CN116736288B - Indoor human body target tracking method and system based on millimeter wave radar - Google Patents
Indoor human body target tracking method and system based on millimeter wave radarInfo
- Publication number
- CN116736288B CN116736288B CN202310447818.7A CN202310447818A CN116736288B CN 116736288 B CN116736288 B CN 116736288B CN 202310447818 A CN202310447818 A CN 202310447818A CN 116736288 B CN116736288 B CN 116736288B
- Authority
- CN
- China
- Prior art keywords
- point
- points
- partition
- target
- distance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/66—Radar-tracking systems; Analogous systems
- G01S13/72—Radar-tracking systems; Analogous systems for two-dimensional tracking, e.g. combination of angle and range tracking, track-while-scan radar
- G01S13/723—Radar-tracking systems; Analogous systems for two-dimensional tracking, e.g. combination of angle and range tracking, track-while-scan radar by using numerical data
- G01S13/726—Multiple target tracking
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The application discloses an indoor human body target tracking method and system based on millimeter wave radar, which relate to the technical field of radar positioning and aim at the problem of low tracking efficiency and accuracy of the human body target tracking method in the prior art, and the application enables an algorithm to cluster according to the proportion of the distance between data instead of a fixed value by specifying coefficients, and the clustering can be adaptively adjusted according to the distance between the target and the radar; through the contour coefficients, the clustering mode with the best effect can be selected, and then the accuracy and the efficiency of indoor human body target detection are improved.
Description
Technical Field
The invention relates to the technical field of radar positioning, in particular to an indoor human body target tracking method and system based on millimeter wave radar.
Background
Lei Daqun target tracking refers to monitoring and tracking the motion of multiple targets in a radar system. In general, radar systems can detect a large number of indoor human targets, which if not classified and tracked, can easily lead to information confusion and target tracking errors, thereby reducing the efficiency and accuracy of the detection system. The clustering algorithm is widely applied to Lei Daqun target tracking, can classify and analyze targets detected by the radar, and improves the accuracy and instantaneity of target tracking. Compared with other target tracking algorithms, the clustering algorithm can better identify the behavior patterns of a plurality of targets, separate the data sets better, and reduce the possibility of misjudgment as much as possible while ensuring the tracking efficiency. The clustering algorithm can effectively group a large number of radar-detected targets and further analyze the behavior patterns and motion laws of each group. This information helps us better grasp the motion changes of the target, thus achieving more accurate, real-time target tracking. Another important application scenario of the clustering algorithm is data preprocessing for target tracking. For example, redundant information in radar data can be deleted by using a clustering technology, so that subsequent target tracking is more accurate. In addition, in the process of target tracking, different grouping strategies can be adopted in real time and applied to a density-based clustering algorithm, so that targets with different modes can be tracked and predicted. In an indoor environment, the human body target tracking is performed by using the prior art, and the problems of low tracking efficiency and low accuracy can occur.
Disclosure of Invention
The invention aims to provide an indoor human body target tracking method and system based on millimeter wave radar, aiming at the problems of low tracking efficiency and accuracy of the human body target tracking method in the prior art.
The technical scheme adopted by the invention for solving the technical problems is as follows:
an indoor human body target tracking method based on millimeter wave radar comprises the following steps:
Firstly, acquiring echo time domain original data of an indoor human body target by utilizing a millimeter wave radar, reconstructing the original data to form a matrix, wherein the matrix size is ADCNum x NumChirp (NTx NRx), and then respectively carrying out two-dimensional Fourier transform on each channel in the matrix to obtain an RD (remote data) spectrum matrix;
Wherein NumChirp denotes the number of chirp, ADCNum denotes the number of sampling points contained in each chirp, NTx denotes the number of transmitting antennas, and NRx denotes the number of receiving antennas;
Step two, static clutter filtering and constant false alarm detection are carried out on the RD spectrum matrix, and then the distance between a target point and the radar is obtained;
measuring angle information of a target point through an angle measurement algorithm, and obtaining the abscissa of the target point through the distance between the target point and the radar and the angle information, so as to construct a data set;
step four, obtaining an reachable distance sequence by utilizing the horizontal coordinates and the vertical coordinates of the target points in the data set;
Step five, clustering and dividing each reachable distance in the reachable distance sequence to obtain all clustering and dividing of the target point;
Step six, calculating the contour coefficients of each cluster partition in all cluster partitions of the target point, and selecting the partition with the largest contour coefficient as a clustering Result, namely the state of the target at the moment;
and seventhly, repeating the second step to the sixth step to track the indoor human body target.
