An ultrasound-based human body detection method and device

Abstract

This invention discloses an ultrasound-based human body detection method and apparatus. The method includes: controlling a smart terminal to emit ultrasound signals to the surrounding environment, collecting and storing a certain number of ultrasound observations; updating the variance of the ultrasound observations in real time, and retaining the minimum variance of the ultrasound observations within a set time period; setting a Kalman filter threshold based on the minimum variance; filtering the variance of the ultrasound observations using Kalman filtering based on the Kalman filter threshold to obtain the filtered variance of the ultrasound observations; determining whether the smart terminal has detected a human body based on the filtered variance of the ultrasound observations, and controlling the working state of the smart terminal based on the detection result. This invention utilizes ultrasound signals transmitted by a smart device for human body detection and uses Kalman filtering to filter environmental noise in the ultrasound signals, thereby reducing the cost of human body detection and improving its accuracy.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of human body sensing technology, and in particular to a human body detection method and device based on ultrasound. Background Technology

[0002] In smart device applications, to improve battery life, human presence detection devices are typically added. These devices detect the presence of a human on the smart device and control its operation based on the detection results.

[0003] In existing technologies, infrared detection technology is commonly used for human body perception detection. However, this traditional infrared sensing often requires the installation of multiple devices at multiple angles, resulting in high detection costs. Furthermore, the angle and accuracy of infrared detection are limited. Summary of the Invention

[0004] To address the problems in the prior art, this invention provides an ultrasound-based human body detection method and device to improve the accuracy and sensitivity of human body detection by intelligent devices and reduce the cost of human body detection.

[0005] In a first aspect, the present invention provides an ultrasound-based human body detection method, comprising:

[0006] S1. Control the intelligent terminal to emit ultrasonic signals to the surroundings, collect and retain a certain number of ultrasonic observation values;

[0007] S2. Update the variance of the ultrasound observations in real time, retain the minimum variance of the ultrasound observations within a set time period, and set the threshold of the Kalman filter based on the minimum variance.

[0008] S3. Based on the threshold of the Kalman filter, the variance of the ultrasonic observation values ​​is filtered using Kalman filtering to obtain the variance of the ultrasonic observation values ​​after filtering.

[0009] S4. Determine whether the smart terminal has detected a human body based on the variance of the filtered ultrasound observations, and control the working state of the smart terminal based on the detection results.

[0010] Optionally, the threshold for the Kalman filter is set based on the minimum variance, including:

[0011] The product of the minimum variance and a first preset multiple is used as the threshold for Kalman filtering.

[0012] Optionally, the value range of the first preset multiple is 10-20 times.

[0013] Optionally, S3 includes:

[0014] S31. Compare the variance of the ultrasonic observations with the threshold of the Kalman filter, and set the value of the process noise covariance in the Kalman filter based on the comparison result;

[0015] S32. Based on the set process covariance value, perform Kalman filtering on the variance of the ultrasonic observation values ​​to obtain the filtered variance of the ultrasonic observation values.

[0016] Optionally, S31 includes:

[0017] When the variance of the ultrasonic observations is less than the threshold of the Kalman filter, the process noise covariance in the Kalman filter is set as the first process noise covariance.

[0018] When the variance of the ultrasonic observation is greater than or equal to the threshold of the Kalman filter, the process noise covariance in the Kalman filter is set as the second process noise covariance.

[0019] The noise covariance of the first process is smaller than that of the noise covariance of the second process.

[0020] Optionally, S4 includes:

[0021] Calculate the mean of the minimum absolute differences of the ultrasound observations, and use the product of the mean of the minimum absolute differences and a second preset multiple as the detection threshold;

[0022] The variance of the filtered ultrasound observations is compared with the detection threshold. If the variance of the filtered ultrasound observations is greater than or equal to the detection threshold, it is determined that the smart terminal has detected a human body and the smart terminal is controlled to be in working state.

[0023] Otherwise, it is determined that the smart terminal has not detected a human body and the smart terminal is controlled to be in a non-working state.

