Human body detection method and device based on millimeter wave radar

Abstract

The application discloses a human body detection method and device based on a millimeter wave radar, which comprises the following steps: acquiring detection data output by a millimeter wave radar module; the detection data at least comprises a current detection distance value and a current signal strength; obtaining a dynamic distance reference value according to the current detection distance value; determining a moving state of a detected object according to the current detection distance value, the dynamic distance reference value and a preset detection distance value; calculating an equivalent trigger number according to the moving state; and outputting a detection result as a human state if the equivalent trigger number is greater than a preset threshold value. The application can make the human detection equipment have an optimal response speed while keeping the accuracy of the detection result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar detection, in particular to a human body detection method and device based on millimeter wave radar. BACKGROUND

[0002] Millimeter wave radar refers to a technology or device that uses the Doppler effect of electromagnetic waves with a wavelength of 1-10mm to realize the detection and positioning of moving targets. If millimeter wave radar technology is applied in the field of unmanned detection, it will have the advantages of large detection range, high sensitivity, independence from environmental light, penetration of smoke and general shielding, digital quantization of detection results, and configuration of detection distance and sensitivity for different application scenarios.

[0003] The detection module of the millimeter wave radar can set appropriate detection distance and signal strength threshold according to the actual application scenario. When the detection signal strength is higher than the set signal strength threshold and the detection target is located within the set detection range, the detection module will output relevant detection information data through a serial port, including at least detection distance and signal strength.

[0004] Because the millimeter wave radar detection module has micro-motion recognition capability and high sensitivity, it may respond to the swing of non-human objects within the detection range or human activity at the edge of the detection range in the application, resulting in inaccurate human body detection results obtained by the millimeter wave radar.

[0005] Currently, there is no effective solution to the above problems. SUMMARY

[0006] The embodiments of the present application provide a human body detection method and device based on millimeter wave radar, which improves the accuracy of human body detection results based on millimeter wave radar.

[0007] According to an aspect of an embodiment of the present application, a human body detection method based on millimeter wave radar is provided, comprising:

[0008] Obtaining detection data output by a millimeter wave radar module; wherein the detection data at least includes a current detection distance value and a current signal strength, and the detection data is temporarily stored in a first-in-first-out cache queue;

[0009] Calculating a dynamic distance reference value according to the detection distance data items in the cache queue;

[0010] Filtering noise data according to the current detection distance, the dynamic distance reference value and the current signal strength;

[0011] Determining the movement state of the detected object according to the current detection distance value, the dynamic distance reference value and a preset detection distance value.

[0012] According to the movement state of the detected object, the equivalent trigger times are calculated and processed, and if the equivalent trigger times are greater than a preset threshold, it is determined that the detection result is a person state;

[0013] If there is no active detection data output from the millimeter wave radar module within a specified time, it is determined that the detection result is a no person state.

[0014] The dynamic distance reference value is calculated according to the detection distance data items in the cache queue, comprising:

[0015] The detection distance data items in the cache queue and the storage time corresponding to the detection distance data items are obtained;

[0016] According to the storage time of the detection distance data items, the corresponding weight coefficient is obtained;

[0017] The sum of the products of all detection distance data items in the cache queue and the weight coefficients is calculated to obtain a distance data sum;

[0018] The sum of the weight coefficients of all detection distance data items in the cache queue is calculated to obtain a weight coefficient sum;

[0019] The distance data sum is divided by the weight coefficient sum to obtain a dynamic distance reference value.

[0020] According to the current detection distance and the dynamic distance reference value and the current signal strength, the noise data is filtered, comprising:

[0021] If the current detection distance value and the dynamic distance reference value have a large jump difference or the current signal strength is less than a preset signal strength, it is determined that the current detection data is noise data, and the current detection data can be discarded.

[0022] The movement state of the detected object at least includes a close movement state, a far movement state, a peripheral activity state, an edge activity state, a general activity state and a no person activity state.

[0023] According to another aspect of the embodiment of the application, a human body detection device based on a millimeter wave radar is provided, at least comprising:

[0024] A microprocessor module, a millimeter wave radar module, an IO module or a network communication module, such as Figure 2 As shown.

[0025] The microprocessor module in the device reads the detection data from the millimeter wave radar module and runs the computer program to realize the human body detection method based on the millimeter wave radar in the first aspect.

[0026] The human body detection method based on the millimeter wave radar provided in the present application can be realized by the microprocessor module in the human body detection device based on the millimeter wave radar. The microprocessor stores a computer program that can be run. When the computer program is run, each step recorded in the above-mentioned method embodiments can be realized, for example, acquiring the detection data collected by the millimeter wave radar, calculating the dynamic distance reference value, filtering the noise data, determining the moving state of the measured object, calculating the equivalent trigger number, judging and outputting whether there is a person, etc.

