An object activity perception system, method, apparatus, and storage medium

Abstract

This invention discloses an object activity sensing system, method, device, and storage medium. The system includes: an ultrasonic transmitting module for continuously transmitting ultrasonic signals; an ultrasonic receiving module for continuously receiving and processing spatially propagating sound wave signals to obtain multi-channel ultrasonic signals; an ultrasonic preprocessing module; an ultrasonic positioning module for calculating the motion trajectory of a moving object; an ultrasonic activity detection module for detecting whether there is a moving object within a specific range; and a decision module for determining whether there is an object within a given distance D1 and whether there is a moving object within a given distance D2, where D2>D1. This invention can simultaneously detect the presence of stationary objects at close range and moving objects at a distant distance. It is fast and convenient to implement, can directly use the audio hardware built into electronic devices without increasing hardware costs, and is highly efficient. It has significant application value and enormous economic value in fields such as smart homes and consumer electronics.

Description

Technical Field

[0001] This invention belongs to the field of communication technology, and specifically relates to an object activity sensing system, method, device, and storage medium. Background Technology

[0002] Detecting the presence and activity of objects can be achieved through various sensors and information, such as measuring changes in the distance and orientation of a specific object from the sensor, and changes in sensor signals (sound / light / electrical / magnetic) caused by the object's movement. Typical sensors include cameras, LiDAR, millimeter-wave radar, and ultrasonic radar. In the fields of smart homes and consumer electronics, these sensors are not standard equipment. Achieving object presence and activity detection requires redesigning the product and will increase product costs.

[0003] Therefore, the present invention discloses an object activity sensing system, method, device, and storage medium that does not require changes to the existing product structure and does not increase hardware costs. Summary of the Invention

[0004] To address the technical problems existing in the prior art, the present invention aims to provide an object activity sensing system, method, device, and storage medium.

[0005] To achieve the above objectives and technical effects, the technical solution adopted by this invention is as follows:

[0006] An object motion sensing system, comprising:

[0007] An ultrasonic transmitting module is used to continuously transmit pre-designed ultrasonic signals.

[0008] The ultrasonic receiving module is used to continuously receive and process sound wave signals propagating in space to obtain multi-channel ultrasonic signals.

[0009] The ultrasonic preprocessing module is used to preprocess the received multi-channel ultrasonic signals to remove direct wave signals and signals reflected from stationary objects;

[0010] The ultrasonic positioning module is used to calculate the motion trajectory of a moving object based on the signal received from the ultrasonic preprocessing module.

[0011] The ultrasonic activity detection module is used to detect whether there are moving objects within a specific range based on the signals received from the ultrasonic preprocessing module.

[0012] The decision module is used to make a comprehensive decision based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module, to determine whether there is an object within a given distance D1 and whether there is a moving object within a given distance D2, and D2>D1.

[0013] Furthermore, the ultrasonic signal emitted by the ultrasonic transmitting module is a combination of a single-frequency signal and a swept-frequency signal, with a frequency greater than or equal to 20kHz, and is emitted periodically. The ultrasonic transmitting module is a consumer-grade speaker. Even further, the built-in speaker of a large television screen can be used as the ultrasonic transmitting module, thus eliminating the need to modify the existing product design and saving costs.

[0014] Furthermore, the ultrasonic signal emitted by the ultrasonic transmitting module is a combined signal x(t) formed by a single-frequency signal and a swept-frequency signal (also known as a frequency-modulated continuous wave signal, FMCW), which is periodically emitted.

[0015] Furthermore, T1 = 40ms, T2 = 80ms, and s = 125000.

[0016] Furthermore, the ultrasonic receiving module is a set of consumer-grade microphones capable of simultaneously receiving acoustic signals with a maximum frequency of 24kHz (sampling frequency up to 48kHz). The ultrasonic receiving module can be composed of multiple microphones integrated into the large-screen TV, thus eliminating the need to modify existing product designs and saving costs.

[0017] Furthermore, the ultrasonic receiving module acquires sound waves (including audible sound waves and ultrasonic waves) propagating in space to obtain multi-channel acoustic signals. These signals are then passed through a bandpass filter to filter out frequency components below a given threshold, i.e., audible sound waves, retaining only the ultrasonic signal y(t) within a given frequency range. Specifically, the start and end frequencies of the bandpass filter are 18kHz and 23kHz, respectively.

