Recording shielding method and device based on ultrasonic interference

Abstract

The invention discloses a sound recording shielding method and device based on ultrasonic interference, and the method comprises the steps: detecting and recognizing sound recording equipment in a target region, and obtaining the type information of the sound recording equipment; configuring a first ultrasonic interference signal and a second ultrasonic interference signal based on the type information; and performing recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal. The technical problem of poor recording interference effect caused by lack of accurate shielding methods for different types of recording devices in the prior art is solved, and the technical effects of customizing the ultrasonic interference signal according to the type of the recording device and realizing efficient and accurate recording content shielding are achieved.

Description

Technical Field

[0001] The present invention relates to the field of recording shielding, and particularly to a recording shielding method and device based on ultrasonic interference. Background Art

[0002] With the continuous development of information technology, recording devices have been widely used in various fields. However, in some occasions that require confidentiality, such as business negotiations and confidential meetings, unauthorized recording behaviors may lead to information leakage, bringing serious security risks and economic losses. Therefore, recording shielding is required in these occasions. Currently, common recording shielding technologies in the market include white noise interference method, sound wave cancellation method, electromagnetic interference method, etc. However, most of these methods adopt a unified interference strategy and cannot perform targeted shielding according to the characteristics of different types of recording devices. Moreover, due to significant differences in the sound pickup principle, frequency response range, and signal processing mechanism among various recording devices, general interference methods often have difficulty effectively shielding all recording devices. For example, the interference signal for traditional analog recording devices may have limited effect on digital recording devices; the interference method optimized for a specific frequency range may be insufficient for wide-band recording devices. Therefore, the prior art has the defect of lacking a precise shielding method for different types of recording devices, resulting in poor recording interference effect. Summary of the Invention

[0003] The present application provides a recording shielding method and device based on ultrasonic interference, aiming to solve the technical problem in the prior art that there is a lack of a precise shielding method for different types of recording devices, resulting in poor recording interference effect.

[0004] In the first aspect disclosed by the present application, a recording shielding method based on ultrasonic interference is provided. The method includes: detecting and identifying the recording device in the target area and obtaining the type information of the recording device; based on the type information, configuring a first ultrasonic interference signal and a second ultrasonic interference signal, wherein the first ultrasonic interference signal is inaudible to the human ear but can be captured by the recording device, and the second ultrasonic interference signal is a signal used to generate a non-linear mixing effect inside the recording device; performing recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal.

[0005] Optionally, the recording device that detects and identifies the target area obtains the type information of the recording device, including: scanning the electromagnetic signals in the target area through a radio frequency scanner; performing spectrum analysis on the obtained electromagnetic signals to extract multiple signal features, where the multiple signal features include frequency features, intensity features, and modulation features; according to the frequency features, the intensity features, and the modulation features, performing device matching in a preset recording device feature database, and determining the type information of the recording device according to the matching result.

[0006] Optionally, the performing device matching in a preset recording device feature database according to the frequency features, the intensity features, and the modulation features, and determining the type information of the recording device according to the matching result includes: obtaining a device feature matching formula; determining a first sample device in the preset recording device feature database, and extracting the first frequency feature, the first intensity feature, and the first modulation feature of the first sample device; based on the device feature matching formula, according to the frequency feature, the intensity feature, and the modulation feature of the recording device, and the first frequency feature, the first intensity feature, and the first modulation feature of the first sample device, obtaining the first device matching degree between the recording device and the first sample device; according to the method of obtaining the first device matching degree, continuing to obtain the device matching degrees between other sample devices in the preset recording device feature database and the recording device, and obtaining multiple device matching degrees; selecting the maximum device matching degree among the multiple device matching degrees as the matching sample device; in the preset recording device feature database, extracting the type information of the matching sample device as the type information of the recording device.

[0007] Optionally, the device feature matching formula is: Wherein, represents the device matching degree, represents the frequency feature, represents the first frequency feature, represents the similarity between the frequency feature and the first frequency feature, represents the intensity feature, represents the first intensity feature, represents the similarity between the intensity feature and the first intensity feature, represents the modulation feature, represents the first modulation feature, represents the similarity between the modulation feature and the first modulation feature, is the weight coefficient, and .

[0008] Optionally, configuring the first ultrasonic interference signal and the second ultrasonic interference signal based on the type information includes: constructing an interference signal configuration model, where the interference signal configuration model includes a first signal configuration channel and a second signal configuration channel; inputting the type information into the first signal configuration channel to obtain the first ultrasonic interference signal; and inputting the type information into the second signal configuration channel to obtain the second ultrasonic interference signal.

[0009] Optionally, constructing an interference signal configuration model, where the interference signal configuration model includes a first signal configuration channel and a second signal configuration channel, includes: obtaining test interference data of multiple sample recording devices, where the test interference data includes recording interference data and device interference data; establishing a first correspondence library between recording device types and first ultrasonic interference parameters based on the recording interference data of multiple sample recording devices; establishing a second correspondence library between recording device types and second ultrasonic interference parameters based on the device interference data of multiple sample recording devices; training a first deep learning network as the first signal configuration channel according to the first correspondence library; training a second deep learning network as the second signal configuration channel according to the second correspondence library; and combining the first deep learning network and the second deep learning network to form the interference signal configuration model.

