CN111444926B - Regional population counting method, device and equipment based on radar and storage medium - Google Patents
Regional population counting method, device and equipment based on radar and storage medium Download PDFInfo
- Publication number
- CN111444926B CN111444926B CN202010200465.7A CN202010200465A CN111444926B CN 111444926 B CN111444926 B CN 111444926B CN 202010200465 A CN202010200465 A CN 202010200465A CN 111444926 B CN111444926 B CN 111444926B
- Authority
- CN
- China
- Prior art keywords
- convolution
- data
- radar data
- radar
- layers
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- 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
- G01S13/88—Radar or analogous systems specially adapted for specific applications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Probability & Statistics with Applications (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The embodiment of the invention discloses a regional population counting method, a regional population counting device, regional population counting equipment and a storage medium based on radar. The regional population counting method based on radar comprises the following steps: acquiring and storing first radar data without living bodies in a first preset time of a target area; establishing a first monitoring background model according to the first radar data; acquiring and storing second radar data of the target area; and counting a first number of people in the target area according to the first monitoring background model and the second radar data. The embodiment of the invention realizes the work environment requirement and the calculation resource requirement when the number of people in the statistical area is reduced.
Description
Technical Field
The embodiment of the invention relates to an image processing technology, in particular to a regional population counting method, a regional population counting device, regional population counting equipment and a storage medium based on radar.
Background
Along with the development of technologies such as the Internet of things and 5G, the intelligent architecture, the intelligent home, the intelligent security and the like are also rapidly developed. For the above aspects, the information on the number of people in the region of interest is a very important type of basic data information. The information can be used for crowd abnormity monitoring, regional security, public resource scheduling and other application aspects. The statistical problem of regional population is long coming into the field of view of researchers, and a plurality of researchers (engineering application users) develop a plurality of population statistical systems based on different principles according to different sensors, and common population statistical methods are mainly a method based on section detection and a method based on regional detection.
The method based on the section detection mainly comprises a three-roller brake method, an infrared correlation method and a gravity induction method. The three-roller gate method is to count the number of people by using a physical and mechanical method. The method can accurately count the number of the entrance and the exit, and the side surface reflects the number density of people in a certain area, but is not suitable for areas with large people flow and has poor experience; the infrared correlation method is to count the number of people by detecting the blocking condition of infrared rays generated by human bodies passing through the correlation area. The number of people passing through a certain section can be detected, but the direction cannot be detected, and the accuracy is poor when the people flow is large and is influenced by the environmental temperature; the gravity sensing method estimates the number of people and counts the number of people by detecting the gravity change of a certain section, but the gravity sensing method is limited in detection precision and inconvenient to install and maintain due to the influence of environmental factors.
The method based on regional detection mainly comprises an optical sensor vision method, an active infrared imaging method, a radio frequency identification method and a wireless local area network method. The optical sensor vision method is that the regional people counting can be carried out by analyzing the human images. However, the method is affected by illumination and geographical environmental conditions, and involves privacy problems, requiring more calculation and storage resources; the active infrared imaging method can acquire information similar to an optical sensor, and can acquire higher angular resolution and count regional population. However, the detection distance is shortened under strong light due to the influence of the ambient temperature, and more calculation resources are also needed; the radio frequency identification method is to count the number of people by detecting RFID tags carried by human bodies. This approach lacks convenience in many random life scenarios; the wireless lan method estimates the number of people in an area by detecting channel state information (for calculating power or other information). Because of the physical characteristics of the undetectable human body, the performance is limited, and the human body is easily influenced by external signals.
In summary, the existing statistical method for regional population has various problems, especially the high requirement for the working environment of the detection device and the high occupation of computing resources are common problems of the method, and are also the problems to be solved urgently.
Disclosure of Invention
The embodiment of the invention provides a regional population counting method, a device, equipment and a storage medium based on radar, which are used for realizing the reduction of the working environment requirement and the calculation resource requirement when the regional population is counted.
To achieve the object, an embodiment of the present invention provides a method for regional population statistics based on radar, including:
acquiring and storing first radar data without living bodies in a first preset time of a target area;
establishing a first monitoring background model according to the first radar data;
acquiring and storing second radar data in a second preset time of the target area;
and counting a first number of people in the target area according to the first monitoring background model and the second radar data.