[0018] This invention discloses a method for sensing object activity, which uses an object activity sensing system as described above for sensing, and includes the following steps:

[0019] The ultrasonic transmitting module continuously transmits pre-set ultrasonic signals, and the ultrasonic receiving module continuously receives ultrasonic signals in the space. Then, the signals are sent to the ultrasonic preprocessing module for processing, and multi-channel sinusoidal signals and single-frequency signals are output. These signals are then sent to the ultrasonic positioning module and the ultrasonic activity detection module for processing, respectively.

[0020] The ultrasonic positioning module estimates the position of the moving object based on the received multi-channel sinusoidal wave signals, tracks the motion trajectory of the moving object, and outputs the result to the decision module; the ultrasonic activity detection module detects whether there is any object moving within a given distance range D2 based on the received multi-channel single-frequency signals, and outputs the result to the decision module.

[0021] The decision module makes a comprehensive decision based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module, and outputs whether there is an object within a given distance D1 and whether there is a moving object within a given distance D2.

[0022] Furthermore, the processing steps of the ultrasound preprocessing module include:

[0023] The ultrasonic preprocessing module preprocesses the received ultrasonic signal y(t) by removing the direct wave signal from the input signal using the following formula, thus obtaining the signal y′(t):

[0024] y′(t)=y(t)-Γ(y(t))

[0025] Where Γ(.) is the direct wave space transfer function, which can be measured in advance in a quiet room;

[0026] Subsequently, background noise filtering is performed, that is, filtering the reflected wave signals from surrounding stationary objects from the signal y′(t) to obtain the signal y″(t):

[0027] y″(t)=ISFT(Y′(k)-Θ(k))

[0028] Where: Y′(k) is the frequency domain signal corresponding to y′(t), Θ(k) represents the frequency domain signal of the reflected wave from a stationary object, and ISFT(.) represents the inverse short-time Fourier transform;

[0029] Next, the single-frequency signal y″ is separated from each group of periodic signals y″(t) according to time. sf (t) and the sweep frequency signal y″ fm (t);

[0030] For the sweep frequency signal y″ fm (t), and the corresponding transmitted signal x fm Multiplying (t) yields the difference frequency signal y″ if (t), that is:

[0031] y″ if (t)=x fm (t)×y″ fm (t);

[0032] Then, the difference frequency signal y″ if (t) is low-pass filtered with a cutoff frequency of 5kHz to obtain a sinusoidal signal y″. if (t);

[0033] Finally, the multi-channel single-frequency signals y″ are output respectively. sf (t) and the sinusoidal signal y″ if (t), and respectively transmitted to the ultrasonic activity detection module and the ultrasonic positioning module.

[0034] Furthermore, the ultrasonic positioning module estimates the position of the moving object and tracks its motion trajectory based on the received multi-channel sinusoidal wave signals, including the following steps:

[0035] Step 1: Process the multi-channel sinusoidal wave signals containing distance and azimuth information using the 2D-Music algorithm to estimate the distance and azimuth of the moving object at each signal cycle moment, and finally form a series of two-dimensional maps of the distance-azimuth profile (Distance-AOA Profile);

[0036] Search the current two-dimensional distance-azimuth profile map, and find the distance and azimuth corresponding to the point with the maximum energy as the candidate for the target position at the current moment (the plane coordinates can be converted according to the distance and azimuth);

[0037] Step 2: If the maximum energy >= E1 and the distance between the target position (plane coordinates) at the current moment and the target position at the previous moment <= G1, then output the candidate position and energy information of the target at the current moment; otherwise, go to Step 3;

[0038] Step 3: If the distance between the target position at the current moment and the target position at the previous moment > G1, it is considered that the estimation of the target position at the current moment is incorrect, and output the position of the target at the previous moment + offset O; otherwise, go to Step 4;

[0039] Step 4: If the maximum energy < E1 and the ultrasonic activity detection module outputs that there is no object activity at the current moment, it is considered that the target is stationary at the current moment, and output the target position at the previous moment.

[0040] Further, the distance between positions is the Euclidean distance between two plane coordinate points.

[0041] Further, the coordinate offset O refers to the change amount of the plane coordinates, and can be calculated by the following formula:

[0042] O=(O x , O y ); O x =δ×sin(θ); O y =δ×cos(θ); θ = θ2 - θl

[0043] Where: θ2 is the target azimuth angle at the current moment, θ1 is the target azimuth angle at the previous moment, δ is the preset offset distance, which can be set according to the target moving speed, and is preferably 6 cm.

