An integrated wearable ultrasonic gesture recognition circuit and system

Abstract

The application discloses an integrated wearable ultrasonic gesture recognition circuit and system, which comprises a wearable ultrasonic sensing probe, an ultralow-power driving control circuit and a microcontroller, the ultralow-power driving control circuit is connected with the wearable ultrasonic sensing probe and the microcontroller, the microcontroller controls the ultralow-power driving control circuit to output a control signal, the wearable ultrasonic sensing probe emits ultrasonic waves to deep muscle tissue according to the control signal and receives ultrasonic echo signals, the ultralow-power driving control circuit converts the ultrasonic echo signals into echo data, and the microcontroller processes the echo data to recognize a gesture, so that the circuit and system can realize gesture recognition and solve the technical problems of high power consumption, large size and difficulty in adapting to changes in skin coupling impedance of an ultrasonic device.

Description

Technical Field

[0001] This invention belongs to the field of human-computer interaction and ultrasonic electronics technology, and relates to an integrated wearable ultrasonic gesture recognition circuit and system. Background Technology

[0002] With the widespread adoption of wearable devices, human-computer interaction (HMI) technology is evolving from traditional touch control to more refined gesture control and biosignal-based control. Currently, the mainstream biosignal interaction schemes mainly rely on surface electromyography (sEMG). However, sEMG signals are easily affected by changes in skin impedance (such as sweating), electromagnetic interference, and electrode contact status. Furthermore, its signal source mainly comes from superficial muscle potentials, making it difficult to accurately distinguish the fine activities of deep muscle groups, thus limiting its application in complex gesture recognition.

[0003] In contrast, sonomyography (SMG) uses ultrasound waves to detect the mechanical deformation of muscles and tendons, offering higher spatial resolution and directly reflecting the physical contraction state of muscles. Therefore, it has a significant advantage in gesture recognition accuracy. However, most existing ultrasound imaging or sensing devices utilize the hardware architecture of medical ultrasound, which presents the following technical bottlenecks: First, there are power consumption and size issues caused by high-voltage driving. Traditional ultrasound probes typically require high-voltage pulses of 50V to 100V for excitation, which necessitates bulky and inefficient charge pump circuits. This not only increases the PCB area but also leads to significant power consumption and heat generation, making it difficult for devices to meet the requirements of lightweight design and long battery life (e.g., all-day wear) in wearable scenarios.

[0004] Secondly, the analog front-end circuitry is complex. Weak ultrasonic echo signals typically require high-gain (>40dB), low-noise analog amplifiers (LNAs) and variable-gain amplifiers (VGAs) for conditioning, which further increases the system's power consumption and hardware cost.

[0005] Finally, there's the issue of acoustic impedance mismatch and frequency drift. In wearable mode, the coupling state between the probe and the skin (contact pressure, wetness / dryness) is dynamically changing. The load effect of the skin tissue causes the probe's resonant frequency to drift. Traditional fixed-frequency driving methods cannot adapt to this change, leading to decreased transmission efficiency and a deteriorated echo signal-to-noise ratio.

[0006] Therefore, there is an urgent need for a low-power wearable ultrasonic gesture recognition system that can eliminate high-voltage circuits, simplify the analog front end, and adapt to changes in coupling state. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide an integrated wearable ultrasound gesture recognition circuit and system. This circuit and system can realize gesture recognition and solve the technical problems of high power consumption, large size and difficulty in adapting to changes in skin coupling impedance of ultrasound equipment.

[0008] To achieve the above objectives, this invention discloses an integrated wearable ultrasonic gesture recognition circuit, comprising a wearable ultrasonic sensing probe, an ultra-low power drive control circuit, and a microcontroller. The ultra-low power drive control circuit is connected to the wearable ultrasonic sensing probe and the microcontroller. The microcontroller controls the ultra-low power drive control circuit to output a control signal. The wearable ultrasonic sensing probe emits ultrasonic waves to deep muscle tissue and receives ultrasonic echo signals according to the control signal. The ultra-low power drive control circuit converts the ultrasonic echo signals into echo data. The microcontroller processes the echo data to recognize the gesture.