[0010] Optionally, performing recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal includes: configuring a first interference duration interval for the first ultrasonic interference signal and a second interference duration interval for the second ultrasonic interference signal; randomly selecting a first execution duration from the first interference duration interval; randomly selecting a second execution duration from the second interference duration interval; after executing the first ultrasonic interference signal according to the first execution duration, executing the second ultrasonic interference signal according to the second execution duration; and repeating the duration random selection and interference signal execution to continuously perform recording interference on the target area.

[0011] Another aspect disclosed in this application provides a recording shielding device based on ultrasonic interference. The device includes: a type information acquisition unit for detecting and identifying a recording device in a target area and obtaining the type information of the recording device; an interference signal configuration unit for configuring a first ultrasonic interference signal and a second ultrasonic interference signal based on the type information, where the first ultrasonic interference signal is inaudible to the human ear but can be captured by the recording device, and the second ultrasonic interference signal is a signal used to generate a nonlinear mixing effect inside the recording device; and a recording interference execution unit for performing recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal.

[0012] The recording shielding method and device based on ultrasonic interference proposed by the present invention first detect and identify the recording devices in the target area, obtain the type information of the recording devices, determine the specific objects to be interfered with and their characteristics, and provide basic data for subsequent targeted configuration of interference signals; secondly, based on the obtained type information of the recording devices, configure the first ultrasonic interference signal and the second ultrasonic interference signal. Among them, the first ultrasonic interference signal is a signal that is inaudible to the human ear but can be captured by the recording device, ensuring that the interference process will not affect normal interpersonal communication; the second ultrasonic interference signal is a signal used to generate a non-linear mixing effect inside the recording device, interact with the first signal inside the recording device, and produce a more effective interference effect; then, according to the configured first ultrasonic interference signal and the second ultrasonic interference signal, perform recording interference on the target area, so that the recording device cannot normally capture effective sound information, thereby realizing the recording shielding of the target area and improving the effect of recording interference. Description of the Drawings

[0013] Figure 1 FIG. is a schematic flow chart of a recording shielding method based on ultrasonic interference provided by an embodiment of the present application; Figure 2 FIG. is a schematic structural diagram of a recording shielding device based on ultrasonic interference provided by an embodiment of the present application.

[0014] Description of the Reference Numerals: 11, type information acquisition unit; 12, interference signal configuration unit; 13, recording interference execution unit. Detailed Embodiment

[0015] The general idea of the technical solution provided by the present application is as follows: The embodiments of the present application provide a recording shielding method and device based on ultrasonic interference, which can accurately identify the type of recording device and configure customized ultrasonic interference signals according to its characteristics, so as to achieve efficient shielding of recording content.

[0016] Specifically, first, detect and identify the recording devices in the target area, obtain their type information, and lay a foundation for subsequent precise interference; secondly, based on the obtained type information of the recording devices, configure two ultrasonic interference signals in a targeted manner, namely, the first ultrasonic interference signal that is inaudible to the human ear but can be captured by the recording device, and the second ultrasonic interference signal used to generate a non-linear mixing effect inside the recording device; then, use the configured ultrasonic interference signals to perform precise recording interference on the target area, so that the recording device cannot normally record effective information.

[0017] Through the above technical solution, the technical problem in the prior art that there is a lack of an accurate shielding method for different types of recording devices, resulting in poor recording interference effect, is solved, and a more effective and accurate recording shielding solution is provided for various occasions that require confidentiality.

[0018] After introducing the basic principle of the present application, the various non-limiting implementation manners of the present application will be specifically introduced below in conjunction with the accompanying drawings of the specification.

[0019] Example 1, as Figure 1 shown, the embodiment of the present application provides a recording shielding method based on ultrasonic interference, and the method includes: S100: Detect and identify the recording devices in the target area, and obtain the type information of the recording devices.

[0020] Specifically, first, determine the recording devices existing in the target area and their specific type information. In actual application scenarios, different types of recording devices have different hardware structures, signal processing characteristics, and frequency response characteristics. Therefore, accurately identifying the type of recording device is crucial for subsequent configuration of targeted interference signals. Specifically, the target area is scanned by using specific detection means to discover possible recording devices. Among them, the detection means may include, but are not limited to, electromagnetic signal detection, radio spectrum analysis, non-linear node detection and other technologies. When a recording device is detected in the target area, the obtained signal characteristics are further analyzed and compared with the preset recording device characteristic database, so as to identify the specific type of the recording device, such as the recording function of a smart phone, a professional recording pen, a concealed recording device, etc.

[0021] By detecting and identifying the recording devices in the target area, necessary parameter basis is provided for subsequent configuration of targeted ultrasonic interference signals, ensuring the accuracy and effectiveness of the interference effect.

[0022] S200: Based on the type information, configure a first ultrasonic interference signal and a second ultrasonic interference signal, where the first ultrasonic interference signal is inaudible to the human ear but can be captured by the recording device, and the second ultrasonic interference signal is a signal used to generate a non-linear mixing effect inside the recording device.

[0023] Specifically, after obtaining the type information of the recording device, targeted ultrasonic interference signals are customized and configured, including a first ultrasonic interference signal and a second ultrasonic interference signal. By configuring two ultrasonic interference signals with different functions, a dual interference effect on the recording device is achieved.