[0044] Further, G1 is set to 8 cm and E1 is set to 0.65.

[0045] Furthermore, the algorithm adopted by the ultrasonic positioning module depends on the movement of an object. When the object is stationary, it is impossible to measure the distance to it. When a moving object stops at a certain position, if the distance at the previous moment before it stops is less than D1, then the distance and azimuth at that moment are used as the distance and azimuth at the stationary moment.

[0046] Furthermore, the ranging accuracy of the ultrasonic positioning module is limited by factors such as the transmitted signal intensity, the object material, and the object distance. The reliable ranging range is 3m. When the distance exceeds 3m, the ranging accuracy drops sharply and the stability is poor. Therefore, in the present invention, D1 < 3m. The present invention uses an ultrasonic activity detection module to determine the activity of objects in a larger range.

[0047] Furthermore, the ultrasonic activity detection module uses the principle of the Doppler frequency shift effect caused by a moving object on the ultrasonic signal to detect the object activity. The steps for the ultrasonic activity detection module to detect whether there is an object activity within a given distance range D2 based on the received multi-channel single-frequency signals include:

[0048] Step1: Perform a frequency-domain transformation on the multi-channel single-frequency signal y″ sf (t), and extract the spectral energy within the range of 200Hz above and below each channel's single-frequency signal as the basic feature; among them, the frequency of the single-frequency signal is preferably 18kHz;

[0049] Step2: Calculate the spectral energy difference between the first channel and other channels, and form a feature vector with the basic feature of the first channel;

[0050] Step3: Send the above feature vector into a pre-trained activity detection model, and calculate whether there is an object activity within the given distance D2 at the current moment. If there is, output 1, otherwise output 0.

[0051] Furthermore, the deep learning model is distance-dependent. That is to say, this model depends on the distance D2. If the distance D2 changes, it is necessary to re-adjust the training data and re-train the model. The deep learning model is a deep residual network based on CNN, with 97 nodes in the input layer and 2 nodes in the output layer.

[0052] Furthermore, the steps for the decision-making module to make a comprehensive decision based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module and output whether there is an object within the given distance D1 and whether there is a moving object within the given distance D2 include:

[0053] Step1: If the distance of the moving object output by the ultrasonic positioning module < D1, then output that there is an object within the given distance D1 and there is an object activity within the given distance D2; otherwise, go to Step2;

[0054] Step 2: If the ultrasonic activity detection module outputs that there is an object moving, then it outputs that there is no object within the given distance D1 and there is an object moving within the given distance D2; otherwise, it outputs that there is no object within the given distance D1 and there is no object moving within the given distance D2.

[0055] The present invention also discloses an object activity sensing device, including a processor, a memory, and a program, wherein the program is stored in the memory, and the processor calls the program stored in the memory to execute an object activity sensing method as described above.

[0056] The present invention also discloses a computer-readable storage medium configured to store a program configured to perform an object activity sensing method as described above.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0058] This invention discloses an object activity sensing system, method, device, and storage medium. It continuously emits pre-set ultrasonic signals through an ultrasonic transmitting module and continuously receives ultrasonic signals from space through an ultrasonic receiving module, sending them to an ultrasonic preprocessing module for processing. This outputs multi-channel quasi-sine wave signals and single-frequency signals, which are then sent to an ultrasonic positioning module and an ultrasonic activity detection module for processing, respectively. This enables the estimation of the position of moving objects, the calculation and tracking of object trajectories, and the detection of object activity within a given distance range D2. Finally, a decision module performs a comprehensive decision, outputting whether there are objects within a given distance range D1 and whether there are moving objects within a given distance range D2. It can simultaneously detect the presence of stationary objects at close range and moving objects at a distance. Implementation is fast and convenient, and it can directly utilize the built-in audio hardware of electronic devices without modifying existing product structures or increasing hardware costs. It is highly efficient and has significant application value and substantial economic value in fields such as smart homes and consumer electronics. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of the present invention;

[0060] Figure 2 This is a schematic diagram of the ultrasonic signal frequency design of the present invention;

[0061] Figure 3 This is a flowchart of the decision-making module of the present invention. Detailed Implementation

[0062] The present invention will now be described in detail so that its advantages and features can be more easily understood by those skilled in the art, thereby providing a clearer and more explicit definition of the scope of protection of the present invention.