[0024] Optionally, the value range of the second preset multiple is 20-30 times.

[0025] Secondly, the present invention also provides an ultrasound-based human body detection device, comprising:

[0026] The signal transmission and acquisition module is used to control the intelligent terminal to transmit ultrasonic signals to the surroundings, and to collect and retain a certain number of ultrasonic observation values.

[0027] The threshold setting module for Kalman filtering is used to update the variance of the ultrasound observations in real time, retain the minimum variance of the ultrasound observations within a set time period, and set the threshold of Kalman filtering based on the minimum variance.

[0028] The filtering module is used to filter the variance of the ultrasound observations using Kalman filtering based on the threshold of the Kalman filter, so as to obtain the variance of the filtered ultrasound observations.

[0029] The human body detection module is used to determine whether the smart terminal has detected a human body based on the variance of the filtered ultrasound observation values, and to control the working state of the smart terminal based on the detection results.

[0030] Optionally, the filtering module is specifically used for:

[0031] The variance of the ultrasonic observations is compared with the threshold of the Kalman filter, and the value of the process noise covariance in the Kalman filter is set according to the comparison result.

[0032] Based on the set process covariance value, the variance of the ultrasonic observations is subjected to Kalman filtering to obtain the variance of the filtered ultrasonic observations.

[0033] Optionally, the human body detection module is specifically used for:

[0034] Calculate the mean of the minimum absolute differences of the ultrasound observations, and use the product of the mean of the minimum absolute differences and a second preset multiple as the detection threshold;

[0035] The variance of the filtered ultrasound observations is compared with the detection threshold. If the variance of the filtered ultrasound observations is greater than or equal to the detection threshold, it is determined that the smart terminal has detected a human body and the smart terminal is controlled to be in working state.

[0036] Otherwise, it is determined that the smart terminal has not detected a human body and the smart terminal is controlled to be in a non-working state.

[0037] This invention reduces the cost of human body detection by utilizing the ultrasonic transceiver device of the smart terminal itself. It sets the threshold of Kalman filtering by using the variance of the observed values ​​and filters the variance of the ultrasonic observed values ​​using Kalman filtering. Then, it determines whether the smart terminal has detected a human body based on the variance of the filtered ultrasonic observed values. Through filtering, the accuracy of ultrasonic human body detection is further improved.

[0038] Furthermore, compared to mainstream infrared human body sensing methods, ultrasound offers higher sensitivity and better cost-effectiveness. It provides more accurate detection in dark environments, even in low light, and can detect minute human movements such as head tilting. Ultrasound is essentially a sound wave, a mechanical wave, and therefore poses no electromagnetic radiation hazard. Ultrasound also has a wider detection range. Traditional infrared sensing often requires installation of multiple devices at multiple angles, while ultrasound offers a broader detection range, reducing costs and simplifying logical operations. Attached Figure Description

[0039] Figure 1 A flowchart illustrating an ultrasound-based human body detection method provided in an embodiment of the present invention;

[0040] Figure 2 This is a schematic diagram of human body detection before filtering provided in an embodiment of the present invention;

[0041] Figure 3 A schematic diagram of the variance before and after filtering provided in an embodiment of the present invention;

[0042] Figure 4 This is a schematic diagram of the human body detection results before filtering, provided in an embodiment of the present invention. Detailed Implementation

[0043] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0044] Example

[0045] Figure 1 A flowchart of an ultrasound-based human body detection method provided in this embodiment of the invention specifically includes the following steps:

[0046] S1. Control the intelligent terminal to emit ultrasonic signals to the surrounding area, and collect and retain a certain number of ultrasonic observation values.

[0047] In this embodiment, the smart device can include any electronic device with an ultrasonic transmitting module and an ultrasonic receiving module. For example, the smart device can be a whole-house smart device such as lighting, switch sensing, television, or smart screen.

[0048] Specifically, the ultrasonic transmitting module can be a built-in speaker of the smart device, and the ultrasonic receiving module can be a built-in microphone of the smart device.

