A target detection and tracking method and device, electronic equipment and storage medium

Abstract

Embodiments of the present application disclose a target detection and tracking method and device, electronic equipment and storage medium, belonging to the field of target detection, comprising: determining a determined track of a living body target according to the position, speed and continuous frame number of a continuous point cloud of the living body target based on a real-time detection radar point cloud; detecting and tracking the living body target according to the associated point cloud of the determined track in the real-time detection radar point cloud; determining the determined track of the living body target as a disappeared track according to the change of the real-time detection radar point cloud around the last position of the determined track and a preset first determination rule; and detecting and tracking the living body target according to the energy of the real-time detection radar point cloud around the last position of the disappeared track. The present application can accurately detect and track the living body target in a deep sleep state with low complexity.

Description

Technical Field

[0001] This invention relates to the field of target detection technology, and in particular to a target detection and tracking method, apparatus, electronic device, and storage medium. Background Technology

[0002] There are four main types of technologies commonly used for indoor live target detection and tracking. The first is live target detection methods based on optical sensors. Optical sensors can acquire high-resolution images and videos with rich semantic information. However, they are susceptible to environmental factors such as light and dust, and also suffer from large data volumes, high computational complexity, and privacy leaks, making them unsuitable for certain situations. The second is detection methods based on infrared sensors, which can be divided into active and passive infrared sensing. Active infrared sensing can obtain high-resolution infrared images containing rich information, but it is susceptible to environmental factors such as light and temperature, and requires complex hardware and software processing. Passive infrared sensing has a relatively simple sensing principle and low power consumption, but its detection sensitivity is low, the amount of information acquired is small, and it is more susceptible to the influence of surrounding environmental factors. The third is detection methods based on WiFi. These methods mainly detect live targets by analyzing the strength of the WiFi received signal and channel status. However, these methods struggle to obtain specific physical information about the target and are easily interfered with by other wireless devices, increasing the difficulty of WiFi-based live target detection. Fourth is the radar sensor-based method. By extracting the radar echo signal after it is reflected from an object, multi-dimensional information such as distance, angle, and speed can be obtained. While acquiring rich information, radar has the advantages of being less affected by environmental factors, operating in all weather conditions, and not infringing on personal privacy. It can be deployed and applied even in private places such as bedrooms, hotels, and bathrooms.

[0003] Existing technologies based on radar sensors for detecting live targets indoors suffer from several problems, including indoor interference leading to inaccurate acquisition of live target signals, difficulty in accurately detecting stationary live targets (e.g., targets are easily missed when they are in deep sleep), and high method complexity. Summary of the Invention

[0004] This invention provides a target detection and tracking method, apparatus, electronic device, and storage medium, which can accurately detect and track living targets with low complexity by eliminating interference.

[0005] In a first aspect, embodiments of the present invention provide a target detection and tracking method, comprising: determining a definite track representing a live target's track from the tracks of at least one suspected live target acquired based on a real-time detection radar point cloud, the position and velocity of the track of at least one suspected live target in an indoor environment, and the number of consecutive frames of the real-time detection radar point cloud containing associated point clouds of the suspected live target's track; acquiring associated point clouds of the definite track from the real-time detection radar point cloud, and detecting and tracking the live target based on the associated point clouds; for definite tracks in the real-time detection radar point cloud where the associated point clouds are not present, determining the definite track as a vanished track based on changes in the real-time detection radar point cloud in the space surrounding the last position of the definite track and a preset first determination rule; and detecting and tracking the live target based on the energy of the real-time detection radar point cloud in the space surrounding the last position of the vanished track.

[0006] In a second aspect, embodiments of the present invention provide a target detection and tracking device, comprising:

[0007] The system includes a trajectory acquisition module, configured to determine a definite trajectory representing a live target from the trajectories of at least one suspected live target acquired based on the position and velocity of the trajectory of at least one suspected live target in the room, and the number of consecutive frames of the real-time detection radar point cloud containing associated point clouds of the suspected live target's trajectory. The real-time detection radar point cloud is a radar point cloud obtained by real-time detection of the room using radar sensors. A first tracking module is configured to acquire associated point clouds of the definite trajectory from the real-time detection radar point cloud and detect and track the live target based on the associated point clouds. A disappearing trajectory determination module is configured to determine a disappearing trajectory for definite trajectories where the associated point clouds are absent from the real-time detection radar point cloud, based on changes in the real-time detection radar point cloud around the last position of the definite trajectory and a preset first determination rule. A second tracking module is configured to detect and track the live target based on the energy of the real-time detection radar point cloud around the last position of the disappearing trajectory.

[0008] Thirdly, embodiments of the present invention also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the target detection and tracking method as described in any of the embodiments of the present invention.

[0009] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the target detection and tracking method as described in any of the embodiments of the present invention.

[0010] This invention provides a target detection and tracking method, apparatus, electronic device, and storage medium. By determining the definite trajectory of a live target based on the position, speed, and the number of frames of a real-time monitoring radar point cloud with continuous associated point clouds, the invention detects and tracks the live target accordingly. This effectively eliminates indoor interference signals and accurately detects and tracks live targets. Furthermore, when no associated point cloud with a definite trajectory is found in the real-time radar point cloud, and a live target is suspected to exist around the definite trajectory, the invention classifies the definite trajectory as a vanished trajectory and detects and tracks the live target based on the energy of the radar point cloud in the space surrounding the vanished trajectory. This low-complexity detection and tracking method continues to detect and track potentially live targets, avoiding the loss of detection of stationary live targets in existing technologies, improving detection and tracking accuracy, and simplifying computational complexity. Attached Figure Description

[0011] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart illustrating the target detection and tracking method provided in an embodiment of the present invention;

[0013] Figure 2 This is another flowchart illustrating the target detection and tracking method provided in this embodiment of the invention;

[0014] Figure 3 This is another flowchart illustrating the target detection and tracking method provided in this embodiment of the invention;

[0015] Figure 4 This is a schematic diagram of beamforming in the target detection and tracking method provided in this embodiment of the invention;