[0063] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form to prepare for the more detailed descriptions that follow.

[0064] like Figure 1-3 As shown, an object activity sensing system includes:

[0065] An ultrasonic transmitting module is used to continuously transmit pre-designed ultrasonic signals.

[0066] The ultrasonic receiving module is used to continuously receive and process sound wave signals propagating in space to obtain multi-channel ultrasonic signals.

[0067] The ultrasonic preprocessing module is used to preprocess the received multi-channel ultrasonic signals to remove direct wave signals and signals reflected from stationary objects;

[0068] The ultrasonic positioning module is used to calculate the motion trajectory of a moving object based on the signal received from the ultrasonic preprocessing module.

[0069] The ultrasonic activity detection module is used to detect whether there are moving objects within a specific range based on the signals received from the ultrasonic preprocessing module.

[0070] The decision-making module is used to make a comprehensive decision based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module, to determine whether there is an object within a given distance D1 and whether there is a moving object within a given distance D2. Specifically, D2>D1.

[0071] As one specific implementation, the ultrasonic transmitting module is a consumer-grade speaker capable of emitting sound wave signals with a frequency greater than or equal to 20kHz (referred to as low-frequency ultrasound). Furthermore, the built-in speaker of a large television screen can be used as the ultrasonic transmitting module, thus eliminating the need to modify existing product designs and saving costs.

[0072] The ultrasonic transmitting module emits an ultrasonic signal that is a combined signal x(t) consisting of a single-frequency signal and a swept-frequency signal (also known as a frequency-modulated continuous wave signal, FMCW), which is periodically emitted.

[0073] In a more specific implementation, T1 = 40ms, T2 = 80ms, and s = 125000.

[0074] As one specific implementation, the ultrasonic receiving module is a set of consumer-grade microphones capable of simultaneously receiving acoustic signals with a maximum frequency of 24kHz (sampling frequency up to 48kHz). The ultrasonic receiving module can be composed of multiple microphones integrated into the large-screen TV, thus eliminating the need to modify existing product designs and saving costs.

[0075] The ultrasonic receiving module acquires sound waves (including audible sound waves and ultrasound) propagating in space to obtain multi-channel acoustic signals. These signals are then passed through a bandpass filter to filter out frequency components below a given threshold, i.e., audible sound waves, retaining only the ultrasound signal y(t) within a given frequency range. Specifically, the start and end frequencies of the bandpass filter are 18kHz and 23kHz, respectively.

[0076] A method for sensing object activity includes the following steps:

[0077] 1) The ultrasonic transmission module continuously transmits pre-designed ultrasonic signals.

[0078] 2) The ultrasonic receiving module continuously receives ultrasonic signals propagating in space and transmits them to the ultrasonic preprocessing module.

[0079] 3) The received multi-channel ultrasonic signals are preprocessed using an ultrasonic preprocessing module to remove direct wave signals and signals reflected from stationary objects. The main steps are as follows:

[0080] Step 1: For the input multi-channel ultrasound signal y(t), remove the direct wave signal from the input signal using the following formula to obtain the signal y′(t):

[0081] y′(t)=y(t)-Γ(y(t))

[0082] Wherein: Γ(·) is the direct wave space transfer function, which can be measured in advance in a quiet room;

[0083] Step 2: Background noise filtering, that is, filtering the reflected wave signals from surrounding stationary objects from the signal y′(t) to obtain the signal y″(t):

[0084] y″(t)=ISFT(Y(k)-Θ(k))

[0085] Where: Y′(k) is the frequency domain signal corresponding to y′(t), Θ(k) represents the frequency domain signal of the reflected wave from a stationary object, and ISFT(.) represents the inverse short-time Fourier transform; it should be noted that Θ(k) is obtained by collecting a signal when there is no object activity in the current environment, and then calculating the average value of the frequency domain signals corresponding to all periodic signal segments in the signal segment;

[0086] Step 3: Separate the single-frequency signal y″ from each group of periodic signals y″(t) based on time.sf (t) and the sweep frequency signal y″ fm (t);

[0087] Step 4: Target the sweep frequency signal y″ fm (t), and the corresponding transmitted signal x″ fm Multiplying (t) yields the difference frequency signal y″ if (t, i.e.:

[0088] y″ if (t)=x fm (t)×y″ fm (t);

[0089] Step 5: Adjust the difference frequency signal y″ if (t) is low-pass filtered with a cutoff frequency of 5kHz to obtain a sinusoidal signal y″. if (t);

[0090] Step 6: Output the single-frequency signals y″ of the multiple channels respectively. sf (t) and the sinusoidal signal y″ if (t), and then transmitted to the ultrasonic activity detection module and the ultrasonic positioning module for processing respectively.

