Sound system with DSP sound effect enhancement processing

Abstract

The invention discloses a sound system with DSP sound effect enhancement processing, which belongs to the technical field of audio processing and comprises an audio input module, a data acquisition module, a DSP processing unit, an audio output module and a user interaction module. By acquiring characteristic data (such as audiogram and heart rate) of a hearing impaired person or a child in real time, a DSP processing unit dynamically calls a targeted algorithm (weak hearing frequency band compensation and voice rhythm adjustment), and a multi-frequency band equalization and noise suppression technology is combined, so that an optimized audio signal adaptive to user requirements is output. The problem that a traditional sound system cannot adapt to special crowds in a personalized mode is solved, the sound effect definition and user experience are remarkably improved, and the system is suitable for education, medical treatment and family scenes.

Description

Technical Field

[0001] The present invention belongs to the technical field of audio processing, and particularly relates to an audio system with DSP sound effect enhancement processing. Background Art

[0002] With the aggravation of population aging and the growth of the digital demand for children's education, audio processing technologies for specific populations are becoming increasingly important in scenarios such as medical rehabilitation, home entertainment, and educational assistance. For example, hearing-impaired people need to perform gain compensation on specific frequency bands according to their hearing loss characteristics and suppress environmental noise to improve speech intelligibility; when children are studying or entertaining, it is necessary to guide their attention and improve the information reception effect by enhancing the speech signal and adjusting the speech rhythm. However, the general processing mode of traditional audio systems is difficult to meet the above-mentioned subdivision requirements, and there is an urgent need for an intelligent sound effect enhancement solution driven by characteristic data. Summary of the Invention

[0003] In view of the above pain points, the present invention provides an intelligent audio system based on DSP. By real-time acquiring the characteristic data of hearing-impaired people or children and dynamically invoking the matching sound effect enhancement algorithm, personalized audio processing is realized, and the auditory experience and information reception efficiency of special populations are significantly improved.

[0004] The solution of the present invention is as follows: An audio system with DSP sound effect enhancement processing, characterized by comprising an audio input module, a data acquisition module, a DSP processing unit, an audio output module, and a user interaction module; The audio input module is used for receiving audio signals; The data acquisition module is used for acquiring the relevant characteristic data of the special population, and the special population includes hearing-impaired people and children; the special population includes hearing-impaired people and children; The DSP processing unit is connected to the data acquisition module to acquire the relevant characteristic data of the special population, and performs targeted sound effect enhancement processing on the audio signal according to the relevant characteristic data; The audio output module is used for outputting the audio signal processed by the DSP processing unit; The user interaction module is used for the user to input the relevant characteristic data of the special population and set and adjust the sound effect enhancement processing.

[0005] Preferably, the audio input module includes a wired audio input interface and a wireless audio input interface, and can receive audio signals from different sound sources.

[0006] Preferably, the DSP processing unit stores a sound effect enhancement algorithm library for different special population characteristic data. The algorithm library includes a hearing compensation algorithm set for hearing-impaired people and a children's speech enhancement algorithm set for children. The DSP processing unit calls the corresponding sound effect enhancement algorithm for processing according to the obtained relevant characteristic data, as follows: When the obtained characteristic data is hearing detection data, the algorithm in the hearing compensation algorithm set is called; When the obtained characteristic data is the physiological data of children, the algorithm in the children's speech enhancement algorithm set is called.

[0007] Preferably, the audio output module includes a speaker, and the speaker is connected to the DSP processing unit for playing the processed audio signal.

[0008] Preferably, the data acquisition module includes an interface for data interaction with an external device, and the relevant characteristic data stored in the external device is obtained through the interface.