[0016] Figure 5 This is a schematic diagram of the target detection and tracking device provided in an embodiment of the present invention;

[0017] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0019] There are four main types of technologies commonly used for indoor live target detection and tracking. The first is live target detection methods based on optical sensors. Optical sensors can acquire high-resolution images and videos with rich semantic information. However, they are susceptible to environmental factors such as light and dust, and also suffer from large data volumes, high computational complexity, and privacy leaks, making them unsuitable for certain situations. The second is detection methods based on infrared sensors, which can be divided into active and passive infrared sensing. Active infrared sensing can obtain high-resolution infrared images containing rich information, but it is susceptible to environmental factors such as light and temperature, and requires complex hardware and software processing. Passive infrared sensing has a relatively simple sensing principle and low power consumption, but its detection sensitivity is low, the amount of information acquired is small, and it is more susceptible to the influence of surrounding environmental factors. The third is detection methods based on WiFi. These methods mainly detect live targets by analyzing the strength of the WiFi received signal and channel status. However, these methods struggle to obtain specific physical information about the target and are easily interfered with by other wireless devices, increasing the difficulty of WiFi-based live target detection. Fourth is the radar sensor-based method. By extracting the radar echo signal after it is reflected from an object, multi-dimensional information such as distance, angle, and speed can be obtained. While acquiring rich information, radar has the advantages of being less affected by environmental factors, operating in all weather conditions, and not infringing on personal privacy. It can be deployed and applied even in private places such as bedrooms, hotels, and bathrooms.

[0020] Existing technologies based on radar sensors for indoor live target detection suffer from several drawbacks. Indoor interference signals can prevent accurate acquisition of live target signals, and stationary live targets cannot be detected, such as when the target is in deep sleep. On the other hand, radar sensor-based machine learning methods suffer from high complexity.

[0021] Figure 1 This is a schematic flowchart illustrating a target detection and tracking method provided in an embodiment of the present invention. This method can be executed by a target detection and tracking device provided in this embodiment, which can be implemented using software and / or hardware. In a specific embodiment, the device can be integrated into an electronic device, such as a computer or server. The following embodiments will illustrate this using the integration of the device into an electronic device as an example. (Reference) Figure 1 The method may specifically include the following steps:

[0022] Step 101: Based on the position and velocity of the trajectory of at least one suspected living target indoors, obtained from the real-time detection radar point cloud, and the number of consecutive frames of the real-time detection radar point cloud associated with the trajectory of the suspected living target, a definite trajectory representing the living target is determined from the trajectory of at least one suspected living target. The real-time detection radar point cloud is a radar point cloud obtained by real-time detection of the indoor environment using radar sensors. This helps to avoid interference from objects such as fans, natural wind, curtains, bed curtains, and mosquito nets when detecting living targets.

[0023] The specific trajectory is determined based on the movement characteristics of a live target indoors. When a definite trajectory exists indoors, it can be determined that a live target exists indoors.

[0024] Specifically, the aforementioned live targets include humans, animals, etc.

[0025] Specifically, the associated point cloud of the track of the suspected live target can be the real-time detection radar point cloud corresponding to the suspected live target that can be associated with the track of the suspected live target.

[0026] Specifically, the aforementioned radar sensor can be a millimeter-wave radar. A 2-transmit, 4-receive multiple-input multiple-output (MIMO) radar module can be used for experiments.

[0027] Specifically, the process of obtaining radar point clouds by real-time indoor detection using radar sensors can include: preprocessing, range-angle imaging, suspected live target detection and point cloud imaging.

[0028] Specifically, the radar sensor continuously emits electromagnetic wave signals into space. These signals are scattered or reflected by objects and received by the radar receiver. After passing through a signal amplifier, mixer, and ADC sampling, a discrete echo signal containing information in the range, time, and angle dimensions is obtained. The echo received by the radar module can be represented as y(m,n,k), where m is the slow time dimension, representing the m-th linear frequency modulated continuous wave signal, n is the fast time dimension, representing the n-th sampling point, and k is the antenna dimension, representing the received signal of the k-th channel.

[0029] The aforementioned preprocessing may include performing a Fast Fourier Transform on the received radar echo in the fast time dimension to obtain the range-slow time-antenna dimension signal. Where r∈[1,N], it represents range cell sampling; then range-dimensional clutter suppression is performed, which can be achieved using algorithms such as moving average algorithm, intra-frame cancellation method, and singular value decomposition, to obtain the clutter-suppressed signal y. c (m,r,k).

[0030] Specifically, preprocessing can suppress static clutter associated with static objects such as the ground, tables, and beds, thereby improving the accuracy of live target detection.

[0031] Specifically, the aforementioned range-angle imaging may include: utilizing the angle measurement capability of radar sensors, based on algorithms such as Angle-FFT, Bartlett's algorithm, Capon's algorithm, and MUSIC algorithm, to measure the angle of y. c Angle estimation is performed on (m,r,k) to obtain the distance-angle spectrum M[r,θ], where θ∈[1,V] represents the azimuth angle value and V represents the maximum angle range.

[0032] Specifically, the aforementioned suspected live target detection and point cloud imaging may include: using the Order Statistics Constant False Alarm Rate (OS-CFAR) algorithm for target detection, and using a cross window instead of a rectangular window to speed up the calculation. The detection result contains radar polar coordinate information: radial distance and azimuth angle. Therefore, coordinate transformation is required to obtain an indoor planar point cloud map, which is parallel to the ground where people walk.

[0033] Specifically, a single target will generate multiple scattering points in the point cloud map, and the point cloud in the map is not entirely generated by the target. The point cloud density clustering and point-to-track association process can include: based on the characteristic that the point cloud density generated by noise and weak interference is relatively sparse, the DBSCAN equal-density clustering algorithm is used to segment the point cloud map, reduce the influence of noise and some interference, form different clusters, and extract the weighted centroids of the point clouds through the distance-angle spectrum energy values ​​corresponding to the point clouds. The point cloud clusters that are successfully clustered for multiple consecutive frames are the starting points of the track.