[0009] Furthermore, the wearable ultrasonic sensor probe uses a PZT-5 piezoelectric ceramic sheet as the transducer material; the backing layer of the wearable ultrasonic sensor probe is made of a PCB substrate; and a matching layer is coated on the surface of the wearable ultrasonic sensor probe 100.

[0010] Furthermore, the pulse emission unit in the ultra-low power drive control circuit 200 includes a power supply VCC, an energy storage capacitor C1, an N-channel enhancement-mode MOSFET Q1, and a current-limiting resistor R1. One end of the energy storage capacitor C1 is connected to the power supply VCC, and the other end of the energy storage capacitor C1 is grounded. The drain D of the N-channel enhancement-mode MOSFET Q1 is connected to the negative terminal of the wearable ultrasonic sensing probe 100, the source S of the N-channel enhancement-mode MOSFET Q1 is grounded, and the gate G of the N-channel enhancement-mode MOSFET Q1 is connected to the PWM output pin of the microcontroller. The current-limiting resistor R1 is connected in series between the power supply VCC and the drain D of the N-channel enhancement-mode MOSFET Q1.

[0011] Furthermore, the pulse emission unit also includes a first protection diode D1 and a second protection diode D2, which are connected in parallel in reverse at both ends of the wearable ultrasonic sensing probe.

[0012] Furthermore, the passive receiving unit in the ultra-low power drive control circuit includes a passive high-pass filter, wherein the passive high-pass filter consists of a series capacitor. and parallel resistors composition.

[0013] Furthermore, when the N-channel enhancement-mode MOS transistor Q1 is turned on, the negative potential of the wearable ultrasonic sensor probe is pulled down to ground, and a negative step voltage is generated by utilizing the capacitance characteristics of the wearable ultrasonic sensor probe itself.

[0014] Furthermore, during the initialization phase, the pulse transmitting unit adjusts the DC operating point of the circuit through a sliding rheostat and outputs pulses of different widths in sequence to detect the echo intensity. The pulse width that generates the maximum echo intensity is locked as the working pulse width to match the probe resonant frequency under the current skin load.

[0015] Furthermore, the microcontroller runs a lightweight convolutional neural network accelerated by lookup tables to classify the echo data in order to recognize gestures.

[0016] Furthermore, during the training of the lightweight convolutional neural network, random temporal offsets and amplitude scaling transformations are applied to the original training data to obtain the final training data, which is then used to train the lightweight convolutional neural network.

[0017] This invention discloses an integrated wearable ultrasonic gesture recognition system, including a peripheral communication module, an external device, and an integrated wearable ultrasonic gesture recognition circuit. The microcontroller is connected to the external device via the peripheral communication module.

[0018] The present invention has the following beneficial effects: In practical operation, the integrated wearable ultrasound gesture recognition circuit and system described in this invention uses a microcontroller to control an ultra-low power drive control circuit to output a control signal. The wearable ultrasound sensing probe emits ultrasound waves into deep muscle tissue and receives ultrasound echo signals according to the control signal. This solves the technical problems of high power consumption, large size, and difficulty in adapting to changes in skin coupling impedance of ultrasound devices. The ultra-low power drive control circuit converts the ultrasound echo signals into echo data, and the microcontroller processes the echo data to recognize gestures, making it highly practical.

[0019] Furthermore, this invention achieves broadband frequency tracking technology through pulse width adjustment, effectively solving the resonant frequency drift problem caused by skin coupling, and still ensuring the echo signal-to-noise ratio under low-voltage drive.

[0020] Furthermore, the convolution acceleration method based on lookup tables in this invention overcomes the shortcomings of insufficient computing power of low-power microcontrollers and enables real-time inference of complex deep learning algorithms. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a structural diagram of the present invention; Figure 2 The circuit schematic of the ultra-low power drive control circuit 200; Figure 3 A schematic diagram showing the matching of the transmission spectrum and the probe resonant frequency under different pulse widths; Figure 4 This is a schematic diagram of the convolution operation process accelerated by lookup tables; Figure 5 This is a flowchart of the method of the present invention.

