CN109034356B - Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume - Google Patents

Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume Download PDF

Info

Publication number
CN109034356B
CN109034356B CN201810810986.7A CN201810810986A CN109034356B CN 109034356 B CN109034356 B CN 109034356B CN 201810810986 A CN201810810986 A CN 201810810986A CN 109034356 B CN109034356 B CN 109034356B
Authority
CN
China
Prior art keywords
insect
radar
rcs
insects
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
Application number
CN201810810986.7A
Other languages
Chinese (zh)
Other versions
CN109034356A (en
Inventor
王锐
胡程
蔡炯
龙腾
曾涛
张天然
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Institute of Technology BIT
Original Assignee
Beijing Institute of Technology BIT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Institute of Technology BIT filed Critical Beijing Institute of Technology BIT
Priority to CN201810810986.7A priority Critical patent/CN109034356B/en
Publication of CN109034356A publication Critical patent/CN109034356A/en
Application granted granted Critical
Publication of CN109034356B publication Critical patent/CN109034356B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06M—COUNTING MECHANISMS; COUNTING OF OBJECTS NOT OTHERWISE PROVIDED FOR
    • G06M11/00—Counting of objects distributed at random, e.g. on a surface
    • G06M11/02—Counting of objects distributed at random, e.g. on a surface using an electron beam scanning a surface line by line, e.g. of blood cells on a substrate

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention discloses an insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume, and provides an effective means for accurately counting the density of each height layer of migratory flying insects. Compared with the existing insect radar density statistical method, the method associates the same insect appearing at different scanning wave positions through the nearest neighbor method, effectively prevents the same insect from being counted for many times, and reduces the false alarm rate of detection; the radar detection volume is calculated through Gaussian beams, different radar detectable regions corresponding to insects of different RCSs are considered, and the volume calculation is more reasonable.

Description

Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume
Technical Field
The invention belongs to the technical field of insect radars, and relates to an insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume.
Background
Insect migration is a very important biological phenomenon in entomology. The research on the migratory flight of insects can help human beings to improve the understanding on nature, and enhance the early warning, prevention and control of plant diseases and insect pests. The method has important significance for researching the insect migration by acquiring characteristic parameters such as time, type, direction, height, density and the like of the insect migration.
In the process of insect migration, layering phenomenon occurs. In biology, it is often desirable to know the distribution density of insects in various height layers to understand the migratory flight habits and even the life habits of insects. The insect radar is used as a powerful tool for detecting high-altitude insect migration, and the work of insect density statistics can be completed quickly and efficiently.
The scanning type insect radar can rapidly scan a large airspace so as to obtain the information of the quantity, the density, the distribution and the like of insects in the large airspace. When the insect density is counted by using Radar data, the traditional method is that the Radar rotates 360 degrees under a certain pitch angle, the number of the insects detected by the Radar is counted, the detectable volume of the Radar is calculated by the average Radar scattering Cross Section (Radar-Cross Section RCS) of the insects and is used as the total volume, and finally the insect number is divided by the total volume to obtain the density.
However, this approach has two problems: firstly, in scanning radar detection, the detectable volume of the radar is very complicated to calculate, and the farthest detection distances of the radar are different for different RCS targets, so that the detectable volume difference is very large. The traditional method only adopts average RCS to calculate the detectable volume uniformly, and the error is very large; secondly, counting the number of insects: since the RCS of a large insect is large, the detection distance of the radar is long, and the farthest detection distance of the radar is correspondingly large. Thus, during radar scanning, large insects stay in the detectable detection volume for a longer time and are detected multiple times at multiple wave positions, which results in multiple statistics for one insect.
Disclosure of Invention
In view of the above, the invention provides an insect density statistical method based on nearest neighbor method correlation and gaussian beam volume, which can improve the precision value of insect density and provide an effective means for accurately counting the density of each height layer of migratory flying insects.
Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume:
firstly, observing migrating insects by using a two-dimensional scanning insect radar to obtain insect radar echoes under each wave position; performing pulse Doppler PD processing on the insect radar echo under each wave position to obtain radar data under each wave position; associating radar data under each wave position by adopting a nearest neighbor method to obtain an insect track; obtaining the number of insects in each height layer and each RCS section of the scattering cross section area of the insect radar through insect track division;
step two, acquiring insect radar scanning volumes of different height layers in different RCS sections by Gaussian beam rotation integration;
and step three, obtaining the insect density corresponding to each RCS section and each height layer according to the insect radar scanning volume and the quantity of the insects in each RCS section and each height layer.
