An indoor multi-target tracking method, device, equipment and storage medium

Abstract

The application discloses an indoor multi-target tracking method, device and equipment and a storage medium. The method comprises the following steps: receiving indoor discrete echo signals collected by a radar, and performing long-short time preprocessing on the indoor discrete echo signals to obtain an indoor measurement signal set; obtaining an estimated trajectory of at least one indoor target, and associating each measurement signal in the indoor measurement signal set with the estimated trajectory of the at least one indoor target to obtain an associated trajectory of the at least one indoor target; updating the associated trajectory by using a Kalman filtering algorithm to obtain an updated trajectory of the at least one indoor target; and tracking the at least one indoor target according to the updated trajectory. According to the technical scheme, the indoor multi-target can be stably and real-timely tracked, and the detection probability of the indoor target and the accuracy of target tracking are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an indoor multi-target tracking method, apparatus, device and storage medium. Background Technology

[0002] With the rapid development of digitalization and automation technologies, the intelligence of indoor environments has become increasingly important. In this trend, millimeter-wave radar has attracted much attention due to its ability to acquire high-resolution target information, its immunity to environmental factors such as light and temperature, and its advantages of privacy protection and low cost, making it an ideal sensor for indoor multi-target tracking.

[0003] Unlike traditional point target tracking, indoor human targets appear as multiple scattering points in front of millimeter-wave radar, generating a series of point clouds. The tracking problem for this is also known as extended target tracking. Existing research on extended target tracking is generally divided into two categories: The first category uses clustering techniques to convert a series of point clouds into different measurement points, thus reconstructing the extended target tracking problem into a point target tracking problem, and using the traditional point target tracking framework for tracking. However, this method is only suitable for simple scenarios. In complex scenarios, there are many targets, which easily generate a lot of clutter and multipath effects. When the target distance is close, the point clouds of multiple targets are easily clustered into the same cluster, leading to trajectory association errors.

[0004] The second type is group tracking. This method preserves the initial point cloud and performs correlation tracking on the point cloud, which is beneficial for trajectory maintenance in various scenarios. However, due to the limitation of sensor resolution, the correlation gate setting cannot be too small. When multiple people are nearby, it may also be associated with points generated by other targets, which can easily lead to point cloud merging. It may also be associated with noise points, resulting in trajectory association errors or target loss, thus reducing the accuracy of indoor multi-target tracking. Summary of the Invention

[0005] This invention provides an indoor multi-target tracking method, apparatus, device, and storage medium, which can solve the problem of reduced accuracy in indoor target tracking due to the easy occurrence of target trajectory association errors or target loss in complex indoor environments.

[0006] According to one aspect of the present invention, an indoor multi-target tracking method is provided, comprising:

[0007] Receive the indoor discrete echo signal collected by the radar, and perform long and short time preprocessing on the indoor discrete echo signal to obtain the indoor measurement signal set.

[0008] The estimated trajectory of at least one indoor target is obtained, and each measurement signal in the indoor measurement signal set is associated with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target;

[0009] The associated trajectory is updated by using the Kalman filter algorithm to obtain the updated trajectory of at least one indoor target;

[0010] At least one target indoors is tracked based on the updated trajectory.

[0011] According to another aspect of the present invention, an indoor multi-target tracking device is provided, the indoor multi-target tracking device comprising:

[0012] The measurement module is used to receive the indoor discrete echo signal collected by the radar and perform long and short time preprocessing on the indoor discrete echo signal to obtain the indoor measurement signal set.

[0013] The trajectory acquisition module is used to acquire the estimated trajectory of at least one indoor target and associate each measurement signal in the indoor measurement signal set with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target.

[0014] The trajectory update module is used to update the associated trajectory using a Kalman filter algorithm to obtain the updated trajectory of at least one indoor target;

[0015] A trajectory tracking module is used to track at least one target indoors based on the updated trajectory.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the indoor multi-target tracking method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the indoor multi-target tracking method according to any embodiment of the present invention.

[0021] This invention, through receiving indoor discrete echo signals acquired by radar and performing long- and short-time preprocessing on these signals to obtain an indoor measurement signal set, acquires the estimated trajectory of at least one indoor target and associates each measurement signal in the indoor measurement signal set with the estimated trajectory of the at least one indoor target to obtain an associated trajectory of the at least one indoor target. The associated trajectory is then updated using a Kalman filter algorithm to obtain an updated trajectory of the at least one indoor target. Based on the updated trajectory, the at least one indoor target is tracked. This invention solves the problem of reduced indoor target tracking accuracy caused by trajectory association errors or target loss in complex indoor environments. It enables robust real-time tracking of multiple indoor targets in complex environments, improving the detection probability and tracking accuracy of indoor targets.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments 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.

[0024] Figure 1 This is a flowchart of an indoor multi-target tracking method according to Embodiment 1 of the present invention;

[0025] Figure 2 This is a schematic diagram of a method for obtaining an indoor measurement signal set according to Embodiment 1 of the present invention;

[0026] Figure 3 This is a schematic diagram of a secondary correlation and update of an estimated trajectory in Embodiment 1 of the present invention;

[0027] Figure 4 This is a schematic diagram of an indoor multi-target tracking method according to Embodiment 1 of the present invention;

[0028] Figure 5 This is a schematic diagram of the structure of an indoor multi-target tracking device according to Embodiment 2 of the present invention;

[0029] Figure 6 This is a schematic diagram of the structure of an electronic device according to Embodiment 3 of the present invention. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0033] Example 1

[0034] Figure 1 This is a flowchart of an indoor multi-target tracking method according to Embodiment 1 of the present invention. This embodiment is applicable to real-time tracking of targets indoors. The method can be executed by the indoor multi-target tracking device in this embodiment, which can be implemented in software and / or hardware. Figure 1 As shown, the specific steps include the following:

[0035] S110 receives the indoor discrete echo signal collected by the radar and performs long and short time preprocessing on the indoor discrete echo signal to obtain the indoor measurement signal set.

[0036] Among them, the radar can be a MIMO (Multiple-Input Multiple-Output) radar with 2 transmit and 4 receive signals to generate indoor discrete echo signals.