[0009] Preferably, the special population is hearing-impaired people, the relevant characteristic data is hearing detection data, and the weak hearing frequency band compensation algorithm in the hearing compensation algorithm set includes the following steps: (a) Characteristic data analysis: Analyze the obtained audiogram data to determine the weak hearing frequency band range of the hearing-impaired people and the attenuation value of each frequency band , where is the frequency; (b) Frequency band gain calculation: According to the obtained attenuation value , calculate the gain value of each weak hearing frequency band , and the calculation formula is , where is the gain adjustment coefficient, and the value range is , which can be adjusted according to the actual situation; (c) Multi-band equalization processing: Use a multi-band equalizer to adjust the gain of the audio signal in the weak hearing frequency band according to the calculated gain value , and at the same time perform smoothing processing on adjacent frequency bands to avoid frequency distortion; (d) Dynamic noise reduction processing: Synchronously enable the environmental noise suppression algorithm to perform narrow-band filtering on common noises in the weak hearing frequency band. This algorithm calculates the center frequency and bandwidth of the noise, and sets the cut-off frequency and of the filter to suppress the noise frequency band.

[0010] Preferably, the special population is children, and the relevant characteristic data is the physiological data of children. Currently, the physiological data only includes the heart rate data of children. The speech enhancement and rhythm adjustment comprehensive algorithm in the children's speech enhancement algorithm set includes the following steps: (a) Heart rate data collection and analysis: Collect the heart rate data of children in real time , and analyze it according to the preset heart rate threshold to judge the attention state of children; The preset heart rate threshold is divided into a low threshold and a high threshold . When , it is determined that the attention is concentrated; When , it is determined that the attention is dispersed; When , it is determined to be in a normal state; (b) Speech frequency band extraction and enhancement: Perform spectrum analysis on the input audio signal to extract the core speech frequency band , generally ; (c) Adjust the gain of the core speech frequency band according to the attention state of children : When it is determined that the attention is concentrated, , where is the default gain value, and the value range is ; When it is determined that the attention is dispersed, , where is the gain adjustment factor, and the value range is , to further enhance the speech signal to attract the attention of children; When it is determined to be in a normal state, ; (d) Background noise suppression: Use the band-pass filtering algorithm to suppress the low-frequency background music (frequency ) and high-frequency noise (frequency ); The transfer function of the band-pass filter is ; (e) Speech rhythm adjustment: Adjust the rhythm parameters of the speech according to the attention state of children: When it is determined that the attention is dispersed, reduce the speech speed by , The value range of , and add a pause time seconds at the end of the sentence, The value range of ; When it is determined that the attention is concentrated, the speech speed can be appropriately increased, and the increase ratio is , The value range of ; When it is determined to be in the normal state, the current speech rate and pause time are maintained.

[0011] Preferably, in the comprehensive algorithm for speech enhancement and rhythm adjustment, when performing speech rhythm adjustment, the specific calculation methods for speech rate adjustment and pause time addition are as follows: (a) When it is determined that the attention is distracted, the new speech rate , where is the original speech rate; the new end-of-sentence pause time , where is the original end-of-sentence pause time; (b) When it is determined that the attention is concentrated, the new speech rate .

[0012] Preferably, the user interaction module is also integrated with an automatic detection unit, which can automatically identify that the current service object is a hearing-impaired person or a child based on the special population-related feature data including hearing detection data, child physiological data, and identity identification information input by the user obtained by the data acquisition module, and trigger the DSP processing unit to call the corresponding sound effect enhancement algorithm; when it is detected that the service object has switched, the user interaction module synchronously updates the display interface and operation logic, simplifies the setting options for child users, and highlights the frequency band adjustment entry for hearing-impaired users.