[0034] Specifically, the aforementioned point cloud density clustering and point-to-track association can include: first, associating point cloud data with flight tracks, with the most common data association algorithm being nearest neighbor association. Flight tracks successfully associated with point clouds will have their track information updated according to a filtering algorithm, while unassociated point clouds will undergo density clustering to determine if any new targets have been generated.

[0035] In an optional embodiment of the present invention, the process of determining the trajectory representing the trajectory of a live target from the trajectories of at least one suspected live target, based on the position and velocity of the trajectory of at least one suspected live target acquired from a real-time detection radar point cloud and the number of consecutive frames of the real-time detection radar point cloud associated with the trajectories of the suspected live targets, includes:

[0036] The trajectory of at least one suspected live target is identified as a transient trajectory; the current displacement of the transient trajectory from its initial position to its current position and the current velocity of the transient trajectory are calculated; and transient trajectories with a corresponding consecutive frame number greater than a first preset frame number, a current displacement greater than a preset displacement, and a current velocity greater than a preset velocity are identified as defined trajectories.

[0037] Optionally, a transient trajectory where the current displacement is greater than a preset displacement and the current velocity is greater than a preset velocity can be defined as a fixed trajectory.

[0038] Specifically, objects such as fans, natural wind, curtains, bed curtains, and mosquito nets are in a relatively fixed position and have a velocity of 0 most of the time. They also do not exist for long periods of time as point clouds associated with tracks that have a certain velocity and can produce a certain displacement. Therefore, by using a first preset number of frames, a preset displacement, and a preset velocity, the tracks of interfering objects can be filtered out from the tracks of suspected live targets, resulting in a more accurate track for identifying live targets.

[0039] Optionally, when a suspected live target is detected, the system is initialized to S0, and the starting position L of the track is recorded. s The number of frames N that the trajectory is associated with in the point cloud F Current position L of the flight path n Current speed V n When the following conditions are met:

[0040] N F >T N

[0041] V n >T V

[0042] abs|L n -L s |>T L

[0043] At that time, the track state is changed from S0 to S1. Where T... N For frame parameters, T V For the velocity parameter, T L This is the distance parameter.

[0044] Step 102: Obtain the associated point cloud of the determined track from the real-time detection radar point cloud, and detect and track the live target based on the associated point cloud. This enables accurate tracking of the live target and timely acquisition of the state of the associated point cloud of the real-time detection radar point cloud where there is no determined track.

[0045] Specifically, the associated point cloud of the determined track can be a real-time detection radar point cloud that can be associated with the determined track and corresponds to the live target corresponding to the determined track.

[0046] In an optional embodiment of the present invention, the process of detecting and tracking a live target based on the associated point cloud includes:

[0047] The current velocity and current position of the determined trajectory are estimated based on the position of the associated point cloud at the current moment, resulting in the estimated current velocity and estimated current position of the determined trajectory; and based on the two-dimensional uniform linear motion model, the Kalman filter algorithm is used to calculate the current velocity and current position of the determined trajectory according to the estimated current velocity and estimated current position respectively.

[0048] Specifically, the speed of people walking indoors is approximately 0.5 m / s to 2 m / s, which is considered low-speed movement and mostly conforms to uniform motion. The point cloud image also contains two-dimensional measurement information of the target; therefore, a two-dimensional uniform linear motion model is selected for the tracking model. Since there is a certain error between the radar sensor's measurements and the target's actual state, a Kalman filter algorithm is used for filtering, primarily for trajectory prediction and updating. This algorithm can effectively remove random noise, obtain a more accurate target position, and achieve precise target tracking.

[0049] Step 103: When there are no associated point clouds in the real-time detection radar point cloud, the determined track is determined as a vanished track based on the changes in the real-time detection radar point cloud around the last position of the determined track and the preset first judgment rule. This can facilitate the detection and tracking of live targets based on the energy of the real-time detection radar point cloud around the last position of the vanished track.

[0050] In an optional specific embodiment of the present invention, the process of determining the trajectory as a vanished trajectory based on the real-time detection of changes in the radar point cloud around the final position of the trajectory and a preset first determination rule includes:

[0051] If the duration of unrelated point clouds in the real-time monitoring radar point cloud exceeds a first preset duration, the determined track will be judged as a disappeared track.

[0052] Step 104: Based on the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappeared track, live targets are detected and tracked. This combines steps 101-103 to achieve accurate and effective tracking and detection of both moving and stationary live targets with a low-complexity detection and tracking method. This avoids the loss of detection of stationary live targets in existing technologies, improves detection and tracking accuracy, and simplifies computational complexity.

[0053] Specifically, if a live target is located in the space around the last position of the vanished track, the energy of the real-time radar detection point cloud in the space around the last position of the vanished track will necessarily be greater than the energy when there is no live target. Therefore, the live target can be detected and tracked based on the energy of the real-time detection radar point cloud.

[0054] Specifically, the space surrounding the last position of the vanished track can be defined as an active zone, and the energy of the real-time detection radar point cloud in the active zone can be monitored. Specifically, the area for energy monitoring (range-angle spectrum) of the surrounding space can be centered on the last position of the vanished track.

[0055] In an optional embodiment of the present invention, the process of detecting and tracking a live target based on the energy of the real-time detection radar point cloud of the space surrounding the last position of the vanished track includes: determining the activity level of the space surrounding the last position of the vanished track based on the energy of the real-time detection radar point cloud of the space surrounding the last position of the vanished track; and determining whether a live target exists in the space surrounding the last position of the vanished track based on the activity level.

[0056] Optionally, an initial activity value can be set. If the maximum energy value of the real-time detection radar point cloud around the last position of the disappeared track is greater than the preset third energy threshold, the activity value is increased until the preset activity upper limit value is reached; otherwise, the activity value is decreased until it is 0.