Further, the specific steps of the fourth step are as follows:
Step four, core distances of all the target points in the data set are obtained according to the abscissa of the target points, then the target point corresponding to the maximum core distance is selected as an initial observation point, the rest target points are all set as unobserved points, and then the initial reachable distances of all the unobserved points are set as the core distances of the initial observation points;
Step four, respectively obtaining a distance value sequence A between the observation point and all the unobserved points and a core distance value B of the observation point, respectively comparing each value in the distance value sequence A with B according to the sequence order, sequentially reserving the maximum value in each comparison to obtain an M sequence, respectively comparing each value in the current reachable distance of the unobserved point with the value of the corresponding position in the M sequence, and if the value in the current reachable distance of the unobserved point is larger than the corresponding value in the M sequence, replacing the corresponding value in the reachable distance of the unobserved point with the corresponding value in the M sequence, otherwise, not replacing the value;
And fourthly, finding out the point with the smallest reachable distance in the unobserved points, taking the point as a new observation point, marking the previous observation point as observed, and then not comparing the points, repeating the fourth and the second steps until all the points are observed, finally recording the target points which are taken as the observation points each time, sorting the target points, and constructing a reachable distance table according to the reachable distance corresponding to each target point in the sorting.
Further, the specific steps of the fifth step are as follows:
Traversing an reachable distance table, wherein a target point with the reachable distance change exceeding a specified coefficient xi is a mutation point, and representing a division by every two mutation points, if the mutation point is one, only one division is needed by the data set, namely the data set itself, and the division comprises a starting point and an ending point;
Step five, sorting the partitions obtained through the mutation points according to the starting points in the partitions from small to large, defaulting that all the partitions do not conflict, taking all the partitions except the partition corresponding to the first starting point as undetermined partitions, and taking the partition corresponding to the first starting point as reference partitions;
Step five, starting with the partition corresponding to the minimum starting point in the ordering without judgment, judging whether the partition is overlapped with the reference partition from small to large, if so, taking the partition overlapped with the reference partition in the ordering as conflict partition, if not, merging the partition with the reference partition as new reference partition,
Repeating the fifth step until the last division in the sequence is judged, and taking the finally obtained reference division as a clustering division result;
and step five, deleting the last conflict-causing partition from the finally obtained reference partitions to obtain new reference partitions, taking all partitions after the last conflict-causing partition as undetermined partitions, and executing step five three until all partitions are judged, wherein all cluster partition results are obtained as final cluster partition results.
Further, the specific steps of the step six are as follows:
Calculating the contour coefficients of all cluster partitions of the target point, selecting the partition with the largest contour coefficient as a clustering Result, and determining the state of the target at the moment according to the clustering Result and the radial speed of the target point relative to the radar, namely the state of the target at the moment;
and the radial speed of the target point relative to the radar is obtained by carrying out static clutter filtering and constant false alarm detection on the RD spectrum matrix in the second step.
Further, the contour coefficients are expressed as:
Where a i represents the average distance of the i-th point to other points in the same cluster and b i represents the minimum average distance of the point to other points in different clusters.
Further, the minimum point number in the cluster division is 5.
An indoor human body target tracking system based on millimeter wave radar comprises an indoor human body target echo time domain original data acquisition module, a data set construction module and an indoor human body target tracking module;
The indoor human body target echo time domain original data acquisition module is used for acquiring echo time domain original data of an indoor human body target by utilizing a millimeter wave radar, reconstructing the original data to form a matrix, and then respectively carrying out two-dimensional Fourier transform on each channel in the matrix to obtain an RD spectrum matrix, wherein NumChirp represents the number of chirp, ADCNum represents the number of sampling points contained in each chirp, NTx represents the number of transmitting antennas, and NRx represents the number of receiving antennas;
the data set construction module is used for obtaining the distance between the target point and the radar after static clutter filtering and constant false alarm detection are carried out on the RD spectrum matrix, measuring the angle information of the target point through an angle measurement algorithm, and obtaining the abscissa of the target point through the distance between the target point and the radar and the angle information, so as to construct a data set;
The indoor human body target tracking module is used for obtaining an reachable distance sequence by utilizing the horizontal and vertical coordinates of the target points in the data set, then clustering and dividing each reachable distance in the reachable distance sequence to obtain all cluster divisions of the target points, then calculating the contour coefficients of each cluster division in all cluster divisions of the target points, and selecting the division with the largest contour coefficient as a clustering Result, namely the state of the target at the moment;
the indoor human body target is tracked by the indoor human body target echo time domain original data acquisition module, the data set construction module and the indoor human body target tracking module.