[0037] Specifically, the system receives indoor discrete echo signals acquired by radar and performs long-time preprocessing and short-time preprocessing on the indoor discrete echo signals. In the preprocessing stage, clutter is estimated and suppressed using echo signals of different durations. The duration of long-time preprocessing is longer than that of short-time preprocessing. Long-time preprocessing can obtain the target component of the weak signal, while short-time preprocessing can obtain the component of the strong signal. Based on this, strong and weak measurement signals are generated, and thus an indoor measurement signal set is obtained.

[0038] Optionally, the indoor discrete echo signal is subjected to long-time and short-time preprocessing to obtain an indoor measurement signal set, including:

[0039] The indoor discrete echo signal is subjected to short-time preprocessing to obtain a short-time echo signal;

[0040] The indoor discrete echo signal is subjected to long-time preprocessing to obtain a long-time echo signal;

[0041] The short-time range angle spectrum is obtained based on the short-time echo signal, and the long-time range angle spectrum is obtained based on the long-time echo signal;

[0042] The first algorithm is used to detect the short-time range angle spectrum to obtain a set of two-dimensional point cloud measurement signals;

[0043] The second algorithm is used to detect the long-term distance angle spectrum to obtain a set of effective peak measurement signals;

[0044] The indoor measurement signal set is determined based on the effective peak measurement signal set and the two-dimensional point cloud measurement signal set.

[0045] The first algorithm can be the CFAR (Constant False Alarm Rate) algorithm, and the second algorithm can be the effective peak algorithm.

[0046] Specifically, the radar continuously transmits electromagnetic wave signals into the indoor space. These signals are scattered or reflected by objects and then received by the radar. After passing through a signal amplifier, mixer, and ADC sampling, a discrete echo signal containing fast time dimension, slow time dimension, and antenna dimension information is obtained, which 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. A fast Fourier transform is performed on the received radar echo in the fast time dimension to obtain the range-slow time dimension-antenna dimension echo signal. Where r∈[1,N], it represents distance cell sampling; it should be noted that the echo signal The echo signal contains static clutter from objects such as the ground, walls, and furniture, which can easily drown out the signal from the human target. Therefore, preprocessing of the echo signal is necessary, i.e., clutter suppression. Short-time preprocessing estimates the clutter pair using a single frame signal. Clutter suppression is performed to obtain the short-time echo signal y. ts (m,r,k) This signal can highlight moving targets with a high signal-to-noise ratio.

[0047] Specifically, long-term preprocessing is used to estimate clutter pairs from multiple frames of signal. Clutter suppression is performed to obtain the long-time echo signal y. tl (m,r,k) This signal can retain the high signal-to-noise ratio moving target signal while also retaining the weak target signal such as resting target or occluded target.

[0048] Specifically, the Capon algorithm is used to analyze the short-time echo signal y. ts Angle estimation is performed on each distance cell in (m,r,k) to obtain the short-time range-angle spectrum M. s (r,θ); Capon algorithm is used to process long-time echo signal y tl Angle estimation is performed on each distance cell in (m,r,k) to obtain the long-term distance angle spectrum M. l (r,θ), where θ represents the azimuth angle of the signal.

[0049] Specifically, the short-time range angular spectrum M s The range angle spectrum (r, θ) mainly contains the range angle spectrum corresponding to moving human targets with high signal-to-noise ratio. The first algorithm, such as the CFAR algorithm, is used to analyze the short-time range angle spectrum M. s The target is detected at (r, θ), and its signal-to-noise ratio (SNR) is estimated to obtain the measurement information p. s (r,θ,SNR) converts the polar coordinates of the point cloud to Cartesian coordinates, thus obtaining the set of two-dimensional point cloud measurement signals. Where x and y represent the horizontal and vertical coordinates of the target in the room corresponding to the two-dimensional point cloud measurement signal, respectively.

[0050] Specifically, long-term distance angular spectrum M l The spectrum (r, θ) includes not only moving human targets with high signal-to-noise ratio (SNR) but also resting or occluded human targets with low SNR. Because the SNR levels of the targets within the spectrum vary significantly, traditional CFAR detection algorithms are difficult to tune. Therefore, a second algorithm, such as the effective peak algorithm, is chosen for the long-term angular spectrum M. l (r,θ) is used for detection to obtain the effective peak set p. l (r,θ,a), where a represents the effective peak at M l The amplitude in (r,θ) is converted from the effective peak polar coordinates to Cartesian coordinates, thus obtaining the set of effective peak measurement signals.

[0051] Specifically, the set of two-dimensional point cloud measurement signals obtained by the first algorithm and the set of effective peak measurement signals obtained by the second algorithm together constitute the indoor measurement signal set.

[0052] Commonly used clutter suppression algorithms include moving target indication (MTI) and intra-frame cancellation. These algorithms can suppress static clutter in the signal and preserve the moving target signal component. However, in indoor daily activities, people are often in a static state, such as standing, sitting, or lying down. In these situations, the human target signal component is easily confused with the surrounding static clutter. In multi-person scenarios, people are easily obstructed, resulting in a weak target signal. Therefore, using traditional clutter suppression algorithms can easily lead to the loss of signals from these static or weak targets. Short-time preprocessing is performed on the indoor discrete echo signal to obtain a short-time echo signal. Long-time preprocessing is then performed on the indoor discrete echo signal to obtain a long-time echo signal. The long-time range-angle spectrum is obtained based on the long-time echo signal. A first algorithm is used to detect the short-time range-angle spectrum to obtain a two-dimensional point cloud measurement signal set. A second algorithm is used to detect the long-time range-angle spectrum to obtain a set of effective peak measurement signals. The indoor measurement signal set is determined based on the set of effective peak measurement signals and the set of two-dimensional point cloud measurement signals. This method can meet the needs of multi-target tracking in various complex scenarios, retain the signal components of various human targets as much as possible, and improve the accuracy of obtaining the set of two-dimensional point cloud measurement signals and the set of effective peak measurement signals.

[0053] Optionally, the indoor measurement signal set is determined based on the effective peak measurement signal set and the two-dimensional point cloud measurement signal set, including:

[0054] Cluster the measurement signals in the two-dimensional point cloud measurement signal set to obtain a strong measurement signal set;

[0055] The measurement signals in the effective peak measurement signal set are matched with the measurement signals in the strong measurement signal set. The successfully matched measurement signals are removed from the effective peak measurement signal set to obtain the weak measurement signal set.