Preferably, in the step one, the specific method for obtaining the insect track by correlating the radar data under each wave position by using the nearest neighbor method includes:
step 1, determining whether each detection unit in radar data under each wave position has a target;
step 2, determining a tracking space: discretizing the inherent detection distance of each wave position of the insect radar to obtain a series of distance units; forming a line of distance units belonging to the same wave position, and forming a rectangle according to the sequence of the wave positions of all the wave positions of one screen scanned by the insect radar; then, recording the signal-to-noise ratio value of the target at the distance unit corresponding to the detection unit with the target in the step 1; recording the distance unit corresponding to the detection unit without the target as '0', and finally forming a matrix, wherein the matrix is a tracking space;
step 3, determining a time threshold of an associated gate according to a coherent processing interval CPI value required by an angle of the insect radar for scanning a beam width; taking the maximum distance unit number which can be crossed by the insects in one CPI as the distance threshold of the associated door;
searching according to rows, enabling a first non-zero value in the tracking space to serve as a track starting point of the insect, searching non-zero values in the correlation gate according to rows by taking the track starting point as a center, and enabling the nearest non-zero value to serve as a next appearing point track of the insect until no non-zero value exists in a new correlation gate, so that the point track correlation of the insect is finished.
Preferably, in step 1, the radar data is:
the insect radar starts scanning at any elevation angle, PD processing is carried out on the insect radar echo within each CPI time, and a Doppler-distance domain corresponding to the insect radar wave of each wave position at the elevation angle is obtained, namely the radar data.
Preferably, in step 1, the method for determining whether each detection unit has a target according to radar data includes:
detecting one detection unit in the radar data under any wave position, and sequencing sampling values of reference units of the detection unit from small to large; determining a sampling value of a kth reference unit as a total background clutter power level estimation according to the false alarm rate; if the power of the detection unit is larger than the constant false alarm threshold, the target in the detection unit is considered to exist; otherwise, judging that no target exists;
according to the method, all detection units in all radar data are traversed.
Preferably, the CPI value is set to the time required for the insect radar to sweep through an angle of half the beamwidth.
Preferably, in the step one, the number of insects in each RCS segment in each altitude layer obtained by insect track division is:
obtaining the RCS and the ground-distance height of each insect according to the elevation angle of the insect radar, the ground-distance height of the insect radar, the maximum signal-to-noise ratio of the echo of each insect and the distance between the insect and the insect radar; then dividing a height layer and dividing an RCS section of the insect; and counting the number of insects in each height layer and each RCS section according to the obtained insect track.
Preferably, in the second step, the specific implementation method for obtaining the insect radar scanning volumes of different height layers in different RCS sections by gaussian beam rotation integration includes:
step 2.1, according to
Figure BDA0001739148990000041
Obtaining the farthest detection distance r of the kth RCS segment0Wherein σ iskDenotes the RCS median, SNR, of the k-th RCS segmentminC is a constant associated with the insect radar system for the minimum detectable signal-to-noise ratio of the insect radar receiver;
step 2.2, according to the wave beam width theta of the insect radar3dbAnd the farthest detection range r of the kth RCS segment0Determining the detection range of the insect radar in the kth RCS section by using an insect radar main lobe Gaussian beam formula;
2.3, cutting Gaussian beams according to the elevation angle of the insect radar and the correspondingly divided height layers to obtain the scanning volume of the insect radar in each height layer in the kth RCS section;
and 2.4, traversing all the RCS sections by adopting the methods of the steps 2.1 to 2.3 to obtain the insect radar scanning volumes of different height layers in different RCS sections.
Preferably, the specific method for cutting the gaussian beam is to approximately convert the gaussian beam into cutting on the central axis of the gaussian beam by cutting on the height layer.
Preferably, in step one, the insect radar scans in a multi-elevation mode without overlapping elevation beams.
Has the advantages that:
the insect density statistical method based on the nearest neighbor method correlation and the Gaussian beam volume provides an effective means for accurately counting the density of each height layer of the migratory flying insects. Compared with the existing insect radar density statistical method, the method associates the same insect appearing at different scanning wave positions through the nearest neighbor method, effectively prevents the same insect from being counted for many times, and reduces the false alarm rate of detection; the radar detection volume is calculated through Gaussian beams, different radar detectable regions corresponding to insects of different RCSs are considered, and the volume calculation is more reasonable.
Drawings
Fig. 1 shows the beam shape of the radar as it scans different RCS targets.