[0057] In an optional embodiment of the present invention, the process of determining the activity level of the space surrounding the last position of the vanished track based on the energy of the real-time detection radar point cloud of the space surrounding the last position of the vanished track includes: determining the static level of the space surrounding the last position of the vanished track based on the energy of the real-time detection radar point cloud of the space surrounding the last position of the vanished track; and determining the activity level of the space surrounding the last position of the vanished track based on the energy of the real-time detection radar point cloud of the space surrounding the last position of the vanished track and the static level.

[0058] In an optional embodiment of the present invention, the process of determining the activity level of the space surrounding the last position of the vanished track based on the energy and stationarity of the real-time detection radar point cloud in the space surrounding the last position of the vanished track includes: setting initial values ​​for stationarity and activity; when the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished track is not less than a preset first energy threshold, determining whether the activity level is less than a preset upper limit value; if it is less, increasing the activity level; if it is not less, keeping the activity level unchanged; and when the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished track is less than the preset first energy threshold, determining whether the stationarity value is less than a preset upper limit value; if it is less, increasing the stationarity value; if it is not less, decreasing the activity level and resetting the stationarity value to the initial value.

[0059] Optionally, initialize the stillness and activity levels by setting the first energy threshold to T. e When the maximum energy value E detected in the active regionn ≥T e When E... n <T e If the current inactivity level has not reached the inactivity level limit, the inactivity level increases. If the inactivity level reaches the limit, the activity level decreases and the inactivity level is set to zero.

[0060] Specifically, the initial value for the degree of stillness can be set to 0.

[0061] In an optional embodiment of the present invention, the process of determining the activity level of the space surrounding the last position of the vanished track based on the energy and stationarity of the real-time detection radar point cloud of the space surrounding the last position of the vanished track includes: setting initial values ​​for stationarity and activity; when the maximum energy value of the real-time detection radar point cloud of the space surrounding the last position of the vanished track is less than a preset second energy threshold, calculating the accumulated energy of the real-time detection radar point cloud of the space surrounding the last position of the vanished track from a second preset time period to the current time; when the accumulated energy is not less than a preset first energy threshold, determining whether the activity level is less than a preset upper limit of activity level; if it is less, increasing the activity level; if it is not less, keeping the activity level unchanged; and when the accumulated energy is less than the preset first energy threshold, determining whether the stationarity value is less than a preset upper limit of stationarity; if it is less, increasing the stationarity; if it is not less, decreasing the activity level and resetting the stationarity value to the initial value.

[0062] Optionally, an initial activity value can be set. If the maximum energy value of the real-time detection radar point cloud around the last position of the disappeared track is greater than the preset third energy threshold, the activity value is increased until the preset activity upper limit value is reached; otherwise, the activity value is decreased until it is 0.

[0063] Optionally, the second energy threshold mentioned above can be set to T. a When the maximum energy E in the active region n <T a At the same time, inter-frame incoherent accumulation of signals in the active region is performed to improve the signal-to-noise ratio of the target. An exponential moving average algorithm is used to incoherently accumulate the signal M[r,θ].

[0064] M A [r,θ]=α*M A [r,θ]+(1-α)*M[r,θ] (6)

[0065] Here, α is the accumulation parameter, which is related to the number of accumulated signal frames.

[0066] In an optional embodiment of the present invention, the activity limit value includes a first activity limit value and a second activity limit value, and the inactivity limit value includes a first inactivity limit value and a second inactivity limit value; the first activity limit value is greater than the second activity limit value, and the first inactivity limit value is greater than the second inactivity limit value; wherein the first activity limit value and the first inactivity limit value are preset activity limit values ​​and inactivity limit values ​​respectively when the last position of the disappeared flight path is located at the non-boundary point of the indoor space; the second activity limit value and the second inactivity limit value are preset activity limit values ​​and inactivity limit values ​​respectively when the last position of the disappeared flight path is located at the boundary point of the indoor space.

[0067] Specifically, based on prior boundary knowledge, vanishing tracks can be divided into non-boundary region vanishing tracks S. 21 And the disappearance track S in the boundary area 22 Therefore, the active area is divided into the non-boundary active area A1 and the boundary active area A2. Multi-level management of the active area, flexibly setting different characteristics based on its location, can improve the accuracy of personnel presence detection results in complex indoor environments. For example, lowering the activity limit of A2 while simultaneously lowering the stillness limit to accelerate the activity level clearing rate can reduce long-term false alarms at the boundary in special circumstances.

[0068] In an optional specific embodiment of the present invention, the process of determining whether there is a live target in the space around the last position of the disappeared track based on the activity level includes: if the activity level value at the current moment is not 0, then it is determined that there is a live target in the space around the last position of the disappeared track; otherwise, it is determined that there is no live target in the space around the last position of the disappeared track.

[0069] Optionally, if there is no definite flight path indoors, and there is an active area with non-zero activity, then it is determined that there is a live target indoors; if there is no definite flight path indoors, and there is no active area or the activity level of the active area is zero, then it is determined that there is no live target indoors.

[0070] The target detection and tracking method in another embodiment is further described below, such as Figure 2 As shown, it can include the following steps:

[0071] Step 201: Based on the position and velocity of the trajectory of at least one suspected live target in the room, and the number of consecutive frames of the real-time detection radar point cloud of the associated point cloud of the trajectory of the suspected live target, the determined trajectory representing the trajectory of the live target is determined from the trajectory of at least one suspected live target. The real-time detection radar point cloud is the radar point cloud obtained by real-time detection of the room using radar sensors.

[0072] Step 202: Obtain the associated point cloud of the determined trajectory from the real-time detection radar point cloud, and detect and track the live target based on the associated point cloud.

[0073] Step 203: When there are no associated point clouds in the real-time detection radar point cloud, the determined track is determined as a deleted track based on the changes in the real-time detection radar point cloud around the last position of the determined track and the preset second determination rule, or the disappeared track is determined as a deleted track based on the changes in the real-time detection radar point cloud around the last position of the disappeared track and the second determination rule.