[0024] Among them, the frequency range of the first ultrasonic interference signal is set in a frequency band that cannot be perceived by the human ear but can be captured by a recording device. Although this signal has no impact on the human auditory system, it can be captured and recorded by the sensor of the recording device, thereby introducing a large amount of noise interference into the recorded content. The second ultrasonic interference signal utilizes the non-linear characteristics of the internal microphone element and amplifier circuit of the recording device and is designed to be a signal that can generate a non-linear mixing effect inside the recording device. When two or more ultrasonic signals of different frequencies act on the recording device simultaneously, due to the non-linear characteristics of the device itself, these signals will mix inside the device, generating new difference frequency and sum frequency components, further interfering with the recorded content.

[0025] For example, the look-up table configuration method is used to achieve the precise configuration of the interference signal. Specifically, a mapping database of recording device types and optimal interference signal parameters is established in advance. This database stores the interference parameter sets corresponding to various common recording devices, including but not limited to the frequency, amplitude, and waveform parameters of the first ultrasonic interference signal, as well as the frequency combinations, phase relationships, and modulation methods of the second ultrasonic interference signal. After identifying the type of recording device in the target area, directly query this mapping database, quickly extract the optimal interference parameter set corresponding to this type of device, and configure the first and second ultrasonic interference signals accordingly. This look-up table configuration method has the characteristics of fast response speed and high efficiency, and is particularly suitable for real-time interference scenarios for recording devices of known types.

[0026] By configuring the first ultrasonic interference signal and the second ultrasonic interference signal, the most effective ultrasonic interference signal combination can be generated for a specific type of recording device, laying a foundation for subsequent interference implementation, thereby achieving precise shielding for different recording devices.

[0027] S300: Perform recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal.

[0028] Specifically, after configuring the first ultrasonic interference signal and the second ultrasonic interference signal, the two configured ultrasonic interference signals are actually transmitted to the target area, thereby achieving effective interference on the recording devices in the target area. In the specific implementation process, first, the configured first ultrasonic interference signal and the second ultrasonic interference signal are converted into acoustic energy through an ultrasonic transducer array. The ultrasonic transducer array usually consists of multiple precisely calibrated ultrasonic transmitting units, and these units can be independently controlled and work together to ensure that the interference signal can accurately cover the target area.

[0029] To improve the interference effect and reduce the possibility of being circumvented, an intelligent scheduling strategy can be adopted to control the transmission order, duration, and intensity of the two interference signals. For example, the two interference signals can be transmitted alternately, or simultaneously according to the characteristics of the recording device to form a composite interference effect. At the same time, the directivity and energy distribution of the interference signal can be adjusted in real time according to factors such as the spatial characteristics of the target area and the ambient noise level to ensure the maximization of the interference effect. Due to the strong directivity and attenuation characteristics of ultrasonic waves, the formation and projection of the ultrasonic beam will be precisely controlled during interference, so that the interference signal acts concentratedly on the target area while minimizing the impact on non-target areas to achieve precise interference.

[0030] By performing recording interference based on the first ultrasonic interference signal and the second ultrasonic interference signal, it is possible to effectively interfere with the recording devices in the target area, making it impossible to normally record clear and distinguishable sound content, thereby achieving the secure shielding of the recorded content and protecting sensitive information from being illegally obtained by recording.

[0031] Furthermore, detecting and identifying the recording devices in the target area and obtaining the type information of the recording devices includes: S110: Scanning the electromagnetic signals in the target area with a radio frequency scanner; S120: Performing spectrum analysis on the obtained electromagnetic signals, extracting multiple signal features, and the multiple signal features include frequency features, intensity features, and modulation features; S130: According to the frequency features, the intensity features, and the modulation features, performing device matching in a preset recording device feature database, and determining the type information of the recording device according to the matching result.

[0032] Specifically, when obtaining the type information of the recording device, first, a dedicated radio frequency scanner is used to scan the full-band electromagnetic signals in the target area. The radio frequency scanner can be configured to cover a frequency range of 300 MHz to 6 GHz, and this range includes the operating bands of the vast majority of recording devices. During the scanning process, the radio frequency scanner continuously samples the electromagnetic environment in the target area at a preset frequency step to obtain the original data such as the electromagnetic signal intensity and phase at different frequency points. To improve the scanning efficiency and detection sensitivity, the radio frequency scanner can adopt a directional antenna array and use beam scanning technology to conduct fine grid detection on the target area to ensure that no possible electromagnetic signal source is missed.

[0033] Then, perform in-depth signal processing and feature extraction on the acquired electromagnetic signals to extract multiple signal features. Convert the time-domain signals into frequency-domain representations through fast Fourier transform to obtain the electromagnetic spectrum distribution map in the target area. On this basis, further extract frequency features, intensity features, and modulation features. Among them, the frequency features include parameters such as the center frequency, bandwidth, frequency offset, harmonic components, and their ratios of the signal, which can reflect the characteristics of the radio frequency oscillator and modulation circuit of the recording device; the intensity features include parameters such as the power spectral density, peak intensity, average intensity, and their time-varying characteristics of the signal, which can reflect the transmission power and distance information of the recording device; the modulation features include parameters such as the modulation mode, modulation depth, symbol rate, and pulse characteristics of the signal, which can reflect the data transmission protocol and coding method of the recording device.