[0027] The network module and the IO module in the device of the present application are used to output the human detection result of the device of the present application to external equipment or external systems, so as to realize the functional application and functional value of the device of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0028] The accompanying drawings, which are included to provide a further understanding of the present application and constitute a part of this application, illustrate embodiments of the present application and serve to explain the present application but do not limit the present application. In the drawings:

[0029] Figure 1 is a flowchart of an optional human body detection method based on a millimeter wave radar provided according to an embodiment of the present application;

[0030] Figure 2 is a structural schematic diagram of an optional human body detection device based on a millimeter wave radar provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should fall within the scope of protection of the present application.

[0032] The following will be described with reference to Figure 1 , Figure 1 is a flowchart of a human body detection method based on a millimeter wave radar provided according to an embodiment of the present application. It should be noted that the embodiments of the present application can be applied to any applicable scene.

[0033] Figure 1 The flow of the human body detection method based on the millimeter wave radar provided by the embodiment of the present application shown in the figure includes:

[0034] In step S101, detection data output by a millimeter wave radar module is acquired.

[0035] In the embodiment of the present application, the detection data at least includes a current detection distance value and a current signal strength, and the detection data is temporarily stored in a first-in-first-out cache queue.

[0036] In step S102, a dynamic distance reference value is calculated according to the detection distance value data items in the cache queue.

[0037] In step S102, the dynamic distance reference value is calculated in the following manner:

[0038] The detection distance data items in the cache queue and the time corresponding to the detection distance data items are obtained.

[0039] The corresponding weight coefficient is obtained according to the storage time of the detection distance data item.

[0040] The sum of the products of all detection distance data items in the cache queue and the weight coefficients is calculated to obtain a distance data sum.

[0041] The sum of the weight coefficients of all detection distance data items in the cache queue is calculated to obtain a weight coefficient sum.

[0042] The distance data sum is divided by the weight coefficient sum to obtain the dynamic distance reference value.

[0043] In step S103, noise data is filtered according to the current detection distance, the dynamic distance reference value and the current signal strength.

[0044] In step S103, the noise data is filtered in the following manner:

[0045] If the current detection distance value has a large jump compared to the dynamic distance reference value, or the current signal strength is less than a preset signal strength, the current detection data is determined to be noise data, and the current detection data is discarded.

[0046] If the difference between the current detection distance value and the dynamic distance reference value and the current signal strength are within a normal range, the current detection distance value is put into a data cache pool to establish a new dynamic distance reference value for the next noise filtering process.

[0047] Steps S101-S103 constitute a so-called noise filtering algorithm.

[0048] In step S104, the moving state of the detected object is determined according to the current detection distance value, the dynamic distance reference value and a preset detection range value.

[0049] In the embodiment of the present application, the moving state at least includes an approaching moving state, an away moving state, a peripheral activity state, an edge activity state, a general activity state and a no one activity state.

[0050] Step S105, the equivalent trigger times are calculated according to the moving state of the detected object, and if the equivalent trigger times are greater than a preset threshold, it is determined that the detection result is a person state.

[0051] Step S106, if there is no active detection data output from the millimeter wave radar module within a specified time, it is determined that the detection result is a no-person state.

[0052] Steps S103-S106 constitute a so-called moving behavior analysis algorithm.

[0053] The purpose of using the moving behavior analysis algorithm is to eliminate false outputs as much as possible. Although the noise filtering algorithm can effectively filter noise data, it cannot distinguish between the swing of non-human objects within the detection range and the detection data of human activity at the edge of the detection position, because it is the detection data generated by the detection module actually detecting object activity.

[0054] The processing of non-human object swing and detection data of human activity at the edge of the detection range, the traditional method is generally to accumulate the trigger times of the effective output of the detection module within a specified time, and when the accumulated trigger times reach a specified value, it is considered as a real person state, otherwise it is maintained as a no-person state. This method is simple, but if the set cumulative trigger times are small, false triggers may occur frequently, and if the set cumulative trigger times are high, the device response time will be longer, and even if the set cumulative trigger times are high, some continuous non-human object swing or detection of human activity at the edge of the detection range will still cause unexpected person state output.

[0055] The switching of the presence-absence detection device from the no-person state to the person state may serve as a trigger condition for the linkage response of some devices in the system (such as turning on the light or alarming), so it not only requires accuracy and reliability, but also requires fast response.

[0056] In order to eliminate such unexpected presence-absence state false output as much as possible while maintaining good device response speed, the present application performs moving behavior analysis processing on the detection data output by the detection module, and according to the moving behavior type corresponding to the trigger data, a different weight calculation method is given, so that the device can realize reliable and fast response.

[0057] Generally, during the switching process of the detection device from the no-person state to the person state, the human body moves from far to near, so this feature can be used to increase the incremental weight of the equivalent trigger times during this process.