[0023] Among them, 100 is a wearable ultrasonic sensing probe, 200 is an ultra-low power drive control circuit, and 300 is a microcontroller. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] In the description of this invention, it should be understood that the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0026] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0027] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this invention generally indicates that the preceding and following objects have an "or" relationship.

[0028] It should be understood that although terms such as first, second, third, etc., may be used in the embodiments of the present invention to describe the preset range, these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from one another. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0029] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0031] The accompanying drawings illustrate various structural schematic diagrams according to embodiments disclosed in this invention. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0032] This invention discloses an integrated wearable ultrasound gesture recognition circuit and system, which solves the problems of high power consumption, large size and difficulty in adapting to changes in skin coupling impedance caused by the reliance on high voltage drive in existing wearable ultrasound devices. It is particularly suitable for wearable devices such as smartwatches, wristbands or dedicated myoelectric armbands.

[0033] Example 1 refer to Figure 1The integrated wearable ultrasonic gesture recognition circuit includes a wearable ultrasonic sensor probe 100, an ultra-low power drive control circuit 200, and a microcontroller 300. The ultra-low power drive control circuit 200 is connected to the wearable ultrasonic sensor probe 100 and the microcontroller 300. The wearable ultrasonic sensor probe 100 is a patch structure that fits tightly against the skin surface of the human forearm. It is used to emit ultrasonic waves into deep muscle tissue and receive ultrasonic echo signals reflected back from muscle fiber bundles, tendons, and bone interfaces. The deep muscle tissue can be the flexor digitorum superficialis and flexor digitorum profundus muscles.

[0034] When the present invention is in operation, it emits ultrasonic waves into deep muscle tissue and receives ultrasonic echo signals through a wearable ultrasonic sensing probe 100. The ultrasonic echo signals are converted into echo data by an ultra-low power drive control circuit 200. The echo data is processed by a microcontroller 300 to recognize gestures. At the same time, the ultra-low power drive control circuit 200 outputs drive signals and sends them to the wearable ultrasonic sensing probe 100.

[0035] In this embodiment, the microcontroller 300 serves as the control core of the system. It is preferably a low-power MCU (such as the STM32H7 series or nRF52 series) based on the ARM Cortex-M4 or M33 core, which is responsible for timing control, data acquisition, signal processing, and the operation of the deep learning-based gesture recognition algorithm.

[0036] In this embodiment, the wearable ultrasonic sensing probe 100 uses a PZT-5 (lead zirconate titanate) piezoelectric ceramic sheet as the transducer material. It should be noted that PZT-5 material has a high electromechanical coupling coefficient (d33) and dielectric constant, which allows the wearable ultrasonic sensing probe 100 to generate sufficient mechanical deformation under low-voltage drive, ensuring transmission sensitivity. To further improve transmission efficiency, the backing layer of the wearable ultrasonic sensing probe 100 innovatively uses a PCB substrate directly. This embodiment utilizes the characteristic that the acoustic impedance of FR-4 material (approximately 3 MRayl) is much smaller than that of PZT-5 material (approximately 30 MRayl), forming a high acoustic impedance mismatch interface between the two.

[0037] During operation, based on the principle of sound wave reflection, most of the sound wave energy propagating to the back side is reflected back to the front side, i.e., the human tissue side, thus effectively compensating for the energy loss caused by the reduction in driving voltage. In addition, the surface of the wearable ultrasound sensing probe 100 is also coated with a matching layer. The acoustic impedance of the matching layer is between that of PZT-5 material and human soft tissue (approximately 1.5 MRayl), typically 5-7 MRayl, to further improve the sound wave transmittance.

[0038] refer to Figure 2 In this embodiment, the pulse emission unit in the ultra-low power drive control circuit 200 includes an energy storage capacitor C1, an N-channel enhancement-mode MOSFET Q1, a current-limiting resistor R1, a first protection diode D1, and a second protection diode D2.

[0039] The energy storage capacitor C1 is a low-ESR ceramic capacitor, and its capacitance value is between 10nF and 100nF. One end of the energy storage capacitor C1 is connected to the 5V power supply VCC, and the other end is grounded. The energy storage capacitor C1 is used to provide instantaneous high-current discharge and maintain voltage stability at the moment of transmission.