[0056] The indoor measurement signal set is obtained based on the strong measurement signal set and the weak measurement signal set.

[0057] Among them, strong measurement signals represent measurements with a high signal-to-noise ratio, and are considered to have a higher reliability in representing human targets. In contrast, weak measurement signals are considered to have a lower reliability in representing human targets, and can serve as a supplement to subsequent target tracking and processing.

[0058] Specifically, The targets contained within have a high signal-to-noise ratio and can be directly used for subsequent target tracking processing. Therefore, the density-based clustering DBSCAN algorithm can be used to process the two-dimensional point cloud measurement signal set. Clustering of the measurement signals in the data generates a point cloud cluster set. That is, a set of strong measurement signals.

[0059] Specifically, The target signal-to-noise ratio levels included vary considerably, not only including those... Repeated measurement signals may contain non-human target information such as multipath propagation, which can easily affect the results of subsequent target tracking processing if used directly. Therefore, measurement signals in the effective peak measurement signal set can be matched with measurement signals in the strong measurement signal set based on the nearest neighbor criterion, and the successfully matched measurement signals can be removed from the effective peak measurement signal set. Remove from the middle to obtain a new set of effective peaks. That is, a set of weak measurement signals.

[0060] Specifically, the obtained strong measurement signal set and the obtained weak measurement signal set are used to construct the indoor measurement signal set. In a specific example, Figure 2 This is a schematic diagram of a method for obtaining an indoor measurement signal set according to Embodiment 1 of the present invention, as shown below. Figure 2 As shown, after short-time preprocessing of the indoor discrete echo signal, a short-time echo signal is obtained. The short-time echo signal is then processed to obtain a short-time range-angle spectrum. The CFAR detection algorithm is used to detect the short-time range-angle spectrum, resulting in a two-dimensional point cloud measurement signal set. DBSCAN clustering is then performed on the two-dimensional point cloud measurement signal set to obtain a strong measurement signal set. Simultaneously, long-time preprocessing of the indoor discrete echo signal is performed to obtain a long-time echo signal. The long-time echo signal is then processed to obtain a long-time range-angle spectrum. The effective peak detection algorithm is used to detect the long-time range-angle spectrum, resulting in an effective peak measurement signal set. The measurement signals in the effective peak measurement signal set are matched with the measurement signals in the strong measurement signal set using the nearest neighbor criterion. Successfully matched measurement signals are removed from the effective peak measurement signal set, resulting in a weak measurement signal set. The strong and weak measurement signal sets together constitute the indoor measurement signal set.

[0061] S120: Obtain the estimated trajectory of at least one indoor target, and associate each measurement signal in the indoor measurement signal set with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target.

[0062] The target is an indoor human body, which may include human bodies in motion or at rest. The estimated trajectory of at least one target is the estimated trajectory obtained by predicting the trajectory state of the target, and the associated trajectory of at least one target is the trajectory after the estimated trajectory is successfully associated with the measurement signal.

[0063] Specifically, a two-dimensional uniform linear motion model can be used to track indoor targets. Since there is a certain error between radar measurements and the actual target state, a Kalman filter algorithm can be used for state estimation to obtain the estimated trajectory of at least one indoor target. After obtaining the estimated trajectory of at least one indoor target, each measurement signal in the indoor measurement signal set is attempted to be associated with the estimated trajectory of at least one indoor target. The estimated trajectory that has been associated with the measurement signals is called the associated trajectory. It should be noted that if there are remaining unassociated strong and weak measurement signals in the indoor measurement signal set, the unassociated strong measurement signals are selected as one of the trajectory starting conditions, while the weak measurement signals are not involved. When the strong measurement signals satisfy the N / M criterion, that is, when N frames of strong measurement signals are successfully clustered in M ​​frames, it is considered the starting point of the estimated trajectory of a new target, and trajectory association with the next frame of discrete echo signals is performed.

[0064] Optionally, each measurement signal in the indoor measurement signal set is associated with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target, including:

[0065] Obtain the indoor prior boundary, and filter the indoor measurement signal set according to the indoor prior boundary to obtain the filtered indoor measurement signal set.

[0066] Each measurement signal in the filtered set of indoor measurement signals is associated with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target.

[0067] Among them, the indoor prior boundary includes the indoor outer boundary and the indoor inner boundary.

[0068] Specifically, before performing indoor multi-target tracking, a single human target is used to circle the indoor boundary, and the measurement signal closest to the radar is taken as the target to be tracked, thus generating a circumference trajectory. This trajectory is then used as a reference for the indoor prior boundary, generating the inner and outer boundaries of the indoor prior boundary. The area enclosed by the outer boundary is slightly larger than the circumference trajectory, serving to filter the measurement signals. That is, measurement signals outside the outer boundary are considered multipath and are not processed further. The inner boundary is the area within the circumference trajectory. Measurement signals within this area have high confidence and directly participate in subsequent tracking processing. Therefore, the indoor measurement signal set can be filtered based on the indoor prior boundary to obtain a filtered indoor measurement signal set, which can simplify the indoor measurement signals and improve the efficiency of target tracking processing.

[0069] Specifically, each measurement signal in the filtered set of indoor measurement signals is attempted to be associated with the estimated trajectory of at least one indoor target, and the successfully associated estimated trajectory is called the associated trajectory.

[0070] By acquiring the indoor prior boundary and filtering the indoor measurement signal set according to the indoor prior boundary, a filtered indoor measurement signal set is obtained. Each measurement signal in the filtered indoor measurement signal set is associated with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target. This can avoid the multipath effect that is prone to exist in indoor scenes detected by millimeter-wave radar, especially at indoor edges such as walls. At the same time, the radar detection range often exceeds the indoor activity range, and multipath and other non-human targets may also appear beyond the indoor boundary area. This effectively reduces the impact of non-human targets on the accuracy of human target tracking.

[0071] Optionally, each measurement signal in the filtered set of indoor measurement signals is associated with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target, including:

[0072] Each strong measurement signal in the filtered indoor measurement signal set is initially associated with the estimated trajectory of at least one indoor target. If the initial association is successful, the first associated trajectory of at least one indoor target is obtained.

[0073] If the initial association fails, and the unassociated estimated trajectory is a definite trajectory, then the estimated trajectory will be associated a second time with each weak measurement in the filtered set of indoor measurement signals.