[0074] In an optional specific embodiment of the present invention, the process of determining the determined track as a deleted track based on the changes in the real-time detection radar point cloud of the space surrounding the last position of the determined track and the preset second determination rule includes: when a new determined track is determined based on the real-time detection radar point cloud of the space surrounding the last position of the determined track, a judgment is made on whether the new determined track and the determined track are successfully matched based on the distance between the new determined track and the determined track. If they are successfully matched, the determined track is determined as a deleted track.

[0075] Optionally, if the distance between the newly determined track and the established track is less than a preset distance, then the newly determined track is considered to have successfully matched the old determined track.

[0076] Specifically, based on the speed of the determined trajectory obtained earlier, the determined trajectory can be further subdivided into dynamic trajectories S. 11 and static track S 12 .

[0077] Optionally, when a new determined track is determined based on the real-time detection radar point cloud of the space surrounding the last position of the determined track, the determined track is identified as the old determined track; the old determined track is scored based on the distance between the new determined track and the old determined track, the dynamic and static state of the old determined track, whether there is a related point cloud at the last position of the old determined track, and the energy value of the related point cloud at the last position; and the new determined track is judged to be successfully matched with the old determined track based on the score and a preset score threshold.

[0078] In an optional specific embodiment of the present invention, the process of determining a lost track as a deleted track based on the changes in the real-time detection radar point cloud of the space surrounding the last position of the lost track and the second determination rule includes: when a new determined track is determined based on the real-time detection radar point cloud of the space surrounding the last position of the lost track, a judgment is made on whether the new determined track and the lost track are successfully matched based on the distance between the new determined track and the lost track; if they are successfully matched, the determined track is determined as a deleted track.

[0079] Optionally, if the distance between the newly determined track and the disappeared track is less than a preset distance, then the newly determined track and the disappeared track are considered to have successfully matched.

[0080] Optionally, when a new trajectory is determined based on the real-time detection radar point cloud of the space surrounding the last position of the vanished trajectory, the vanished trajectory is scored according to the distance between the new trajectory and the vanished trajectory, the dynamic and static state of the vanished trajectory, whether there is a related point cloud at the last position of the vanished trajectory, and the energy value of the related point cloud at the last position. The new trajectory is then judged to be successfully matched with the vanished trajectory based on the score and a preset score threshold.

[0081] Optionally, when a new track is determined based on the real-time detection radar point cloud of the space surrounding the last position of the vanished track, if the duration of the new track in the space surrounding the last position of the vanished track is greater than a third preset duration, it is determined that the new track and the vanished track have successfully matched.

[0082] Step 204: Delete the deleted track, and detect and track the same live target based on the newly determined track that corresponds to the live target of the deleted track and the real-time detection radar point cloud.

[0083] In an optional specific embodiment of the present invention, the target detection and tracking method of the present invention further includes: if the number of consecutive frames of the real-time detection radar point cloud of the associated point cloud that does not have a transient track is greater than a second preset number of frames, then the transient track is determined to be a deleted track, and the deleted track is deleted.

[0084] Specifically, indoor environments are complex and diverse, influenced by factors such as interference and multipath effects, which may cause the same living target to generate multiple tracks. The behavior of living targets indoors is also varied. When a person transitions from walking to a stationary state (e.g., sleeping), or walks behind a barrier, the signal-to-noise ratio is low, resulting in a vanishing track at that location. When the person leaves or moves around in that location, a new, definite track is formed. This invention aims to eliminate false tracks and redundant old tracks, ensuring a one-to-one correspondence between tracks and the actual living target.

[0085] The following further describes the target detection and tracking method in another embodiment, such as... Figure 3 As shown, it may include the following steps:

[0086] Step 301: Based on the position and velocity of the trajectory of at least one suspected live target in the room, and the number of consecutive frames of the real-time detection radar point cloud of the associated point cloud of the trajectory of the suspected live target, the determined trajectory representing the live target is determined from the trajectory of at least one suspected live target. The real-time detection radar point cloud is the radar point cloud obtained by real-time detection of the room using radar sensors.

[0087] Step 302: Obtain the associated point cloud of the determined trajectory from the real-time detection radar point cloud, and detect and track the live target based on the associated point cloud.

[0088] Step 303: Obtain the attitude information of the live target corresponding to the determined track based on the pitch direction energy of the real-time detected radar point cloud.

[0089] In an optional specific embodiment of the present invention, the process of obtaining the attitude information of the live target corresponding to the track based on the pitch direction energy of the real-time detection radar point cloud includes: obtaining the horizontal direction energy and the downward direction energy of the real-time detection radar point cloud based on the beamforming method; and determining whether the live target is in a low attitude based on the ratio of the horizontal direction energy to the downward direction energy and a preset ratio threshold.

[0090] Optional, such as Figure 4 As shown, for d h and d l Beamforming is performed in each direction to obtain d h Energy P in the direction h and d l Energy P in the direction l When the target is walking or standing, d h and d l The target can be detected in both directions; however, when the target is at a low altitude, only d... l There is a direction and a goal. Therefore, in N h If it exists continuously within a frame:

[0091]

[0092] The target is then considered to be in a low-profile state. Where N... h For the set parameters, T h These are the low-attitude threshold parameters.

[0093] The altitude information of the corresponding live target is further extracted from the determined flight path. To reduce computational complexity, the altitude information is calculated using a beamforming-based method to identify whether the target is in a low posture (such as lying down or falling over).

[0094] Step 304: For the determined track of the unrelated point cloud in the real-time detection radar point cloud, based on the changes in the real-time detection radar point cloud around the last position of the determined track and the preset first judgment rule, the determined track is judged as a disappeared track.

[0095] Step 305: Detect and track the live target based on the attitude information of the live target corresponding to the vanished track and the energy of the real-time detection radar point cloud in the space surrounding the last position of the vanished track.

[0096] Optionally, the process of detecting and tracking a live target based on the attitude information of the live target corresponding to the disappearing track and the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappearing track includes: setting an initial value for the aforementioned activity based on the attitude information of the live target corresponding to the last position of the disappearing track.