[0013] Compared with the prior art, the advantages of the present invention are as follows: (1) Precise sound effect processing to enhance the auditory experience of special populations: For hearing-impaired people, based on the weak hearing frequency band compensation algorithm analyzed from the audiogram, dynamic multi-band equalization gain is achieved (gain adjustment coefficient k ∈ [0.8, 1.2]). Combining narrowband noise suppression technology, while compensating for hearing attenuation, environmental noise interference is reduced, and the clarity of the speech signal is increased by 30% - 50%; for children, through the speech enhancement and rhythm adjustment algorithm driven by heart rate data, when the attention is distracted, the gain of the core speech frequency band is increased by 20% - 50%, and the speech rate is reduced by 10% - 20%, significantly enhancing the speech intelligibility and attractiveness, which is applicable to the educational content playback scenario; (2) Intelligent interaction design to reduce the usage threshold: The automatic detection unit realizes the non-intrusive identification of the service object based on multi-source data (hearing detection data, heart rate data, identity identification), without the need for the user to manually select the mode; the interface dynamic adaptation function simplifies the setting options for children (such as hiding complex frequency band adjustments), highlights the key adjustment entry for hearing-impaired people (such as the weak hearing frequency band gain slider), and the operation efficiency is increased by more than 60%; (3) Algorithm innovation improves processing efficiency and stability: The adjacent frequency band smoothing process in the hearing-impaired frequency band compensation algorithm avoids frequency distortion. The dynamic noise reduction algorithm performs narrowband filtering on the noise center frequency and bandwidth, and the noise suppression efficiency is ≥ 80%. The band-pass filtering of the children's voice enhancement algorithm accurately retains the voice frequency band of 200 Hz - 5 kHz, while suppressing the low-frequency background music, and the signal-to-noise ratio is increased by more than 40%. Description of the Drawings

[0014] Figure 1 It is a schematic structural diagram of an audio system with DSP sound effect enhancement processing. Specific Embodiments

[0015] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0016] Embodiment 1: The special population is children In this embodiment, children aged 6 - 12 are the service objects. Through the full process details of the hardware architecture, data flow, algorithm execution, and interface adaptation, it shows how the system realizes the closed-loop processing of "heart rate perception - voice enhancement - rhythm adjustment - intelligent interaction".

[0017] I. Realization of the system hardware architecture and data input layer 1. Audio input module: (a) Hardware configuration: Integrate a Bluetooth 5.2 wireless module (supporting AAC / SBC decoding) and a 3.5 mm wired input interface, which can be connected to devices such as mobile phones, tablets, and learning machines, and is compatible with digital audio signals with a sampling rate of 48 kHz; built-in MEMS microphone array (3 microphones, with a spacing of 5 cm), supporting beamforming technology, and real-time collecting voice signals in the environment to suppress non-directional noise; (b) Data interaction: Wireless transmission, receiving children's stories and English enlightenment audio through the Bluetooth protocol, with a transmission rate of 1 Mbps and a delay of ≤ 50 ms; Wired input: When connecting a microphone, use an ADC analog-to-digital conversion chip (PCM3060, signal-to-noise ratio 98 dB) to convert the analog signal into 16-bit digital audio.

[0018] 2. Data acquisition module: (a) Collection of children's physiological data: Connect to a children's smart bracelet (such as the Huawei Children's Watch 4X) via BLE low-power Bluetooth to obtain real-time heart rate data (HR) with a sampling frequency of 1 Hz. The data format is JSON (including timestamp, heart rate value, device ID). The bracelet is equipped with a PPG photoplethysmogram sensor, which supports motion state recognition (stationary / active) to assist in judging the validity of heart rate data (such as excluding abnormal heart rates during exercise). (b) Input of identity identification information: When the user uses it for the first time, input the basic information of the child through an 8-inch capacitive touch screen: name, age (automatically associated with the heart rate threshold range, such as for children aged 6 - 8, HR_low = 75 beats per minute; for children aged 8 - 12, HR_low = 70 beats per minute), and the model of the worn device (synchronously matching the bracelet communication protocol). Support for the face recognition quick-switching mode: The front 2 million-pixel camera collects facial data, and the local AI model (lightweight MobileNet) is used to identify the child's identity, and the historical setting parameters are automatically retrieved.

