reduction control parameters are input into the generative adversarial network architecture, and feature extraction and learning are performed through the neural network to form a noise generative adversarial network.

[0070] Among them, the second vibration signal is used as a real sample, and after the characteristics of the second vibration signal are learned through the neural network in the noise generative adversarial network, a reverse second vibration signal similar to the second vibration signal is generated. Through adversarial training, the generated reverse second vibration signal is matched with the original second vibration signal as much as possible, while being different from the real second vibration signal, ensuring that the generated reverse signal is equal to the second vibration signal in frequency and amplitude, but opposite in phase.

[0071] Since the reverse second vibration signal has an opposite phase to the second vibration signal, the two will produce an interference effect when superimposed, causing the signals to cancel each other out. After interference, the part of the first vibration signal that is the same as the second vibration signal is effectively removed, thereby obtaining a clear and accurate biological sound signal.

[0072] It is understandable that by using a generative adversarial network, a reverse second vibration signal with the same frequency and amplitude as the second vibration signal, as well as a reverse second vibration signal with the opposite phase to the second vibration signal can be actively generated, and the second vibration signal can be removed by signal superposition, thereby filtering out the signal identical to the second vibration signal in the first vibration signal to obtain the target biological sound signal, thereby obtaining the biological sound signal more completely and accurately, restoring the vibration characteristics of the biological sound to the greatest extent, effectively improving the signal-to-noise ratio of signal acquisition, and improving the acquisition quality and accuracy of the biological sound signal. At the same time, increasing the sampling frequency can capture more subtle signal changes and improve the resolution of the data.

[0073] The above description is only an exemplary embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the technical concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A wearable device, characterized in that: The wearable device comprises: A housing, the housing having a mounting cavity and a detection surface facing away from the mounting cavity, the detection surface being configured to conduct biological sound; A bio-sound sensor, wherein the bio-sound sensor is disposed on the bottom wall of the installation cavity, a signal collection end of the bio-sound sensor is disposed toward the bottom wall of the installation cavity, and the bio-sound sensor can collect the bio-sound signal conducted by the detection surface; and At least two auxiliary sensors, each of which is disposed on a side wall of the mounting cavity and is electrically connected to the bio-sound sensor, and a signal collection end of the auxiliary sensor is disposed facing away from the bio-sound sensor; The biological sound sensor collects the biological sound signal transmitted by the detection surface in a first direction, and the signal collection end of the auxiliary sensor is directed away from the biological sound sensor in a second direction, and the second direction is set at an angle to the first direction.

2. The wearable device according to claim 1, wherein: The signal detection directions of at least two of the auxiliary sensors are arranged perpendicular to each other; And / or, the second direction and the first direction are arranged perpendicular to each other.

3. The wearable device according to any one of claims 1 or 2, characterized in that: The bio-sound sensor includes one of a bone conduction sensor, a microphone, an acceleration sensor, an inertial sensor, a piezoelectric ceramic, a pressure sensor, and a vibration detection sensor; And / or, the auxiliary sensor is an acceleration sensor; And / or, the auxiliary sensor is a vibration detection sensor.

4. The wearable device according to any one of claims 1 or 2, characterized in that: The shell is provided with a detection hole, the detection hole is connected to the installation cavity and passes through the detection surface; Part of the bio-sound sensors are disposed in the detection hole and protrude from the detection surface.

5. A signal processing method, the signal processing method being implemented by the wearable device according to any one of claims 1 to 4, characterized in that: The steps of the signal processing method include: Acquire a first vibration signal collected by the bio-sound sensor within a preset time period; Acquire a second vibration signal collected by the auxiliary sensor within the preset time period; A filtering strategy is used to obtain the target biological sound signal; The steps of the filtering strategy include: determining whether there is a signal in the first vibration signal that is the same as the second vibration signal; If yes, filtering out the signal identical to the second vibration signal from the first vibration signal to obtain the target biological sound signal; If not, the collected first vibration signal is used as the target biological sound signal.

6. The signal processing method according to claim 5, characterized in that: The wearable device includes a plurality of bio-sound sensors, and the plurality of bio-sound sensors are arranged in an array on the bottom wall of the installation cavity. The signal processing method further includes: Acquire a plurality of the first vibration signals collected by the plurality of the bio-sound sensors within the preset time period; using the filtering strategy to compare the second vibration signal with each of the first vibration signals to obtain a plurality of the target biological sound signals; Within the preset time period, selecting a plurality of preset time points, and calculating the signal average of the target biological sound signal collected at the plurality of preset time points; Mapping each of the preset time points with a signal average value of each of the target biological sound signals; and determining the regional biological sound signal corresponding to the area to be monitored based on the signal average value.

7. The signal processing method according to claim 6, characterized in that: The step of using the filtering strategy to compare the second vibration signal with each of the first vibration signals to obtain a plurality of the target biological sound signals comprises: Acquire the second vibration signal of each of the auxiliary sensors; comparing each of the second vibration signals with the first vibration signal in sequence; determining whether the first vibration signal contains a signal identical to each of the second vibration signals; If yes, removing the signal identical to the second vibration signal from the first vibration signal to obtain the target biological sound signal; If not, the collected first vibration signal is used as the target biological sound signal.

8. The signal processing method according to claim 5, characterized in that: After acquiring the second vibration signal collected by the auxiliary sensor within the preset time period, the signal processing method further includes: Determining whether the bio-sound sensor is in a stable state within the preset time period; If yes, acquiring the first vibration signal through the bio-sound sensor and acquiring the second vibration signal through the auxiliary sensor, and increasing the sampling frequency of the bio-sound sensor in the stable state; If not, stop collecting the first vibration signal through the bio-sound sensor, and stop collecting the second vibration signal through the auxiliary sensor.

9. The signal processing method according to claim 8, characterized in that: The step of determining whether the bio-sound sensor is in a stable state within the preset time period comprises: storing the collected second vibration signal to form noise vibration data; Finding a maximum value in the noise and vibration data as an abnormal vibration value; The time when the abnormal vibration value appears within the preset time length is regarded as the abnormal point; Determine whether the number of abnormal points is less than or equal to the preset number of abnormal points within the preset time length; If yes, the bio-sound sensor is in a stable state within the preset time period; If not, the bio-sound sensor is not in a stable state within the preset time period.

10. The signal processing method according to any one of claims 5 to 9, characterized in that: The step of filtering out the signal identical to the second vibration signal in the first vibration signal comprises: generating corresponding noise reduction control parameters according to the signal characteristic index of the second vibration signal; Establishing a noise generation adversarial network according to the noise reduction control parameters; When the first vibration signal contains a signal identical to the second vibration signal, the noise generation adversarial network generates and outputs a corresponding reverse second vibration signal; The frequency and amplitude of the reverse second vibration signal are equal to those of the second vibration signal, and the phase of the reverse second vibration signal is opposite to that of the second vibration signal; The reverse second vibration signal is input to the bio-sound sensor to remove the second vibration signal.

Images