Further, the specific steps of obtaining the reachable distance sequence by utilizing the horizontal and vertical coordinates of the target point in the data set in the indoor human body target tracking module are as follows:
step 1, obtaining core distances of all target points in a data set according to the abscissa of the target points, then selecting the target point corresponding to the maximum core distance as an initial observation point, setting the rest target points as unobserved points, and setting the initial reachable distances of all the unobserved points as the core distances of the initial observation points;
Step 2, respectively solving a distance value sequence A between the observation point and all the unobserved points and a core distance value B of the observation point, respectively comparing each value in the distance value sequence A with B according to the sequence order, sequentially reserving the maximum value in each comparison to obtain an M sequence, respectively comparing each value in the current reachable distance of the unobserved point with the value of the corresponding position in the M sequence, and if the value in the current reachable distance of the unobserved point is larger than the corresponding value in the M sequence, replacing the corresponding value in the reachable distance of the unobserved point with the corresponding value in the M sequence, otherwise, not replacing the value;
And 3, finding out the point with the smallest reachable distance in the unobserved points, taking the point as a new observation point, marking the previous observation point as observed, and then not comparing the points, repeating the step 2 until all the points are observed, finally recording the target points which are taken as the observation points each time, sorting the target points, and constructing a reachable distance table according to the reachable distance corresponding to each target point in the sorting.
Further, the specific steps of clustering and dividing each reachable distance in the reachable distance sequence in the indoor human body target tracking module to obtain all the clustered and divided target points are as follows:
step A, traversing an reachable distance table, wherein a target point with the reachable distance change exceeding a specified coefficient xi is a mutation point, and representing a division by every two mutation points, if the mutation point is one, only one division is performed by the data set, namely the data set itself, and the division comprises a starting point and an ending point;
Step B, sorting the partitions obtained through the mutation points according to the starting points in the partitions from small to large, defaulting that all the partitions do not conflict, taking all the partitions except the partition corresponding to the first starting point as undetermined partitions, and taking the partition corresponding to the first starting point as reference partitions;
step C, starting with the partition corresponding to the minimum starting point in the ordering without judgment, judging whether the partition is overlapped with the reference partition from small to large, if so, taking the partition overlapped with the reference partition in the ordering as conflict partition, if not, merging the partition with the reference partition as new reference partition,
Step D, repeating the step C until the last division in the sequence is judged, and taking the finally obtained reference division as a clustering division result;
And E, deleting the last conflict-causing partition from the finally obtained reference partitions to obtain new reference partitions, taking all partitions after the last conflict-causing partition as undetermined partitions, and then executing the step C until all partitions are judged, wherein all cluster partition results obtained are final cluster partition results.
Further, the contour coefficients are expressed as:
Where a i represents the average distance of the i-th point to other points in the same cluster and b i represents the minimum average distance of the point to other points in different clusters.
The beneficial effects of the invention are as follows:
the method and the system enable the algorithm to cluster according to the ratio of the distance between the data instead of a fixed value through the specified coefficients, enable the clustering to be adaptively adjusted according to the distance between the target and the radar, select the clustering mode with the best effect through the contour coefficients, and further improve the accuracy and the efficiency of indoor human body target detection.
Under the indoor tracking condition, the specific number of targets cannot be known in advance, so that different target numbers need to be clustered respectively by applying a K-means clustering algorithm, and then the optimal target number is judged according to additional judging parameters. In order to reduce the dependence of the K-means algorithm on initial values, the random values can be taken for clustering, so that the algorithm is complicated and redundant, the hierarchical clustering algorithm can lead to a plurality of obtained clustering results, the algorithm is long in operation time, in the density clustering method, the DBSCAN can overcome the defect that the K-means clustering center has a large influence on the clustering results to a certain extent, but the clustering effect also depends on the preset clustering radius, and under the indoor environment, the discrete degrees of the millimeter wave echo points at a far and near distance are different, and the fixed clustering radius can lead to the clustering result being difficult to meet the indoor tracking requirement.
