Human body orientation and action form recognition method, system and equipment and medium

Abstract

The invention discloses a human body orientation and action form identification method, system and device and a medium, and relates to the technical field of human body orientation and action form identification, and the method comprises the steps: obtaining analog signals outputted by a sensor under different conditions; converting the preprocessed analog signals into waveform signals, and analyzing each waveform signal in a time domain and a frequency domain to obtain first waveform data corresponding to each waveform signal; training the preset initial model until the preset initial model reaches a preset training ending condition; acquiring second waveform data corresponding to the to-be-detected signal, and inputting the second waveform data into the identification model for processing to obtain a target distance and a target form of the to-be-detected signal; by means of the mode, while action form and distance recognition is achieved, the cost and power consumption of the system are greatly reduced, and the problems that an existing video image form recognition mode is high in cost, complex in algorithm, difficult to maintain in the later period and the like are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of human body position and motion form recognition, and more specifically, to a human body position and motion form recognition method, system, equipment and medium. Background Art

[0002] Human beings are the main body of social development and progress, and the position and motion morphology detection technology targeting the human body has naturally become a hot topic in related research fields. Human position and morphology recognition is widely used in the field of intelligent control. The recognition methods include ultrasonic detection, wireless radio frequency detection, active infrared detection, etc. These methods are to detect the position or motion information of the human body by transmitting a certain signal and detecting the signal received by the receiving device. At present, video images are the most widely used for morphological recognition. The advantage of human detection through image videos is that the amount of information contained in the image is huge, so the detection accuracy is particularly high, and there are many ways to perform subsequent information processing; but this method has the problem of huge data volume. On the one hand, when there is a demand for a single aspect, there will be too much useless information, which makes the signal processing algorithm complicated and increases the cost and power consumption of the system. On the other hand, the use of the overall equipment also faces the problems of high cost, complex algorithm and difficult later maintenance. Summary of the invention

[0003] The purpose of the present invention is to provide a method, system, device and medium for recognizing human body position and motion form, so as to solve the problems existing in the above-mentioned background technology.

[0004] The above technical objectives of the present invention are achieved through the following technical solutions:

[0005] In a first aspect, the present application provides a method for recognizing human body position and motion form, comprising the following specific steps:

[0006] Based on the preset pyroelectric infrared sensor, the analog signal output by the sensor is obtained under different distances and different human body shapes;

[0007] Preprocessing each analog signal, converting the preprocessed analog signal into a waveform signal, and analyzing each waveform signal in the time domain and the frequency domain to obtain first waveform data corresponding to each waveform signal;

[0008] The preset initial model is trained using each first waveform data under different distances and different human body shapes until the preset initial model reaches a preset training end condition, and the preset initial model that reaches the preset training end condition is determined as a recognition model;

[0009] The signal to be measured output by the pyroelectric infrared sensor within the monitoring range is obtained, and the second waveform data corresponding to the signal to be measured is obtained. The second waveform data is input into the recognition model for processing to obtain the target distance and target shape of the signal to be measured.

[0010] The beneficial effects of the present invention are as follows: in this scheme, firstly, the analog signal output by the pyroelectric infrared sensor is obtained under different distances and different human body shapes. The human body shape can include relatively basic morphological movements such as running, stepping, squatting, jumping, waving, kicking, etc., and then the analog signals of different distances corresponding to these movements are converted into waveform signals, and the waveform signals are analyzed in the time domain and frequency domain. Through this analysis process, deeper features at various distances and various shapes can be analyzed from the waveform signals. By first extracting the features, the learning of the model can also be made more efficient, and the features of each movement extracted are the first waveform data; secondly, the first waveform data at different distances and different shapes are used to extract the basic The preset initial model of deep learning is trained to enable the preset initial model to learn and recognize the subtle differences between these features, and when the preset initial model reaches the preset training end condition, the preset initial model that reaches the preset training end condition is determined as the recognition model; finally, in actual use, the test signal output within the monitoring range is obtained by the pyroelectric infrared sensor, and the test signal is subjected to the above-mentioned signal conversion and analysis process to obtain the second waveform data of the test signal, and the second waveform data is the feature extracted from the test signal; the second waveform data is then input into the recognition model for processing, so that the target distance and target shape corresponding to the test signal can be identified, thereby completing the recognition of the human body orientation and movement shape.