[0034] Subsequently, based on the extracted frequency features, intensity features, and modulation features, perform matching and identification in the preset recording device feature database. The preset recording device feature database stores electromagnetic signal feature templates of various common recording devices, and each template contains the frequency features, intensity features, and modulation features of this type of device. The matching process uses a similarity calculation method that fuses multiple features, comprehensively considers the matching degree of features in each dimension, and calculates the matching degree between the target signal and each template in the database. Finally, select the device type with the highest matching degree as the identification result, which serves as the type information of the recording device and provides an accurate basis for subsequent interference signal configuration.

[0035] Further, the device matching in the preset recording device feature database according to the frequency features, the intensity features, and the modulation features, and determining the type information of the recording device according to the matching result includes: S131: Obtain the device feature matching formula; S132: Determine the first sample device in the preset recording device feature database, and extract the first frequency feature, the first intensity feature, and the first modulation feature of the first sample device; S133: Based on the device feature matching formula, according to the frequency feature, the intensity feature, and the modulation feature of the recording device, and the first frequency feature, the first intensity feature, and the first modulation feature of the first sample device, obtain the first device matching degree between the recording device and the first sample device; S134: According to the method of obtaining the first device matching degree, continue to obtain the device matching degrees between other sample devices in the preset recording device feature database and the recording device, and obtain multiple device matching degrees; S135: Select the maximum device matching degree among the multiple device matching degrees as the matching sample device; S136: In the preset recording device feature database, extract the type information of the matched sample device as the type information of the recording device.

[0036] As an optional implementation, first, obtain a device feature matching formula, which is used to calculate the similarity between a recording device and a sample device. By comprehensively considering the importance differences of three different types of features, namely frequency feature, intensity feature, and modulation feature, in device recognition, and reasonably regulating the contributions of various features through weight coefficients, the device feature matching formula is obtained. Then, select a sample device from the preset recording device feature database, denoted as the first sample device, as the comparison object. The selection process can be based on a preset traversal order or a likelihood ranking estimated according to prior knowledge. After determining the first sample device, extract the features of this sample device from the preset recording device feature database, including its first frequency feature, first intensity feature, and first modulation feature. These features are reference features obtained through test analysis of known devices during the database establishment stage and serve as the benchmark data for subsequent matching calculations. Next, apply the obtained device feature matching formula to the device features of the recording device in the extracted target area and the features of the first sample device, and calculate the similarity between the two, that is, the first device matching degree. During the calculation process, calculate the similarity between the frequency feature, intensity feature, and modulation feature and the first frequency feature, first intensity feature, and first modulation feature respectively, and then perform weighted fusion through the weight coefficients in the device feature matching formula to obtain a comprehensive matching degree score as the first device matching degree. This matching degree score quantitatively characterizes the similarity between the recording device and the first sample device and provides a basis for subsequent matching judgments.

[0037] Then, calculate the matching degrees between the recording device and the remaining sample devices in the database one by one in the same way as obtaining the first device matching degree. By traversing all the sample devices in the database, multiple device matching degrees are finally obtained, and then find the maximum value among them. The sample device with the maximum matching degree is considered to be the device type most similar to the recording device and is therefore selected as the matched sample device. Subsequently, according to the determined matched sample device, extract the type information of the matched sample device from the preset recording device feature database, including key attributes such as the brand, model, function parameters, and hardware architecture of the device. These information can comprehensively describe the characteristics of the recording device. The extracted type information is used as the recognition result of the recording device and provides an accurate reference basis for subsequent interference signal configuration.

[0038] Precise matching and type recognition of the recording device based on multi-dimensional features lay the foundation for subsequent interference signal configuration.

[0039] Furthermore, the device feature matching formula is: Among them, represents the device matching degree, represents the frequency feature, represents the first frequency feature, represents the similarity between the frequency feature and the first frequency feature, represents the intensity feature, represents the first intensity feature, represents the similarity between the intensity feature and the first intensity feature, represents the modulation feature, represents the first modulation feature, represents the similarity between the modulation feature and the first modulation feature, is the weight coefficient, and .

[0040] The above device feature matching formula adopts the form of weighted linear combination to comprehensively evaluate the similarity degree between the recording device and the sample device in terms of frequency feature, intensity feature and modulation feature. Among them, the frequency feature similarity quantifies the frequency feature of the recording device and the first frequency feature of the sample device The degree of closeness between them. In actual calculation, mathematical methods such as cosine similarity and the reciprocal of the normalization of Euclidean distance can be used to calculate the similarity of two frequency feature vectors. The frequency feature similarity reflects the matching degree of two devices in terms of working frequency, frequency band characteristics, etc. The intensity feature similarity measures the intensity feature of the recording device and the first intensity feature of the sample device The degree of similarity between them. The intensity feature usually shows the statistical characteristics of signal energy distribution, and its similarity calculation can adopt methods such as histogram comparison or statistical moment difference comparison. The intensity feature similarity can reflect the transmission power characteristics and near and far field performance of the device. The modulation feature similarity evaluates the modulation feature of the recording device and the first modulation feature of the sample device The consistency between them. The modulation feature usually contains information such as signal modulation method and symbol feature, and its similarity calculation can be based on feature comparison in the modulation domain or result comparison of modulation recognition algorithms. The weight coefficient respectively represents the importance degree of the frequency feature, intensity feature and modulation feature in the final matching judgment. The sum of the three is 1, ensuring the normalization of the matching degree score. According to different application scenarios and device type distributions, these weight coefficients can be dynamically adjusted to optimize the matching performance.