Fig. 2(a) is a schematic diagram of a radar gaussian beam cut by a certain height layer, and the region between the upper and lower boundary lines of the height layer is in the shape after cutting.
Fig. 2(b) is a conversion of the cut section in the height layer direction into a cut section on the central axis in the radar beam.
Fig. 2(c) is a plurality of trapezoids further subdivided in the central axis direction with respect to the converted cross section.
FIG. 3 is a statistical density map based on measured insect radar scan data.
Detailed Description
The invention is described in detail below by way of example with reference to the accompanying drawings.
The method comprises the steps of observing migratory flying insects by using a two-dimensional scanning insect radar, correlating radar data and counting the number of insects by using a nearest neighbor method, calculating the scanning volume of the radar by using Gaussian beam rotation integral, and dividing the number of the insects by the corresponding volume to obtain the insect density.
Scanning radars typically discretize elevation angles from 0 to 90, with the radar performing airspace scanning at each elevation angle by continuous rotation of the azimuth. It should be noted that the present invention counts insect density in a scanning mode with a fixed elevation angle or a radar at multiple elevation angles and no overlapping of beams at each elevation angle.
Step one, counting the number of insects
The key of the quantity statistics is the detection and the correlation of the target, and the radar data needs to be processed carefully. And for the subsequent volume calculation of Gaussian beams and the calculation of insect density of each height layer, the detected insects need to be classified more finely according to the height layers and the RCS of the insects. The following is the procedure for insect population statistics:
step 1.1 PD treatment
In general, the main lobe beamwidth of a radar is: the angle between the two half-power points (3db) of the beam. Setting a Coherent Process Interval (CPI) as the time required by the radar to scan an angle of half beam width at a certain elevation angle, performing Fast Fourier Transform (FFT) on a slow time for a plurality of frames of one-dimensional range profiles in each CPI, converting a time-range domain into a Doppler-range domain (PD plane), and obtaining each wave position and a Doppler-range domain corresponding to the radar wave.
And step 1.2, judging the target by adopting an ordered statistic constant false alarm rate threshold (OS-CFAR) detection method.
After PD processing, the target is detected by using OS-CFAR.
Firstly, detecting a certain detection unit in a Doppler-distance domain under any wave position, and sequencing sampling values of a reference unit of the detection unit:
x(1)≤x(2)≤…≤x(R) (1)
wherein x is(1)≤x(2)≤…≤x(R)R is the number of the reference units.
The k-th sample is then taken as the total background clutter power level estimate Z, where k is determined based on the false alarm rate, i.e. k is a measure of the total background clutter power level
Z＝x(k) (2)
In the Doppler-distance domain, if the power of the detection unit is greater than a constant false alarm threshold Z, the detection unit is considered to have a target, and the distance unit corresponding to the detection unit and the signal-to-noise ratio value of the target are recorded; where the signal-to-noise ratio is obtained from the power of the unit targeted by the target divided by the average power of the entire PD plane. And traversing all the PD planes in step 1.1 according to the method of step 1.2.
Step 1.3 associating insect stippling by nearest neighbor method
Since we set the time for the radar to scan half the beamwidth as CPI, half the beam is overlapped between adjacent wave positions, that is, a part of the insects is repeatedly detected. In addition, the RCS of large insects is large and may be detectable at multiple wave positions. In order to prevent repeated insect quantity counting, a nearest neighbor method is adopted to associate targets with adjacent wave positions appearing in a close distance unit, and the targets are regarded as one target. How the association is made is described below.
First the tracking space needs to be determined. Discretizing the inherent detection distance of each wave position of the radar to obtain a series of distance units; and forming a line of distance units belonging to the same wave position, and forming a rectangle according to the sequence of the wave positions of all the wave positions of one screen scanned by the radar. Then, according to the records in the step 1.2, recording the signal-to-noise ratio value of the corresponding target at the distance unit where the target exists, which is recorded in each wave position, wherein the distance unit which is not marked stores '0', and finally forming a matrix, wherein the matrix is a tracking space, the number of rows is time, and the time is increased from top to bottom; the number of columns is distance, increasing from left to right.
After the tracking space is determined, an associated door is required to be arranged. The angle at which the radar sweeps across one beamwidth requires two CPIs, so 2 is the time threshold for the associated gate; the maximum number of range cells that an insect can cross within a CPI is the range threshold of the associated door.