[0074] If the secondary association is successful, a second associated trajectory of at least one target indoors will be obtained;

[0075] The first and second associated trajectories are identified as associated trajectories.

[0076] In this set of filtered indoor measurement signals, the specific human body being measured is uncertain. Therefore, it is necessary to associate the estimated trajectory of at least one target with the measurement signals. The first associated trajectory is the estimated trajectory that successfully associates with a strong measurement signal during the initial association. The second associated trajectory is the estimated trajectory that successfully associates with a weak measurement signal during the secondary association. The determined trajectory is the trajectory with high confidence, that is, the trajectory with a confidence level greater than a preset confidence threshold, and is considered to have a high probability of representing a human target.

[0077] Specifically, each strong measurement signal in the filtered set of indoor measurement signals is initially associated with the estimated trajectory of at least one indoor target. According to the nearest neighbor criterion, the strong measurement signal is assigned to the nearest estimated trajectory. If the estimated trajectory is successfully associated with the strong measurement signal, the first associated trajectory of at least one indoor target is obtained.

[0078] Specifically, if the initial association fails and the unassociated estimated trajectory is determined to be a definite trajectory, then it is secondary-associated with each weak measurement in the filtered indoor measurement signal set. For each weak measurement signal, it is assigned to the nearest unassociated estimated trajectory according to the nearest neighbor criterion. If the secondary association is successful, the estimated trajectory with successful secondary association is determined as the second associated trajectory. It should be noted that if the initial association fails and the unassociated estimated trajectory is a transient trajectory, then no secondary association is performed. That is, there is no associated measurement signal in the discrete echo signal of the current frame for this estimated trajectory. Here, a transient trajectory refers to a trajectory with low confidence, that is, a trajectory with a confidence level less than or equal to a preset confidence threshold. It is considered to reflect a low confidence level for human targets and may be a multipath or other interfering target. It should be noted that the trajectories of all new targets are initialized as transient trajectories.

[0079] Specifically, both the first and second associated trajectories are associated trajectories of the target.

[0080] By first associating the strong measurement signal with the estimated trajectory of at least one indoor target to obtain the first associated trajectory, and then associating the unassociated estimated trajectory with the weak measurement signal to obtain the second associated trajectory, the accuracy of detection and tracking of weak targets can be improved, and association errors in trajectory proximity situations such as trajectory intersection and trajectory occlusion can be reduced.

[0081] Optionally, each strong measurement signal in the filtered set of indoor measurement signals is initially associated with the estimated trajectory of at least one indoor target. If the initial association is successful, a first associated trajectory of at least one indoor target is obtained, including:

[0082] Based on the nearest neighbor criterion, each strong measurement signal in the filtered set of indoor measurement signals is assigned to the nearest estimated trajectory;

[0083] If the estimated trajectory has a unique strong measurement signal that has been assigned, then the estimated trajectory with the unique strong measurement signal is determined to be the first associated trajectory of at least one indoor target;

[0084] If the estimated trajectory has at least two assigned strong measurement signals, the strong measurement signal with the highest signal-to-noise ratio is determined as the unique strong measurement signal of the estimated trajectory, and the estimated trajectory of the unique strong measurement signal is determined as the first associated trajectory of at least one indoor target.

[0085] Specifically, based on the nearest neighbor criterion, each strong measurement signal in the filtered set of indoor measurement signals is assigned to the nearest estimated trajectory. For an estimated trajectory that is successfully assigned a strong measurement signal, if the estimated trajectory is assigned only one strong measurement signal, then the estimated trajectory is determined to be uniquely associated with this unique strong measurement signal. That is, the estimated trajectory with the unique strong measurement signal is determined to be the first associated trajectory of at least one indoor target. If the estimated trajectory has at least two assigned strong measurement signals, that is, the estimated trajectory is assigned multiple strong measurement signals, then the strong measurement signal with the highest signal-to-noise ratio among the multiple assigned strong measurement signals is determined to be the unique strong measurement signal of the estimated trajectory, and the estimated trajectory with the unique strong measurement signal is determined to be the first associated trajectory of at least one indoor target.

[0086] It should be noted that the nearest neighbor criterion is also used when performing secondary correlation between the estimated trajectory and each weak measurement in the filtered indoor measurement signal set. The method for determining the second correlation trajectory is similar to that for determining the first correlation trajectory, and will not be elaborated here.

[0087] By assigning each strong measurement signal in the filtered indoor measurement signal set to the nearest estimated trajectory based on the nearest neighbor criterion; if the estimated trajectory has a unique strong measurement signal already assigned, then the estimated trajectory of the unique strong measurement signal is determined as the first associated trajectory of at least one indoor target; if the estimated trajectory has at least two assigned strong measurement signals, then the strong measurement signal with the highest signal-to-noise ratio is determined as the unique strong measurement signal of the estimated trajectory, and the estimated trajectory of the unique strong measurement signal is determined as the first associated trajectory of at least one indoor target. This method can more accurately obtain the measurement signals associated with the estimated trajectory of at least one target, thereby improving the efficiency of obtaining associated trajectories.

[0088] S130, the associated trajectory is updated using the Kalman filter algorithm to obtain the updated trajectory of at least one indoor target.

[0089] Among them, the Kalman filter algorithm can effectively remove random noise, obtain a more accurate position of the human target, and achieve precise target tracking. The updated trajectory is the target trajectory updated by the association estimate of at least one target using the Kalman filter algorithm.

[0090] Specifically, after obtaining the associated trajectories, the Kalman filter algorithm is used to update the associated trajectories to obtain the updated trajectory of at least one indoor target.

[0091] Optionally, the associated trajectory is updated using a Kalman filter algorithm to obtain an updated trajectory for at least one indoor target, including:

[0092] Acquire the velocity and amplitude of the measurement signal associated with the track;

[0093] Based on the speed and amplitude of the measurement signal, the measurement signal is divided into dynamic measurement signal and micro-motion measurement signal;

[0094] The Kalman filter algorithm is used to update the associated trajectory of the dynamic measurement signal to obtain the updated trajectory of at least one indoor target.

[0095] By extending the associated trajectory of the micro-motion measurement signal by a preset period, an updated trajectory of at least one indoor target can be obtained.