[0041] The device matching degree calculated by the device feature matching formula is a scalar value between 0 and 1. The closer its value is to 1, the higher the similarity between the recording device and the sample device, and the more likely they belong to the same type.

[0042] Further, configuring the first ultrasonic interference signal and the second ultrasonic interference signal based on the type information includes: S210: Construct an interference signal configuration model, where the interference signal configuration model includes a first signal configuration channel and a second signal configuration channel; S220: Input the type information into the first signal configuration channel to obtain the first ultrasonic interference signal; S230: Input the type information into the second signal configuration channel to obtain the second ultrasonic interference signal.

[0043] In a feasible manner, first, construct a two-channel interference signal configuration model as a processing unit for converting the recording device type information into specific interference signal parameters. This model adopts a parallel processing architecture and includes two independent but cooperative signal configuration channels. The first signal configuration channel is specifically responsible for generating the parameter configuration of the first ultrasonic interference signal, which is an interference signal inaudible to the human ear but capturable by the recording device; the second signal configuration channel is specifically responsible for generating the parameter configuration of the second ultrasonic interference signal, which is used to generate a non-linear mixing effect inside the recording device. Although the two configuration channels have different functional positions, they maintain consistency in the model architecture and both adopt a mapping conversion mechanism, which can effectively convert the type information of the recording device into the corresponding interference signal parameter set.

[0044] When configuring the first ultrasonic interference signal and the second ultrasonic interference signal, the obtained type information of the recording device is used as input and imported into the first signal configuration channel for processing. Inside the first signal configuration channel, the mapping relationship from the device type to the signal parameters is implemented. According to the input type information, the first ultrasonic interference signal parameters suitable for this type of recording device are generated, including key parameters such as frequency, amplitude, waveform, and duration. Since the first ultrasonic interference signal mainly targets the frequency bands that can be captured by the recording device but are inaudible to the human ear, this channel will pay special attention to the upper limit characteristic of the frequency response of the recording device and optimize the frequency setting of the interference signal to ensure that the generated interference signal can be effectively captured by the target recording device while maintaining the characteristic of being imperceptible to the human ear. Through this configuration channel, a highly targeted and optimized first ultrasonic interference signal is finally obtained. Similarly, the type information of the recording device is input into the second signal configuration channel. The second signal configuration channel focuses on generating ultrasonic signal parameters that can cause the internal non-linear mixing effect of the recording device. Inside this channel, the professional mapping relationship from the device type to the non-linear mixing interference signal parameters is implemented. According to the differences in the non-linear characteristics of different types of recording devices, the ultrasonic frequency combination, phase relationship, and modulation parameters that can produce the best mixing effect are calculated. The second signal configuration channel will specifically consider the non-linear transfer function characteristic of the recording device and optimize the frequency combination of the interference signal so that the generated difference frequency components can effectively cover the core frequency band of human speech, thereby interfering with the recording of speech content to the greatest extent. Through this configuration channel, the second ultrasonic interference signal optimized for non-linear mixing interference is obtained.

[0045] Through the configuration conversion from the recording device type information to the dual-channel ultrasonic interference signal, it provides a parameter basis for the subsequent implementation of recording interference, ensuring the pertinence and effectiveness of interference against different types of recording devices.

[0046] Furthermore, an interference signal configuration model is constructed. The interference signal configuration model includes a first signal configuration channel and a second signal configuration channel, and includes: S211: Obtain the test interference data of multiple sample recording devices. The test interference data includes recording interference data and device interference data; S212: Based on the recording interference data of multiple sample recording devices, establish a first correspondence library between the recording device type and the first ultrasonic interference parameters; S213: Based on the device interference data of multiple sample recording devices, establish a second correspondence library between the recording device type and the second ultrasonic interference parameters; S214: According to the first correspondence library, train the first deep learning network as the first signal configuration channel; S215: According to the second correspondence library, train the second deep learning network as the second signal configuration channel; S216: Combine the first deep learning network and the second deep learning network to form the interference signal configuration model.

[0047] Specifically, when constructing the interference signal configuration model, first, conduct systematic interference tests on various types of sample recording devices to obtain comprehensive test interference data. During the test process, a strictly controlled experimental environment is adopted. Apply ultrasonic interference signals with different parameters to each type of recording device and record the interference effects in detail. The obtained test interference data is divided into two categories: recording interference data and device interference data. The recording interference data mainly reflects the interference effects of ultrasonic signals with different parameters on the quality of recorded content, including quantization indicators such as the degree of signal-to-noise ratio reduction and the level of speech intelligibility degradation; the device interference data mainly reflects the situation of the internal non-linear mixing effect caused by ultrasonic signals in the recording device, including characteristic parameters such as the spectral distribution and relative intensity of the mixing products. These test data constitute the basic data set for subsequent model training, and their quality and comprehensiveness directly affect the performance of the final interference model.