[0049] For example, a smart device can be controlled to emit a 22kHz ultrasonic signal through a speaker, and then a microphone can be used to collect the ultrasonic signal emitted by the speaker and store a certain number of ultrasonic observations. Due to the Doppler effect of sound waves, that is, the frequency of sound waves will change when they encounter a moving object, the change in the received ultrasonic signal can be used to determine whether the smart device has detected a human body.

[0050] See further Figure 2 , Figure 2 This is a schematic diagram of human body detection before filtering, provided in an embodiment of the present invention.

[0051] S2. Update the variance of the ultrasound observations in real time, retain the minimum variance of the ultrasound observations within a set time period, and set the threshold of the Kalman filter based on the minimum variance.

[0052] In this embodiment, the ultrasound observations are first stored in a fixed-size memory pool. As the memory pool is updated and overflows, the variance (Var) and mean (μ) of the ultrasound observations in the memory pool are continuously calculated. Typically, when the smart device detects a human body, the ultrasound observations will change drastically, and the variance of the memory pool will increase accordingly. When the smart device does not detect a human body, the ultrasound observations will not change drastically, and even with significant environmental noise, the variance remains relatively stable.

[0053] However, during the continuous updating of variance, human detection based on changes in observations may introduce errors due to the influence of environmental noise, making it difficult to define the thresholds for whether a person is present or not. Therefore, this embodiment employs Kalman filtering to filter the variance in order to eliminate the influence of environmental noise on the detection results.

[0054] Kalman filtering is an algorithm that uses the state equations of a linear system to make an optimal estimate of the system state based on ultrasonic observations of the system's input and output. Its function is filtering, aiming to reduce the impact of noise and interference on data measurements.

[0055] The Kalman equation used in this embodiment is as follows:

[0056] (1) Prediction section:

[0057] The system covariance at time K is equal to the system covariance at time k-1 plus the process noise covariance, i.e.:

[0058]

[0059] (2) Kalman gain equation:

[0060] Kalman gain = System estimate covariance at time k / (System estimate covariance at time k + Observation noise covariance), that is:

[0061]

[0062] (3) Update the optimal solution:

[0063] The optimal value of the state at time k = the predicted value of the state + Kalman gain * (measured value - predicted value of the state), that is:

[0064] P k =P k-1 +K k (z k -Pk-1 )

[0065] P k-1 =P k

[0066]

[0067] Specifically, during the variance update process, it is necessary to retain the minimum variance value Var over a certain period of time. min During this period, the variance of Var is reduced by a certain multiple of the minimum variance during this period. min *K var Set the threshold for Kalman filtering.

[0068] Optional, K var The setting range is 10-20 times, for the following reasons:

[0069] Variance represents the dispersion of data. Variance data in unoccupied environments is relatively stable, mainly concentrated around a single value, with small variance. Environmental noise or pedestrian movement in the distance only slightly increases the data fluctuation; this fluctuation can be encompassed by using a factor of 10-20 of the minimum variance. By using K... var Within the above range, the variance after Kalman filtering can filter out detected environmental noise, such as... Figure 3 Line segment 1 represents the original variance data before filtering, and line segment 2 represents the variance data after filtering.

[0070] S3. Based on the threshold of the Kalman filter, the variance of the ultrasonic observations is filtered using Kalman filtering to obtain the variance of the ultrasonic observations after filtering.

[0071] Specifically, S3 includes:

[0072] S31. Compare the variance of the ultrasonic observations with the threshold of the Kalman filter, and set the value of the process noise covariance in the Kalman filter based on the comparison result.

[0073] Wherein, when the variance Var of the ultrasonic observations is less than the threshold Var of the Kalman filter. min *K var The process noise covariance in the Kalman filter is set as the first process noise covariance;

[0074] When the variance Var of the ultrasound observations is greater than or equal to the threshold Var of the Kalman filter min *K var The process noise covariance in the Kalman filter is set as the second process noise covariance;

[0075] The noise covariance of the first process is smaller than that of the noise covariance of the second process.