[0019] 3. User interaction module: Core components: The main control chip uses STM32H743VI (ARM Cortex-M7, main frequency 480 MHz), integrated with an automatic detection unit (based on the FreeRTOS real-time operating system) to parse heart rate data and identity information in real time. The touch screen supports 10-point touch with a resolution of 1280×800, and the display interface is developed using the QT framework, supporting dynamic element loading (such as switching button icons and color schemes according to the service object).

[0020] II. Automatic detection unit: Dual verification mechanism for child users 1. Multi-dimensional data fusion recognition logic The automatic detection unit determines that the current service object is a child through the following three-layer verification: (a) Device association verification: The data acquisition module detects that the connected external devices include children's exclusive devices (such as smart bracelets with device ID prefix "KID_", learning machines with model "LEARNING_PAD"). Example: When the bracelet device ID is "KID_20230815" and the age input is "8 years old", the child mode is pre-activated. (b) Physiological data feature matching: Heart rate data preprocessing: Remove noise through 5-point moving average filtering. When the HR value is collected continuously 3 times and > HR_high (120 beats per minute, dynamically adjusted based on age), it is determined as a valid attention dispersion signal. Motion state calibration: If the acceleration data synchronously transmitted by the bracelet shows that the child is in a stationary state (all three-axis accelerations < 0.2g), the heart rate data is considered valid to avoid misjudgment during exercise. (c) Identity verification: The user manually selects the "Children's Mode" or matches a registered children's account through face recognition (the account information includes age, hearing normal / abnormal mark, etc.).

[0021] 2. Algorithm Trigger and Module Communication When the above three-layer verification passes, the automatic detection unit sends a control instruction (16-bit binary data, the high 4 bits are the mode identifier "0010" indicating the children's mode, and the low 12 bits are the user ID) to the DSP processing unit through the SPI bus; After receiving the instruction, the DSP processing unit switches the algorithm call pointer through the internal register and loads the corresponding program from the address space of the "Children's Voice Enhancement Algorithm Set" (stored in a 256KB Flash memory).

[0022] III. DSP Processing Unit: Full-process Algorithm Implementation (a) Attention State Analysis Driven by Heart Rate Data Threshold Dynamic Calibration (Based on Children's Age): Real-time Data Processing Example: The real-time heart rate of a 10-year-old child is collected as HR = 130 beats per minute (exceeding HR_high = 120 for 5 consecutive times), and the acceleration data shows static (x = 0.1g, y = 0.05g, z = 0.12g), which is determined to be in a distracted state.

[0023] (b) Precise Extraction of the Core Voice Frequency Band: Using spectrum analysis technology, perform a 1024-point FFT transformation on the input audio signal, calculate the power spectral density (PSD), and extract the frequency band with an energy ratio > 80% as the core voice frequency band, which is default set to [200Hz, 5kHz] (covering the fundamental frequency and main overtones of children's voices); for dialect audio (such as Cantonese), adaptively adjust the lower frequency limit of the frequency band to 150Hz through the dynamic time warping (DTW) algorithm to ensure the retention of low-frequency components such as the vowels "a" and "o".

[0024] (c) Gain Adjustment Associated with Attention State: Gain calculation model, default gain (determined through subjective listening tests, the best clarity of children's voices is within the range of 3 - 6dB); when distracted, the gain adjustment factor (take the middle value of 0.4 to balance the enhancement effect and auditory comfort), actual gain: ; Use an IIR band-pass filter (3rd order, passband ripple ≤ 0.3dB) to enhance the gain of the target frequency band, and the filter coefficients are designed through Matlab FDAtool to ensure linear phase.

[0025] (d) Multi-band noise suppression strategy: implemented by bandpass filter, the transfer function strictly follows the definition of claim 7(d), and segmented processing is adopted: low frequency suppression, The frequency band (such as the bass drum beats in the background music) is applied with a Butterworth low-pass filter (4th order, cut-off frequency 180Hz), with a stop-band attenuation of 25dB; high-frequency suppression, Butterworth high-pass filter (4th order, cut-off frequency 8.2kHz) is applied to frequency bands (such as air conditioning noise and high-frequency current noise) with a stop-band attenuation of 22dB; real-time noise monitoring: through the energy detection algorithm, when the energy of the noise frequency band exceeds that of the voice band by 30%, the suppression depth is automatically increased to 30dB.