Drawings
FIG. 1 is a flow chart of the present application;
FIG. 2 is a graph of reachable distances;
FIG. 3 is a diagram of raw data;
FIG. 4 is a schematic diagram of a clustering result calculated by the method of the present application;
FIG. 5 is a schematic diagram of a clustering result of a conventional DBSCAN algorithm;
fig. 6 is a schematic diagram of a fourth step.
Detailed Description
It should be noted that, in particular, the various embodiments of the present disclosure may be combined with each other without conflict.
The first embodiment specifically describes an indoor human body target tracking method based on millimeter wave radar according to the present embodiment with reference to fig. 1, and includes the following steps:
Firstly, acquiring echo time domain original data of an indoor human body target by utilizing a millimeter wave radar, reconstructing the original data to form a matrix, wherein the matrix size is ADCNum x NumChirp (NTx NRx), and then respectively carrying out two-dimensional Fourier transform on each channel in the matrix to obtain an RD (remote data) spectrum matrix;
Wherein NumChirp denotes the number of chirp, ADCNum denotes the number of sampling points contained in each chirp, NTx denotes the number of transmitting antennas, and NRx denotes the number of receiving antennas;
Step two, static clutter filtering and constant false alarm detection are carried out on the RD spectrum matrix, and then the distance between a target point and the radar is obtained;
Measuring angle information of a target point through an angle measurement algorithm, and then combining the distance between the target point and the radar with the angle information to obtain the abscissa of the target point, so as to construct a data set;
step four, obtaining an reachable distance sequence by utilizing the horizontal coordinates and the vertical coordinates of the target points in the data set;
Step five, clustering and dividing each reachable distance in the reachable distance sequence to obtain all clustering and dividing of the target point;
Step six, calculating the contour coefficient S of each cluster partition in all cluster partitions of the target point, and selecting the partition with the largest contour coefficient as a clustering Result, namely the state of the target at the moment;
and seventhly, repeating the second step to the sixth step to track the indoor human body target.
Clustering algorithms are a common machine learning technique that can divide a data set into several categories, each of which contains a set of highly similar samples. In Lei Daqun target tracking, a clustering algorithm is mainly used for distinguishing the motion laws and behavior patterns of targets of different categories. The most common algorithms in the clustering algorithm include a K-means clustering algorithm, a hierarchical clustering algorithm, a density clustering algorithm and the like. These algorithms have respective advantages and application ranges.
The K-means clustering algorithm is a clustering algorithm based on a central point. In this algorithm, the parameter k is first entered, representing the classification of the data set into k different categories. Then, k samples are randomly selected in the dataset as the center of k categories. Then, all samples in the dataset are assigned to the k different categories, respectively, according to the distances between the samples in the dataset and the k samples. The sum of the distances between samples assigned to the same class is minimal, and the assignment results satisfy the minimal distance principle. Finally, the center point of each category is recalculated according to the distribution result. This process is repeated until the center point is no longer changed or other stop conditions are met.
Hierarchical clustering algorithms are a top-down or bottom-up clustering method. In the algorithm, for a data set, it is first treated as a single cluster, then each cluster is split into several clusters in turn, and this process is repeated until a predetermined condition is met. The distance between samples in the hierarchical clustering process can be defined differently, and single connectivity, full connectivity, average connectivity and the like are common. Wherein, single connectivity refers to the shortest distance between two clusters, full connectivity refers to the longest distance between two clusters, and average connectivity refers to calculating the average distance between all samples in two clusters.
The density clustering algorithm is a density-based clustering algorithm. In this algorithm, clusters are defined as dense areas of sample points, and distances between clusters are defined as distances of sample points between clusters. By calculating the density of sample points, the data set can be divided into clusters. Wherein the density of a sample point represents the number of samples contained within a circle centered at that point and having a radius. In the density clustering algorithm, two parameters, namely a radius and a density threshold, are adopted when clustering and discriminating clusters. The density clustering algorithm can overcome the problem of strong initial value dependence in the K-means algorithm, so that the density clustering algorithm is widely applied to the field of target tracking.