[0058] For example, when the object movement is first detected, the equivalent trigger number can be increased by 1. If the current detection distance is continuously detected to be significantly less than the dynamic distance reference value, it can be considered that the detected object is in a continuous approaching movement behavior. The equivalent detection number can be increased by n or 2*n, 2^n based on the original value, where n is the approaching movement detection number. The greater the value of the approaching movement detection number n, the stronger the acceleration characteristics of the equivalent trigger number. When the approaching movement number n is 5, the actual accumulated equivalent detection number is 15, 21 or 31, respectively. Compared with the ordinary cumulative increase of 1, the response time is only about 1 / 3-1 / 6, effectively accelerating the response process of the device from no one state to someone state.

[0059] If the current detection distance is significantly greater than the dynamic distance reference value, it means that the detected object is in a moving away behavior. The accumulated equivalent trigger number can be decreased by m or 2*m, 2^m to prevent the device from triggering to someone state when the human body moves away. Here, m is the moving away detection number of the detected object.

[0060] If the current detection distance is significantly greater than the preset detection distance value, it is considered to be a peripheral activity behavior, and the equivalent detection trigger number is not calculated and the presence or absence of the state is not determined.

[0061] If the current detection distance has only a small change compared to the dynamic distance reference value and is near the preset detection distance edge, it means that the detection module detects human or object activity, which may be a small swing of the human body at the detection distance edge position (stationary activity, passing by or other behavior) or non-human objects (such as curtains, plants, etc.) in the indoor periphery. It cannot be determined whether a real person is moving in the detection range. In this case, this trigger data can be ignored, and the equivalent trigger number remains unchanged. However, the approaching movement detection number n can be decreased to reduce the acceleration sensitivity of subsequent approaching movement behavior and prevent accidental approaching movement behavior data from causing the detection device to accelerate into someone state trigger process. If the current detection distance value is within the detection range and not at the detection edge, it can be considered as a general activity behavior, and the equivalent detection trigger number can be increased by 1. If the approaching movement detection number n is insufficient, the equivalent detection trigger number can be further increased to ensure that the device can reliably enter someone state when a human body is moving in the detection range.

[0062] Sometimes non-human objects (such as curtains, plants, etc.) under the influence of air flow will produce a large amplitude swing, which may be misjudged as approaching movement and moving away, but at this time the approaching movement and moving away are basically reciprocating symmetric, the increments and decrements of the equivalent trigger times caused by them can basically offset each other, and the swing of such non-human objects is not normal, as long as it stops swinging for a short time, the equivalent trigger times will be quickly cleared, so the false trigger caused by the disturbance of such non-human objects can also be effectively inhibited in the present application.

[0063] When the detection module continuously does not detect human activity, the equivalent trigger times can be subjected to weight decreasing operation at certain interval period, for example, if the detection module does not detect detection data for 0.2 seconds, the equivalent trigger times are subjected to m or 2*m, 2^m operation, until the equivalent trigger times are zero, where m is the number of unattended state accumulation, which can share a variable with the number of moving away accumulation.

[0064] The number of unattended state accumulation or the number of moving away accumulation m and the number of approaching movement accumulation n are in mutual exclusive relationship, that is, when m is a non-zero value, n must be zero, and when n is a non-zero value, m must be zero.

[0065] When there is no human detection device in the presence of a person, the human body in the detection range may sometimes be in a relatively static state for a short time and unable to detect human activity, in which case the device cannot be simply switched to an unattended state, otherwise it is easy to make the device frequently switch between the presence and unattended states, thereby affecting the use experience of the device. In this case, a continuous unattended time parameter can be set as a judgment condition, and only when the continuous unattended time is greater than this value, the device will really switch from the presence to the unattended state, and once the detection module outputs new presence activity trigger data within this time, the continuous unattended time will be reset to zero.

[0066] The noise filtering algorithm described in the present application has the following characteristics:

[0067] Compare the current detection distance value with the distance reference value, if there is a large amplitude jump difference or the current signal strength is weak, it can be determined as noise data and discarded;

[0068] The movement behavior analysis algorithm described in the present application has the following characteristics:

[0069] According to the relationship among the current detection distance value, the dynamic distance reference value and the detection edge distance value, the current detection data can be classified into approaching movement, moving away, peripheral activity, edge activity, general activity and unattended activity.

[0070] Then according to the current detection data belongs to the active behavior type is given different weights to the equivalent trigger times increase and decrease calculation, and according to the value of the equivalent trigger times determine the output of the state of no one.

[0071] The present application can not only effectively suppress noise data, but also effectively suppress the false trigger caused by the disturbance of some non-human objects and human edge activities after adopting noise filtering algorithm and mobile behavior analysis algorithm, so that the presence-absence detection device can realize the accuracy of detection results and has better response speed.