[0040] The N-channel enhancement-mode MOSFET Q1 is an N-channel MOSFET with low on-resistance (Rds_on < 100mΩ) and low turn-on voltage (Vth < 2V). The drain (D) of the N-channel enhancement-mode MOSFET Q1 is connected to the negative terminal of the wearable ultrasonic sensor probe 100, the source (S) of the N-channel enhancement-mode MOSFET Q1 is grounded, and the gate (G) of the N-channel enhancement-mode MOSFET Q1 is connected to the PWM output pin of the microcontroller 300.

[0041] A current-limiting resistor R1 is connected in series between the power supply VCC and the drain D of the N-channel enhancement-mode MOSFET Q1. The resistance of the current-limiting resistor R1 is typically between 1kΩ and 10kΩ. The current-limiting resistor R1 is used to charge the parasitic capacitance of the wearable ultrasonic sensor probe 100 through the power supply VCC during the off-state of Q1, restoring it to a 5V potential; simultaneously, it limits the short-circuit current flowing from the power supply to ground during the on-state of the N-channel enhancement-mode MOSFET Q1, thus protecting the power supply.

[0042] The pulse transmitting unit operates as follows: In the idle state, the microcontroller 300 outputs a low level, and the N-channel enhancement-type MOS transistor Q1 is turned off. At this time, the 5V power supply VCC charges the wearable ultrasonic sensor probe 100 through the current-limiting resistor R1. The potentials at both the positive and negative terminals of the wearable ultrasonic sensor probe 100 are 5V, the voltage difference is 0V, and it is in a balanced state.

[0043] At the moment of transmission, the microcontroller 300 outputs an extremely short high-level pulse, and the N-channel enhancement-mode MOSFET Q1 is instantly turned on, rapidly pulling the negative potential of the wearable ultrasonic sensor probe 100 down to 0V.

[0044] Since the voltage across the energy storage capacitor C1 cannot change instantaneously, a negative step voltage with an amplitude of -5V is instantaneously generated inside the wearable ultrasonic sensor probe 100. This step voltage excites the piezoelectric ceramic sheet to produce mechanical vibration, emitting broadband ultrasonic waves with a center frequency of approximately 2MHz-5MHz into the human tissue; the first protection diode D1 and the second protection diode D2 are connected in parallel in reverse across the wearable ultrasonic sensor probe 100.

[0045] When the echo signal arrives, if the voltage amplitude of the echo signal exceeds the forward voltage (approximately 0.7V) of the first protection diode D1 and the second protection diode D2, the first protection diode D1 and the second protection diode D2 will be clamped to prevent the high voltage echo from damaging the subsequent circuit.

[0046] In this embodiment, the passive receiving unit in the ultra-low power drive control circuit 200 includes a passive high-pass filter, wherein the passive high-pass filter consists of a series capacitor. and parallel resistors Composition, cutoff frequency of a passive high-pass filter Set it to between 400kHz and 600kHz, for example, take =100pF, =3.3kΩ.

[0047] The reason for setting this cutoff frequency is that the electromyographic (EMG) signals and motion artifacts generated during muscle contraction are mainly concentrated below 500Hz, while the ultrasound echo signal is usually around 4MHz.

[0048] It should be noted that the passive high-pass filter can effectively filter out low-frequency interference and prevent baseline drift from affecting subsequent quantization. The conditioned signal is directly sent to the built-in ADC input pin of the microcontroller 300.

[0049] In this embodiment, reference Figure 3 In practical wearable applications, the coupling state between the wearable ultrasonic sensor probe 100 and the skin is complex and dynamically changing. The dryness or wetness of the skin, the thickness of the stratum corneum, and the contact pressure during wear all alter the acoustic radiation load of the wearable ultrasonic sensor probe 100. This load variation causes a frequency shift in the series resonant frequency of the wearable ultrasonic sensor probe 100, typically manifesting as a shift towards lower frequencies.

[0050] When the system always uses a fixed transmission frequency (or fixed pulse width), the transmission efficiency will drop sharply when the resonant frequency of the wearable ultrasonic sensor probe 100 deviates, causing the already weak echo signal to weaken further and the signal-to-noise ratio to deteriorate.