[0091] 4) The ultrasonic positioning module uses a 2D-MUSIC algorithm (multiple signal classification algorithm) to estimate the distance and orientation of the moving object at each signal cycle based on the multi-channel sinusoidal wave signal received from the ultrasonic preprocessing module. This results in a series of two-dimensional distance-azimuth profiles (AOA profiles). By performing a joint spatiotemporal search on these two-dimensional distance-azimuth profiles, the tracking and calculation of the object's movement trajectory are achieved. The main steps are as follows:

[0092] Step 1: Search the 2D distance-azimuth contour map at the current moment, find the distance and azimuth corresponding to the point with the highest energy as candidates for the target position at the current moment (the distance and azimuth can be converted into planar coordinates).

[0093] Step 2: If the maximum energy is greater than or equal to E1, and the distance between the current target position (planar coordinates) and the previous target position is less than or equal to G1, then output the current target candidate position and energy information; otherwise, proceed to Step 3; where the distance between positions is the Euclidean distance between two planar coordinate points.

[0094] Step 3: If the distance between the current target position and the previous target position is greater than G1, then the current target position is considered to be incorrectly estimated, and the previous target position plus coordinate offset O is output; otherwise, proceed to Step 4; where coordinate offset O refers to the change in planar coordinates, which can be calculated using the following formula:

[0095] O = (O x , O y ); O x = δ × sin(θ); O y = δ × cos(θ); θ = θ2 - θ1

[0096] Where: θ2 is the target azimuth angle at the current moment, θ1 is the target azimuth angle at the previous moment, and δ is the preset offset distance, which can be set according to the target movement speed, preferably 6 cm;

[0097] Step4: If the maximum energy < E1 and the ultrasonic activity detection module outputs that there is no object activity at present, it is considered that the target is stationary at the current moment, and the target position at the previous moment is output;

[0098] G1 is preferably set to 8 cm, and E1 is preferably set to 0.65;

[0099] The algorithm adopted by the ultrasonic positioning module depends on the object movement. When the object is stationary, it is impossible to measure the distance to it. When a moving object stops at a certain position, if the distance at the previous moment before stopping is less than D1, then the distance and azimuth at that moment are used as the distance and azimuth at the stationary moment.

[0100] The ranging accuracy of the ultrasonic positioning module is limited by factors such as the transmitted signal intensity, object material, object distance, etc. The reliable ranging range is 3 m. When the distance exceeds 3 m, the ranging accuracy drops sharply and the stability is poor. Therefore, the present invention uses an ultrasonic activity detection module to determine object activities in a larger range.

[0101] 5) The ultrasonic activity detection module uses the principle of Doppler frequency shift effect caused by moving objects on ultrasonic signals to detect object activities. Specifically, the ultrasonic activity detection module extracts Doppler frequency shift features from the multi-channel single-frequency signals transmitted by the ultrasonic preprocessing module, and then uses a pre-trained activity detection model to judge whether there are object activities within a given distance D2. The main steps are as follows:

[0102] Step1: Perform frequency domain transformation on the multi-channel single-frequency signal y″ sf (t), and extract the spectral energy within the range of 200 Hz above and below the frequency of each channel single-frequency signal as the basic feature; the frequency of the single-frequency signal is 18 kHz;

[0103] Step 2: Calculate the spectral energy difference between the first channel and other channels, and form a feature vector together with the basic features of the first channel. The channels actually correspond to the microphones used for signal acquisition, and the number of channels corresponds to the number of microphones. The more channels there are, the higher the algorithm accuracy. The first channel is designated artificially and only serves as a reference point without special requirements. In this invention, due to the adoption of a linear 4-microphone array, the number of channels is 4, and the first channel refers to the leftmost microphone.