[0072] Finally, it should be noted that: the above-described embodiments, only for the specific embodiments of the present application, in order to illustrate the technical solutions of the present application, rather than limit it, the scope of protection of the present application is not limited to this, although the foregoing detailed description of the present application, those skilled in the art should be understood: any familiar with the technical field of the technical personnel in the technical range of the present application disclosed, it still can be modified or easily thought of changes to the technical solution recorded in the foregoing examples, or part of the technical features of the equivalent replacement, and these modifications, changes or replacement, without the corresponding technical solutions of the essence of the present application deviate from the spirit and scope of the technical solutions of the embodiment, all should be covered in the scope of protection of the present application. Therefore, the scope of protection of the present application should be said to the scope of protection of the claims.

[0073] In addition, although the operations of the method of the present application are described in a particular order in the drawings, this does not require or imply that the operations must be performed in that particular order, or that all of the illustrated operations must be performed to achieve the desired result. Additionally or alternatively, certain steps can be omitted, combined into a single step, and / or split into multiple steps.

Claims

1. A human body detection method based on millimeter wave radar, comprising: acquiring detection data output by a millimeter wave radar module; wherein the detection data at least includes a current detection distance value and a current signal strength, and the detection data is temporarily stored in a first-in-first-out cache queue; calculating a dynamic distance reference value according to the detection distance data items in the cache queue; filtering noise data according to the current detection distance, the dynamic distance reference value, and the current signal strength; determining a moving state of a detected object according to the current detection distance value, the dynamic distance reference value, and a preset detection distance value; calculating and processing an equivalent trigger number according to the moving state of the detected object, and determining that the detection result is a person state if the equivalent trigger number is greater than a preset threshold value; determining that the detection result is a no person state if the millimeter wave radar module has not output active detection data within a specified time; wherein the calculation of the dynamic distance reference value according to the detection distance data items in the cache queue comprises: acquiring the detection distance data items in the cache queue and the storage time corresponding to the detection distance data items; obtaining a corresponding weight coefficient according to the storage time of the detection distance data items; calculating the sum of the product of all detection distance data items in the cache queue and the weight coefficient to obtain a distance data sum; calculating the sum of the weight coefficients of all detection distance data items in the cache queue to obtain a weight coefficient sum; dividing the distance data sum by the weight coefficient sum to obtain the dynamic distance reference value; and during the switching process of the detection result from the no person state to the person state, the detected object moves from far to near, and the increment weight of the equivalent trigger number in this process is increased; wherein: when the detected object is first detected to move, the equivalent trigger number is increased by 1; If the current detection distance is continuously detected to be obviously smaller than the dynamic distance reference value, it is determined that the continuously approaching movement behavior of the detected object, and the equivalent detection times are increased by n or 2×n on the basis of the original value, the operation of n, which is the approaching movement detection times. If the current detection distance is obviously greater than the dynamic distance reference value, it is determined that the detected object is moving away, and the accumulated equivalent trigger times are decremented by m or 2×m, , m is the number of times of detecting the moving away of the detected object. if the current detection distance is significantly greater than the preset detection distance value, it is determined as a peripheral activity behavior, and the calculation of the equivalent detection trigger number and the determination of the presence or absence of a person state are not performed; if the current detection distance has only a small change compared with the dynamic distance reference value and is near the edge of the preset detection distance, it is confirmed that the detected object is detected to move in the detection distance edge position or to have a small swing of the detected object in the indoor periphery, and this trigger data is ignored, so that the equivalent trigger number remains unchanged, and the approaching movement detection number n is decremented to reduce the acceleration sensitive characteristics of the subsequent approaching movement behavior, so as to prevent the detection device from accelerating into the person state trigger process due to the data of the occasional approaching movement behavior. 2.The human body detection method based on millimeter wave radar according to claim 1, wherein the filtering of noise data according to the current detection distance, the dynamic distance reference value, and the current signal strength comprises: if there is a large jump difference between the current detection distance value and the dynamic distance reference value or the current signal strength is less than a preset signal strength, it is determined that the detection data is noise data, and the current detection data is discarded.

3. The method of human body detection based on millimeter wave radar according to claim 1, wherein the moving state of the detected object comprises at least approaching moving state, moving away moving state, peripheral activity state, edge activity state, general activity state and no activity state.

4. A device of human body detection based on millimeter wave radar, applied to the method of human body detection based on millimeter wave radar according to claim 1, comprising at least: microprocessor module, millimeter wave detection module, IO module or network communication module.

5. The device of human body detection based on millimeter wave radar according to claim 4, wherein the microprocessor module is used for: reading the detection data from the millimeter wave radar module and running the program for realizing the method of human body detection based on millimeter wave radar according to claim 1.

6. The device of human body detection based on millimeter wave radar according to claim 4, wherein the IO module and the network module are used for: outputting the human detection result to the external device or system. ​ ​ ​

Images