[0096] The dynamic measurement signal is a measurement signal with a velocity greater than a velocity threshold and an amplitude greater than an amplitude threshold, while the micro-motion measurement signal is a measurement signal with a velocity less than or equal to a velocity threshold and / or an amplitude less than or equal to an amplitude threshold. Measurement signals include strong and weak measurement signals. The preset period can be set according to actual conditions.

[0097] Specifically, the velocity and amplitude of the measurement signals associated with the associated trajectory are obtained by means of a rangefinder or spectrum analysis. The measurement signals include strong measurement signals and weak measurement signals. If the associated trajectory is the first associated trajectory, the corresponding measurement signal is a strong measurement signal. If the associated trajectory is the second associated trajectory, the corresponding measurement signal is a weak measurement signal.

[0098] Specifically, based on the speed and amplitude of the measurement signal, the measurement signal is divided into dynamic measurement signal and micro-motion measurement signal. The measurement signal with a speed greater than the speed threshold and an amplitude greater than the amplitude threshold is identified as a dynamic measurement signal, and the measurement signal with a speed less than or equal to the speed threshold and / or an amplitude less than or equal to the amplitude threshold is identified as a micro-motion measurement signal.

[0099] Specifically, if the measurement signal is a dynamic measurement signal, the associated trajectory of the dynamic measurement signal is updated by using a Kalman filter to obtain an updated trajectory; if the measurement signal is a micro-motion measurement signal, the associated trajectory of the micro-motion measurement signal is extended by a preset period to obtain an updated trajectory.

[0100] In a specific example Figure 3 This is a schematic diagram of a secondary correlation and update of an estimated trajectory according to Embodiment 1 of the present invention, as shown below. Figure 3As shown, the estimation of at least one indoor target is initially associated with strong measurement signals in the indoor measurement signal set based on the nearest neighbor criterion. Strong measurement signals are assigned to the nearest trajectory. For an estimated trajectory successfully assigned a strong measurement signal, if the estimated trajectory is assigned only one strong measurement signal, it is determined that the estimated trajectory is associated with the unique strong measurement signal, resulting in a first associated trajectory. If the estimated trajectory has at least two assigned strong measurement signals, the strong measurement signal with the highest signal-to-noise ratio is determined as the unique strong measurement signal for that estimated trajectory, resulting in a first associated trajectory. For estimated trajectories that fail to be initially associated, and for which no determination is made... The trajectory is then correlated a second time with weak measurement signals in the indoor measurement signal set, also based on the nearest neighbor criterion and the high signal-to-noise ratio priority correlation method, to obtain a second correlated trajectory. The first and second correlated trajectories are collectively referred to as the correlated trajectory. The velocity and amplitude of the measurement signals correlated with the correlated trajectory are obtained. Based on the velocity and amplitude of the measurement signals, the measurement signals are divided into dynamic measurement signals and micro-motion measurement signals. The correlated trajectory associated with the dynamic measurement signals is updated according to the Kalman filter algorithm to obtain the updated trajectory. The correlated trajectory associated with the micro-motion measurement signals is extended according to a preset period to obtain the updated trajectory.

[0101] S140, track at least one target indoors based on the updated trajectory.

[0102] Specifically, after obtaining the updated trajectory for each target, each target indoors is tracked according to the updated trajectory. Alternatively, the updated trajectory can be evaluated based on the prior boundaries of the indoor environment to determine if it is a false trajectory. If it is a false trajectory, it is deleted, and the target corresponding to it is tracked according to the updated trajectory after deleting the false trajectory. Another approach is to set trajectory management rules, manage the updated trajectories according to these rules, and track the target corresponding to it based on the managed updated trajectory.

[0103] Optionally, tracking at least one indoor target based on the updated trajectory includes:

[0104] If the updated trajectory is a false trajectory, then the false trajectory is converted into a deleted trajectory;

[0105] The updated trajectories, excluding the false trajectories, are classified according to their confidence levels to obtain transient trajectories and definitive trajectories;

[0106] If no associated measurement signal exists in the transient trajectory within the first preset time period, the transient trajectory is converted into a deleted trajectory.

[0107] If the trajectory is determined to have no associated measurement signal within a second preset time period, the determined trajectory is converted into a deleted trajectory, wherein the first preset time period is less than the second preset time period;

[0108] The deleted trajectory is deleted, and at least one indoor target is tracked based on the transient trajectory and the determined trajectory.

[0109] False trajectories are those generated due to erroneous tracking of non-existent human targets. Multipath propagation or noise in indoor scenes can severely impact multi-target tracking results, causing target trajectory loss or false trajectories. Deleted trajectories are those that need to be removed and will not be processed further. The first and second preset times can be set according to actual conditions, with the first preset time being shorter than the second preset time.

[0110] Specifically, if false trajectories exist in the updated trajectory, they are identified as deletion trajectories and deleted. The updated trajectories, excluding false trajectories, are classified according to their confidence level, resulting in transient and definitive trajectories. Transient trajectories represent those with low confidence levels, i.e., confidence levels less than or equal to a preset confidence threshold. Definitive trajectories represent those with high confidence levels, i.e., confidence levels greater than the preset confidence threshold. It should be noted that when the correlation between a transient trajectory and the measurement signal is higher than a preset correlation level and meets the preset trajectory length, the transient trajectory can be converted into a definitive trajectory.

[0111] It is important to note that, further, the determined trajectory can be categorized into dynamic and static trajectories based on the position change and velocity magnitude. If the rate of position change is greater than a preset rate of change and the velocity is greater than a preset velocity, the determined trajectory is identified as a dynamic trajectory; otherwise, it is a static trajectory. To reduce the probability of a static trajectory being associated with a measurement signal, the association distance of the static trajectory can be reduced. Furthermore, the updated trajectory can be categorized into unobstructed and obstructed trajectories based on the positional relationship between trajectories. If the first updated trajectory precedes the second updated trajectory, the first updated trajectory is an unobstructed trajectory, and the second updated trajectory is an obstructed trajectory. Generally, multipath propagation tends to occur behind the target, and obstructed trajectories have lower signal-to-noise ratios. Obstructed trajectories are also more susceptible to the influence of multipath propagation from other trajectories, leading to trajectory loss or the generation of false trajectories. Therefore, when an updated trajectory is obstructed or has been obstructed in a short period of time, it is an important basis for determining false trajectories.