Under the indoor human body tracking condition, the specific number of targets cannot be known in advance, so that different target numbers need to be clustered respectively by applying a K-means clustering algorithm, and then the optimal target number is judged according to additional judgment parameters. In order to reduce the dependence of the K-means algorithm on initial values, the random values can be taken for clustering, so that the algorithm is complicated and redundant, the hierarchical clustering algorithm can lead to a plurality of obtained clustering results, the algorithm is long in operation time, in the density clustering method, the DBSCAN can overcome the defect that the K-means clustering center has a large influence on the clustering results to a certain extent, but the clustering effect also depends on the preset clustering radius, and under the indoor environment, the discrete degrees of the millimeter wave echo points at a far and near distance are different, and the fixed clustering radius can lead to the clustering result being difficult to meet the indoor tracking requirement. The present application improves upon the shortcomings of these clustering algorithms.
The technical scheme of the application specifically comprises three stages of core distance and reachable distance calculation, clustering division and contour coefficient calculation.
Stage one, core distance and reachable distance calculation
The data set D is assumed to need clustering, and parameters and conditions are set as follows, wherein each cluster at least comprises MinPts points, and the distance adopts Euclidean distance. The neighborhood of a point q is denoted as N ε (q), where epsilon denotes the neighborhood radius.
The core distance of a data point q is defined as the minimum neighborhood radius epsilon 0 for the neighborhood of that point, which contains the MinPts points. Data points with smaller core distances are typically more dense and relatively concentrated data points. For each data point q belonging to the data set D, only the distances of other points need to be calculated, the distances are ordered from small to large, and the distance with the MinPts small is taken out, namely the core distance corresponding to the point.
The reachable distance is the distance between one data point p and another data point q, and also the core distance of q and the core distance of p need to be considered. If p to q have a connected path and the core distance maximum of all data points on the path is not higher than the core distance of q, then the reachable distance between p and q is the core distance of q.
The calculation of the reachable distance is carried out according to the following steps:
(1) Finding out the point corresponding to the maximum core distance of all points in the data set D as an observation point, which means that the point is the most likely noise point;
(2) Setting the reachable distances of points contained in the core area of the observation point as the core distance of the observation point, and setting the reachable distances of points not in the core area as the distance between the point and the observation point, wherein the distance is denoted as m;
(3) Finding out points with the reachable distance larger than m in the points in the core area, and setting the points as m;
(4) Finding out the point with the smallest current reachable distance from the rest points, taking the point as a new observation point, repeating the steps (2) - (4) until the rest points in the step (4) are zero, ending, and taking note that in the processing process, the points which are taken as the observation points each time are required to be recorded and ordered, and constructing a reachable distance table according to the reachable distance corresponding to each point.
Stage two, cluster partitioning
In stage one, the reachable distances and their corresponding sequences are obtained, and it is first clear that the reachable distance corresponding to each point represents the close relationship between the point and the point before it, so that the smaller the reachable distance, the closer the point to the point before it, and vice versa. According to the characteristics, it can be known that a region with a mutation of a reachable distance exists between two clusters, and the region between the two mutation regions is one cluster.
In the cluster division stage, a coefficient ζ needs to be specified, which determines in what case a region can be defined as "mutation". After the value of ζ is given according to the actual situation, all mutation points are found out by traversing the reachable distance table, and every two mutation points are combined to obtain a cluster. The clusters obtained by this method are usually coincident with each other and therefore require further processing for cluster division. The specific method comprises the steps of finding out sub-clusters which cover all data sets as far as possible and are not overlapped with each other, and listing the sub-clusters as one of the clustering methods, wherein the steps are as follows:
(1) Setting a specified coefficient xi, and traversing an reachable distance table, wherein the reachable distance table is used as a mutation point when the reachable distance is changed beyond the specified coefficient xi, and each two mutation points represent a division, and if the mutation point is one, only one division is used for the data set, namely the data set per se.