Performing trace point association in a tracking space:
searching according to rows, and searching a first non-zero value in a tracking space from a first row to make the first non-zero value as a track starting point of an insect;
setting a correlation gate by taking the starting point of the flight path as the center, searching non-zero values in the correlation gate according to rows, wherein the non-zero value closest to the correlation gate is the next appearing point path of the insect;
and thirdly, repeating the step II until no nonzero value exists in the new association door, and finishing the trace point association of the insect. Recording the maximum non-zero value and the distance unit where the maximum non-zero value is located in the tracking process, and then clearing all the associated traces in the tracking process;
and fourthly, repeating the step one until no nonzero value exists in the tracking space and the point trace correlation of all the insects is finished.
Step 1.4 Classification statistics of insect numbers
According to the radar elevation angle, the radar ground clearance, the maximum signal-to-noise ratio of each insect echo and the distance between the insects and the radar, the RCS and the ground clearance of each insect can be calculated, and the insects are classified and counted according to the RCS and the ground clearance.
Firstly, the height h of each insect from the ground is calculatedinsect. This can be determined by the ground height h of the radarradarElevation angle beta of radar, distance r of insect from radarinsectAnd (6) obtaining. The calculation formula is as follows:
hinsect＝hradar+rinsectsinβ (3)
the RCS of each insect then needs to be found. It can utilize SNR of insect echoinsectDistance r of insect from radarinsectAnd the data are calculated through a deformation form of a radar equation.
The radar equation is:
Figure BDA0001739148990000071
wherein, PrAnd G is the transmission gain, lambda is the wavelength, sigma is the radar cross section, R is the target distance, and L is the system loss.
From the equation (4), the signal-to-noise ratio of the target echo can be expressed as:
Figure BDA0001739148990000081
wherein P isNIs the receiver noise power, which is typically constant.
Order to
Figure BDA0001739148990000082
Then
Figure BDA0001739148990000083
By using equation (6), the RCS of the insect can be obtained:
Figure BDA0001739148990000084
in the formula, σinsectFor an insect RCS, the coefficient C is a constant related to the radar system. Therefore, the RCS of the insect can be roughly estimated by the formula (7).
Then, the height layer is divided, and the RCS section of the insect is divided. The minimum ground height of the detected insects is hminMaximum distance height of hmaxTo h is aligned withmin～hmaxUniformly quantizing and dividing height layers; the minimum RCS of the detected insects is sigmaminMaximum RCS is σmaxTo σmin～σmaxAnd (4) uniformly quantizing and dividing the RCS section.
And finally, counting the number of insects in each height layer and each RCS section according to the insect tracks obtained in the step 1.3. For convenience of description, the number of insects in the kth RCS segment recorded in the nth height layer is Numk,n。
Second, calculation of radar detectable volume
As shown in fig. 1, the detectable volume of the radar is different for targets of different RCS; the radar detectable volumes are different for different height layers.
The RCS values of all targets within the same RCS segment are collectively considered to be the median value of the RCS segment. And the RCS median is used to calculate the detectable volumes of the targets in the various height slices.
Suppose σkDenotes the mean RCS value of the k-th RCS segment, with RCS being σkThe farthest detection distance of the target of (1) is r0The minimum detectable signal-to-noise ratio of the radar receiver is SNRminThe maximum detection distance r can be obtained from the formula (6)0Comprises the following steps:
Figure BDA0001739148990000091
c is a constant associated with the radar system.
Then according to the beam width theta of the radar3dbAnd the detection range of the radar can be determined by utilizing a radar main lobe Gaussian beam formula. The following describes the specific derivation process:
in parabolic antennas, the target deviatesThe central axis of the main lobe of the antenna deviates from the angle theta, the gain G of the antenna in the directionθComprises the following steps:
Figure BDA0001739148990000092
wherein G is0The antenna gain in the direction of the central axis.
Minimum detectable signal-to-noise ratio (SNR) of radar receiverminThe relation between the farthest detection distance R and the antenna gain G can be obtained by the radar equation of the formula (5):
Figure BDA0001739148990000093
g is to be0And GθIn the formula (10), the following compounds are obtained:
Figure BDA0001739148990000094
wherein r isθIndicating the farthest detection range of the radar at the angle of departure theta.
Combining formula (9) with formula (11) to obtain:
Figure BDA0001739148990000095
the detectable region of the radar can be determined by equation (12), and is in the shape of a gaussian beam.
Then, according to the elevation angle of the radar and the division of the height layer, a gaussian beam is cut, as shown in fig. 2 (a). Since the cross section formed by cutting the gaussian beam for each height layer is irregular, it is difficult to directly calculate the rotation volume. An approximation calculation method is proposed herein that approximates the volume in the elevation direction using the volume in the pitch direction.