[0011] In this solution, the distance and motion form recognition method provided is compared to the most widely used video image form recognition method. When only some simple actions need to be recognized, such as running, stepping, squatting, jumping, waving, kicking, etc., this method can realize motion form and distance recognition while greatly reducing the system cost and power consumption, and effectively solves the problems of high cost, complex algorithm, and difficult post-maintenance faced by the current video image form recognition method.

[0012] Based on the above technical solution, the present invention can also be improved as follows.

[0013] Furthermore, for each waveform signal, the waveform signal is analyzed in the time domain, specifically:

[0014] According to the preset time step, the waveform signal is subjected to first-order difference processing to obtain the first-order difference sequence corresponding to the time series of each time step in the waveform signal;

[0015] Each first-order difference sequence is smoothed and filtered, and the first-order difference sequence after smoothing and filtering is detected for zero crossing to obtain the signal peak value, wave crest value and waveform frequency corresponding to each time series. The first waveform data includes the signal peak value, wave crest value and waveform frequency.

[0016] The beneficial effect of adopting the above further scheme is: in the time domain, by performing first-order difference processing on the waveform signal, characteristic data such as signal peak value, peak value and waveform frequency can be obtained from the first-order difference sequence after smoothing filtering, and it is accurate to each time step, so that the waveform characteristics in each waveform signal can be displayed more completely.

[0017] Furthermore, the above first-order difference sequence is specifically:

[0018] b(j)=(a(j+1)-a(j-1)) / 2, where: 1<j<n;

[0019] Where a(j+1) represents the time series of the j+1th time step, a(j-1) represents the time series of the j-1th time step, b(j) represents the first-order difference sequence of the time series a(j), and n represents the total number of time series in the waveform signal.

[0020] Furthermore, for each waveform signal, the waveform signal is analyzed in the frequency domain, specifically:

[0021] The waveform signal is subjected to windowing and framing processing using a preset window function to obtain an analysis frame corresponding to each window in the waveform signal;

[0022] For each analysis frame, a short-time Fourier transform is performed on the analysis frame to obtain a target spectrum corresponding to the analysis frame, and a high-frequency component in the target spectrum is determined as a high-frequency signal feature. The first waveform data includes the high-frequency signal feature.

[0023] The beneficial effect of adopting the above further scheme is: in the frequency domain, by windowing and framing the waveform signal and performing short-time Fourier transform on each analysis frame, the characteristics of the waveform signal in the frequency domain can be displayed more completely, and the high-frequency component is selected as one of the first waveform data, so that the waveform characteristics of the waveform signal are more recognizable.

[0024] Furthermore, the above short-time Fourier transform is specifically:

[0025] Where: 0≤k≤N-1;

[0026] Where, X n(k) represents the analysis frame after short-time Fourier transform, w(nm) represents the preset window function, x(m)w(nm) represents each analysis frame, n represents the time index, N represents the number of sampling points, and k represents the frequency index.

[0027] Furthermore, the above preprocessing includes analog signal filtering processing, analog signal amplification processing and analog signal conversion processing.

[0028] In a second aspect, the present application provides a human body position and motion form recognition system, which is applied to a human body position and motion form recognition method according to any one of the first aspects, comprising:

[0029] The first module is used to obtain the analog signal output by the preset pyroelectric infrared sensor under different distances and different human body shapes;

[0030] The second module is used to preprocess each analog signal, convert the preprocessed analog signal into a waveform signal, and analyze each waveform signal in the time domain and the frequency domain to obtain first waveform data corresponding to each waveform signal;

[0031] The third module is used to train the preset initial model using the first waveform data at different distances and different human body shapes until the preset initial model reaches a preset training end condition, and determine the preset initial model that reaches the preset training end condition as the recognition model;

[0032] The fourth module is used to obtain the signal to be measured output by the pyroelectric infrared sensor within the monitoring range, obtain the second waveform data corresponding to the signal to be measured, and input the second waveform data into the recognition model for processing to obtain the target distance and target shape of the signal to be measured.