Further, the acquiring and storing the second radar data in the second preset time of the target area includes:
and acquiring and storing second radar data of the target area within a second preset time, and additionally storing second radar data without living bodies within the second preset time of the target area as third radar data.
Further, the counting the first person number of the target area according to the first monitoring background model and the second radar data includes:
after the third radar data are acquired, if the continuous living body-free time of the target area exceeds a third preset time, a second monitoring background model is established according to the third radar data;
and counting the number of people in a second area of the target area according to the second monitoring background model and second radar data.
Further, the establishing a first monitoring background model according to the first radar data includes:
merging each frame of data of the first radar data by using a fixed adjacent distance unit to obtain first merged data;
after the logarithm of the first combined data is obtained by means of modulo, carrying out mean and variance iteration based on the first monitoring background model to obtain first background data;
and establishing a first monitoring background model according to the first background data.
Further, the counting the first person number of the target area according to the first monitoring background model and the second radar data includes:
obtaining a first three-dimensional feature map according to the first monitoring background model and the second radar data;
obtaining a first STFT feature vector according to the second radar data;
and placing the first three-dimensional feature map and the first STFT feature vector into a pre-trained neural network model to count a first number of people in the target area.
Further, the obtaining a first three-dimensional feature map according to the first monitoring background model and the second radar data includes:
performing clutter suppression processing on each frame of data of the second radar data to obtain fourth radar data;
merging each frame of data of the fourth radar data by a fixed adjacent distance unit to obtain second merged data;
detecting a potential target peak of the second combined data in a fixed scale, and obtaining a first two-dimensional characteristic diagram after linear amplification and amplitude limiting operation;
noise removal based on a preset noise threshold is carried out on the second combined data, and a second two-dimensional characteristic diagram is obtained after linear amplification and amplitude limiting operation;
performing outlier iterative detection on each frame of data of the second radar data according to the first monitoring background model, and obtaining a third two-dimensional characteristic diagram after linear amplification and amplitude limiting operation;
and obtaining a first three-dimensional characteristic diagram according to the first two-dimensional characteristic diagram, the second two-dimensional characteristic diagram and the third two-dimensional characteristic diagram.
Further, the obtaining the first STFT feature vector according to the second radar data includes:
performing short-time Fourier transform on the fourth radar data to obtain a first power spectrum;
normalizing the first power spectrum to obtain a first power spectrum vector;
and adding the sum of the values of the first power spectrums to the first power spectrum vector to obtain a first STFT feature vector.
In one aspect, an embodiment of the present invention further provides a device for regional people counting based on radar, where the device includes:
the data acquisition module is used for acquiring and storing first radar data without living bodies in a first preset time of a target area;
the model building module is used for building a first monitoring background model according to the first radar data;
the data acquisition module is also used for acquiring and storing second radar data in a second preset time of the target area;
and the people counting module is used for counting the first people in the target area according to the first monitoring background model and the second radar data.
On the other hand, the embodiment of the invention also provides a regional people counting device based on radar, which comprises: one or more processors; and a storage means for storing one or more programs that, when executed by the one or more processors, cause the one or more processors to implement a method as provided by any of the embodiments of the present invention.
In yet another aspect, embodiments of the present invention further provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method as provided by any of the embodiments of the present invention.
According to the embodiment of the invention, the first radar data without living bodies in the first preset time of the target area are acquired and stored; establishing a first monitoring background model according to the first radar data; acquiring and storing second radar data in a second preset time of the target area; according to the first monitoring background model and the second radar data, the first person number of the target area is counted, the problem that the existing area person number counting method has high requirements on the working environment of the detection equipment and high occupation of computing resources is solved, and the effects of reducing the working environment requirements and the computing resource requirements when the area person number is counted are achieved.
Drawings
FIG. 1 is a schematic flow chart of a method for regional people counting based on radar according to an embodiment of the invention;
fig. 2 is a schematic flow chart of a regional people counting method based on radar according to a second embodiment of the present invention;
FIG. 3 is a schematic flow chart of a method for regional people counting based on radar according to a third embodiment of the present invention;
fig. 4 is a schematic flow chart of a regional people counting method based on radar according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of a regional people counting device based on radar according to a fourth embodiment of the present invention;
fig. 6 is a schematic structural diagram of a regional people counting device based on radar according to a fifth embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are for purposes of illustration and not of limitation. It should be further noted that, for convenience of description, only some, but not all of the structures related to the present invention are shown in the drawings.
Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although a flowchart depicts steps as a sequential process, many of the steps may be implemented in parallel, concurrently, or with other steps. Furthermore, the order of the steps may be rearranged. The process may be terminated when its operations are completed, but may have additional steps not included in the figures. The processes may correspond to methods, functions, procedures, subroutines, and the like.
Furthermore, the terms "first," "second," and the like, may be used herein to describe various directions, acts, steps, or elements, etc., but these directions, acts, steps, or elements are not limited by these terms. These terms are only used to distinguish one direction, action, step or element from another direction, action, step or element. For example, a first module may be referred to as a second module, and similarly, a second module may be referred to as a first module, without departing from the scope of the present application. Both the first module and the second module are modules, but they are not the same module. The terms "first," "second," and the like, are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more of the described features. In the description of the embodiments of the present invention, the meaning of "plurality" is at least two, for example, two, three, etc., unless explicitly defined otherwise.
Example 1
As shown in fig. 1, a first embodiment of the present invention provides a method for regional people counting based on radar, which includes:
s110, acquiring and storing first radar data of no living body in a first preset time of a target area.
S120, a first monitoring background model is established according to the first radar data.
In this embodiment, a low-power UWB pulse radar system is used to perform data acquisition, and the radar system is placed at different preset positions, so as to acquire radar data. First, first radar data of a target area are acquired through the radar system, the target area can be multiple, the first radar data is radar data of no living body in a first preset time in the target area, the first preset time can be 20 seconds, and the radar data of the target area which is a static picture for 20 seconds are acquired through the radar system. And uploading the first radar data to the processing terminal through a data line or a wireless communication module after the first radar data are acquired. At this time, the processing terminal can establish a first monitoring background model according to the first radar data.
S130, acquiring and storing second radar data in a second preset time of the target area.
And S140, counting the first number of people in the target area according to the first monitoring background model and the second radar data.
In this embodiment, after the establishment of the first monitoring background model is completed, radar data can be formally acquired for people counting, second radar data in a second preset time of the target area is acquired and uploaded to the processing terminal for storage, and then the processing terminal can count the first people in the target area according to the first monitoring background model and the second radar data, wherein the second preset time can be 3 seconds, and both the first radar data and the second radar data can be stored, so that post-searching and analysis are facilitated.
According to the embodiment of the invention, the first radar data without living bodies in the first preset time of the target area are acquired and stored; establishing a first monitoring background model according to the first radar data; acquiring and storing second radar data in a second preset time of the target area; according to the first monitoring background model and the second radar data, the first person number of the target area is counted, the problem that the existing area person number counting method has high requirements on the working environment of the detection equipment and high occupation of computing resources is solved, and the effects of reducing the working environment requirements and the computing resource requirements when the area person number is counted are achieved.
Example two
As shown in fig. 2, a second embodiment of the present invention provides a method for counting regional population based on radar, and the second embodiment of the present invention is further explained based on the first embodiment of the present invention, where the method includes:
s210, acquiring and storing first radar data of a living body in a first preset time of a target area.
S220, merging the fixed adjacent distance units on each frame of data of the first radar data to obtain first merged data.
S230, carrying out mean and variance iteration based on the first monitoring background model after the logarithm of the first combined data is obtained by means of modulo, so as to obtain first background data.
S240, a first monitoring background model is established according to the first background data.
In this embodiment, after the first radar data is obtained, a series of signal preprocessing needs to be performed on the first radar data, specifically, each frame of data of the first radar data is combined by fixing adjacent distance units frame by frame, so as to obtain first combined data of each frame of data, then, after the logarithm of the first combined data is obtained by modulo calculation, mean and variance iteration based on a first monitoring background model is performed so as to obtain first background data, and finally, a first monitoring background model is established according to the first background data.
S250, obtaining a first three-dimensional characteristic diagram according to the first monitoring background model and the second radar data.
And S260, obtaining a first STFT feature vector according to the second radar data.
S270, placing the first three-dimensional feature map and the first STFT feature vector into a pre-trained neural network model to count a first number of people in the target area.