[0048] Then, conduct in-depth analysis and data mining on the obtained recording interference data to extract the mapping relationship between the recording device type and the optimal first ultrasonic interference parameters. Statistical methods such as multivariate correlation analysis and parameter sensitivity analysis are used in the analysis process to determine the optimal parameter combinations of interference frequencies, amplitudes, waveforms, etc. corresponding to different types of recording devices. Based on the analysis results, establish a structured first correspondence library, which is indexed by the recording device type and stores the corresponding optimal first ultrasonic interference parameter set in an associated manner. This correspondence library not only contains discrete parameter correspondence records but also integrates the internal association rules between parameters, providing a high-quality structured knowledge source for subsequent deep learning model training. At the same time, use a similar method, but analyze and process the device interference data to establish a mapping relationship library between the recording device type and the optimal second ultrasonic interference parameters. Since the second ultrasonic interference signal is mainly used to generate non-linear mixing effects, the analysis process pays special attention to the non-linear characteristics of different types of recording devices and their sensitivity to mixing parameters. Through professional methods such as non-linear system identification and harmonic analysis, determine the ultrasonic signal parameter combinations that can produce the strongest non-linear mixing effects on various types of recording devices. Based on the analysis results, establish a second correspondence library, which accurately records the optimal second ultrasonic interference parameters corresponding to each type of recording device, providing a professional knowledge basis for non-linear mixing effects for subsequent model training.

[0049] Subsequently, based on the established first correspondence library, the first signal configuration channel is trained using deep learning techniques. The deep learning network adopts a multi-layer neural network structure, including an input layer, multiple hidden layers, and an output layer. The input of the network is the type feature vector of the recording device, and the output is the parameter vector of the first ultrasonic interference signal. During the training process, the data in the first correspondence library is used as training samples, and the network parameters are optimized through the backpropagation algorithm, enabling the network to accurately map the relationship from the device type to the optimal interference parameters. To improve the generalization ability of the model, regularization techniques and early stopping strategies are introduced during the training process to avoid overfitting problems. After training, the first deep learning network can quickly output the optimized parameters of the first ultrasonic interference signal for the input device type information, realizing the intelligent conversion from the device type to the interference parameters. At the same time, a similar deep learning method is adopted, but based on the established second correspondence library, the second deep learning network for non-linear mixing interference is trained. This network also adopts a multi-layer neural network structure, but the learning ability for the non-linear characteristics of the recording device is specifically optimized in the network design. The input layer of the network receives the type features of the recording device, and the output layer generates the parameter vector of the second ultrasonic interference signal. During the training process, the data in the second correspondence library is used as training samples, and the network weights are adjusted through optimization algorithms such as gradient descent, enabling the network to accurately predict the optimal non-linear mixing interference parameters. To enhance the robustness of the model, cross-validation and data augmentation techniques are adopted during the training process to ensure that the model can still perform well when facing unknown device types. After training, the second deep learning network can output the best parameters of the second ultrasonic interference signal for a given device type, providing parameter support for realizing effective non-linear mixing interference. Subsequently, the trained first deep learning network and the trained second deep learning network are integrated to form a complete interference signal configuration model. The integration process adopts a parallel architecture, and the two deep learning networks independently process the same input (the type information of the recording device), and respectively output the parameters of the first acoustic interference signal and the second ultrasonic interference signal. The integrated interference signal configuration model has a dual-channel parallel processing ability, and can simultaneously generate two optimized interference signal parameters for the recording content and the device non-linear characteristics, providing intelligent configuration support for realizing all-round and high-efficiency recording interference.

[0050] Through the complete process from obtaining experimental data to deep learning model training and then to system integration, an interference signal configuration model with intelligent learning and precise configuration capabilities is constructed, providing technical support for configuring interference signals.

[0051] Further, the recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal includes: S310: Configure the first interference duration range for the first ultrasonic interference signal and the second interference duration range for the second ultrasonic interference signal; S320: Randomly select a first execution duration from the first interference duration range; S330: Randomly select a second execution duration from the second interference duration range; S340: After executing the first ultrasonic interference signal according to the first execution duration, execute the second ultrasonic interference signal according to the second execution duration; S350: Repeat the duration random selection and interference signal execution to continuously perform recording interference on the target area.

[0052] In an alternative embodiment, first, configure the first interference duration range for the first ultrasonic interference signal and the second interference duration range for the second ultrasonic interference signal. Among them, the first interference duration range is for the first ultrasonic interference signal and is usually set to , , where and are the minimum and maximum execution durations of the first ultrasonic interference signal respectively; similarly, the second interference duration range is for the second ultrasonic interference signal and is set to , . The specific setting of the duration range needs to consider the sampling characteristics of the recording device, the signal processing mechanism, and the requirement for the persistence of the interference effect. A reasonable duration range configuration can not only ensure the effectiveness of each interference but also provide sufficient variation space for subsequent randomization strategies, enhancing the unpredictability of the interference.