As shown in fig. 2(b), the gaussian beam is first cut approximately on the elevation layer to be cut on the central axis of the gaussian beam. The cutting conversion method for a layer with a certain height comprises the following steps: the intersection point of the upper and lower boundary lines of the height layer and the central axis of the Gaussian beam is Q, P, a perpendicular line perpendicular to the central axis is made after Q, P points are passed, and the region formed by the two points and the Gaussian beam is the converted cutting region. Wherein Q is the intersection point of the upper boundary line of the height layer and the central axis of the Gaussian beam, and P is the intersection point of the lower boundary line of the height layer and the central axis of the Gaussian beam.
The converted cross section is then further subdivided in the direction of the central axis, and the subdivided shape is approximately trapezoidal, as shown in fig. 2 (c). And (4) integrating the rotation of each trapezoid, and summing to obtain the rotation volume of the cutting section in the central axis direction, wherein the rotation volume is approximate to the rotation volume of the cutting section on the height layer.
The detectable volumes need to be calculated separately for different elevation angles. When the radar detects the target of the kth RCS section under the mth elevation angle, the detectable volume under the nth height layer is
Figure BDA0001739148990000101
Adding up the corresponding volumes at the respective elevation angles, i.e.
Figure BDA0001739148990000102
Where M is the number of all elevation angles, Vk,nRepresenting the detectable volume under the nth height layer for the target of the kth RCS segment after the radar completes the two-dimensional scan.
Third, density determination
First, the density of insects within each RCS segment, in each height level, was calculated as:
Figure BDA0001739148990000104
where rhok,nRepresenting the density of insects in the kth RCS segment in the nth altitude layer.
Then, summing up the insects of each RCS section of each height layer to obtain the insect density of each height layer:
Figure BDA0001739148990000103
Ρnrepresents the density of insects in the nth height layer, and K represents the number of divided RCS segments.
The following examples illustrate the implementation steps:
in order to verify the density statistical method, based on actually measured data of the Ku-band entomology radar, the insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume is adopted to complete insect density statistics. The radar parameters used in the experiment are shown in table 1.
TABLE 1 Radar parameters
Figure BDA0001739148990000111
The radar scans 180 degrees in azimuth under the elevation angles of 30 degrees, 45 degrees, 60 degrees and 75 degrees respectively. And preprocessing the original radar echo data to obtain a one-dimensional range profile of each frame.
Step one, counting the number of insects:
1. detecting association processing
The following detection association processing is respectively carried out on the scanning data under the elevation angles of 30 degrees, 45 degrees, 60 degrees and 75 degrees:
firstly, PD processing and OS-CFAR detection are carried out on one-dimensional range profile data of every 20 frames, and the signal-to-noise ratio of a target detected in each wave position and a distance unit where the target is located are recorded;
and then, associating the insect traces by using a nearest neighbor method. The distance threshold for the associated gate is set to 3 distance units and the time threshold is set to 2 CPIs. And recording the maximum value of the echo signal-to-noise ratio of the insects and the distance unit where the insects are located when the maximum signal-to-noise ratio is reached in the tracking process of each target.
2. Statistical classification of insect numbers
And calculating the distance height and RCS of each insect according to the radar elevation angle, the radar ground height, the maximum signal-to-noise ratio of each insect echo and the distance between the insect and the radar.
Height layer and insect RCS division: the height range of the insects from the ground is 120-600 m, each 10m is divided into one layer, 120-130 m is the first layer, and the rest is done in the same way; the RCS range of the insects is-70 dbsm to-10 dbsm, each 1dbsm is divided into a section, and-70 dbsm to-69 dbsm are divided into a first section, and so on.
Counting the number Num of insects in each RCS section in each height layerk,nAs shown in table two.
TABLE 2 categorical insect numbers
Figure BDA0001739148990000121
Step two, calculating the radar detection volume:
when the elevation angles are respectively calculated to be 30 degrees, 45 degrees, 60 degrees and 75 degrees, the radar detects the target of each RCS section, and the detectable volume in each height layer is as follows:
firstly, determining the farthest detection distance when the radar detects the target of each RCS section according to the median value of each RCS section;
then according to the beam width theta of the radar3dbBy utilizing a radar main lobe Gaussian beam formula, the range of the radar for detecting the target of each RCS section can be determined, and the shape of the target is Gaussian beam;
cutting Gaussian beams according to the elevation angle of the radar and the division of the height layer;
and converting the cutting on the height layer into the cutting on the central axis of the Gaussian beam, approximating the converted section by a smaller trapezoid (the height of the trapezoid is 1m), integrating the rotation of each trapezoid, and summing to obtain the rotation volume of the cutting section in the direction of the central axis, wherein the rotation volume is approximated to the rotation volume of the cutting section on the height layer.