[0112] Specifically, if there is no associated measurement signal in the transient trajectory within the first preset time period, it is converted into a deletion trajectory and deleted. If it is determined within the second preset time period that there is no associated measurement signal in the trajectory, it is converted into a deletion trajectory and deleted.

[0113] Specifically, the deleted trajectory is removed, and the target is continuously tracked based on the transient trajectory and the determined trajectory.

[0114] By dividing the updated trajectory into transient trajectories, deterministic trajectories, and false trajectories for trajectory management, and then tracking the target based on the updated trajectory after trajectory management, the quality and reliability of the trajectory corresponding to the target can be improved.

[0115] Optionally, if the updated trajectory is a false trajectory, then the false trajectory is converted into a deleted trajectory, including:

[0116] Obtain the outer and inner boundaries corresponding to the indoor prior boundary. If the updated trajectory is between the outer and inner boundaries, then the first preset coefficient threshold is determined as the coefficient threshold.

[0117] If the updated trajectory is not between the outer boundary and the inner boundary, then the second preset coefficient threshold is determined as the coefficient threshold, wherein the first preset coefficient threshold is greater than the second preset coefficient threshold.

[0118] If the distance between any two updated trajectories is less than or equal to the first distance threshold and greater than the second distance threshold, then obtain the velocity direction cross-correlation coefficient between any two updated trajectories.

[0119] If the cross-correlation coefficient of the velocity direction is greater than the coefficient threshold, then either of the two updated trajectories is determined to be a false trajectory.

[0120] If the distance between any two updated trajectories is less than or equal to the second distance threshold, then either of the two updated trajectories is determined to be a false trajectory.

[0121] The false trajectory is converted into a deleted trajectory.

[0122] The first preset coefficient threshold and the second preset coefficient threshold can be set according to actual conditions, with the first preset coefficient threshold being greater than the second preset coefficient threshold. Similarly, the first spacing threshold and the second spacing threshold can be set according to actual conditions, with the first spacing threshold being greater than the second spacing threshold.

[0123] Specifically, the outer and inner boundaries corresponding to the indoor prior boundary are obtained. If the updated trajectory is in the low confidence region between the outer and inner boundaries, the first preset coefficient threshold is set as the coefficient threshold. If the updated trajectory is not in the low confidence region between the outer and inner boundaries, the second preset coefficient threshold is set as the coefficient threshold. The first preset coefficient threshold is greater than the second preset coefficient threshold.

[0124] Specifically, the first spacing threshold can be understood as the minimum centroid distance between people. If the spacing between any two updated trajectories is less than or equal to the first spacing threshold and greater than the second spacing threshold, it is considered that a false trajectory may occur. Dynamic time warping algorithm or longest common subsequence algorithm can be used to obtain the velocity direction cross-correlation coefficient between the two updated trajectories. If the velocity direction cross-correlation coefficient is greater than the coefficient threshold, it indicates that the motion characteristics between the two updated trajectories are highly similar, and it is considered that a false trajectory has been generated. Then, any one of the two updated trajectories is determined to be a false trajectory. If the two updated trajectories are between the outer boundary and the inner boundary, the coefficient threshold is the first preset coefficient threshold. If the two updated trajectories are not between the outer boundary and the inner boundary, the coefficient threshold is the second preset coefficient threshold.

[0125] Specifically, if the distance between any two updated trajectories is less than or equal to the second distance threshold, it indicates that the two updated trajectories overlap. In this case, either of the two updated trajectories is directly identified as a false trajectory, and the updated trajectory identified as a false trajectory is converted into a deletion trajectory and deleted.

[0126] It should be noted that when a target's trajectory is lost during tracking, relaxing the trajectory initiation conditions and generating a new target trajectory in a timely manner can improve the situation to some extent. However, relaxing the trajectory initiation conditions increases the probability of generating false trajectories. Simultaneously, because the signal-to-noise ratio (SNR) of targets in weak measurement signals varies, they may include non-human targets such as multipath propagation, which can also generate false trajectories. Therefore, trajectory quality needs to be evaluated to determine the existence of false trajectories and suppress them. In secondary correlation, the premise for estimating the trajectory correlation with weak measurement signals is the inability to correlate with strong measurement signals. The estimated trajectories correlated with strong measurement signals are mostly low SNR targets, i.e., occluded targets or resting targets. Resting targets, in an unoccluded state, can generate weak measurement signals with higher SNR by extending the observation time, resulting in high-quality updated trajectories. However, occluded targets may lose some echo signals, resulting in low SNR, which is easily confused with multipath propagation generated by other targets, leading to correlation errors and the generation of false trajectories. Therefore, generally speaking, the generation of false tracks is related to occluded tracks. Specifically, when the updated track corresponding to a target is occluded and loses its signal, it may be associated with weak multipath measurement signals behind other targets, causing the gap between the two updated tracks to gradually decrease and continuously track the same human target. All of the above situations can cause false tracks to appear, leading to a decrease in the accuracy of indoor multi-target tracking.

[0127] Therefore, if the distance between any two updated trajectories is less than or equal to a first distance threshold and greater than a second distance threshold, the velocity direction cross-correlation coefficient between any two updated trajectories is obtained; if the velocity direction cross-correlation coefficient is greater than a coefficient threshold, then any one of the two updated trajectories is determined to be a false trajectory; if the distance between any two updated trajectories is less than or equal to the second distance threshold, then any one of the two updated trajectories is determined to be a false trajectory; converting the false trajectory into a deletion trajectory enables timely identification of false trajectories and their deletion, greatly reducing the probability of false trajectories appearing and improving the effectiveness and accuracy in complex multi-target tracking.