[0053] Then, use a random number generation algorithm to randomly select a duration value within the first interference duration range , as the first execution duration . The random selection process can adopt a uniform distribution or other suitable probability distributions to ensure the randomness and uniformity of the duration value within the set range. This random selection mechanism makes the execution duration of each interference different, increasing the complexity and unpredictability of the interference pattern, and effectively preventing the recording device or recording software from filtering or compensating the interference signal through pattern recognition and adaptive algorithms. At the same time, use a similar random selection method, but for the second interference duration range , , randomly select a duration value as the second execution duration The random selection of the second execution duration is independent of the first execution duration, which further increases the randomness and complexity of the interference sequence. By double randomizing the execution durations of the two interference signals, the entire interference process presents highly irregular time characteristics, significantly improving the anti-recognition and anti-avoidance capabilities of the interference.

[0054] Then, the configured first ultrasonic interference signal is executed according to the selected first execution time. The first stage of interference is carried out; after the first stage of interference is completed, the configured second ultrasonic interference signal is immediately transmitted according to the selected second execution time. The first ultrasonic interference signal is mainly used to directly interfere with the recording content, while the second ultrasonic interference signal focuses on inducing the nonlinear mixing effect inside the recording device. The alternating execution of the two ultrasonic interference signals forms a multi-level and all-round interference mechanism. The process from S320 to S340 is regarded as an interference cycle, and multiple interference cycles are repeatedly executed to achieve continuous recording interference in the target area. In each cycle, a random duration is reselected, so that the entire interference process is manifested as a time-varying and unpredictable interference sequence on a macro scale. This continuous and randomly changing interference pattern can effectively prevent the recording device or post-processing software from identifying and filtering out the interference signal through time domain analysis or frequency domain analysis. The overall duration or number of cycles of continuous interference can be set according to actual needs, and it can also be dynamically determined whether to continue the interference according to the external control signal.

[0055] By realizing an ultrasonic interference sequence with time-varying characteristics and randomness, the effectiveness and difficulty of avoiding recording interference are improved, providing a more reliable technical means for protecting sensitive sound information.

[0056] Embodiment 2 is based on the same inventive concept as the recording shielding method based on ultrasonic interference in the above embodiment. Figure 2 As shown, the embodiment of the present application provides a recording shielding device based on ultrasonic interference, the device comprising: A type information acquisition unit 11 is used to detect and identify a recording device in a target area and acquire type information of the recording device; An interference signal configuration unit 12, configured to configure a first ultrasonic interference signal and a second ultrasonic interference signal based on the type information, wherein the first ultrasonic interference signal is a signal that is inaudible to human ears but can be captured by the recording device, and the second ultrasonic interference signal is a signal used to generate a nonlinear mixing effect inside the recording device; The recording interference execution unit 13 is configured to perform recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal.

[0057] Further, the execution steps of the type information acquisition unit 11 include: Scanning the electromagnetic signals in the target area through a radio frequency scanner; Performing spectrum analysis on the acquired electromagnetic signals, and extracting a plurality of signal features, where the plurality of signal features include frequency features, intensity features, and modulation features; According to the frequency features, the intensity features, and the modulation features, performing device matching in a preset recording device feature database, and determining the type information of the recording device according to the matching result.

[0058] Further, the execution steps of the type information acquisition unit 11 further include: Obtaining a device feature matching formula; Determining a first sample device in the preset recording device feature database, and extracting a first frequency feature, a first intensity feature, and a first modulation feature of the first sample device; Based on the device feature matching formula, according to the frequency feature, the intensity feature, and the modulation feature of the recording device, and the first frequency feature, the first intensity feature, and the first modulation feature of the first sample device, obtaining a first device matching degree between the recording device and the first sample device; According to the manner of obtaining the first device matching degree, continuing to obtain the device matching degrees between other sample devices in the preset recording device feature database and the recording device, and obtaining a plurality of device matching degrees; Selecting the maximum device matching degree among the plurality of device matching degrees as the matching sample device; In the preset recording device feature database, extracting the type information of the matching sample device as the type information of the recording device.

[0059] Further, the device feature matching formula is: Wherein, represents the device matching degree, represents the frequency feature, represents the first frequency feature, represents the similarity between the frequency feature and the first frequency feature, represents the intensity feature, represents the first intensity feature, represents the similarity between the intensity feature and the first intensity feature, represents the modulation feature, represents the first modulation feature, Indicates the similarity between the modulation feature and the first modulation feature, is the weight coefficient, and .

[0060] Furthermore, the execution steps of the interference signal configuration unit 12 include: Construct an interference signal configuration model, which includes a first signal configuration channel and a second signal configuration channel; Input the type information into the first signal configuration channel to obtain a first ultrasonic interference signal; Input the type information into the second signal configuration channel to obtain a second ultrasonic interference signal.

[0061] Furthermore, the execution steps of the interference signal configuration unit 12 further include: Obtain the test interference data of multiple sample recording devices, where the test interference data includes recording interference data and device interference data; Based on the recording interference data of multiple sample recording devices, establish a first correspondence library between the recording device type and the first ultrasonic interference parameter; Based on the device interference data of multiple sample recording devices, establish a second correspondence library between the recording device type and the second ultrasonic interference parameter; According to the first correspondence library, train a first deep learning network as the first signal configuration channel; According to the second correspondence library, train a second deep learning network as the second signal configuration channel; Combine the first deep learning network and the second deep learning network to form the interference signal configuration model.