[0104] Step 3: Send the above feature vector into a pre-trained activity detection model, and calculate whether there is an object activity within a given distance D2 at the current moment. If there is, output 1; otherwise, output 0. It should be noted that the activity detection model adopts a deep learning model and is distance-dependent. That is to say, this model is dependent on the distance D2. If the distance D2 changes, it is necessary to re-adjust the training data and re-train the model. The deep learning model is a deep residual network based on CNN, with 97 nodes in the input layer and 2 nodes in the output layer.

[0105] 6) The decision module makes a comprehensive judgment based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module to determine whether there is an object within a given distance D1 and whether there is an active object within a given distance D2. The main steps are as follows:

[0106] Step 1: If the distance of the active object output by the ultrasonic positioning module < D1, output that there is an object within the given distance D1 and there is an object activity within the given distance D2; otherwise, go to Step 2;

[0107] Step 2: If the ultrasonic activity detection module outputs that there is an object activity, output that there is no object within the given distance D1 and there is an object activity within the given distance D2; otherwise, output that there is no object within the given distance D1 and there is no object activity within the given distance D2.

[0108] An object activity perception device includes a processor, a memory, and a program. The program is stored in the memory, and the processor calls the program stored in the memory to execute an object activity perception method as described above.

[0109] A computer-readable storage medium is configured to store a program, and the program is configured to execute an object activity perception method as described above.

[0110] Embodiment 1

[0111] An object activity sensing system is deployed on a PC / laptop. After the system starts working, when a user sits approximately 0.6m in front of the PC / laptop, the screen automatically turns on. When there is no object activity within 1m in front of the PC / laptop for a preset duration (e.g., 5 minutes), the system sends a message to the operating system to turn off the screen. In this scenario, D1 = 0.6m and D2 = 1m.

[0112] Example 2

[0113] An object activity detection system is deployed on a large television screen. Once the system is operational, if someone (mainly a child) approaches the television and remains at close range for more than a preset time, the system will alert the user that they are too close and provide a distance value. If no one moves within a preset distance in front of the television screen within the preset time frame, a message is sent to the operating system to turn off the television screen. In this scenario, the distance threshold for child proximity detection is set to 1.5m, i.e., D1 = 1.5m; the effective distance threshold for object activity detection is set to 5m, i.e., D2 = 5m.

[0114] By utilizing the built-in hardware (speakers, microphones) of the large TV screen, the system can detect the presence of objects at different distances, ultimately enabling reminders for children viewing the screen at close range and automatic screen off / on functionality. It detects whether there are objects within a specific distance range, and further detects whether there is movement within that specific distance range. The former serves as a reminder for children viewing the screen at close range, while the latter enables the automatic screen off / on functionality.

[0115] Any parts or structures not specifically described in this invention can be made using existing technologies or products, and will not be elaborated upon here.

[0116] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. An object motion sensing method, characterized by, The object motion sensing system is used for sensing, and the object motion sensing system includes: An ultrasonic transmitting module is used to continuously transmit pre-designed ultrasonic signals. The ultrasonic receiving module is used to continuously receive and process sound wave signals propagating in space to obtain multi-channel ultrasonic signals. The ultrasonic preprocessing module is used to preprocess the received multi-channel ultrasonic signals to remove direct wave signals and signals reflected from stationary objects; The ultrasonic positioning module is used to calculate the motion trajectory of a moving object based on the signal received from the ultrasonic preprocessing module. The ultrasonic activity detection module is used to detect whether there are moving objects within a specific range based on the signals received from the ultrasonic preprocessing module. The decision module is used to make a comprehensive decision based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module, to determine whether there is an object within a given distance D1 and whether there is a moving object within a given distance D2, and D2>D1; The object activity sensing method includes the following steps: The ultrasonic transmitting module continuously transmits pre-set ultrasonic signals, and the ultrasonic receiving module continuously receives ultrasonic signals in the space. Then, the signals are sent to the ultrasonic preprocessing module for processing, and multi-channel sinusoidal signals and single-frequency signals are output. These signals are then sent to the ultrasonic positioning module and the ultrasonic activity detection module for processing, respectively. The ultrasonic positioning module estimates the position of the moving object and calculates its trajectory based on the received multi-channel sinusoidal wave signals, and outputs the result to the decision module; the ultrasonic activity detection module detects whether there is any moving object within a given distance range D2 based on the received multi-channel single-frequency signals, and outputs the result to the decision module. The decision module makes a comprehensive decision based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module, and outputs whether there is an object within a given distance D1 and whether there is a moving object within a given distance D2.