[0128] In a specific example Figure 4 This is a schematic diagram of an indoor multi-target tracking method according to Embodiment 1 of the present invention, as shown below. Figure 4 As shown, after obtaining the estimated trajectory of at least one indoor target and the set of indoor measurement signals, the estimated trajectory is first initially associated with the strong measurement signals in the indoor measurement signals. If the initial association fails and the estimated trajectory is a definite trajectory, the unassociated estimated trajectory is secondarily associated with the weak indoor measurement signals to obtain the associated trajectory. Simultaneously, the N / M criterion is applied to the strong measurement signals of the unassociated estimated trajectory to obtain the starting point of the new target's estimated trajectory. Trajectory association is then performed on the discrete echo signals of the next frame. The associated trajectory is updated to obtain the updated trajectory. Furthermore, based on prior boundaries, trajectory management is performed on the updated trajectory. Specifically, trajectory management includes dividing and delineating the updated trajectory. The updated trajectory can be divided into transient trajectories, definitive trajectories, and deleted trajectories. Definitive trajectories include dynamic trajectories, static trajectories, unobstructed trajectories, and obstructed trajectories. Trajectory management also includes determining whether the updated trajectory is a false trajectory; if it is, it is converted into a deleted trajectory. If the correlation between the transient trajectory and the measurement signal is higher than a preset correlation and meets a preset trajectory length, the transient trajectory can be converted into a definitive trajectory. If the transient trajectory has no associated measurement signal within a first preset time period, it is converted into a deleted trajectory. If the definitive trajectory has no associated measurement signal within a second preset time period, it is converted into a deleted trajectory. Based on the updated trajectory after trajectory management, at least one indoor target is tracked, improving the quality and reliability of indoor multi-target tracking.

[0129] The technical solution of this embodiment receives indoor discrete echo signals acquired by radar and performs long and short time preprocessing on the indoor discrete echo signals to obtain an indoor measurement signal set; obtains the estimated trajectory of at least one indoor target, and associates each measurement signal in the indoor measurement signal set with the estimated trajectory of at least one indoor target to obtain an associated trajectory of at least one indoor target; updates the associated trajectory using a Kalman filter algorithm to obtain an updated trajectory of at least one indoor target; and tracks at least one indoor target based on the updated trajectory. This solves the problem of reduced indoor target tracking accuracy caused by trajectory association errors or target loss in complex indoor environments. It enables robust real-time tracking of multiple indoor targets in complex indoor environments, improving the detection probability and tracking accuracy of indoor targets.

[0130] Example 2

[0131] Figure 5 This is a schematic diagram of an indoor multi-target tracking device according to Embodiment 2 of the present invention. This embodiment is applicable to real-time target tracking indoors. The device can be implemented using software and / or hardware, and can be integrated into any device that provides indoor multi-target tracking functionality, such as... Figure 5 As shown, the indoor multi-target tracking device specifically includes: a measurement acquisition module 210, a trajectory acquisition module 220, a trajectory update module 230, and a trajectory tracking module 240.

[0132] The measurement module 210 is used to receive the indoor discrete echo signal collected by the radar and perform long and short time preprocessing on the indoor discrete echo signal to obtain the indoor measurement signal set.

[0133] The trajectory acquisition module 220 is used to acquire the estimated trajectory of at least one indoor target and associate each measurement signal in the indoor measurement signal set with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target.

[0134] The trajectory update module 230 is used to update the associated trajectory using a Kalman filter algorithm to obtain the updated trajectory of at least one indoor target;

[0135] The trajectory tracking module 240 is used to track at least one target indoors based on the updated trajectory.

[0136] Optionally, the measurement module is specifically used for:

[0137] The indoor discrete echo signal is subjected to short-time preprocessing to obtain a short-time echo signal;

[0138] The indoor discrete echo signal is subjected to long-time preprocessing to obtain a long-time echo signal;

[0139] The short-time range angle spectrum is obtained based on the short-time echo signal, and the long-time range angle spectrum is obtained based on the long-time echo signal;

[0140] The first algorithm is used to detect the short-time range angle spectrum to obtain a set of two-dimensional point cloud measurement signals;

[0141] The second algorithm is used to detect the long-term distance angle spectrum to obtain a set of effective peak measurement signals;

[0142] The indoor measurement signal set is determined based on the effective peak measurement signal set and the two-dimensional point cloud measurement signal set.

[0143] Optionally, the measurement module is specifically used for:

[0144] Cluster the measurement signals in the two-dimensional point cloud measurement signal set to obtain a strong measurement signal set;

[0145] The measurement signals in the effective peak measurement signal set are matched with the measurement signals in the strong measurement signal set. The successfully matched measurement signals are removed from the effective peak measurement signal set to obtain the weak measurement signal set.

[0146] The indoor measurement signal set is obtained based on the strong measurement signal set and the weak measurement signal set.

[0147] Optionally, the trajectory obtaining module is specifically used for:

[0148] Obtain the indoor prior boundary, and filter the indoor measurement signal set according to the indoor prior boundary to obtain the filtered indoor measurement signal set.

[0149] Each measurement signal in the filtered set of indoor measurement signals is associated with the estimated trajectory of at least one indoor target to obtain the associated trajectory of at least one indoor target.

[0150] Optionally, the trajectory obtaining module is specifically used for:

[0151] Each strong measurement signal in the filtered indoor measurement signal set is initially associated with the estimated trajectory of at least one indoor target. If the initial association is successful, the first associated trajectory of at least one indoor target is obtained.

[0152] If the initial association fails, and the unassociated estimated trajectory is a definite trajectory, then the estimated trajectory will be associated a second time with each weak measurement in the filtered set of indoor measurement signals.

[0153] If the secondary association is successful, a second associated trajectory of at least one target indoors will be obtained;

[0154] The first and second associated trajectories are identified as associated trajectories.

[0155] Optionally, the trajectory update module is specifically used for:

[0156] Acquire the velocity and amplitude of the measurement signal associated with the track;

[0157] Based on the speed and amplitude of the measurement signal, the measurement signal is divided into dynamic measurement signal and micro-motion measurement signal;

[0158] The Kalman filter algorithm is used to update the associated trajectory of the dynamic measurement signal to obtain the updated trajectory of at least one indoor target.

[0159] By extending the associated trajectory of the micro-motion measurement signal by a preset period, an updated trajectory of at least one indoor target can be obtained.

[0160] Optionally, the trajectory tracking module is specifically used for:

[0161] If the updated trajectory is a false trajectory, then the false trajectory is converted into a deleted trajectory;

[0162] The updated trajectories, excluding the false trajectories, are classified according to their confidence levels to obtain transient trajectories and definitive trajectories;

[0163] If no associated measurement signal exists in the transient trajectory within the first preset time period, the transient trajectory is converted into a deleted trajectory.