[0026] (e) Dynamic adjustment algorithm of speech rhythm: Mathematical model of speech rate adjustment: Distraction scenario: Original speed , reduce the proportion (Take the middle value between the upper limit of 20% and the lower limit of 10%). New speed: ; Add pause time at the end of sentence (Based on children's language acquisition research, 0.3-0.8 second pauses can improve speech segmentation perception), original pause , new pause: ; The TD-PSOLA time-domain pitch synchronization superposition algorithm is adopted to detect the pitch period in the speech signal (the pitch frequency of children's speech is about 200-500Hz), insert a silent frame (length t seconds) at the end of the sentence, and time-stretch the speech segment to keep the pitch unchanged and avoid "voice change" distortion.

[0027] 4. User Interaction Module: Dynamic Adaptation of Scenario-based Interface 1. Exclusive interface for children's mode: remove professional options such as "Advanced Equalizer" and "Noise Reduction Parameters", and only retain 3 core function buttons: Story mode: The preset speech speed is reduced by 15% and the gain is 6dB, which is suitable for reading long texts; Children's song mode: the preset speech speed is increased by 10% (when the attention is focused), the gain is 5dB, and the interaction is enhanced with the music with a brisk rhythm; Eye protection mode: The screen brightness automatically adjusts according to the ambient light (20-80%) to reduce visual fatigue.

[0028] Real-time status visualization: heart rate waveform is displayed in the upper left corner (timeline 5 minutes, green curve indicates normal range, red highlights indicate distraction period); speech enhancement status area: dynamic icons display the current processing strategy (such as "speech speed↓18%" and "gain↑7dB" text labels, accompanied by rotating sound wave animation).

[0029] 2. Operational Logic Design: Interaction Process Optimization: Long - press the volume + / - keys to quickly switch between preset modes (story / nursery rhyme) to avoid complex operations for children; when entering the identity identifier, use graphical selection (such as dragging a cartoon avatar to the "child" icon) instead of traditional text input. Feedback Mechanism: When attention dispersion is detected, the touch screen vibrates for feedback (a slight vibration for 50 ms), and there is a voice prompt. During the algorithm adjustment process, the speaker plays a 0.5 - second prompt tone (a 440 - Hz pure tone) to inform the user that the current state is enhanced.

[0030] Embodiment 2: Automatic Identification of Service Objects and Algorithm Trigger This embodiment focuses on the technical feature of the "automatic detection unit" and details how it realizes the non - intrusive identification of service objects, precise algorithm invocation, and dynamic adaptation of the interaction interface based on hearing detection data, children's physiological data, and identity identifier information.

[0031] I. Hardware Architecture and Data Input of the Automatic Detection Unit 1. Configuration of the Data Acquisition Layer: Multimodal Data Interface (a) Hearing Detection Data: Connect a pure - tone audiometer (such as Interacoustics AZ20) through a USB 2.0 interface, supporting the reading of audiogram data that complies with the ANSI S3.6 standard (.csv format, including air - conduction thresholds at each frequency point from 500 Hz to 8 kHz). (b) Children's Physiological Data: Connect a children's smartwatch (such as Xiaotiancai Z6) through a BLE 4.2 module to receive real - time heart rate data (HR, unit: beats per minute) and device type identifier ("KID_WATCH"). (c) Identity Identifier Information: The user inputs account information through a 10 - inch capacitive touch screen, including "user type" (child / hearing - impaired person / ordinary user), age (used to verify the validity of children's physiological data), and the associated ID of the historical audiogram file.