In this embodiment, the first three-dimensional feature map may be obtained from a first monitoring background model and second radar data, the first STFT (short-time Fourier transform ) feature vector may be obtained from second radar data, the pre-trained neural network model may be a convolutional neural network model, and specifically includes nine 2D convolutional layers, four pooling layers, three full-connection layers and one classification layer, and the classification layer may be a Softmax classification layer, where the 2D convolutional layers and the full-connection layers use ReLU (Rectified Linear Unit, linear rectification function) activation functions, but the convolution kernel and the convolution operation step length are different, and the pooling kernel size and the operation step length of the pooling layer are lm=lm and stride=nm, respectively. The data flow direction of the method is through a 2D convolution layer, and the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layer are respectively L1 x L1, M1 and stride=N1; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L2 x L2, M2 and stride=N2; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L3 x L3, M3 and stride=N3; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L4 x L4, M4 and stride=N4; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L5 x L5, M5 and stride=N5; then realizing feature vector stitching of feature map decreasing sum of the first three-dimensional feature map and the first STFT feature vector, and enabling the completed data to flow to three full-connection layers, wherein the number of neurons of the three full-connection layers is F1, F2 and F3 respectively; and finally, the Softmax classification layer is used for processing to obtain the first number of people in the target area.
S280, acquiring and storing second radar data of the target area within a second preset time, and additionally storing second radar data without living bodies within the second preset time of the target area as third radar data.
S290, after the third radar data are acquired, if the continuous living body-free time of the target area exceeds a third preset time, a second monitoring background model is established according to the third radar data.
S300, counting the number of people in a second area of the target area according to the second monitoring background model and second radar data.
In this embodiment, when the second radar data of the target area is acquired, the second radar data without living body in the second preset time of the target area may be additionally stored as the third radar data, and the second preset time may be 1-3 seconds, for example, 2 seconds, so that after the first person number of the target area is obtained through statistics, the second radar data may be further optimized, and when the third radar data is acquired, if the time without living body in the target area continuously exceeds the third preset time, that is, when the third radar data continuously acquires exceeds the third preset time, the second monitoring background model may be further built according to the third radar data, and it is required to be stated that the third radar data is cleared as long as the living body exists in the third preset time, and the radar data is restarted to be stored and restarted. The third preset time may be 5-15 seconds, for example, 10 seconds, and after the second monitoring background model is obtained, the number of people in the second area of the target area is counted according to the second monitoring background model and the second radar data. Preferably, the steps S290-S300 in the first embodiment of the present invention can be repeatedly performed to continuously optimize the monitoring background model and count the most realistic people.
Example III
As shown in fig. 3 and fig. 4, a third embodiment of the present invention provides a method for counting regional population based on radar, and the third embodiment of the present invention is further explained based on the second embodiment of the present invention, as shown in fig. 3, step S250 in the method specifically includes:
s251, clutter suppression processing is carried out on each frame of data of the second radar data to obtain fourth radar data.
S252, merging the fixed adjacent distance units for each frame data of the fourth radar data to obtain second merged data.
S253, detecting a potential target peak of the second combined data in a fixed scale, and obtaining a first two-dimensional characteristic diagram after linear amplification and amplitude limiting operation.
S254, performing noise removal based on a preset noise threshold on the second combined data, and obtaining a second two-dimensional characteristic diagram after linear amplification and amplitude limiting operation.
S255, carrying out outlier iterative detection on each frame of data of the second radar data according to the first monitoring background model, and obtaining a third two-dimensional characteristic map after linear amplification and amplitude limiting operation.
S256, obtaining a first three-dimensional feature map according to the first two-dimensional feature map, the second two-dimensional feature map and the third two-dimensional feature map.
In this embodiment, the specific process of obtaining the first three-dimensional feature map from the first monitoring background model and the second radar data may be: the method comprises the steps of firstly carrying out clutter suppression processing on each frame of data of second radar data frame by frame to obtain fourth radar data, then carrying out merging of fixed adjacent distance units frame by frame on each frame of data of the fourth radar data to obtain second merged data, carrying out fixed-scale potential target peak detection on the second merged data after clutter suppression processing and merging of the fixed adjacent distance units to obtain second merged data, obtaining a first two-dimensional feature map after linear amplification and amplitude limiting operation, continuously carrying out noise removal based on a preset noise threshold on the second merged data, obtaining a second two-dimensional feature map after linear amplification and amplitude limiting operation, carrying out frame by frame outlier iterative detection on each frame of the second radar data according to a first monitoring background model, and obtaining a third two-dimensional feature map after linear amplification and amplitude limiting operation. And finally, obtaining a first three-dimensional feature map from the first two-dimensional feature map, the second two-dimensional feature map and the third two-dimensional feature map.