The partitions obtained through mutation comprise a starting point and an ending point, the partitions are ordered from small to large according to the starting point, all the partitions are defaulted without conflict, and all the partitions except the partition corresponding to the first starting point are used as undetermined partitions. And the partition corresponding to the first starting point is taken as a reference partition,
(2) Starting with the partition corresponding to the minimum starting point in the sequence without judgment, judging whether the partition coincides with the reference partition from small to large, if so, taking the partition coinciding with the reference partition in the sequence as the conflict partition, if not, merging the partition with the reference partition as the new reference partition, and repeating the steps until the last partition in the sequence is judged. And taking the finally obtained reference partition as a clustering partition result.
Deleting the last conflict-causing partition from the finally obtained reference partitions to obtain new reference partitions, taking all the conflict-causing partitions with the last conflict-causing partition as undetermined partitions, and repeating the steps until all the partitions are judged, wherein all the obtained clustering partition results are final clustering partition results.
Stage three, contour coefficient calculation
In the second stage, a plurality of different clustering modes of a data set can be obtained, and the most suitable clustering mode is required to be selected according to the contour coefficients of different clusters.
The contour coefficient (silhouette value) of a point in a cluster is defined as
Wherein a i represents the average distance of the i-th point to other points in the same cluster;
b i represents the minimum average distance of the point to other points in different clusters.
The contour coefficients of all points in the clusters are obtained, and the contour coefficients are averaged to be used as the contour coefficients of the whole clustering mode, so that the contour coefficients of different clustering modes can be obtained, and the clustering mode with the largest contour coefficient is the clustering mode which is most suitable for the data set.
Taking one frame of radar actual data as an example, it is known that there are 3 targets in the current frame, as shown in fig. 2. The minpts=5 is set and,
Specific flow
(1) The reachable distance table of all points was found as in fig. 3.
(2) The small block clusters are obtained according to the mutation values and are respectively [13,19], [13,100], [34,98], [46,98], [57,62],
[81,100],[106,120],[106,208],[106,218],[123,131],[208,218],[232,253];
(3) The small block clusters are combined to obtain different classification results as follows:
{[13,19],[34,98],[106,120],[123,131],[208,218],[232,253]};
{[13,19],[34,98],[106,208],[208,218],[232,253]};
{[13,19],[34,98],[106,218],[232,253]};
{[13,19],[34,98],[123,131],[208,218],[232,253]};
{[13,19],[46,98],[106,120],[123,131],[208,218],[232,253]};
{[13,19],[57,62],[81,100],[106,120],[123,131],[208,218],[232,253]};
{[13,100],[106,120],[123,131],[208,218],[232,253]};
{[13,100],[106,208],[208,218],[232,253]};
{[13,100],[106,218],[232,253]};
{[13,100],[123,131],[208,218],[232,253]};
(4) Respectively solving contour coefficients for the small blocks of clusters, and taking out the clusters corresponding to the largest contour coefficients as
{[13,100],[106,218],[232,253]};
(5) The final cluster is { [13,100], [106,218], [232,253] };
Compared with the DBSCAN clustering method, the method can consider a plurality of clustering modes and obtain the cluster with the best performance.
Assuming that the radar emits FMCW continuous waves, a total of frames are emitted throughout the measurement, and for each frame its parameters are as follows, a total of NumChirp chirp, each chirp having ADCNum sampling points, a total of NTx transmit antennas and NRx receive antennas. In the clustering process, the minimum point number in one cluster is set as MinPts, and the threshold coefficient is zeta.