[0097] Specifically, if the attitude information of the living target corresponding to the last position of the vanished track is low attitude, then the initial value of the above activity level is set to a value greater than 0; otherwise, the initial value of the above activity level is set to 0.

[0098] Specifically, combining the attitude information of the trajectory can improve the accuracy of detecting and tracking live targets. For example, if the last attitude information of the live target corresponding to the disappearing trajectory is low attitude, then it can be more certain that there is a live target in the surrounding space of the disappearing trajectory.

[0099] Figure 5 This is a structural diagram of a target detection and tracking device provided in an embodiment of the present invention. This device is suitable for executing the target detection and tracking method provided in an embodiment of the present invention. Figure 5 As shown, the device may specifically include:

[0100] The trajectory acquisition module 501 is used to determine the trajectory representing the trajectory of a live target from the trajectory of at least one suspected live target, based on the position and speed of the trajectory of at least one suspected live target in the room acquired by real-time detection radar point cloud and the number of consecutive frames of the real-time detection radar point cloud associated with the trajectory of the suspected live target. The real-time detection radar point cloud is a radar point cloud obtained by real-time detection of the room using radar sensors.

[0101] Optionally, the aforementioned trajectory acquisition module 501 can be specifically used to determine the trajectory of at least one suspected live target as a transient trajectory, calculate the current displacement of the transient trajectory from its initial position to its current position, and the current velocity of the transient trajectory; and determine transient trajectories with a corresponding consecutive frame number greater than a first preset frame number, a current displacement greater than a preset displacement, and a current velocity greater than a preset velocity as determined trajectories. This helps to avoid interference from objects such as fans, natural wind, curtains, bed curtains, and mosquito nets that may affect the detection of live targets.

[0102] The first tracking module 502 is used to obtain the associated point cloud of the determined track from the real-time detection radar point cloud, and to detect and track the live target based on the associated point cloud, which is conducive to the accurate tracking of the live target and timely acquisition of the state of the associated point cloud of the real-time detection radar point cloud that does not have a determined track.

[0103] In an optional embodiment of the present invention, the first tracking module 502 described above can be specifically used to estimate the current speed and current position of the determined track based on the position of the associated point cloud at the current time, so as to obtain the estimated current speed and estimated current position of the determined track; and based on a two-dimensional uniform linear motion model, to calculate the current speed and current position of the determined track according to the estimated current speed and estimated current position using a Kalman filter algorithm.

[0104] The disappearing track determination module 503 is used to determine the track of unrelated point clouds in the real-time detection radar point cloud. Based on the changes in the real-time detection radar point cloud around the last position of the determined track and the preset first determination rule, the determined track is determined as a disappearing track. This can facilitate the detection and tracking of living targets based on the energy of the real-time detection radar point cloud around the last position of the disappearing track.

[0105] Optionally, the aforementioned disappearing track determination module 503 can be specifically used to determine a track as a disappearing track if the duration of the unrelated point cloud in the real-time monitoring radar point cloud exceeds a first preset duration, which is beneficial for detecting and tracking live targets based on the energy of the real-time monitoring radar point cloud in the space surrounding the last position of the disappearing track.

[0106] The second tracking module 504 is used to detect and track live targets based on the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappeared track. Combined with the aforementioned modules 501-503, this low-complexity detection and tracking method accurately and effectively tracks and detects both moving and stationary live targets, avoiding the loss of detection for stationary live targets in existing technologies, improving detection and tracking accuracy, and simplifying computational complexity.

[0107] Optionally, the second tracking module 504 can be specifically used to determine the activity level of the space around the last position of the vanished track based on the energy of the real-time detection radar point cloud around the last position of the vanished track; and to determine whether there is a live target in the space around the last position of the vanished track based on the activity level.

[0108] Optionally, the second tracking module 504 can be specifically used to determine the static degree of the space around the last position of the vanished track based on the energy of the real-time detection radar point cloud of the space around the last position of the vanished track; and to determine the activity degree of the space around the last position of the vanished track based on the energy and static degree of the real-time detection radar point cloud of the space around the last position of the vanished track.

[0109] Optionally, the second tracking module 504 can be specifically used to: set initial values ​​for stationarity and activity; when the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished track is not less than a preset first energy threshold, determine whether the activity value is less than a preset upper limit of activity; if it is less, increase the activity value; if it is not less, keep the activity value unchanged; and when the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished track is less than a preset first energy threshold, determine whether the stationarity value is less than a preset upper limit of stationarity; if it is less, increase the stationarity value; if it is not less, decrease the activity value and reset the stationarity value to the initial value.

[0110] Optionally, the second tracking module 504 can be specifically used to: set initial values ​​for stationarity and activity; when the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished track is less than a preset second energy threshold, calculate the accumulated energy of the real-time detection radar point cloud in the space surrounding the last position of the vanished track from the second preset time period to the current time; when the accumulated energy is not less than a preset first energy threshold, determine whether the value of activity is less than a preset upper limit of activity; if it is less, increase the value of activity; if it is not less, keep the value of activity unchanged; and when the accumulated energy is less than the preset first energy threshold, determine whether the value of stationarity is less than a preset upper limit of stationarity; if it is less, increase the value of stationarity; if it is not less, decrease the value of activity and reset the value of stationarity to the initial value.

[0111] Optionally, the activity limit includes a first activity limit and a second activity limit, and the inactivity limit includes a first inactivity limit and a second inactivity limit; the first activity limit is greater than the second activity limit, and the first inactivity limit is greater than the second inactivity limit; wherein the first activity limit and the first inactivity limit are preset activity limit and inactivity limit, respectively, when the last position of the disappeared flight path is located at a non-boundary point indoors; the second activity limit and the second inactivity limit are preset activity limit and inactivity limit, respectively, when the last position of the disappeared flight path is located at a boundary point indoors.

[0112] Optionally, the second tracking module 504 can be specifically used to determine that if the activity value at the current moment is not 0, there is a live target in the space around the last position of the disappeared track; otherwise, it is determined that there is no live target in the space around the last position of the disappeared track.