As shown in fig. 4, step S260 in the method specifically includes:
and S261, performing short-time Fourier transform on the fourth radar data to obtain a first power spectrum.
S262, normalizing the first power spectrum to obtain a first power spectrum vector.
And S263, adding the sum of the numerical values of the first power spectrums to the first power spectrum vector to obtain a first STFT feature vector.
In this embodiment, the specific process of obtaining the first STFT feature vector from the second radar data may be: and (3) performing short-time Fourier transform on the fourth radar data obtained in the step S251 to obtain a first power spectrum, normalizing the first power spectrum to obtain a first power spectrum vector, and finally adding the sum of the values of the first power spectrum to the first power spectrum vector to obtain a first STFT feature vector.
Example IV
As shown in fig. 5, a fourth embodiment of the present invention provides a radar-based regional people counting device 100, and the radar-based regional people counting device 100 provided in the third embodiment of the present invention can execute the radar-based regional people counting method provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method. The radar-based regional people counting device 100 comprises a data acquisition module 200, a model building module 300 and a people counting module 400.
Specifically, the data acquisition module 200 is configured to acquire and store first radar data of a target area without a living body within a first preset time; the model building module 300 is configured to build a first monitoring background model according to the first radar data; the data acquisition module 200 is further configured to acquire and store second radar data within a second preset time of the target area; the people counting module 400 is used for counting the first people in the target area according to the first monitoring background model and the second radar data.
In this embodiment, the data acquisition module 200 is specifically configured to acquire and store second radar data within a second preset time of the target area, and store the second radar data without a living body within the second preset time of the target area as third radar data. The model building module 300 is specifically configured to combine each frame data of the first radar data by using a fixed adjacent distance unit to obtain first combined data; after the logarithm of the first combined data is obtained by means of modulo, carrying out mean and variance iteration based on the first monitoring background model to obtain first background data; and establishing a first monitoring background model according to the first background data. The people counting module 400 is specifically configured to obtain a first three-dimensional feature map according to the first monitoring background model and the second radar data; obtaining a first STFT feature vector according to the second radar data; and placing the first three-dimensional feature map and the first STFT feature vector into a pre-trained neural network model to count a first number of people in the target area. The people counting module 400 is specifically further configured to perform clutter suppression processing on each frame of data of the second radar data to obtain fourth radar data; merging each frame of data of the fourth radar data by a fixed adjacent distance unit to obtain second merged data; detecting a potential target peak of the second combined data in a fixed scale, and obtaining a first two-dimensional characteristic diagram after linear amplification and amplitude limiting operation; noise removal based on a preset noise threshold is carried out on the second combined data, and a second two-dimensional characteristic diagram is obtained after linear amplification and amplitude limiting operation; performing outlier iterative detection on each frame of data of the second radar data according to the first monitoring background model, and obtaining a third two-dimensional characteristic diagram after linear amplification and amplitude limiting operation; and obtaining a first three-dimensional characteristic diagram according to the first two-dimensional characteristic diagram, the second two-dimensional characteristic diagram and the third two-dimensional characteristic diagram. The people counting module 400 is specifically further configured to perform short-time fourier transform on the fourth radar data to obtain a first power spectrum; normalizing the first power spectrum to obtain a first power spectrum vector; and adding the sum of the values of the first power spectrums to the first power spectrum vector to obtain a first STFT feature vector.
Further, the radar-based regional population statistics apparatus 100 further includes a secondary statistics module 500, where the secondary statistics module 500 is configured to, after acquiring the third radar data, establish a second monitoring background model according to the third radar data if the continuous living body-free time of the target region exceeds a third preset time; and counting the number of people in a second area of the target area according to the second monitoring background model and second radar data.
Example five
Fig. 6 is a schematic structural diagram of a regional people counting computer device 12 based on radar according to a fifth embodiment of the present invention. FIG. 6 illustrates a block diagram of an exemplary computer device 12 suitable for use in implementing embodiments of the present invention. The computer device 12 shown in fig. 6 is merely an example and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.
As shown in FIG. 6, the computer device 12 is in the form of a general purpose computing device. Components of computer device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, a bus 18 that connects the various system components, including the system memory 28 and the processing units 16.