[0032] 2. Core Processing Module: The hardware of the automatic detection unit uses an NXP i.MX 6UL processor (ARM Cortex - A7, main frequency 800 MHz), integrated with a dedicated data fusion chip (Maxim MAX20305), supporting parallel processing of three - way data (audiogram parsing, heart rate analysis, identity verification); it has an 8 - MB DDR3 cache to store the pre - trained recognition model (based on Support Vector Machine - SVM, with training data including 1000 groups of hearing - impaired person and 500 groups of children's characteristic data).

[0033] II. Service Object Identification Logic and Data Fusion Algorithm 1. Three - layer data verification mechanism (a) Initial screening of identity identification: When the "user type" field in the user input or historical account is "child" or "hearing - impaired", the corresponding data verification is triggered: Child candidate: age ≤ 12 years old and the device type contains the prefix "KID_" (e.g., the first 6 digits of the bracelet MAC address are "00:1A:7D", and the manufacturer defines it as a child device); Hearing - impaired candidate: age ≥ 18 years old and there is a valid audiogram file (file creation time ≤ 365 days to avoid expired data). (b) Physiological / feature data matching Child mode verification: The heart rate data HR obtained from the smartwatch needs to meet: Data validity: The continuous 5 - time sampling values are between 50 - 150 beats per minute (normal heart rate range for children), and the device model is bound to the child account; Motion state calibration: It is judged that the child is in a stationary state (the tri - axis acceleration is all < 0.3g) through the watch acceleration sensor to exclude abnormal heart rate interference during exercise. Hearing - impaired mode verification: Analyze the audiogram data, extract the weak - hearing frequency band range [fx, fy] and the attenuation value A(f), and it needs to meet: At least one frequency band attenuation value ≥ 15dB (WHO mild hearing loss standard); Data integrity: It contains the thresholds of five standard test frequency points of 500Hz, 1kHz, 2kHz, 4kHz, and 8kHz. (c) Conflict resolution and priority determination: When two types of data are input simultaneously (such as a child wearing a hearing aid), it is processed according to the following rules: It is preferentially identified as "child" (the child mode contains a more strict attention - guiding strategy, and the physiological data has higher real - time performance); The interface prompts "Multiple feature data detected, currently running in child mode. Please switch manually if hearing compensation is required."

[0034] 2. Implementation of intelligent recognition algorithm (a) Feature vector construction: Child mode: [age, heart rate value, device type identifier, motion state]; Hearing - impaired mode: [attenuation values of each frequency band, audiogram update time, hearing aid wearing status (yes / no)].[[]] (b) Classification model inference: Calculate the posterior probability through the SVM model. When the probability of the child mode > 0.8 or the probability of the hearing - impaired mode > 0.7, it is determined as an effective recognition; Example: Input the child's age of 8 years old, heart rate of 110 beats per minute (normal state), and device type "KID_WATCH", and the model outputs the probability of the child mode of 0.92, triggering the corresponding algorithm call.

[0035] III. Trigger mechanism and instruction interaction of the DSP processing unit 1. Algorithm call instruction format The automatic detection unit sends 16-bit control instructions to the DSP processing unit via the SPI bus. The data format is as follows: [4-bit mode identifier][12-bit user ID / feature data check code] Mode identifier: 0010 = child mode, 0100 = hearing-impaired mode, 0000 = normal mode Example: When the child mode is detected, the instruction "0010_000000000123" is sent (the first 4 bits are for the child mode, and the last 12 bits are for the child account ID); When the hearing-impaired mode is detected, the instruction "0100_000000004567" is sent (the first 4 bits are for the hearing-impaired mode, and the last 12 bits are for the audiogram file check code).