[0113] In another specific embodiment of the present invention, the target detection and tracking device of the present invention further includes: a track deletion determination module and a deletion and continuation tracking module, wherein the track deletion determination module is used to determine the determined track as a deleted track based on the changes in the real-time detected radar point cloud around the last position of the determined track and a preset second determination rule when there is no associated point cloud in the real-time detection radar point cloud, or to determine the disappeared track as a deleted track based on the changes in the real-time detected radar point cloud around the last position of the disappeared track and the second determination rule; the deletion and continuation module is used to delete the deleted track and detect and track the same living target based on the newly determined track that is the same living target as the deleted track and the real-time detected radar point cloud.

[0114] Optionally, the above-mentioned track deletion determination module can also be used to determine the transient track as a track to be deleted if the number of consecutive frames of the real-time detection radar point cloud of the associated point cloud that does not have a transient track is greater than the second preset number of frames, and then delete the track.

[0115] Optionally, the above-mentioned track deletion determination module can be specifically used to determine whether a new track is successfully matched with the disappearing track based on the distance between the new track and the disappearing track when a new track is determined based on the real-time detection radar point cloud of the space surrounding the last position of the disappearing track. If the track is successfully matched, the determined track is determined to be a deleted track.

[0116] Specifically, indoor environments are complex and diverse, influenced by factors such as interference and multipath effects, which may cause the same living target to generate multiple tracks. The behavior of living targets indoors is also varied. When a person transitions from walking to a stationary state (e.g., sleeping), or walks behind a barrier, the signal-to-noise ratio is low, resulting in a vanishing track at that location. When the person leaves or moves around in that location, a new, definite track is formed. This invention aims to eliminate false tracks and redundant old tracks, ensuring a one-to-one correspondence between tracks and the actual living target.

[0117] In another specific embodiment of the present invention, the target detection and tracking device of the present invention further includes: an attitude acquisition module, used to acquire attitude information of a live target corresponding to a determined track based on the energy value of a real-time detected radar point cloud.

[0118] Optionally, the attitude acquisition module can be specifically used to acquire the horizontal and downward energy of the real-time detection radar point cloud based on the beamforming method; and to determine whether the living target is in a low attitude based on the ratio of the horizontal energy to the downward energy and a preset ratio threshold.

[0119] Optionally, the aforementioned second tracking module can be specifically used to detect and track a living target based on the attitude information of the living target corresponding to the vanished track and the energy of the real-time detection radar point cloud in the space surrounding the last position of the vanished track.

[0120] Specifically, combining the attitude information of the trajectory can improve the accuracy of detecting and tracking live targets. For example, if the last attitude information of the live target corresponding to the disappearing trajectory is low attitude, then it can be more certain that there is a live target in the surrounding space of the disappearing trajectory.

[0121] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is merely an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the functional modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0122] This invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the target detection and tracking method provided in any of the above embodiments.

[0123] This invention also provides a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the target detection and tracking method provided in any of the above embodiments.

[0124] The following is for reference. Figure 6 It shows a schematic diagram of the structure of a computer system 600 suitable for implementing an electronic device according to embodiments of the present invention. Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0125] like Figure 6 As shown, the computer system 600 includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 602 or programs loaded from storage section 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the system 600. The CPU 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0126] The following components are connected to I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to I / O interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 610 as needed so that computer programs read from it can be installed into storage section 608 as needed.

[0127] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611. When the computer program is executed by central processing unit (CPU) 601, it performs the functions defined above in the system of this invention.

[0128] It should be noted that the computer-readable medium shown in this invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0129] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0130] The modules and / or units described in the embodiments of the present invention can be implemented in software or hardware. The described modules and / or units can also be housed in a processor; for example, a processor can be described as including a trajectory acquisition module, a first tracking module, a vanishing trajectory determination module, and a second tracking module. The names of these modules do not necessarily limit the functionality of the module itself.

[0131] In another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiments; or it may exist alone and not assembled into the device. The aforementioned computer-readable medium carries one or more programs that, when executed by a device, enable the device to perform the following: determining a definite track representing a live target's track from the tracks of at least one suspected live target, based on the position and velocity of the track of at least one suspected live target acquired from a real-time detection radar point cloud, and the number of consecutive frames of a real-time detection radar point cloud containing associated point clouds of continuously existing suspected live target tracks; acquiring associated point clouds of the definite track from the real-time detection radar point cloud, and detecting and tracking the live target based on the associated point clouds; determining a definite track without associated point clouds in the real-time detection radar point cloud as a vanished track based on changes in the real-time detection radar point cloud around the last position of the definite track and a preset first determination rule; and detecting and tracking the live target based on the energy of the real-time detection radar point cloud around the last position of the vanished track.