It should be noted that the detailed description is merely for explaining and describing the technical solution of the present invention, and the scope of protection of the claims should not be limited thereto. All changes which come within the meaning and range of equivalency of the claims and the specification are to be embraced within their scope.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310447818.7A CN116736288B (en) | 2023-04-24 | 2023-04-24 | Indoor human body target tracking method and system based on millimeter wave radar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310447818.7A CN116736288B (en) | 2023-04-24 | 2023-04-24 | Indoor human body target tracking method and system based on millimeter wave radar |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116736288A CN116736288A (en) | 2023-09-12 |
| CN116736288B true CN116736288B (en) | 2026-01-16 |
Family
ID=87901895
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310447818.7A Active CN116736288B (en) | 2023-04-24 | 2023-04-24 | Indoor human body target tracking method and system based on millimeter wave radar |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116736288B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110361725A (en) * | 2018-04-11 | 2019-10-22 | 英飞凌科技股份有限公司 | The system and method for carrying out mankind's detection and identification using millimetre-wave radar |
| CN110361727A (en) * | 2019-07-22 | 2019-10-22 | 浙江大学 | A Multi-target Tracking Method for Millimeter-Wave Radar |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP7623106B2 (en) * | 2019-05-12 | 2025-01-28 | オリジン リサーチ ワイヤレス， インコーポレイテッド | Method, Apparatus and System for Wireless Tracking, Scanning and Surveillance - Patent application |
| GB2590115B (en) * | 2019-09-13 | 2023-12-06 | Motional Ad Llc | Extended object tracking using radar |
| CN111239728B (en) * | 2020-02-26 | 2021-06-15 | 深圳雷研技术有限公司 | Passenger counting method and system based on millimeter wave radar |
| CN112114309B (en) * | 2020-08-10 | 2023-04-07 | 西安电子科技大学 | JPDA multi-target tracking method based on optimal contour coefficient self-adaptive K-means clustering |
| CN112444805A (en) * | 2020-11-01 | 2021-03-05 | 复旦大学 | Distributed multi-target detection, positioning tracking and identity recognition system based on radar |
| CN112505648B (en) * | 2020-11-19 | 2023-06-30 | 西安电子科技大学 | Target Feature Extraction Method Based on Millimeter Wave Radar Echo |
| CN114994656A (en) * | 2021-11-16 | 2022-09-02 | 上海昉恩智能科技有限公司 | Indoor personnel tracking method based on millimeter wave radar |
-
2023
- 2023-04-24 CN CN202310447818.7A patent/CN116736288B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110361725A (en) * | 2018-04-11 | 2019-10-22 | 英飞凌科技股份有限公司 | The system and method for carrying out mankind's detection and identification using millimetre-wave radar |
| CN110361727A (en) * | 2019-07-22 | 2019-10-22 | 浙江大学 | A Multi-target Tracking Method for Millimeter-Wave Radar |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116736288A (en) | 2023-09-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112036074B (en) | Radar signal sorting method and system in high pulse density environment | |
| CN107728140B (en) | A Multi-Target Multi-Channel Parallel Tracking Processing Method for Warning Radar | |
| CN109581312A (en) | A multi-target clustering method for high-resolution millimeter-wave radar | |
| CN106851571A (en) | WiFi localization methods in a kind of quick KNN rooms based on decision tree | |
| CN111929653B (en) | Target detection and tracking method and system based on unmanned ship navigation radar | |
| CN115062683B (en) | 4D millimeter wave radar clustering method and storage medium based on DBSCAN | |
| CN110049549A (en) | More fusion indoor orientation methods and its system based on WiFi fingerprint | |
| CN118091572A (en) | A method for clustering dense target point cloud based on millimeter wave radar | |
| CN113608193A (en) | Radar multi-target distance and speed estimation method based on UNet | |
| CN119087357A (en) | A parameter adaptive multi-density clustering radar signal sorting method | |
| CN116736288B (en) | Indoor human body target tracking method and system based on millimeter wave radar | |
| CN113177097B (en) | Track start judging method based on attribute clustering and space-time constraint | |
| CN113269242B (en) | Target detection trace aggregation method based on peak clustering | |
| CN119414341B (en) | Doppler velocity filtering multipath false target suppression method, device and medium | |
| CN119939286A (en) | Adaptive radar signal sorting and clustering optimization method based on improved density peak clustering | |
| CN118692264B (en) | A method for extracting and eliminating incomplete routes based on ADS-B data | |
| CN120314928A (en) | Target tracking method and system based on DBSCAN and Kmeans | |
| CN117194951B (en) | HRRP target identification method based on joint feature extraction and Stacking model classification | |
| CN108445443A (en) | A kind of fingerprint point clustering method based on KNN | |
| Dai et al. | Clustering of DOA data in radar pulse based on SOFM and CDbw | |
| CN115147385B (en) | Intelligent detection and judgment method for repeated damage in aviation hole exploration video | |
| CN113381827B (en) | A Spectrum Sensing Method for Electromagnetic Signals Based on Deep Clustering Network | |
| CN119513627B (en) | A radar pulse sorting method based on inverse distance weighted online clustering | |
| CN120779388B (en) | A method and system for multi-target tracking of radar signals based on dynamic graph convolutional networks | |
| CN115032692B (en) | River phase reservoir sediment microphase partition method and device based on seismic waveform clustering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |