Millimeter wave radar-based method for discriminating entrance and exit of people and monitoring number of people in room

Abstract

The invention discloses a millimeter-wave radar-based method for judging whether people go out or go into a house and monitoring the number of people in the house, and the method comprises the following steps: S1, measuring the length of a door of a current room, and setting the size of a buffer area according to whether a left wall or a right wall exists after people enter the door; s2, after the buffer area is set, the millimeter wave radar starts to work, and the radar processes the collected point cloud data to identify and track a target; and S3, judging whether the target passes through the buffer area or not according to the track of the target, and completing the judgment of entering and exiting of the personnel. The millimeter-wave radar-based method for discriminating the entrance and exit of people and monitoring the number of people indoors is based on the millimeter-wave radar, personal image data does not need to be collected or processed, and personal privacy information of detected people can be effectively protected.

Description

Technical Field

[0001] The present invention relates to the field of monitoring technology, and in particular to a method for identifying people entering and leaving a house and monitoring the number of people indoors based on millimeter wave radar. Background Art

[0002] Existing indoor personnel entry and exit monitoring technologies are mainly divided into image-based technologies and non-image-based technologies. Image-based technologies use video surveillance and image processing to collect video data through cameras, combine artificial intelligence algorithms to identify target features, and determine the entry and exit status of the target based on its movement trajectory. This type of technology has high recognition accuracy, but has high requirements for external conditions such as light and environment, and has high deployment and operation costs. In addition, image technology involves video acquisition and processing, which may lead to insufficient privacy protection for monitored personnel.

[0003] Non-image technology uses various sensors (such as infrared sensors, laser sensors, pressure sensors, etc.) to collect status information and determine the entry and exit status of people. For example, an invisible detection area is formed by an infrared counter-radiation sensor. When a person passes through, the infrared signal is blocked, and the system records an entry and exit event. Non-image technology is relatively low-cost, privacy-friendly, and suitable for the needs of some specific scenarios. However, this method has limited recognition capabilities when multiple people pass by at the same time, which may lead to missed or miscounted counts. It is also difficult to distinguish individual identities and has certain accuracy limitations.

[0004] Existing algorithms for identifying entry and exit usually use a relatively simple identification method, such as the common "three-line" identification method. Fig.10 As shown. This method realizes the discrimination of people entering and exiting by setting three crossing lines. However, this discrimination method is difficult to ensure sufficient accuracy in practical applications. The reason is that millimeter-wave radar may be blocked by obstacles such as doors in actual environments, causing the radar wave to temporarily lose tracking of the target. When the target leaves the blocking range of the door, the radar wave will re-identify the target, which may lead to misjudgment of the discrimination result. In addition, due to the limited monitoring range of radar waves, if the "three-line" discrimination method is adopted, there will be problems with monitoring blind spots. When the target exceeds the edge of the monitoring range, the radar wave may fail to capture the movement of the target in time, and thus misjudge that no target passes. In this case, the traditional discrimination method cannot effectively deal with the complex occlusion and monitoring blind spot problems in the environment. Therefore, it is necessary to provide a method for distinguishing people entering and exiting a house and monitoring the number of people indoors based on millimeter-wave radar. Summary of the invention

[0005] In order to overcome the defects in the prior art, a method for identifying people entering and leaving a house and monitoring the number of people indoors is provided based on millimeter-wave radar.

[0006] The present invention is achieved through the following scheme:

[0007] A method for identifying people entering and leaving a house and monitoring the number of people indoors based on millimeter wave radar, the method comprising the following steps:

[0008] S1. Measure the length of the current room door and set the buffer size according to whether there is a left wall or a right wall after entering the door;

[0009] S2. After the buffer is set, the millimeter wave radar starts working, and the radar processes the collected point cloud data to identify and track the target;

[0010] S3. Determine whether the target has passed through the buffer zone based on the target's trajectory, and complete the identification of personnel entering and exiting.

[0011] The step S1 is specifically as follows:

[0012] S11. Check the material of the room door. If the room door is a wooden door, set the length of the buffer zone to 1.1 times the door length; if it is a metal door, set the length of the buffer zone to 1.3 times the door length;

[0013] S12, check whether there are left and right walls after entering the door. If there is a left wall but no right wall, set it to left wall lock; if there is no left wall but only a right wall, set it to right wall lock; if there are both, set it to full lock; if there are neither, set it to no lock;

[0014] S13. The width of the buffer zone is set according to the wall. The left and right walls are the boundaries of the buffer zone. If the left and right walls do not exist at the same time, the width and length of the buffer zone are the same.

[0015] The step S2 is specifically as follows:

[0016] S21. The millimeter-wave radar acquires point cloud data of the surrounding environment by transmitting millimeter-wave signals and using the echo signal information. Each frame of point cloud data contains three-dimensional information of a large number of reflection points. To simplify processing and analysis, these three-dimensional point cloud data are projected into a two-dimensional polar coordinate system. The polar coordinate system takes the radar as the origin, and each point is represented by its distance and azimuth from the radar.

[0017] S22. For multi-frame point cloud data, a density clustering algorithm is used to cluster the reflection points in the point cloud. By setting reasonable distance thresholds and density parameters, DBSCAN can adaptively identify multiple targets at different distances and noise levels. After clustering, a unique ID is assigned to each target to achieve independent tracking of the target in subsequent steps.

[0018] S23. In order to track the target's motion trajectory in real time, the Kalman filter algorithm is used to estimate and predict the motion state of each target; the algorithm will update the target state according to the point cloud data returned by the radar, so as to maintain accurate tracking of the target trajectory under noise interference; at the same time, combined with the target ID, the Kalman filter can effectively realize the independent tracking of multiple targets and ensure the consistency and accuracy of the trajectory information.

[0019] The echo signal information includes time delay, intensity and Doppler frequency shift.

[0020] The three-dimensional information includes distance, angle and speed.

[0021] The step S3 is specifically as follows:

[0022] S31. For the scene of entering indoor space, the radar first detects that the target is outside the door. After the target enters the buffer zone, it is first determined whether the target trajectory passes through the left and right judgment lines of the buffer zone. If so, the number of people is increased by 1. If not, it is further determined whether the target trajectory passes through the left and right judgment lines or the final judgment line of the indoor space. If so, the number of people is increased by 1.

[0023] S32. For the scene of leaving the indoor area, the radar first detects that the target moves from the indoor area to the buffer zone at the door. If the target trajectory passes through the left and right judgment lines of the buffer zone, the number of people is directly reduced by 1; if the target first passes through the left and right judgment lines of the indoor area or the final judgment line of the indoor area, and then passes through the initial judgment line of the indoor area, the number of people is also reduced by 1.

[0024] The millimeter-wave radar is installed ceiling-mounted and located on the same side as the door. The specific installation height should be between 2.5m and 3m from the ground.

[0025] The radar wave emission angle is maintained at approximately 45 degrees to the horizontal plane, and the maximum monitoring range can reach 6 meters.

[0026] The beneficial effects of the present invention are:

[0027] 1. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter-wave radar is based on millimeter-wave radar, does not require the collection or processing of personal image data, and can effectively protect the personal privacy information of the detected person.

[0028] 2. Due to the high precision, high resolution and multi-functional detection capabilities of millimeter wave radar waves, the application scenarios of the present invention are more extensive and can work effectively in scenes with low light intensity and multiple targets.

[0029] 3. The buffer zone design of this application can effectively reduce the interference of the surrounding environment, effectively limit the reflection interference of the door, and significantly improve the accuracy of monitoring.

[0030] 4. Through precise target state management, buffer zone design and efficient detection capability of millimeter wave radar, the present invention provides a safe and reliable solution for indoor number monitoring and personnel entry and exit identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a flow chart of the algorithm for distinguishing a person entering a room based on millimeter-wave radar according to the present invention;

[0032] Figure 2 This is a flow chart of the algorithm for distinguishing a person leaving a room based on millimeter-wave radar according to the present invention;

[0033] Figure 3 A diagram illustrating the installation of a millimeter wave radar provided in an embodiment of the present invention;

[0034] Figure 4 An illustration of directly entering a room from a buffer zone provided by an embodiment of the present invention;

[0035] Figure 5 An illustration of a method of passing through a buffer zone and entering a room from a left-right discrimination line provided by an embodiment of the invention;

[0036] Figure 6 An illustration of passing through a buffer zone and entering a room from a final indoor determination line provided by an embodiment of the invention;

[0037] Figure 7 An illustration of an embodiment of the invention providing a method of leaving the indoor final decision line through the buffer zone;

[0038] Figure 8 An illustration of an embodiment of the invention providing a method of leaving a room through a buffer zone from a left-right discrimination line in the room;

[0039] Fig. 9 An illustration of leaving the buffer zone directly from the room provided for an embodiment of the invention;

[0040] Fig.10 It is an illustration of the existing entry and exit identification method. DETAILED DESCRIPTION

[0041] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments:

[0042] In this application, the density clustering algorithm is referred to as DBSCAN, Density-Based Spatial Clustering of Applications with Noise. DBSCAN can classify points that are close to each other and have dense distribution into the same cluster based on the distribution density of points in the point cloud, while effectively ignoring discrete points or noise points; Kalman filtering is an efficient recursive algorithm that combines the current observation value and historical motion data of the target to smooth the target's position information and predict its future position.

[0043] Millimeter-wave radar detects the position, speed and direction of the target by emitting high-frequency millimeter-wave signals and using its short wavelength and high resolution. The signal emitted by the radar will be reflected back to the receiving end when encountering the target. The target distance is calculated by measuring the time delay between the transmitted and received signals, the target speed is estimated by analyzing the frequency offset using the Doppler effect, and the target angle is determined by the phase difference or array antenna. Millimeter-wave radar has high-precision, high-resolution and multi-functional detection capabilities, and has good environmental adaptability. It can work stably under complex conditions such as rain, fog, and dust. It is widely used in fields such as autonomous driving, security monitoring, and industrial detection. It is an ideal choice for intelligent detection. In view of its many advantages, the present invention proposes an indoor number monitoring and personnel entry and exit discrimination method based on millimeter-wave radar. Based on the above problem, the present application effectively solves the above problem by setting a buffer zone and a discrimination zone.

[0044] The millimeter-wave radar of this application is installed on the ceiling, on the same side as the door, and the specific installation height should be between 2.5m and 3m from the ground. The angle of radar wave emission needs to be maintained at about 45 degrees to the horizontal plane, and the maximum monitoring range can reach 6 meters. However, due to interference factors in the indoor environment, the accuracy of radar monitoring may be significantly affected, leading to misjudgment during the detection process, especially when there are objects moving indoors. To address this problem, the algorithm introduces a buffer design to reduce the impact of interference on the monitoring results. The size of the buffer zone is dynamically set according to the opening and closing width of the room door to ensure that the opening and closing range of the door is always within the buffer zone, thereby effectively improving the reliability and accuracy of monitoring.

[0045] like Figure 1 , 2 As shown, the method for indoor number monitoring and personnel entry and exit based on millimeter wave radar disclosed in the present invention includes the following steps:

[0046] S1. Measure the length of the current room door and set the buffer size according to whether there is a left wall or a right wall after entering the door.

[0047] S2. After setting the buffer zone, the millimeter wave radar starts working. The radar processes the collected point cloud data to identify and track the target.

[0048] S3. Determine whether the target has passed through the buffer zone based on the target's trajectory, and complete the identification of personnel entering and exiting.

[0049] Step S1 is specifically as follows:

[0050] S11. Check the material of the room door. If it is a wooden door, set the length of the buffer zone to about 1.1 times the door length; if it is a metal door, set the length of the buffer zone to about 1.3 times the door length.

[0051] S12. Check whether there are left and right walls after entering the door. If there is a left wall but no right wall, set it to left wall lock; if there is a right wall but no left wall, set it to right wall lock; if there are both, set it to full lock; if there are neither, set it to no lock.

[0052] S13. The width of the buffer zone is also set according to the wall. The left and right walls are the boundaries of the buffer zone. If the left and right walls do not exist at the same time, the width and length of the buffer zone are the same.

[0053] Step S2 is specifically as follows:

[0054] S21. Millimeter-wave radar acquires point cloud data of the surrounding environment by transmitting millimeter-wave signals and using information such as the time delay, intensity, and Doppler shift of the echo signal. Each frame of point cloud data contains three-dimensional information of a large number of reflection points, including characteristics such as distance, angle, and speed. To simplify processing and analysis, these three-dimensional point cloud data are projected into a two-dimensional polar coordinate system. The polar coordinate system takes the radar as the origin, and each point is represented by its distance (radius) and azimuth (angle) from the radar.

[0055] S22. For multi-frame point cloud data, a density clustering algorithm (DBSCAN, Density-Based Spatial Clustering of Applications with Noise) is used to cluster the reflection points in the point cloud. DBSCAN can classify points that are close to each other and have dense distribution into the same cluster based on the distribution density of points in the point cloud, while effectively ignoring discrete points or noise points. By setting reasonable distance thresholds and density parameters, DBSCAN can adaptively identify multiple targets at different distances and noise levels. After clustering is completed, a unique ID is assigned to each target to enable independent tracking of the target in subsequent steps.

[0056] S23. In order to track the target's motion trajectory in real time, the Kalman filter algorithm is used to estimate and predict the motion state of each target. Kalman filtering is an efficient recursive algorithm that combines the target's current observation value and historical motion data to smooth the target's position information and predict its future position. The algorithm updates the target state (prediction-observation-correction) based on the point cloud data returned by the radar, thereby maintaining accurate tracking of the target's trajectory under noise interference. At the same time, combined with the target's ID, Kalman filtering can effectively achieve independent tracking of multiple targets and ensure the consistency and accuracy of trajectory information.

[0057] Step S3 is specifically as follows:

[0058] S31. For entering indoor scenes, Figure 4 As shown, the radar first detects that the target is outside the door. After the target enters the buffer zone, it first determines whether the target trajectory passes through the left and right judgment lines of the buffer zone. If it does, the number of people increases by 1; if it does not, it continues to determine whether the target trajectory passes through the left and right judgment lines or the final judgment line in the room. If it does, the number of people increases by 1.

[0059] S32, for leaving the indoor scene, such as Figure 7 As shown, the radar first detects that the target moves from the indoor area to the buffer zone at the door. If the target trajectory passes through the left and right judgment lines of the buffer zone, the number of people is directly reduced by 1; if the target first passes through the left and right judgment lines of the indoor area or the final judgment line of the indoor area, and then passes through the initial judgment line of the indoor area, the number of people is also reduced by 1.

[0060] The present application is further described below in conjunction with specific embodiments:

[0061] like Figure 3 As shown in the figure, the radar wave is installed on the same side as the door, 2 to 3 meters above the ground. The position and size of the buffer zone are adjusted according to the indoor conditions. The length of the buffer zone is the same as the width of the indoor door, aiming to reduce the reflection interference caused by the door switch. In addition, the angle between the radar wave and the ground is maintained at 45 degrees to ensure that the target can be accurately mapped to the coordinate system within the detection range.

[0062] The following is a detailed description of the algorithm flow of the present invention for indoor number monitoring and personnel entry and exit discrimination:

[0063] When the target is indoors, the target has two states: InRoom represents whether the target is indoors, and OutRoom represents whether the target is outdoors. When the target is outdoors, InRoom is False and OutRoom is True. The opposite is true. When the target enters or leaves the room, the corresponding state will also change. The dynamic adjustment of the target state is intended to avoid repeated counting after the target enters or leaves the room. The counting operation will only be performed when the target state meets the counting conditions to ensure monitoring accuracy.

[0064] Scenario 1: Crossing the initial judgment line of the buffer zone.

[0065] like Figure 4 As shown in the figure, the upper right corner is a legend. When the target enters the indoor area from the outdoor, the initial judgment line of the buffer zone is consistent with the position of the door. When the target crosses the judgment line, it means that the target enters the buffer zone, but the number of people in the room will not increase at this time. When the target further crosses the left and right judgment lines of the buffer zone, the system determines that the target has successfully entered the room, and the number of people in the room and the number of people entering the room are increased by 1.

[0066] Scenario 2: Crossing the initial indoor judgment line.

[0067] like Figure 5 As shown in Figure 1, when the target crosses the initial indoor judgment line, its state is marked as "ready to enter the room". At this time, as long as the target subsequently crosses the left and right judgment lines of the buffer zone, it is determined that the target has officially entered the room, and the number of people in the room and the number of people entering the room are increased by 1.

[0068] Scenario 3: Crossing the final judgment line indoors.

[0069] like Figure 6 As shown in Figure 1, when the target crosses the final judgment line of the indoor area, its state is "ready to enter the room". After the target further crosses the initial judgment line of the buffer zone, the number of people indoors and the number of people entering the room are increased by 1.

[0070] Next is the description of the personnel leaving judgment process. When the target is indoors, the target has two states: InRoom represents whether the target is indoors, and OutRoom represents whether the target is outdoor. When the target is indoors, InRoom is True, and OutRoom is False. The opposite is true. When the target enters or leaves the room, the corresponding state will also change.

[0071] Scenario 4: Crossing the final judgment line of the indoor area.

[0072] like Figure 7 As shown in Figure 1, when the target crosses the final judgment line from the indoor area, its status is marked as "ready to leave". After the target continues to cross the initial judgment line of the indoor area, the number of people indoors decreases by 1, and the number of people leaving increases by 1.

[0073] Scenario 5: Crossing the left and right judgment lines of the buffer zone.

[0074] like Figure 8 As shown in FIG. 1 , when the target crosses the left and right judgment lines of the buffer zone from indoors, the state is marked as “preparing to leave”. Then the target crosses the initial judgment line of the indoor area, the indoor number count decreases, and the leaving number count increases.