[0132] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A target detection and tracking method, characterized in that, include: Based on the position and velocity of the trajectory of at least one suspected live target in the room, obtained from the real-time detection radar point cloud, and the number of consecutive frames of the real-time detection radar point cloud of the associated point cloud of the trajectory of the suspected live target, a definite trajectory representing the trajectory of the live target is determined from the trajectory of the at least one suspected live target. The real-time detection radar point cloud is a radar point cloud obtained by real-time detection of the room using radar sensors. The associated point cloud of the determined trajectory is obtained from the real-time detection radar point cloud, and the live target is detected and tracked based on the associated point cloud; For the determined track in the real-time detection radar point cloud that has no associated point cloud, the determined track is determined as a disappeared track based on the changes in the real-time detection radar point cloud in the space surrounding the last position of the determined track and a preset first determination rule. as well as The live target is detected and tracked based on the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappeared flight path; The process of detecting and tracking the live target based on the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappeared flight path includes: The activity level of the space surrounding the last position of the vanished trajectory is determined based on the energy of the real-time detected radar point cloud in the space surrounding the last position of the vanished trajectory; and The activity level is used to determine whether a living target exists in the space surrounding the last location of the disappeared flight path; The process of determining the activity level of the space surrounding the last position of the vanished track based on the energy of the real-time detected radar point cloud in the space surrounding the last position of the vanished track includes: The static degree of the space surrounding the last position of the vanished trajectory is determined based on the energy of the real-time detected radar point cloud in the space surrounding the last position of the vanished trajectory; and The activity level of the space around the last position of the vanished track is determined based on the energy of the real-time detected radar point cloud and the staticity of the space around the last position of the vanished track. The process of determining the activity level of the space surrounding the last position of the vanished trajectory based on the energy of the real-time detected radar point cloud and the stationarity of the space surrounding the last position of the vanished trajectory includes: Initial values ​​are set for the stillness and activity. When the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished track is detected to be not less than a preset first energy threshold, it is determined whether the activity value is less than a preset upper limit value. If it is less, the activity value is increased; if it is not less, the activity value remains unchanged. And when the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished track is detected to be less than a preset first energy threshold, it is determined whether the stillness value is less than a preset upper limit value. If it is less, the stillness value is increased; if it is not less, the activity value is decreased and the stillness value is reset to the initial value. Or Initial values ​​are set for the stillness and activity. When the maximum energy value of the real-time detection radar point cloud in the space surrounding the last position of the vanished trajectory is less than a preset second energy threshold, the accumulated energy of the real-time detection radar point cloud in the space surrounding the last position of the vanished trajectory from a second preset time period to the current time is calculated. When the accumulated energy is not less than a preset first energy threshold, it is determined whether the value of the activity is less than a preset upper limit of activity. If it is less, the value of the activity is increased; if it is not less, the value of the activity remains unchanged. And when the accumulated energy is less than the preset first energy threshold, it is determined whether the value of the stillness is less than a preset upper limit of stillness. If it is less, the value of the stillness is increased; if it is not less, the value of the activity is decreased and the value of the stillness is reset to the initial value.

2. The target detection and tracking method according to claim 1, characterized in that, Also includes: When there is no associated point cloud in the real-time detection radar point cloud, the determined track is determined to be a deleted track based on the changes in the real-time detection radar point cloud in the space surrounding the last position of the determined track and the preset second determination rule; or the disappeared track is determined to be a deleted track based on the changes in the real-time detection radar point cloud in the space surrounding the last position of the disappeared track and the second determination rule. as well as The deleted track is deleted, and the same live target is detected and tracked based on the newly determined track that corresponds to the live target of the deleted track and the real-time detection radar point cloud.

3. The target detection and tracking method according to claim 1, characterized in that, The process of determining the trajectory representing the live target from the trajectories of at least one suspected live target based on the position and velocity of the trajectory of at least one suspected live target acquired from the real-time detection radar point cloud, and the number of consecutive frames of the real-time detection radar point cloud containing the associated point cloud of the suspected live target's trajectory, includes: The trajectory of the at least one suspected live target is determined as a transient trajectory; the current displacement of the transient trajectory from the initial position to the current position and the current velocity of the transient trajectory are calculated; and the transient trajectory corresponding to the number of consecutive frames greater than a first preset number of frames, the current displacement greater than a preset displacement, and the current velocity greater than a preset velocity are determined as the determined trajectory. The process of determining the determined trajectory as a vanished trajectory based on the changes in the real-time radar point cloud surrounding the last position of the determined trajectory and a preset first determination rule includes: If the duration during which no associated point cloud is found in the real-time detection radar point cloud exceeds a first preset duration, the determined trajectory is judged as a vanished trajectory. The target detection and tracking method further includes: if the number of consecutive frames of the real-time detection radar point cloud that does not have the associated point cloud of the transient track is greater than a second preset number of frames, then the transient track is determined to be a deleted track, and the deleted track is deleted.

4. The target detection and tracking method according to claim 1, characterized in that, Also includes: Before determining the determined trajectory as a vanished trajectory based on the changes in the real-time radar point cloud around the last position of the determined trajectory and a preset first determination rule. The attitude information of the live target corresponding to the determined track is obtained based on the pitch direction energy of the real-time detected radar point cloud. The process of detecting and tracking the live target based on the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappeared flight path includes: The live target is detected and tracked based on the attitude information of the live target corresponding to the disappeared trajectory and the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappeared trajectory.

5. The target detection and tracking method according to claim 2, characterized in that, The process of determining the determined track as a deleted track based on the changes in the real-time detected radar point cloud around the last position of the determined track and the preset second determination rule includes: When a new determined track is determined based on the real-time detection radar point cloud of the space surrounding the last position of the determined track, the distance between the new determined track and the determined track is used to determine whether the new determined track and the determined track are successfully matched. If they are successfully matched, the determined track is determined to be a deleted track. The process of determining the disappeared track as a deleted track based on the changes in the real-time radar point cloud around the last position of the disappeared track and the second determination rule includes: When a new track is determined based on the real-time detection radar point cloud of the space surrounding the last position of the disappeared track, the distance between the new track and the disappeared track is used to determine whether the new track and the disappeared track are successfully matched. If they are successfully matched, the track is determined to be a deleted track.

6. A target detection and tracking device, used to execute the target detection and tracking method according to any one of claims 1 to 5, characterized in that, The trajectory acquisition module is used to determine a definite trajectory representing the trajectory of a live target from the trajectories of at least one suspected live target in the room, based on the position and velocity of the trajectory of at least one suspected live target in the room acquired based on real-time detection radar point cloud, and the number of consecutive frames of the real-time detection radar point cloud in which the trajectory of the suspected live target is continuously present. The real-time detection radar point cloud is a radar point cloud obtained by real-time detection of the room using radar sensors. The first tracking module is used to obtain the associated point cloud of the determined trajectory from the real-time detection radar point cloud, and to detect and track the live target based on the associated point cloud; The vanished track determination module is used to determine the determined track as a vanished track based on the changes in the real-time detected radar point cloud around the last position of the determined track and a preset first determination rule. as well as The second tracking module is used to detect and track the living target based on the energy of the real-time detection radar point cloud in the space surrounding the last position of the disappeared flight path.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the target detection and tracking method as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the target detection and tracking method as described in any one of claims 1 to 5.

Images