[0062] Furthermore, the execution steps of the recording interference execution unit 13 include: Configure a first interference duration interval for the first ultrasonic interference signal and a second interference duration interval for the second ultrasonic interference signal; Randomly select a first execution duration from the first interference duration interval; Randomly select a second execution duration from the second interference duration interval; After executing the first ultrasonic interference signal according to the first execution duration, execute the second ultrasonic interference signal according to the second execution duration; Repeat the duration random selection and interference signal execution to continuously perform recording interference on the target area.

[0063] The above are only the preferred embodiments of the present application, and do not limit the patent scope of the present application accordingly. Any equivalent structural or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be similarly included in the patent protection scope of the present application.

Claims

1. A recording shielding method based on ultrasonic interference, characterized in that: The method comprises: Detect and identify recording devices in a target area, and obtain type information of the recording devices; Based on the type information, configure a first ultrasonic interference signal and a second ultrasonic interference signal, wherein the first ultrasonic interference signal is a signal that is inaudible to human ears but can be captured by the recording device, and the second ultrasonic interference signal is a signal used to generate a nonlinear mixing effect inside the recording device; The target area is subjected to recording interference according to the first ultrasonic interference signal and the second ultrasonic interference signal.

2. The method according to claim 1, characterized in that The detecting and identifying the recording device in the target area and obtaining the type information of the recording device includes: scanning the electromagnetic signal in the target area by means of a radio frequency scanner; Performing spectrum analysis on the acquired electromagnetic signal to extract multiple signal features, wherein the multiple signal features include frequency features, intensity features, and modulation features; According to the frequency feature, the intensity feature and the modulation feature, device matching is performed in a preset recording device feature database, and the type information of the recording device is determined according to the matching result.

3. The method according to claim 2, characterized in that The performing device matching in a preset recording device feature database according to the frequency feature, the intensity feature, and the modulation feature, and determining the type information of the recording device according to the matching result, includes: Get the device feature matching formula; Determine a first sample device in the preset recording device feature database, and extract a first frequency feature, a first intensity feature, and a first modulation feature of the first sample device; Based on the device feature matching formula, according to the frequency feature, the intensity feature and the modulation feature of the recording device, and the first frequency feature, the first intensity feature and the first modulation feature of the first sample device, obtaining a first device matching degree between the recording device and the first sample device; According to the method for obtaining the first device matching degree, continue to obtain the device matching degrees of other sample devices in the preset recording device feature database and the recording device, and obtain multiple device matching degrees; Selecting the device with the maximum matching degree among multiple device matching degrees as the matching sample device; The type information of the matching sample device is extracted from a preset recording device feature database as the type information of the recording device.

4. The method according to claim 3, characterized in that The device feature matching formula is: in, Indicates the device matching degree. represents the frequency characteristics, represents the first frequency characteristic, Indicates the similarity between the frequency feature and the first frequency feature, Indicates the intensity characteristics, represents the first intensity feature, represents the similarity between the intensity feature and the first intensity feature, represents the modulation characteristics, represents the first modulation characteristic, represents the similarity between the modulation feature and the first modulation feature, is the weight coefficient, and .

5. The method according to claim 1, characterized in that The configuring the first ultrasonic interference signal and the second ultrasonic interference signal based on the type information includes: Constructing an interference signal configuration model, wherein the interference signal configuration model includes a first signal configuration channel and a second signal configuration channel; Inputting the type information into the first signal configuration channel to obtain a first ultrasonic interference signal; The type information is input into the second signal configuration channel to obtain a second ultrasonic interference signal.

6. The method according to claim 1, characterized in that Constructing an interference signal configuration model, wherein the interference signal configuration model includes a first signal configuration channel and a second signal configuration channel, including: Acquire test interference data of a plurality of sample recording devices, wherein the test interference data includes recording interference data and device interference data; Based on the recording interference data of a plurality of sample recording devices, a first correspondence library between the recording device type and the first ultrasonic interference parameter is established; Based on the device interference data of a plurality of sample recording devices, a second correspondence library between the recording device type and the second ultrasonic interference parameter is established; According to the first correspondence library, training a first deep learning network as the first signal configuration channel; According to the second correspondence library, training a second deep learning network as the second signal configuration channel; The first deep learning network is combined with the second deep learning network to form the interference signal configuration model.

7. The method according to claim 1, characterized in that The recording interference of the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal includes: Configure a first interference duration interval of the first ultrasonic interference signal and a second interference duration interval of the second ultrasonic interference signal; Randomly select a first execution duration from the first interference duration interval; Randomly selecting a second execution duration from the second interference duration interval; After executing the first ultrasonic interference signal according to the first execution duration, executing the second ultrasonic interference signal according to the second execution duration; The random selection of duration and execution of interference signals are repeated to continuously interfere with the recording of the target area.

8. A recording shielding device based on ultrasonic interference, characterized in that: The device comprises: A type information acquisition unit, used to detect and identify the recording device in the target area, and acquire the type information of the recording device; an interference signal configuration unit, configured to configure a first ultrasonic interference signal and a second ultrasonic interference signal based on the type information, wherein the first ultrasonic interference signal is a signal that is inaudible to human ears but can be captured by the recording device, and the second ultrasonic interference signal is a signal for generating a nonlinear mixing effect inside the recording device; A recording interference execution unit is used to perform recording interference on the target area according to the first ultrasonic interference signal and the second ultrasonic interference signal.

Images