[0075] Scenario 6: Crossing the distinction line on both sides of the buffer zone.

[0076] like Fig. 9 As shown in the figure, when the target crosses the judgment lines on both sides of the buffer zone from the indoor area, its status is still marked as "ready to leave". When the target continues to cross the initial judgment line of the indoor area, the number of people indoors decreases by 1, and the number of people leaving increases by 1.

[0077] Although the technical solutions of the present invention have been described and listed in detail, it should be understood that it is obvious to those skilled in the art to make modifications to the above embodiments or adopt equivalent alternatives. These modifications or improvements made without departing from the spirit of the present invention are within the scope of protection required by the present invention.

Claims

1. A method for identifying people entering and leaving a house and monitoring the number of people indoors based on millimeter wave radar, characterized in that: The method comprises the following steps: S1. Measure the length of the current room door and set the buffer size according to whether there is a left wall or a right wall after entering the door; S2. After the buffer is set, the millimeter wave radar starts working, and the radar processes the collected point cloud data to identify and track the target; S3. Determine whether the target has passed through the buffer zone based on the target's trajectory, and complete the identification of personnel entering and exiting.

2. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter wave radar according to claim 1, characterized in that: The step S1 is specifically as follows: S11. Check the material of the room door. If the room door is a wooden door, set the length of the buffer zone to 1.1 times the door length; if it is a metal door, set the length of the buffer zone to 1.3 times the door length; S12, check whether there are left and right walls after entering the door. If there is a left wall but no right wall, set it to left wall lock; if there is no left wall but only a right wall, set it to right wall lock; if there are both, set it to full lock; if there are neither, set it to no lock; S13. The width of the buffer zone is set according to the wall. The left and right walls are the boundaries of the buffer zone. If the left and right walls do not exist at the same time, the width and length of the buffer zone are the same.

3. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter wave radar according to claim 1, characterized in that: The step S2 is specifically as follows: S21. The millimeter-wave radar acquires point cloud data of the surrounding environment by transmitting millimeter-wave signals and using the echo signal information. Each frame of point cloud data contains three-dimensional information of a large number of reflection points. To simplify processing and analysis, these three-dimensional point cloud data are projected into a two-dimensional polar coordinate system. The polar coordinate system takes the radar as the origin, and each point is represented by its distance and azimuth from the radar. S22. For multi-frame point cloud data, a density clustering algorithm is used to cluster the reflection points in the point cloud. By setting reasonable distance thresholds and density parameters, DBSCAN can adaptively identify multiple targets at different distances and noise levels. After clustering, a unique ID is assigned to each target to achieve independent tracking of the target in subsequent steps. S23. In order to track the target's motion trajectory in real time, the Kalman filter algorithm is used to estimate and predict the motion state of each target; the algorithm will update the target state according to the point cloud data returned by the radar, so as to maintain accurate tracking of the target trajectory under noise interference; at the same time, combined with the target ID, the Kalman filter can effectively realize the independent tracking of multiple targets and ensure the consistency and accuracy of the trajectory information.

4. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter-wave radar according to claim 3, characterized in that: The echo signal information includes time delay, intensity and Doppler frequency shift.

5. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter-wave radar according to claim 3, characterized in that: The three-dimensional information includes distance, angle and speed.

6. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter wave radar according to claim 1, characterized in that: The step S3 is specifically as follows: S31. For the scene of entering indoor space, the radar first detects that the target is outside the door. After the target enters the buffer zone, it is first determined whether the target trajectory passes through the left and right judgment lines of the buffer zone. If so, the number of people is increased by 1. If not, it is further determined whether the target trajectory passes through the left and right judgment lines or the final judgment line of the indoor space. If so, the number of people is increased by 1. S32. For the scene of leaving the indoor area, the radar first detects that the target moves from the indoor area to the buffer zone at the door. If the target trajectory passes through the left and right judgment lines of the buffer zone, the number of people is directly reduced by 1; if the target first passes through the left and right judgment lines of the indoor area or the final judgment line of the indoor area, and then passes through the initial judgment line of the indoor area, the number of people is also reduced by 1.

7. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter wave radar according to claim 1, characterized in that: The millimeter-wave radar is installed ceiling-mounted and located on the same side as the door. The specific installation height should be between 2.5m and 3m from the ground.

8. The method for identifying people entering and exiting a house and monitoring the number of people indoors based on millimeter-wave radar according to claim 1, characterized in that: The radar wave emission angle is maintained at approximately 45 degrees to the horizontal plane, and the maximum monitoring range can reach 6 meters.

Images