[0051] See Figure 3 To address this problem, this invention utilizes the spectral characteristics of rectangular pulses. The spectrum of a rectangular pulse is a Sinc function, and its main lobe width is inversely proportional to the pulse width (PW).

[0052] By changing the pulse width, the spectral distribution of the transmitted energy can be effectively altered. Before testing, pulse width calibration is performed. For each test pulse width, the ADC synchronously acquires the echo signal and calculates the energy value (EnergyMetric) of that echo. This energy value is typically represented by the sum of the squares or absolute values ​​of the signal amplitudes. The echo energies under all test pulse widths are compared, and the pulse width corresponding to the maximum energy value is identified and locked as the "Optimal PW" for the current state. During subsequent normal operation, the system is always driven by this optimal pulse width, thereby ensuring that the center of the main lobe of the transmitted spectrum always matches the resonant frequency of the wearable ultrasonic sensing probe 100 under the current load.

[0053] In this embodiment, reference Figure 4 In order to implement complex convolutional neural network (CNN) inference on a microcontroller 300 with limited resources (low clock speed, small memory, and no floating-point unit FPU), this invention adopts an acceleration scheme based on look-up table (LUT).

[0054] During the data preprocessing stage, the microcontroller 300 configures the ADC to direct memory access (DMA) mode to continuously acquire echo data at a sampling rate of 10Msps to 20Msps.

[0055] Based on the speed of sound (approximately 1540 m / s) and the target detection depth (e.g., 2-4 cm subcutaneously), region of interest (ROI) data is extracted within a specific time window.

[0056] To reduce the amount of data, the ROI data was downsampled, ultimately retaining approximately 184 sampling points. Since the subsequent LUT algorithm requires integer input, the floating-point or high-bit-width data was linearly quantized into an 8-bit unsigned integer (uint8, range 0-255).

[0057] Regarding the principle of LUT convolution acceleration, in traditional CNN convolutional layers, the core operation is the multiplication and addition operation (MACs) between the kernel weight and the input data.

[0058] For the low-power microcontroller 300, floating-point multiplication is extremely time-consuming. In this embodiment, a micro convolution kernel with a kernel size of 2 is designed.

[0059] This means that each convolution operation involves only two adjacent input data points. Since the input data has been quantized to 8 bits, the input data pair ( The total number of all possible combinations is only kind.

[0060] In the offline or initialization phase, the present invention pre-calculates the convolution results of the convolution kernel weights and these 65,536 input combinations, and stores the results sequentially in the Flash or SRAM of the microcontroller 300 to form a lookup table of size 64KB to 128KB.

[0061] During the real-time inference phase, the microcontroller 300 does not need to perform any multiplication operations. The microcontroller 300 only needs to read the values ​​of the two sample points in the current window and concatenate them into a 16-bit address index.

[0062] Using this index, the corresponding convolution result can be read directly from the lookup table. This method transforms computationally intensive multiplication operations into extremely low-latency memory access operations, which can theoretically increase inference speed by more than 6 times while significantly reducing computational power consumption.

[0063] The lightweight network consists of 1-2 LUT convolutional layers, 1 adaptive average pooling layer, and 1 fully connected layer. The output layer is processed by the Softmax function to output the probability value of each gesture category.

[0064] In this embodiment, muscle fatigue is the main factor affecting the long-term stability of electromyography / ultrasound gesture recognition. After fatigue, decreased muscle strength leads to weakened echo amplitude, and muscle relaxation causes echo phase delay. To improve the robustness of the model, this embodiment introduces a data augmentation strategy during the training phase, specifically: a) Random time-domain shift, i.e., randomly shifting the training samples by a small time to simulate phase jitter caused by loosening of the garment or muscle deformation; b) Random amplitude scaling: This involves multiplying the training samples by a random coefficient between 0.5 and 1.5 to simulate amplitude changes caused by fatigue. Models trained in this way can learn waveform shape features that are insensitive to absolute amplitude and absolute phase, thus maintaining a high recognition rate even after prolonged wear and fatigue.