[0076] S32. Based on the set process covariance value, perform Kalman filtering on the variance of the ultrasonic observation values ​​to obtain the filtered variance of the ultrasonic observation values.

[0077] Specifically, when Var < Var min *K var The noise covariance of the first process is set to a very small value, and the observed value z is... k Subtracting the mean μ and then performing a Kalman filter will result in a very small output value, effectively filtering out environmental errors; when Var ≥ Var min *K var Set the noise covariance of the second process to a very large value, and input the observed value z. k Performing Kalman filtering will result in a large output value, which can effectively preserve the data of detected human bodies.

[0078] According to the Kalman filter formula used in this embodiment, the prediction and gain components of the Kalman filter are mainly calculated using the observation noise covariance R and the process noise covariance Q. The observation noise covariance R is set to a fixed value of 100. If Var < Var min *K var If the process noise covariance Q is set to 0.01, the gain calculation process will place greater trust in the fixed value of the observation noise covariance R, resulting in a very small final filtered value; if Var ≥ Var min *K var By setting the process noise covariance Q to 100, the calculation of the gain is based more heavily on the process noise covariance Q, resulting in a larger final filtered value. Generally, the second process noise covariance is set to be 10,000 times that of the first process noise covariance, making it easier to distinguish between manned and unmanned states.

[0079] See the figure for further details. The blue curve represents the variance, and the black curve represents the variance after Kalman filtering. It is clear that when no one is around, the Kalman filtering algorithm filters out environmental noise.

[0080] S4. Determine whether the smart terminal has detected a human body based on the variance of the filtered ultrasound observations, and control the working state of the smart terminal based on the detection results.

[0081] Specifically, the mean of the minimum absolute differences of the ultrasound observations is calculated, and the product of the mean of the minimum absolute differences and a second preset multiple is used as the detection threshold Dev. min *K dev ;

[0082] The variance of the filtered ultrasound observations is compared with the detection threshold. If the variance of the filtered ultrasound observations is greater than or equal to the detection threshold, it is determined that the smart terminal has detected a human body and the smart terminal is controlled to be in working state.

[0083] Otherwise, it is determined that the smart terminal has not detected a human body and the smart terminal is controlled to be in a non-working state.

[0084] The second preset multiplier mentioned above is 20-30 times. The reason for choosing this multiplier range is as follows: the mean absolute deviation represents the degree of dispersion of the data around the mean. By continuously updating the mean minimum absolute deviation, the mean minimum absolute deviation is relatively small in a quiet environment and larger in a noisy environment. After Kalman filtering, the data can distinguish between a person in front of the computer and an empty room by using 20-30 times the mean minimum absolute deviation.

[0085] See details Figure 4 Curve 3 represents a certain multiple of the minimum mean square error Dev. min *K dev .

[0086] For example, when the aforementioned smart device is a TV or a smart screen, the method of this embodiment determines whether someone is watching on the TV / smart screen. When no human body is detected, the TV / smart screen will activate the screen-off energy-saving mode to save energy.

[0087] Furthermore, due to varying distances, the data fluctuations also differ. Fluctuations are larger when the distance between the human body and the smart device is short, and smaller when the distance is long. At excessively large distances, the detected data resembles environmental noise. Therefore, a threshold (Var) can be set for the Kalman filter. min *K var To set the computer's sensing range. Multiplier K var The larger the setting, the smaller the machine's sensing distance; multiple K var The smaller the threshold value, the larger the machine's sensing range. Therefore, this embodiment of the invention can adapt to different environments, autonomously set different filtering thresholds to determine the presence of people, and can adjust the threshold value by setting different multiples K. var This allows for setting the detection distance of the machine, further expanding the application scenarios of the embodiments of the present invention.