The system memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) 30 and/or cache memory 32. The computer device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read from or write to non-removable, nonvolatile magnetic media (not shown in FIG. 6, commonly referred to as a "hard disk drive"). Although not shown in fig. 6, a magnetic disk drive for reading from and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In such cases, each drive may be coupled to bus 18 through one or more data medium interfaces. Memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to carry out the functions of embodiments of the invention.
A program/utility 40 having a set (at least one) of program modules 42 may be stored in, for example, memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment. Program modules 42 generally perform the functions and/or methods of the embodiments described herein.
The computer device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), one or more devices that enable a user to interact with the computer device 12, and/or any devices (e.g., network card, modem, etc.) that enable the computer device 12 to communicate with one or more other computing devices. Such communication may occur through an input/output (I/O) interface 22. Moreover, computer device 12 may also communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet, through network adapter 20. As shown, network adapter 20 communicates with other modules of computer device 12 via bus 18. It should be appreciated that although not shown, other hardware and/or software modules may be used in connection with computer device 12, including, but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data backup storage systems, and the like.
The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, for example, implementing the methods provided by embodiments of the present invention:
acquiring and storing first radar data without living bodies in a first preset time of a target area;
establishing a first monitoring background model according to the first radar data;
acquiring and storing second radar data in a second preset time of the target area;
and counting a first number of people in the target area according to the first monitoring background model and the second radar data.
Example six
The sixth embodiment of the present invention further provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method as provided by all the inventive embodiments of the present application:
acquiring and storing first radar data without living bodies in a first preset time of a target area;
establishing a first monitoring background model according to the first radar data;
acquiring and storing second radar data in a second preset time of the target area;
and counting a first number of people in the target area according to the first monitoring background model and the second radar data.
The computer storage media of embodiments of the invention may take the form of any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).
Note that the above is only a preferred embodiment of the present invention and the technical principle applied. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, while the invention has been described in connection with the above embodiments, the invention is not limited to the above embodiments, but may include many other equivalent embodiments without departing from the spirit of the invention, the scope of which is determined by the scope of the appended claims.
Claims (8)
1. A method for radar-based regional population statistics, comprising:
acquiring and storing first radar data without living bodies in a first preset time of a target area;
establishing a first monitoring background model according to the first radar data;
acquiring and storing second radar data in a second preset time of the target area;
counting a first number of people in the target area according to the first monitoring background model and the second radar data;
the counting the first person number of the target area according to the first monitoring background model and the second radar data comprises:
obtaining a first three-dimensional feature map according to the first monitoring background model and the second radar data;
obtaining a first STFT feature vector according to the second radar data;
placing the first three-dimensional feature map and the first STFT feature vector into a pre-trained neural network model to count a first number of people in the target area;
the pre-trained neural network model may be a convolutional neural network model, and specifically includes nine 2D convolutional layers, four pooling layers, three full-connection layers and one classification layer, the classification layer may be a Softmax classification layer, wherein the 2D convolutional layers and the full-connection layers use ReLU activation functions, the convolution kernel and the convolution operation step length are different, and the pooling kernel size and the operation step length of the pooling layers are lm×lm and stride=nm respectively. The data flow direction of the method is through a 2D convolution layer, and the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layer are respectively L1 x L1, M1 and stride=N1; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L2 x L2, M2 and stride=N2; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L3 x L3, M3 and stride=N3; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L4 x L4, M4 and stride=N4; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L5 x L5, M5 and stride=N5; then realizing feature map decreasing of the first three-dimensional feature map and feature vector splicing of the first STFT feature vector, and enabling the completed data to flow to three full-connection layers, wherein the number of neurons of the three full-connection layers is F1, F2 and F3 respectively; finally, the Softmax classification layer is utilized to process and obtain the first number of people in the target area;
the obtaining a first three-dimensional feature map according to the first monitoring background model and the second radar data comprises the following steps:
performing clutter suppression processing on each frame of data of the second radar data to obtain fourth radar data;
merging each frame of data of the fourth radar data by a fixed adjacent distance unit to obtain second merged data;
detecting a potential target peak of the second combined data in a fixed scale, and obtaining a first two-dimensional characteristic diagram after linear amplification and amplitude limiting operation;
noise removal based on a preset noise threshold is carried out on the second combined data, and a second two-dimensional characteristic diagram is obtained after linear amplification and amplitude limiting operation;
performing outlier iterative detection on each frame of data of the second radar data according to the first monitoring background model, and obtaining a third two-dimensional characteristic diagram after linear amplification and amplitude limiting operation;
and obtaining a first three-dimensional characteristic diagram according to the first two-dimensional characteristic diagram, the second two-dimensional characteristic diagram and the third two-dimensional characteristic diagram.