[0065] Example 2 This embodiment discloses an integrated wearable ultrasonic gesture recognition system, including an integrated wearable ultrasonic gesture recognition circuit, a peripheral communication module, and an external device. The integrated wearable ultrasonic gesture recognition circuit includes a wearable ultrasonic sensor probe 100, an ultra-low power drive control circuit 200, and a microcontroller 300. The ultra-low power drive control circuit 200 is connected to the wearable ultrasonic sensor probe 100 and the microcontroller 300. The external device is a host computer or a mobile phone. The microcontroller 300 is connected to the external device via the peripheral communication module.

[0066] Example 3 refer to Figure 5 The integrated wearable ultrasonic gesture recognition method of the present invention includes the following steps: 1) Upon system power-on initialization, the microcontroller 300 configures peripherals such as GPIO, ADC, DMA, and timers, and loads pre-trained neural network model parameters and lookup tables from Flash.

[0067] 2) Perform adaptive frequency calibration. The system transmits a scanning pulse sequence and locks the optimal transmission pulse width by comparing the echo energy. It should be noted that step 2) can be performed at the beginning of each wear or can be manually triggered by the user.

[0068] 3) Entering the loop detection mode, the microcontroller 300 outputs a PWM signal with the optimal pulse width at the set frame rate (e.g., 50Hz).

[0069] 4) The pulse emission unit drives the wearable ultrasonic sensor probe 100 to generate ultrasonic waves that enter human muscle tissue.

[0070] 5) The passive receiving unit receives the echo signal, filters out low-frequency noise, and performs DC bias.

[0071] 6) The ADC automatically acquires echo data via DMA and stores it in the buffer. After acquisition is complete, an interrupt is generated to wake up the CPU.

[0072] 7) The CPU performs DC removal, ROI extraction, downsampling, and 8-bit quantization preprocessing on the data.

[0073] 8) Call the LUT acceleration engine, and the CPU traverses the quantized data to complete the convolution feature extraction by looking up a table.

[0074] 9) The feature map is processed by pooling and fully connected layers to output the gesture classification probability.

[0075] 10) Determine whether the maximum probability exceeds a preset threshold; if so, send the corresponding gesture command to the host computer or smart terminal via Bluetooth; otherwise, determine it as an invalid action or a resting state. The gesture command is a fist, an extended palm, and a double-tap.

[0076] Finally, the system enters a low-power sleep state, waiting for the next timer interrupt to trigger the next frame detection.

[0077] According to actual tests, the average operating current of this invention is only at the mA level under 5V power supply, and it can work continuously for dozens of hours. Moreover, the recognition accuracy remains above 95% even after long-term wear, perfectly solving the pain points of traditional wearable ultrasound devices.

[0078] Example 4 A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the integrated wearable ultrasonic gesture recognition method. For example, the method includes: controlling an ultra-low power drive control circuit 200 to output a control signal via a microcontroller 300; emitting ultrasonic waves to deep muscle tissue and receiving ultrasonic echo signals via a wearable ultrasonic sensing probe 100 according to the control signal; converting the ultrasonic echo signals into echo data via the ultra-low power drive control circuit 200; and processing the echo data via the microcontroller 300 to recognize the gesture. The memory may include main memory, such as high-speed random access memory, or it may also include non-volatile memory, such as at least one disk storage device. The processor, network interface, and memory are interconnected via an internal bus, which may be an industry-standard architecture bus, a peripheral component interconnection standard bus, an extended industry-standard architecture bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory stores the program; specifically, the program may include program code, which includes computer operation instructions. Memory can include main memory and non-volatile memory, and provides instructions and data to the processor.

[0079] Example 5 A computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the integrated wearable ultrasound gesture recognition method. For example, the method includes: controlling an ultra-low power drive control circuit 200 to output a control signal via a microcontroller 300; emitting ultrasound waves to deep muscle tissue and receiving ultrasound echo signals via a wearable ultrasound sensing probe 100 according to the control signal; converting the ultrasound echo signals into echo data via the ultra-low power drive control circuit 200; and processing the echo data via the microcontroller 300 to recognize the gesture. Specifically, the computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. The volatile memory may include random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include read-only memory (ROM), hard disk, flash memory, optical disk, magnetic disk, etc.