[0036] 2. DSP Response and Algorithm Loading (a) Algorithm library storage structure: The internal Flash of the DSP is divided into three areas: normal mode algorithm (default equalizer, 512KB); child voice enhancement algorithm set (1MB, including the sub-algorithms of claims 7-8); hearing-impaired compensation algorithm set (1.5MB, including the sub-algorithms of claim 6); (b) Dynamic loading mechanism: After receiving the mode instruction, the DSP switches the algorithm call pointer through the address mapping table: Child mode: Jump to the address 0x0800_4000 (child algorithm entry); Hearing-impaired mode: Jump to the address 0x0801_0000 (hearing compensation algorithm entry); The loading process takes ≤10ms. During this period, the audio processing switches to the temporary direct-through mode (no sound effect enhancement) to ensure seamless switching.

[0037] IV. Dynamic Adaptation of the User Interaction Module 1. Adaptive Adjustment of the Operation Logic (a) Child mode: The touch operation response time is shortened to 50ms (children move faster), and the tolerance for accidental touches is increased (the click area is expanded by 5mm outside the icon); The complex setting entry is disabled, and long-pressing the power button directly switches to the parental control mode (password required); (b) Hearing-impaired mode: Supports directly entering the frequency band frequency using the remote control numeric keys (e.g., pressing "1" corresponds to the 1kHz frequency band), and the adjustment step value can be set to 0.5dB (fine adjustment) or 2dB (quick adjustment); When the audiogram data update is detected, a prompt "The latest hearing compensation scheme has been loaded. Do you want to listen?" will be automatically popped up.

[0038] V. Abnormal Scenario Handling (a) Data missing handling: When only the identity identifier is detected (e.g., manually select the "Children's Mode" but the smart bracelet is not connected): Automatically enter the "Basic Children's Mode", adopt the default heart rate thresholds (HR_low = 70, HR_high = 120), and prompt "It is recommended to connect the smart watch for a more accurate experience". (b) Identification error correction mechanism: When there are 3 consecutive misjudgments (e.g., misidentifying an adult as a hearing-impaired person): Trigger the manual verification process, display "Please confirm the current user type" on the touch screen, and collect the auditory feedback data of 3 frequency bands (play sounds of different frequencies, and the user presses the button to confirm whether they can hear clearly), and correct the parameters of the identification model.

[0039] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification are only preferred examples of the present invention and do not limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. An audio system with DSP sound effect enhancement processing, characterized in that, It includes an audio input module, a data acquisition module, a DSP processing unit, an audio output module, and a user interaction module; The audio input module is used to receive audio signals; The data acquisition module is used to acquire relevant characteristic data of the special population, where the special population includes hearing-impaired people and children; The special population includes hearing-impaired people and children; The DSP processing unit is connected to the data acquisition module to acquire relevant characteristic data of the special population, and perform targeted sound effect enhancement processing on the audio signal according to the relevant characteristic data; The audio output module is used to output the audio signal processed by the DSP processing unit; The user interaction module is used for the user to input relevant characteristic data of the special population and set and adjust the sound effect enhancement processing.

2. The audio system with DSP audio enhancement processing according to claim 1, characterized in that, The audio input module includes a wired audio input interface and a wireless audio input interface, and can receive audio signals from different sound sources.

3. A sound system with DSP sound effect enhancement processing according to claim 1, characterized in that, The DSP processing unit stores a sound effect enhancement algorithm library for different special population characteristic data. The algorithm library includes a hearing compensation algorithm set for hearing-impaired people and a children's voice enhancement algorithm set for children; The DSP processing unit calls the corresponding sound effect enhancement algorithm for processing according to the acquired relevant characteristic data, specifically as follows: When the acquired characteristic data is hearing detection data, call the algorithm in the hearing compensation algorithm set; When the acquired characteristic data is the physiological data of children, call the algorithm in the children's voice enhancement algorithm set.

4. An audio system with DSP sound effect enhancement processing according to claim 1, characterized in that, The audio output module includes a speaker, and the speaker is connected to the DSP processing unit for playing the processed audio signal.

5. A sound system with DSP sound effect enhancement processing according to claim 1, characterized in that, The data acquisition module includes an interface for data interaction with external devices, and acquires relevant characteristic data stored in the external devices through the interface.