2. The method of claim 1, wherein the acquiring and storing second radar data for a second predetermined time for the target area comprises:
and acquiring and storing second radar data of the target area within a second preset time, and additionally storing second radar data without living bodies within the second preset time of the target area as third radar data.
3. The method of claim 2, wherein counting the first number of people in the target area based on the first monitoring background model and second radar data comprises:
after the third radar data are acquired, if the continuous living body-free time of the target area exceeds a third preset time, a second monitoring background model is established according to the third radar data;
and counting the number of people in a second area of the target area according to the second monitoring background model and second radar data.
4. The method of claim 1, wherein the building a first monitoring context model from the first radar data comprises:
merging each frame of data of the first radar data by using a fixed adjacent distance unit to obtain first merged data;
after the logarithm of the first combined data is obtained by means of modulo, carrying out mean and variance iteration based on the first monitoring background model to obtain first background data;
and establishing a first monitoring background model according to the first background data.
5. The method of claim 1, wherein the deriving a first STFT feature vector from the second radar data comprises:
performing short-time Fourier transform on the fourth radar data to obtain a first power spectrum;
normalizing the first power spectrum to obtain a first power spectrum vector;
and adding the sum of the values of the first power spectrums to the first power spectrum vector to obtain a first STFT feature vector.
6. A radar-based regional population statistics apparatus, comprising:
the data acquisition module is used for acquiring and storing first radar data without living bodies in a first preset time of a target area;
the model building module is used for building a first monitoring background model according to the first radar data;
the data acquisition module is also used for acquiring and storing second radar data in a second preset time of the target area;
the people counting module is used for counting the first people in the target area according to the first monitoring background model and the second radar data;
the counting the first person number of the target area according to the first monitoring background model and the second radar data comprises:
obtaining a first three-dimensional feature map according to the first monitoring background model and the second radar data;
obtaining a first STFT feature vector according to the second radar data;
placing the first three-dimensional feature map and the first STFT feature vector into a pre-trained neural network model to count a first number of people in the target area;
the pre-trained neural network model may be a convolutional neural network model, and specifically includes nine 2D convolutional layers, four pooling layers, three full-connection layers and one classification layer, the classification layer may be a Softmax classification layer, wherein the 2D convolutional layers and the full-connection layers use ReLU activation functions, the convolution kernel and the convolution operation step length are different, and the pooling kernel size and the operation step length of the pooling layers are lm×lm and stride=nm respectively. The data flow direction of the method is through a 2D convolution layer, and the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layer are respectively L1 x L1, M1 and stride=N1; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L2 x L2, M2 and stride=N2; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L3 x L3, M3 and stride=N3; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L4 x L4, M4 and stride=N4; then passing through a pooling layer; then passing through two identical 2D convolution layers, wherein the convolution kernel size, the number of the convolution kernels and the convolution operation step length of the 2D convolution layers are respectively L5 x L5, M5 and stride=N5; then realizing feature map decreasing of the first three-dimensional feature map and feature vector splicing of the first STFT feature vector, and enabling the completed data to flow to three full-connection layers, wherein the number of neurons of the three full-connection layers is F1, F2 and F3 respectively; finally, the Softmax classification layer is utilized to process and obtain the first number of people in the target area;
the obtaining a first three-dimensional feature map according to the first monitoring background model and the second radar data comprises the following steps:
performing clutter suppression processing on each frame of data of the second radar data to obtain fourth radar data;
merging each frame of data of the fourth radar data by a fixed adjacent distance unit to obtain second merged data;
detecting a potential target peak of the second combined data in a fixed scale, and obtaining a first two-dimensional characteristic diagram after linear amplification and amplitude limiting operation;
noise removal based on a preset noise threshold is carried out on the second combined data, and a second two-dimensional characteristic diagram is obtained after linear amplification and amplitude limiting operation;
performing outlier iterative detection on each frame of data of the second radar data according to the first monitoring background model, and obtaining a third two-dimensional characteristic diagram after linear amplification and amplitude limiting operation;
and obtaining a first three-dimensional characteristic diagram according to the first two-dimensional characteristic diagram, the second two-dimensional characteristic diagram and the third two-dimensional characteristic diagram.