[0080] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can 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.) containing computer-usable program code.

[0081] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0082] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0083] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0084] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and disclosure of the invention. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0085] It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

[0086] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An integrated wearable ultrasonic gesture recognition circuit, characterized in that, The device includes a wearable ultrasound sensor probe (100), an ultra-low power drive control circuit (200), and a microcontroller (300). The ultra-low power drive control circuit (200) is connected to the wearable ultrasound sensor probe (100) and the microcontroller (300). The microcontroller (300) controls the ultra-low power drive control circuit (200) to output a control signal. The wearable ultrasound sensor probe (100) emits ultrasound waves to deep muscle tissue and receives ultrasound echo signals according to the control signal. The ultra-low power drive control circuit (200) converts the ultrasound echo signals into echo data. The microcontroller (300) processes the echo data to recognize gestures.

2. The integrated wearable ultrasonic gesture recognition circuit according to claim 1, characterized in that, The wearable ultrasonic sensor probe (100) uses PZT-5 piezoelectric ceramic sheet as the transducer material; the backing layer of the wearable ultrasonic sensor probe (100) is made of PCB substrate; and a matching layer is coated on the surface of the wearable ultrasonic sensor probe (100).

3. The integrated wearable ultrasonic gesture recognition circuit according to claim 1, characterized in that, The pulse emission unit in the ultra-low power drive control circuit (200) includes a power supply VCC, an energy storage capacitor C1, an N-channel enhancement-mode MOSFET Q1, and a current-limiting resistor R1. One end of the energy storage capacitor C1 is connected to the power supply VCC, and the other end of the energy storage capacitor C1 is grounded. The drain D of the N-channel enhancement-mode MOSFET Q1 is connected to the negative terminal of the wearable ultrasonic sensor probe (100). The source S of the N-channel enhancement-mode MOSFET Q1 is grounded. The gate G of the N-channel enhancement-mode MOSFET Q1 is connected to the PWM output pin of the microcontroller (300). The current-limiting resistor R1 is connected in series between the power supply VCC and the drain D of the N-channel enhancement-mode MOSFET Q1.

4. The integrated wearable ultrasonic gesture recognition circuit according to claim 3, characterized in that, The pulse emission unit also includes a first protection diode D1 and a second protection diode D2, which are connected in parallel in reverse at both ends of the wearable ultrasonic sensing probe (100).

5. The integrated wearable ultrasonic gesture recognition circuit according to claim 3, characterized in that, The passive receiving unit in the ultra-low power drive control circuit (200) includes a passive high-pass filter, wherein the passive high-pass filter consists of a series capacitor. and parallel resistors composition.

6. The integrated wearable ultrasonic gesture recognition circuit according to claim 3, characterized in that, When the N-channel enhancement-mode MOS transistor Q1 is turned on, the negative potential of the wearable ultrasonic sensor probe (100) is pulled down to ground, and a negative step voltage is generated by utilizing the capacitance characteristics of the wearable ultrasonic sensor probe (100) itself.

7. The integrated wearable ultrasonic gesture recognition circuit according to claim 3, characterized in that, During the initialization phase, the pulse transmitting unit adjusts the DC operating point of the circuit through a sliding rheostat and outputs pulses of different widths in sequence to detect the echo intensity. The pulse width that produces the maximum echo intensity is locked as the working pulse width to match the probe resonant frequency under the current skin load.

8. The integrated wearable ultrasonic gesture recognition circuit according to claim 1, characterized in that, The microcontroller (300) runs a lightweight convolutional neural network accelerated by lookup tables to classify echo data to recognize gestures.

9. The integrated wearable ultrasonic gesture recognition circuit according to claim 8, characterized in that, During the training of the lightweight convolutional neural network, random temporal offset and amplitude scaling transformations are applied to the original training data to obtain the final training data, which is then used to train the lightweight convolutional neural network.

10. An integrated wearable ultrasonic gesture recognition system, characterized in that, It includes a peripheral communication module, an external device, and an integrated wearable ultrasonic gesture recognition circuit as described in any one of claims 1-9, wherein the microcontroller (300) is connected to the external device via the peripheral communication module.

Images