2. The method of claim 1, wherein, The ultrasonic signal emitted by the ultrasonic transmitting module is a combination of a single-frequency signal and a swept-frequency signal, with a frequency greater than or equal to 20kHz, and is emitted periodically.

3. The object activity sensing method according to claim 1, characterized in that, The processing steps of the ultrasound preprocessing module include: The ultrasonic preprocessing module preprocesses the received ultrasonic signals to remove direct wave signals and signals reflected from stationary objects. Then, the single-frequency signal and the swept-frequency signal are separated from each group of periodic signals; The difference frequency signal is obtained by multiplying the swept frequency signal by the corresponding transmitted signal. Low-pass filtering of the difference frequency signal yields a sinusoidal signal containing distance and azimuth information; Finally, multi-channel single-frequency signals and quasi-sine wave signals are output and transmitted to the ultrasonic activity detection module and ultrasonic positioning module, respectively.

4. The object activity sensing method according to claim 3, characterized in that, The ultrasonic preprocessing module preprocesses the received ultrasonic signals, removing direct wave signals and signals reflected from stationary objects. The steps include: The ultrasonic preprocessing module processes the received ultrasonic signals. Preprocessing is performed by removing the direct wave signal from the input signal using the following formula to obtain the signal. : in, The direct wave space transfer function can be measured in advance in a quiet room; Subsequently, background noise filtering, i.e., from the signal... The reflected wave signals from surrounding stationary objects are filtered to obtain the signal. : in: for The corresponding frequency domain signal, This represents the frequency domain signal of a wave reflected from a stationary object. This represents the inverse short-time Fourier transform.

5. The object movement sensing method according to claim 1, characterized in that, The steps of the ultrasonic positioning module to estimate the position of a moving object and track its motion trajectory based on the received multi-channel sinusoidal wave signals include: Step1: Process the multi-channel quasi-sine wave signal containing distance and azimuth information using the 2D-Music algorithm to obtain a two-dimensional distance-azimuth contour map; Search the current two-dimensional distance-azimuth contour map to find the distance and azimuth corresponding to the point with the maximum energy as the candidate for the target position at the current moment. Step2: If the maximum energy >= E1 and the distance between the target position at the current moment and the target position at the previous moment <= G1, output the candidate position and energy information of the target at the current moment; Otherwise, go to Step3. Step3: If the distance between the target position at the current moment and the target position at the previous moment > G1, it is considered that the target position estimation at the current moment is incorrect, and output the position of the target at the previous moment + offset O; Otherwise, go to Step4. Step4: If the maximum energy < E1 and the ultrasonic activity detection module outputs that there is no object activity at the current moment, it is considered that the target is stationary at the current moment, and output the target position at the previous moment.

6. The object movement sensing method according to claim 1, characterized in that, The steps for the ultrasonic activity detection module to detect whether there is object activity within a given distance range D2 based on the received multi-channel single-frequency signal include: Step1: Perform a frequency-domain transformation on the multi-channel single-frequency signal, and extract the spectral energy within the range of 200 Hz above and below each channel's single-frequency signal as the basic feature. Step2: Calculate the spectral energy difference between the first channel and other channels, and form a feature vector with the basic feature of the first channel. Step3: Send the above feature vector into a pre-trained activity detection model, and calculate whether there is object activity within the given distance D2 at the current moment. If there is, output 1; otherwise, output 0.

7. The object motion sensing method according to claim 1, characterized in that, The steps for the decision-making module to comprehensively make a decision based on the output results of the ultrasonic positioning module and the ultrasonic activity detection module and output whether there is an object within a given distance D1 and whether there is an active object within a given distance D2 include: Step1: If the distance of the active object output by the ultrasonic positioning module < D1, output that there is an object within the given distance D1 and there is an object activity within the given distance D2; Otherwise, go to Step2. Step2: If the ultrasonic activity detection module outputs that there is object activity, output that there is no object within the given distance D1 and there is an object activity within the given distance D2; Otherwise, output that there is no object within the given distance D1 and there is no object activity within the given distance D2.

8. An object motion sensing device, characterized in that, It includes a processor, a memory, and a program. The program is stored in the memory, and the processor calls the program stored in the memory to execute an object activity perception method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium is configured to store a program, and the program is configured to execute an object activity perception method according to any one of claims 1-7.

Images