[0088] This invention reduces the cost of human body detection by utilizing the ultrasonic transceiver device of the smart terminal itself to transmit ultrasonic signals; it sets the threshold of Kalman filtering by using the variance of the observed values, and uses Kalman filtering to filter the variance of the ultrasonic observed values, and then determines whether the smart terminal has detected a human body based on the variance of the filtered ultrasonic observed values. Through filtering, the accuracy of ultrasonic human body detection is further improved.

[0089] Furthermore, compared to mainstream infrared human body sensing, ultrasound offers higher sensitivity and better cost-effectiveness. It provides more accurate detection in dark environments, even at night, and can detect minute human movements such as head tilting. Ultrasound is essentially a sound wave, a mechanical wave, and therefore poses no electromagnetic radiation hazard. Ultrasound also has a wider detection range. Traditional infrared sensing often requires installation of multiple devices at multiple angles, while ultrasound offers a broader detection range, reducing costs and simplifying logical operations.

[0090] Furthermore, embodiments of the present invention also provide an ultrasound-based human body detection device, the device comprising:

[0091] The signal transmission and acquisition module is used to control the intelligent terminal to transmit ultrasonic signals to the surroundings, and to collect and retain a certain number of ultrasonic observation values.

[0092] The threshold setting module for Kalman filtering is used to update the variance of the ultrasound observations in real time, retain the minimum variance of the ultrasound observations within a set time period, and set the threshold of Kalman filtering based on the minimum variance.

[0093] The filtering module is used to filter the variance of the ultrasound observations using Kalman filtering based on the threshold of the Kalman filter, so as to obtain the variance of the filtered ultrasound observations.

[0094] The human body detection module is used to determine whether the smart terminal has detected a human body based on the variance of the filtered ultrasound observation values, and to control the working state of the smart terminal based on the detection results.

[0095] The process of setting the threshold for the Kalman filter based on the minimum variance includes:

[0096] The product of the minimum variance and a first preset multiple is used as the threshold for Kalman filtering.

[0097] The first preset multiple ranges from 10 to 20 times.

[0098] Furthermore, the filtering module is specifically used for:

[0099] The variance of the ultrasonic observations is compared with the threshold of the Kalman filter, and the value of the process noise covariance in the Kalman filter is set according to the comparison result.

[0100] Based on the set process covariance value, the variance of the ultrasonic observations is subjected to Kalman filtering to obtain the variance of the filtered ultrasonic observations.

[0101] Optionally, the variance of the ultrasonic observations is compared with the threshold of the Kalman filter, and the value of the process noise covariance in the Kalman filter is set according to the comparison result, including:

[0102] When the variance of the ultrasonic observation is less than the threshold of the Kalman filter, the process noise covariance in the Kalman filter is set as the first process noise covariance.

[0103] When the variance of the ultrasonic observation is greater than or equal to the threshold of the Kalman filter, the process noise covariance in the Kalman filter is set as the second process noise covariance.

[0104] The noise covariance of the first process is smaller than that of the noise covariance of the second process.

[0105] Furthermore, the human body detection module is specifically used for:

[0106] Calculate the mean of the minimum absolute differences of the ultrasound observations, and use the product of the mean of the minimum absolute differences and a second preset multiple as the detection threshold;

[0107] The variance of the filtered ultrasound observations is compared with the detection threshold. If the variance of the filtered ultrasound observations is greater than or equal to the detection threshold, it is determined that the smart terminal has detected a human body and the smart terminal is controlled to be in working state.

[0108] Otherwise, it is determined that the smart terminal has not detected a human body and the smart terminal is controlled to be in a non-working state.

[0109] The second preset multiple ranges from 20 to 30 times.

[0110] The ultrasound-based human body detection device provided in this embodiment of the invention can execute the ultrasound-based human body detection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method, which will not be described in detail here.