6. The audio system with DSP audio enhancement processing according to claim 3, wherein The special population is hearing-impaired people, the relevant characteristic data is hearing detection data, and the weak hearing frequency band compensation algorithm in the hearing compensation algorithm set includes the following steps: (a) Feature data analysis: Analyze the obtained audiogram data to determine the range of weak hearing frequency bands of the hearing-impaired population and the attenuation value of each frequency band , where is the frequency; (b) Band gain calculation: Based on the attenuation values obtained by analysis , calculate the gain values for each hearing-impaired frequency band . The calculation formula is , where is the gain adjustment coefficient, and its value range is , which can be adjusted according to the actual situation; (c) Multi - band equalization processing: Using a multi - band equalizer to perform gain adjustment on the audio signal in the hearing - impaired frequency band according to the calculated gain value and simultaneously perform smoothing processing on adjacent frequency bands to avoid frequency distortion; (d) Dynamic noise reduction processing: Simultaneously enable the environmental noise suppression algorithm to perform narrowband filtering on common noises in the hearing-impaired frequency band. This algorithm calculates the center frequency and bandwidth , sets the cut-off frequencies and of the filter, and suppresses the noise frequency band.

7. An audio system with DSP sound effect enhancement processing according to claim 3, characterized in that, The special population is children, the relevant characteristic data is the physiological data of children, where the physiological data currently only includes the heart rate data of children, and the voice enhancement and rhythm adjustment comprehensive algorithm in the children's voice enhancement algorithm set includes the following steps: (a) Heart rate data collection and analysis: Real-time collection of children's heart rate data , and analyze according to the preset heart rate threshold to judge the attention state of children; The preset heart rate thresholds are divided into a low threshold and a high threshold . When , it is determined that the attention is concentrated; When , it is determined that the attention is distracted; When , it is determined to be in a normal state; (b) Voice frequency band extraction and enhancement: Perform spectral analysis on the input audio signal to extract the core voice frequency band , generally ; (c) Adjust the gain of the voice core frequency band according to the attention state of the child : When it is determined that the attention is concentrated, , where is the default gain value, and the value range is ; When it is determined that the attention is distracted, , where is a gain adjustment factor, and its value range is , to further enhance the voice signal to attract the child's attention; When it is determined to be in a normal state, ; (d) Background noise suppression: Use the band-pass filtering algorithm to suppress low-frequency background music (frequency ) and high-frequency noise (frequency ); The transfer function of the band-pass filter is ； (e) Voice rhythm adjustment: Adjust the rhythm parameters of the voice according to the attention state of children: When it is determined that the attention is distracted, the speech rate is reduced , The value range of is seconds, The value range of ; When it is determined that the attention is concentrated, the speech rate can be appropriately increased, and the increase ratio is , The value range of ; When it is determined to be in a normal state, keep the current speech rate and pause time.

8. An audio system with DSP sound effect enhancement processing according to claim 7, characterized in that, In the voice enhancement and rhythm adjustment comprehensive algorithm, when performing voice rhythm adjustment, the specific calculation methods for speech rate adjustment and pause time addition are as follows: (a) When it is determined that the attention is distracted, the new speech rate , where is the original speech rate; the new end-of-sentence pause time , where is the original end-of-sentence pause time; (b) When it is determined that the attention is concentrated, the new speech rate .

9. The audio system with DSP audio enhancement processing according to claim 1, characterized in that The user interaction module is also integrated with an automatic detection unit, which can automatically identify the current service object as a hearing-impaired person or a child based on the relevant characteristic data of the special population acquired by the data acquisition module, including hearing detection data, children's physiological data, and the identity identification information input by the user, and trigger the DSP processing unit to call the corresponding sound effect enhancement algorithm; When it is detected that the service object switches, the user interaction module synchronously updates the display interface and operation logic, simplifies the setting options for child users, and highlights the frequency band adjustment entry for hearing-impaired users.

Images