[0033] Furthermore, the second module includes:

[0034] The first submodule is used to perform first-order difference processing on the waveform signal according to a preset time step to obtain a first-order difference sequence corresponding to the time series of each time step in the waveform signal;

[0035] The second submodule is used to perform smoothing and filtering processing on each first-order difference sequence, and perform zero-crossing detection on the first-order difference sequence after smoothing and filtering, to obtain the signal peak value, wave crest value and waveform frequency corresponding to each time series, and the first waveform data includes the signal peak value, wave crest value and waveform frequency;

[0036] The third submodule is used to perform windowing and framing processing on the waveform signal using a preset window function to obtain an analysis frame corresponding to each window in the waveform signal;

[0037] The fourth submodule is used to perform short-time Fourier transform on each analysis frame to obtain a target spectrum corresponding to the analysis frame, and determine the high-frequency components in the target spectrum as high-frequency signal features, and the first waveform data includes the high-frequency signal features.

[0038] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the methods in the first aspect when executing the computer program.

[0039] In a fourth aspect, the present application provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions enable a computer to execute any one of the methods in the first aspect.

[0040] Compared with the prior art, the present invention has at least the following beneficial effects:

[0041] In the present application, firstly, the analog signal output by the pyroelectric infrared sensor is obtained under different distances and different human body shapes. The human body shape can include relatively basic morphological movements such as running, stepping, squatting, jumping, waving, kicking, etc., and then the analog signals of different distances corresponding to these movements are converted into waveform signals, and the waveform signals are analyzed in the time domain and frequency domain. Through this analysis process, deeper features at various distances and various shapes can be analyzed from the waveform signal. By extracting the features first, the learning of the model can also be more efficient, and the extracted features of each action are the first waveform data; secondly, the first waveform data at different distances and different shapes are used to perform a deep learning-based The preset initial model is trained to enable the preset initial model to learn and recognize the subtle differences between these features, and when the preset initial model reaches the preset training end condition, the preset initial model that reaches the preset training end condition is determined as the recognition model; finally, in actual use, the test signal output within the monitoring range is obtained by the pyroelectric infrared sensor, and the test signal is subjected to the above-mentioned signal conversion and analysis process to obtain the second waveform data of the test signal, and the second waveform data is the feature extracted from the test signal; the second waveform data is then input into the recognition model for processing, and the target distance and target shape corresponding to the test signal can be identified, thereby completing the recognition of the human body orientation and movement shape.

[0042] In the present application, the distance and motion form recognition method provided is compared to the most widely used video image form recognition method. When only some simple actions need to be recognized, such as running, stepping, squatting, jumping, waving, kicking, etc., this method can realize motion form and distance recognition while greatly reducing the system cost and power consumption, and effectively solves the problems of high cost, complex algorithm, and difficult post-maintenance faced by the current video image form recognition method.

[0043] In the present application, in the time domain, by performing first-order difference processing on the waveform signal, characteristic data such as signal peak value, peak value and waveform frequency can be obtained from the first-order difference sequence after smoothing filtering, and accurate to each time step, so that the waveform characteristics in each waveform signal can be displayed more completely; in the frequency domain, by performing windowing and framing processing on the waveform signal and performing short-time Fourier transform on each analysis frame, the characteristics of the waveform signal in the frequency domain can be displayed more completely, and the high-frequency component is selected as one of the first waveform data, so that the waveform characteristics of the waveform signal are more recognizable. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0045] Figure 1 A method flow chart of an identification method in an embodiment of the present invention;

[0046] Figure 2 A connection diagram of an identification system in an embodiment of the present invention;

[0047] Figure 3 Schematic diagram of the connection of electronic equipment in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0049] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0051] In the description of the embodiments of the present invention, "plurality" means at least 2.

[0052] Embodiment 1:

[0053] In order to solve the problems that the current video image morphology recognition method has a huge amount of data, and when facing a single demand, there is too much useless information, which makes the signal processing algorithm complicated, and the current video image morphology recognition method has high cost, complex algorithm, and difficulty in later maintenance, this embodiment provides a human body position and action morphology recognition method, such as Figure 1 As shown, the specific steps include:

[0054] S1, based on the preset pyroelectric infrared sensor, obtains the analog signal output by the sensor under different distances and different human body shapes.