[0111] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. An ultrasonic-based human detection method, characterized by, The method comprises the following steps: S1, controlling the intelligent terminal to emit ultrasonic signals to the surroundings, collecting and storing a certain number of ultrasonic observation values; S2, updating the variance of the ultrasonic observation values in real time, retaining the minimum value of the variance of the ultrasonic observation values within a set time period, and setting the threshold value of Kalman filtering according to the minimum value; S3, filtering the variance of the ultrasonic observation values by Kalman filtering according to the threshold value of Kalman filtering to obtain the filtered variance of the ultrasonic observation values; S4, judging whether the intelligent terminal detects a human body according to the filtered variance of the ultrasonic observation values, and controlling the working state of the intelligent terminal according to the detection result; The S3 comprises: S31, comparing the variance of the ultrasonic observation values with the threshold value of Kalman filtering, and setting the value of the process noise covariance in Kalman filtering according to the comparison result; S32, filtering the variance of the ultrasonic observation values by Kalman filtering according to the set value of the process covariance to obtain the filtered variance of the ultrasonic observation values; The S31 comprises: When the variance of the ultrasonic observation values is smaller than the threshold value of Kalman filtering, the process noise covariance in Kalman filtering is set as a first process noise covariance; When the variance of the ultrasonic observation values is greater than or equal to the threshold value of Kalman filtering, the process noise covariance in Kalman filtering is set as a second process noise covariance; The first process noise covariance is smaller than the second process noise covariance.

2. The method of claim 1, wherein, Setting the threshold value of Kalman filtering according to the minimum value comprises: Taking the product of the minimum value and a first preset multiple as the threshold value of Kalman filtering.

3. The method of claim 2, wherein, The value range of the first preset multiple is 10-20 times.

4. The method of claim 1, wherein, The S4 comprises: Calculating the mean of the minimum absolute difference values of the ultrasonic observation values, and taking the product of the mean and a second preset multiple as a detection threshold value; Comparing the filtered variance of the ultrasonic observation values with the detection threshold value, if the filtered variance of the ultrasonic observation values is greater than or equal to the detection threshold value, it is judged that the intelligent terminal detects a human body and the intelligent terminal is controlled to be in a working state; Otherwise, it is judged that the intelligent terminal does not detect a human body and the intelligent terminal is controlled to be in a non-working state.

5. The method of claim 4, wherein, The value range of the second preset multiple is 20-30 times.

6. An ultrasonic-based human detection apparatus, characterized by comprising: The method comprises the following steps: A signal emission and collection module is configured to control the intelligent terminal to emit ultrasonic signals to the surroundings, collect and store a certain number of ultrasonic observation values; A threshold value setting module of Kalman filtering is configured to update the variance of the ultrasonic observation values in real time, retain the minimum value of the variance of the ultrasonic observation values within a set time period, and set the threshold value of Kalman filtering according to the minimum value; A filtering module is configured to filter the variance of the ultrasonic observation values by Kalman filtering according to the threshold value of Kalman filtering to obtain the filtered variance of the ultrasonic observation values; The variance of the ultrasonic observation values is compared with the threshold value of Kalman filtering, and the value of the process noise covariance in Kalman filtering is set according to the comparison result; setting the process noise covariance in the Kalman filtering as a first process noise covariance when the variance of the ultrasonic observation value is less than a threshold value of the Kalman filtering; setting the process noise covariance in the Kalman filtering as a second process noise covariance when the variance of the ultrasonic observation value is greater than or equal to the threshold value of the Kalman filtering; the first process noise covariance is less than the second process noise covariance; performing Kalman filtering on the variance of the ultrasonic observation value according to the set value of the process covariance to obtain a variance of a filtered ultrasonic observation value; the human body detection module is specifically configured to:

7. The apparatus of claim 6, wherein, calculating a mean of minimum absolute difference values of the ultrasonic observation value, and taking a product of the mean of minimum absolute difference values and a second preset multiple as a detection threshold value; comparing the variance of the filtered ultrasonic observation value with the detection threshold value, and if the variance of the filtered ultrasonic observation value is greater than or equal to the detection threshold value, determining that the intelligent terminal detects a human body and controlling the intelligent terminal to be in a working state; otherwise, determining that the intelligent terminal does not detect a human body and controlling the intelligent terminal to be in a non-working state. ​

Images