7. A radar-based regional people counting device, comprising:
one or more processors;
storage means for storing one or more programs,
when executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1-5.
8. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the method according to any of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010200465.7A CN111444926B (en) | 2020-03-20 | 2020-03-20 | Regional population counting method, device and equipment based on radar and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010200465.7A CN111444926B (en) | 2020-03-20 | 2020-03-20 | Regional population counting method, device and equipment based on radar and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111444926A CN111444926A (en) | 2020-07-24 |
| CN111444926B true CN111444926B (en) | 2023-06-13 |
Family
ID=71648970
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010200465.7A Active CN111444926B (en) | 2020-03-20 | 2020-03-20 | Regional population counting method, device and equipment based on radar and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111444926B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113313165B (en) * | 2021-05-27 | 2023-11-24 | 深圳大学 | People counting method, device, equipment and storage medium based on radar |
| CN113311405B (en) * | 2021-05-27 | 2023-06-20 | 深圳大学 | Regional population counting method and device, computer equipment and storage medium |
| CN116449330B (en) * | 2023-06-20 | 2023-10-13 | 精华隆智慧感知科技（深圳）股份有限公司 | Indoor people number estimation method and device, computer equipment and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110363151A (en) * | 2019-07-16 | 2019-10-22 | 中国人民解放军海军航空大学 | Based on the controllable radar target detection method of binary channels convolutional neural networks false-alarm |
-
2020
- 2020-03-20 CN CN202010200465.7A patent/CN111444926B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110363151A (en) * | 2019-07-16 | 2019-10-22 | 中国人民解放军海军航空大学 | Based on the controllable radar target detection method of binary channels convolutional neural networks false-alarm |
Non-Patent Citations (2)
| Title |
|---|
| Runhan Bao 等.Short-Range Moving Human Detection Based-on Cascaded Spatial-Temporal Three-Stages Detector in UWB Radar.《IEEE》.2019,第1-6页. * |
| 朱克凡 ; 王杰贵 ; .基于卷积神经网络的低分辨雷达目标一步识别技术.空军工程大学学报(自然科学版).2019,(05),第87-93页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111444926A (en) | 2020-07-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111444926B (en) | Regional population counting method, device and equipment based on radar and storage medium | |
| US11100320B2 (en) | Image recognition method and apparatus | |
| CN109697435B (en) | People flow monitoring method and device, storage medium and equipment | |
| CN108960090B (en) | Video image processing method and device, computer readable medium and electronic equipment | |
| CN112904331B (en) | Method, device, equipment and storage medium for determining moving track | |
| CN112082553A (en) | Indoor positioning method and positioning device based on WIFI and laser radar and robot | |
| CN113780270A (en) | Target detection method and device | |
| CN113627229A (en) | Object detection method, system, device and computer storage medium | |
| CN113012200A (en) | Method and device for positioning moving object, electronic equipment and storage medium | |
| Salma et al. | Smart parking guidance system using 360o camera and haar-cascade classifier on iot system | |
| CN111812670B (en) | Single photon laser radar space transformation noise judgment and filtering method and device | |
| CN110555352A (en) | interest point identification method, device, server and storage medium | |
| CN114332553A (en) | Image processing method, device, equipment and storage medium | |
| JP7484492B2 (en) | Radar-based attitude recognition device, method and electronic device | |
| Ekici | Detecting damaged buildings from satellite imagery | |
| CN116384148A (en) | Meta universe modeling system of radar countermeasure system | |
| CN111815748A (en) | Animation processing method and device, storage medium and electronic equipment | |
| CN113311405B (en) | Regional population counting method and device, computer equipment and storage medium | |
| CN116385957A (en) | X-ray image contraband detection method, device, equipment and medium | |
| CN113469086B (en) | Method, device, equipment and medium for dividing area in building plan | |
| CN113063420B (en) | Indoor positioning method, device, electronic equipment and storage medium | |
| CN111259702A (en) | User interest estimation method and device | |
| Ye et al. | Lightweight and Deep Appearance Embedding for Multiple Object Tracking | |
| CN115166782A (en) | Navigation scene recognition method, system, device and medium based on satellite distribution | |
| CN115100536A (en) | Building identification method, building identification device, electronic equipment and computer readable medium |
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 |