[0055] Among them, the principle of pyroelectric infrared sensor to detect the human body is based on the interaction between the pyroelectric effect and infrared radiation; when the human body emits infrared rays, the pyroelectric sensor can detect this radiation and convert it into electrical signals; that is, the constant body temperature of the human body (generally around 37°C) will emit infrared rays of a specific wavelength (about 10μm), which are enhanced by the Fresnel filter and gathered to the pyroelectric element. When the pyroelectric element receives infrared radiation from the human body, its surface charge will change, resulting in charge imbalance and release of charges. These released charges can output voltage signals after being processed by the amplifier.

[0056] Specifically, by performing different action forms at different distances, the analog signals output at different distances and different actions can be obtained through the pyroelectric infrared sensor. Further processing of these analog signals can extract the signal characteristics of each analog signal, and using these signal characteristics as identification can realize the recognition of distance and form.

[0057] S2, preprocessing each analog signal, converting the preprocessed analog signal into a waveform signal, and analyzing each waveform signal in the time domain and the frequency domain to obtain the first waveform data corresponding to each waveform signal. The above preprocessing includes analog signal filtering processing, analog signal amplification processing and analog signal conversion processing. After the conversion processing, the analog signal can become a digital signal, and the digital signal can be converted into a waveform through an oscilloscope.

[0058] Optionally, for each waveform signal, the waveform signal is analyzed in the time domain, specifically:

[0059] S21, performing first-order difference processing on the waveform signal according to a preset time step, to obtain a first-order difference sequence corresponding to the time series of each time step in the waveform signal.

[0060] Optionally, the first-order difference sequence is specifically:

[0061] b(j)=(a(j+1)-a(j-1)) / 2, where: 1<j<n;

[0062] Where a(j+1) represents the time series of the j+1th time step, a(j-1) represents the time series of the j-1th time step, b(j) represents the first-order difference sequence of the time series a(j), and n represents the total number of time series in the waveform signal.

[0063] S22, perform smoothing filtering on each first-order difference sequence, and perform zero-crossing detection on the first-order difference sequence after smoothing filtering to obtain the signal peak value, peak value and waveform frequency corresponding to each time series. The first waveform data includes the signal peak value, peak value and waveform frequency, and may also include periodicity, etc.

[0064] Among them, in the time domain, by performing first-order difference processing on the waveform signal, characteristic data such as signal peak value, peak value and waveform frequency can be obtained from the first-order difference sequence after smoothing filtering, and it is accurate to each time step, so that the waveform characteristics in each waveform signal can be displayed more completely.

[0065] Optionally, for each waveform signal, the waveform signal is analyzed in the frequency domain, specifically:

[0066] S23, performing windowing and framing processing on the waveform signal using a preset window function to obtain an analysis frame corresponding to each window in the waveform signal; wherein the preset window function may be a Hanning window.

[0067] S24, for each analysis frame, perform short-time Fourier transform on the analysis frame to obtain a target spectrum corresponding to the analysis frame, and determine the high-frequency components in the target spectrum as high-frequency signal features, and the first waveform data includes the high-frequency signal features.

[0068] Optionally, the short-time Fourier transform is specifically:

[0069] Where: 0≤k≤N-1;

[0070] Where, X n(k) represents the analysis frame after short-time Fourier transform, w(nm) represents the preset window function, x(m)w(nm) represents each analysis frame, n represents the time index, N represents the number of sampling points, and k represents the frequency index.

[0071] Among them, in the frequency domain, by windowing and framing the waveform signal and performing short-time Fourier transform on each analysis frame, the characteristics of the waveform signal in the frequency domain can be displayed more completely, and the high-frequency component is selected as one of the first waveform data to make the waveform characteristics of the waveform signal more recognizable.

[0072] S3, using each first waveform data at different distances and different human body shapes to train the preset initial model until the preset initial model reaches a preset training end condition, and determining the preset initial model that reaches the preset training end condition as the recognition model.

[0073] Among them, during the training process, the loss function after each input is calculated through the input, output and true distance form of the initial model, until the loss function for multiple consecutive times does not exceed the threshold, or reaches the set maximum number of iterations, thereby completing the training of the preset initial model, that is, obtaining the recognition model corresponding to the preset initial model.

[0074] S4, obtaining the signal to be measured output by the pyroelectric infrared sensor within the monitoring range, obtaining second waveform data corresponding to the signal to be measured, and inputting the second waveform data into the recognition model for processing to obtain the target distance and target shape of the signal to be measured.

[0075] Specifically, firstly, the analog signal output by the pyroelectric infrared sensor is obtained under different distances and different human body shapes. The human body shape can include relatively basic morphological movements such as running, stepping, squatting, jumping, waving, kicking, etc., and then the analog signals of different distances corresponding to these movements are converted into waveform signals, and the waveform signals are analyzed in the time domain and frequency domain. Through this analysis process, deeper features of each distance and each shape can be analyzed from the waveform signal. By extracting the features first, the learning of the model can be made more efficient, and the extracted features of each movement are the first waveform data; secondly, the first waveform data at different distances and different shapes are used to perform deep learning-based The preset initial model is trained so that the preset initial model can learn and recognize the subtle differences between these features, and when the preset initial model reaches the preset training end condition, the preset initial model that reaches the preset training end condition is determined as the recognition model; finally, in actual use, the test signal output within the monitoring range is obtained by the pyroelectric infrared sensor, and the test signal is subjected to the above-mentioned signal conversion and analysis process to obtain the second waveform data of the test signal, and the second waveform data is the feature extracted from the test signal; the second waveform data is then input into the recognition model for processing, so that the target distance and target shape corresponding to the test signal can be identified, thereby completing the recognition of the human body orientation and movement shape.

[0076] In actual use, information is exchanged through a pre-set communication method to complete a certain functional task. Simple morphological movements and distance recognition are performed in this way, which can be used for waving to open and close doors, automatically turning on and off lights, etc. in daily life.

[0077] Embodiment 2:

[0078] The present application embodiment provides a human body position and motion form recognition system, which is applied to a human body position and motion form recognition method in any one of the embodiments 1, such as Figure 2 As shown, including:

[0079] The first module is used to obtain the analog signal output by the preset pyroelectric infrared sensor under different distances and different human body shapes.

[0080] The second module is used to preprocess each analog signal, convert the preprocessed analog signal into a waveform signal, and analyze each waveform signal in the time domain and the frequency domain to obtain the first waveform data corresponding to each waveform signal.

[0081] Optionally, the second module may include:

[0082] The first submodule is used to perform first-order difference processing on the waveform signal according to a preset time step to obtain a first-order difference sequence corresponding to the time series of each time step in the waveform signal.

[0083] The second submodule is used to perform smoothing filtering on each first-order difference sequence, and perform zero-crossing detection on the first-order difference sequence after smoothing filtering to obtain the signal peak value, peak value and waveform frequency corresponding to each time series. The first waveform data includes the signal peak value, peak value and waveform frequency.

[0084] The third submodule is used to perform windowing and framing processing on the waveform signal using a preset window function to obtain an analysis frame corresponding to each window in the waveform signal.

[0085] The fourth submodule is used to perform short-time Fourier transform on each analysis frame to obtain a target spectrum corresponding to the analysis frame, and determine the high-frequency components in the target spectrum as high-frequency signal features, and the first waveform data includes the high-frequency signal features.

[0086] The third module is used to train the preset initial model using the first waveform data at different distances and different human body shapes until the preset initial model reaches the preset training end condition, and the preset initial model that reaches the preset training end condition is determined as the recognition model.

[0087] The fourth module is used to obtain the signal to be measured output by the pyroelectric infrared sensor within the monitoring range, obtain the second waveform data corresponding to the signal to be measured, and input the second waveform data into the recognition model for processing to obtain the target distance and target shape of the signal to be measured.

[0088] Embodiment 3:

[0089] The present application embodiment provides an electronic device, such as Figure 3 As shown, it includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, any method in Embodiment 3 is implemented.

[0090] Embodiment 4:

[0091] An embodiment of the present application provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions enable a computer to execute any method in Example 1.

[0092] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for recognizing human body position and motion form, characterized in that: The specific steps include: Based on the preset pyroelectric infrared sensor, the analog signal output by the sensor is obtained under different distances and different human body shapes; Preprocessing each of the analog signals, converting the preprocessed analog signals into waveform signals, and analyzing each of the waveform signals in the time domain and the frequency domain to obtain first waveform data corresponding to each of the waveform signals; Using each of the first waveform data at different distances and different human body shapes to train a preset initial model until the preset initial model reaches a preset training end condition, and determining the preset initial model that reaches the preset training end condition as a recognition model; The signal to be measured output by the pyroelectric infrared sensor within the monitoring range is obtained, and the second waveform data corresponding to the signal to be measured is obtained. The second waveform data is input into the recognition model for processing to obtain the target distance and target shape of the signal to be measured.

2. A method for recognizing human body position and motion form according to claim 1, characterized in that: For each of the waveform signals, the waveform signal is analyzed in the time domain, specifically: According to a preset time step, the waveform signal is subjected to first-order difference processing to obtain a first-order difference sequence corresponding to the time series of each time step in the waveform signal; Each of the first-order difference sequences is subjected to smoothing filtering, and zero-crossing detection is performed on the first-order difference sequences after smoothing filtering to obtain the signal peak value, wave crest value and waveform frequency corresponding to each time series, wherein the first waveform data includes the signal peak value, wave crest value and waveform frequency.

3. A method for recognizing human body position and motion form according to claim 2, characterized in that: The first-order difference sequence is specifically: b(j)=(a(j+1)-a(j-1)) / 2, where: 1<j<n; Where a(j+1) represents the time series of the j+1th time step, a(j-1) represents the time series of the j-1th time step, b(j) represents the first-order difference sequence of the time series a(j), and n represents the total number of time series in the waveform signal.

4. A method for recognizing human body position and motion form according to claim 1, characterized in that: For each of the waveform signals, the waveform signal is analyzed in the frequency domain, specifically: Performing windowing and framing processing on the waveform signal using a preset window function to obtain an analysis frame corresponding to each window in the waveform signal; For each of the analysis frames, a short-time Fourier transform is performed on the analysis frame to obtain a target spectrum corresponding to the analysis frame, and a high-frequency component in the target spectrum is determined as a high-frequency signal feature, and the first waveform data includes the high-frequency signal feature.

5. A method for recognizing human body position and motion form according to claim 4, characterized in that: The short-time Fourier transform is specifically: Where: 0≤k≤N-1; In the formula, X n (k) represents the analysis frame after short-time Fourier transform, w(nm) represents the preset window function, x(m)w(nm) represents each analysis frame, n represents the time index, N represents the number of sampling points, and k represents the frequency index.

6. A method for recognizing human body position and motion form according to claim 1, characterized in that: The preprocessing includes analog signal filtering processing, analog signal amplification processing and analog signal conversion processing.

7. A human body position and motion form recognition system, applied to a human body position and motion form recognition method according to any one of claims 1 to 7, characterized in that: include: The first module is used to obtain the analog signal output by the preset pyroelectric infrared sensor under different distances and different human body shapes; The second module is used to preprocess each of the analog signals, convert the preprocessed analog signals into waveform signals, and analyze each of the waveform signals in the time domain and the frequency domain to obtain first waveform data corresponding to each of the waveform signals; The third module is used to train the preset initial model using each of the first waveform data at different distances and different human body shapes until the preset initial model reaches a preset training end condition, and determine the preset initial model that reaches the preset training end condition as the recognition model; The fourth module is used to obtain the signal to be measured output by the pyroelectric infrared sensor within the monitoring range, obtain the second waveform data corresponding to the signal to be measured, and input the second waveform data into the recognition model for processing to obtain the target distance and target shape of the signal to be measured.

8. A human body position and motion form recognition system according to claim 7, characterized in that: The second module includes: The first submodule is used to perform first-order difference processing on the waveform signal according to a preset time step to obtain a first-order difference sequence corresponding to the time series of each time step in the waveform signal; The second submodule is used to perform smoothing filtering on each of the first-order difference sequences, and perform zero-crossing detection on the first-order difference sequences after smoothing filtering to obtain the signal peak value, wave crest value and waveform frequency corresponding to each time series, wherein the first waveform data includes the signal peak value, wave crest value and waveform frequency; The third submodule is used to perform windowing and framing processing on the waveform signal using a preset window function to obtain an analysis frame corresponding to each window in the waveform signal; The fourth submodule is used to perform short-time Fourier transform on each analysis frame to obtain a target spectrum corresponding to the analysis frame, and determine the high-frequency components in the target spectrum as high-frequency signal features, and the first waveform data includes high-frequency signal features.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

10. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, which enable a computer to execute the method of any one of claims 1-6.

Images