[0164] If the trajectory is determined to have no associated measurement signal within a second preset time period, the determined trajectory is converted into a deleted trajectory, wherein the first preset time period is less than the second preset time period;

[0165] The deleted trajectory is deleted, and at least one indoor target is tracked based on the transient trajectory and the determined trajectory.

[0166] The above-described products can perform the methods provided in any embodiment of the present invention, and have the corresponding functional modules and beneficial effects for performing the methods.

[0167] Example 3

[0168] Figure 6This is a schematic diagram of an electronic device according to Embodiment 3 of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0169] like Figure 6 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0170] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0171] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as indoor multi-target tracking methods.

[0172] In some embodiments, the indoor multi-target tracking method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the indoor multi-target tracking method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the indoor multi-target tracking method by any other suitable means (e.g., by means of firmware).

[0173] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0174] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0175] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0176] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0177] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0178] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0179] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0180] 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 be made according to 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. An indoor multi-target tracking method, characterized in that, include: Receives indoor discrete echo signals collected by radar; The indoor discrete echo signal is subjected to short-time preprocessing to obtain a short-time echo signal; The indoor discrete echo signal is subjected to long-time preprocessing to obtain a long-time echo signal; The short-time range angle spectrum is obtained based on the short-time echo signal, and the long-time range angle spectrum is obtained based on the long-time echo signal; The first algorithm is used to detect the short-time range angle spectrum to obtain a set of two-dimensional point cloud measurement signals; Cluster the measurement signals in the two-dimensional point cloud measurement signal set to obtain a strong measurement signal set; The second algorithm is used to detect the long-term distance angle spectrum to obtain a set of effective peak measurement signals; The measurement signals in the effective peak measurement signal set are matched with the measurement signals in the strong measurement signal set. The successfully matched measurement signals are removed from the effective peak measurement signal set to obtain the weak measurement signal set. The indoor measurement signal set is obtained based on the strong measurement signal set and the weak measurement signal set. Obtain the estimated trajectory of at least one target indoors; Obtain the indoor prior boundary, and filter the indoor measurement signal set according to the indoor prior boundary to obtain the filtered indoor measurement signal set. Each strong measurement signal in the filtered indoor measurement signal set is initially associated with the estimated trajectory of at least one indoor target. If the initial association is successful, the first associated trajectory of at least one indoor target is obtained. If the initial association fails, and the unassociated estimated trajectory is a definite trajectory, then the estimated trajectory will be associated a second time with each weak measurement in the filtered set of indoor measurement signals. If the secondary association is successful, a second associated trajectory of at least one target indoors will be obtained; The first and second associated trajectories are determined as associated trajectories; Acquire the velocity and amplitude of the measurement signal associated with the track; Based on the speed and amplitude of the measurement signal, the measurement signal is divided into dynamic measurement signal and micro-motion measurement signal; The Kalman filter algorithm is used to update the associated trajectory of the dynamic measurement signal to obtain the updated trajectory of at least one indoor target. The associated trajectory, which is linked to the micro-motion measurement signal, is extended by a preset period to obtain the updated trajectory of at least one indoor target; At least one target indoors is tracked based on the updated trajectory.

2. The method according to claim 1, characterized in that, Tracking at least one indoor target based on the updated trajectory includes: If the updated trajectory is a false trajectory, then the false trajectory is converted into a deleted trajectory; The updated trajectories, excluding the false trajectories, are classified according to their confidence levels to obtain transient trajectories and definitive trajectories; If no associated measurement signal exists in the transient trajectory within the first preset time period, the transient trajectory is converted into a deleted trajectory. If the trajectory is determined to have no associated measurement signal within a second preset time period, the determined trajectory is converted into a deleted trajectory, wherein the first preset time period is less than the second preset time period; The deleted trajectory is deleted, and at least one indoor target is tracked based on the transient trajectory and the determined trajectory.

3. An indoor multi-target tracking device, characterized in that, include: The measurement module is used to receive indoor discrete echo signals acquired by radar. The indoor discrete echo signal is subjected to short-time preprocessing to obtain a short-time echo signal; The indoor discrete echo signal is subjected to long-time preprocessing to obtain a long-time echo signal; The short-time range angle spectrum is obtained based on the short-time echo signal, and the long-time range angle spectrum is obtained based on the long-time echo signal; The first algorithm is used to detect the short-time range angle spectrum to obtain a set of two-dimensional point cloud measurement signals; Cluster the measurement signals in the two-dimensional point cloud measurement signal set to obtain a strong measurement signal set; The second algorithm is used to detect the long-term distance angle spectrum to obtain a set of effective peak measurement signals; The measurement signals in the effective peak measurement signal set are matched with the measurement signals in the strong measurement signal set. The successfully matched measurement signals are removed from the effective peak measurement signal set to obtain the weak measurement signal set. The indoor measurement signal set is obtained based on the strong measurement signal set and the weak measurement signal set. The trajectory acquisition module is used to: acquire the estimated trajectory of at least one target indoors; Obtain the indoor prior boundary, and filter the indoor measurement signal set according to the indoor prior boundary to obtain the filtered indoor measurement signal set. Each strong measurement signal in the filtered indoor measurement signal set is initially associated with the estimated trajectory of at least one indoor target. If the initial association is successful, the first associated trajectory of at least one indoor target is obtained. If the initial association fails, and the unassociated estimated trajectory is a definite trajectory, then the estimated trajectory will be associated a second time with each weak measurement in the filtered set of indoor measurement signals. If the secondary association is successful, a second associated trajectory of at least one target indoors will be obtained; The first and second associated trajectories are determined as associated trajectories; The trajectory update module is used to: acquire the velocity and amplitude of the measurement signal associated with the associated trajectory; Based on the speed and amplitude of the measurement signal, the measurement signal is divided into dynamic measurement signal and micro-motion measurement signal; The Kalman filter algorithm is used to update the associated trajectory of the dynamic measurement signal to obtain the updated trajectory of at least one indoor target. The associated trajectory, which is linked to the micro-motion measurement signal, is extended by a preset period to obtain the updated trajectory of at least one indoor target; A trajectory tracking module is used to track at least one target indoors based on the updated trajectory.

4. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the indoor multi-target tracking method according to any one of claims 1-2.

5. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the indoor multi-target tracking method according to any one of claims 1-2.

Images