The elevation angles of 30 °, 45 °, 60 °, 75 ° are numbered 1,2,3,4, respectively.
Noting the m elevation angle, the radar has the volume under the n height layer of the target of the k RCS section as
Figure BDA0001739148990000131
Adding the corresponding volumes at each elevation angle to obtain the volume V detected by the radar in each altitude layer for each RCS targetk,nNamely:
Figure BDA0001739148990000132
step three, calculating insect density:
calculating the density rho of the insects in each RCS section in each height layerk,nNamely:
Figure BDA0001739148990000133
the insect density Pp of each height layer can be obtained by summing the insects of each RCS section of each height layern。
Figure BDA0001739148990000134
The insect density distribution results of the height layers are shown in FIG. 3, and the insect density is maximum at 3.9070 x 10 between 140m and 150m-4Only/m3。
The method can be applied to insect radars to realize the statistics of the insect density of each height layer.
In summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (8)

1. An insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume is characterized in that,
firstly, observing migrating insects by using a two-dimensional scanning insect radar to obtain insect radar echoes under each wave position; performing pulse Doppler PD processing on the insect radar echo under each wave position to obtain radar data under each wave position; associating radar data under each wave position by adopting a nearest neighbor method to obtain an insect track; obtaining the number of insects in each height layer and each RCS section of the scattering cross section area of the insect radar through insect track division;
the specific method for obtaining the insect track by correlating radar data under each wave position by adopting the nearest neighbor method comprises the following steps:
step 1, determining whether each detection unit in radar data under each wave position has a target;
step 2, determining a tracking space: discretizing the inherent detection distance of each wave position of the insect radar to obtain a series of distance units; forming a line of distance units belonging to the same wave position, and forming a rectangle according to the sequence of the wave positions of all the wave positions of one screen scanned by the insect radar; then, recording the signal-to-noise ratio value of the target at the distance unit corresponding to the detection unit with the target in the step 1; recording the distance unit corresponding to the detection unit without the target as '0', and finally forming a matrix, wherein the matrix is a tracking space;
step 3, determining a time threshold of an associated gate according to a coherent processing interval CPI value required by an angle of the insect radar for scanning a beam width; taking the maximum distance unit number which can be crossed by the insects in one CPI as the distance threshold of the associated door;
searching according to rows, enabling a first non-zero value in the tracking space to serve as a track starting point of the insect, searching non-zero values in the associated door according to rows by taking the track starting point as a center, and enabling the non-zero value closest to the track starting point to serve as a next-appearing track of the insect until no non-zero value exists in a new associated door, so that the track association of the insect is finished;
step two, acquiring insect radar scanning volumes of different height layers in different insect RCS sections by Gaussian beam rotation integration;
and thirdly, obtaining the insect density corresponding to each height layer and each insect RCS section according to the insect radar scanning volume and the insect number in each height layer and each insect RCS section.
2. The insect density statistical method according to claim 1, wherein in step 1, the radar data is:
the insect radar starts scanning at any elevation angle, PD processing is carried out on the insect radar echo within each CPI time, and a Doppler-distance domain corresponding to the insect radar wave of each wave position at the elevation angle is obtained, namely the radar data.
3. The insect density statistical method according to claim 1, wherein in the step 1, the method for determining whether the target exists in each detection unit according to the radar data comprises:
step a, detecting one detection unit in the radar data under any wave position, and sequencing sampling values of reference units of the detection unit from small to large; determining a sampling value of a kth reference unit as a total background clutter power level estimation according to the false alarm rate; if the power of the detection unit is larger than the constant false alarm threshold, the target in the detection unit is considered to exist; otherwise, judging that no target exists;
and c, traversing all detection units in all radar data according to the method in the step a.
4. The insect density statistic method according to claim 1, wherein the CPI value is set as a time required for the insect radar to sweep through an angle of half a beam width.
5. The insect density statistical method according to claim 1, wherein in the first step, the number of insects in each height layer and each RCS section of insects obtained by dividing the insect track is:
obtaining the RCS and the ground-distance height of each insect according to the elevation angle of the insect radar, the ground-distance height of the insect radar, the maximum signal-to-noise ratio of the echo of each insect and the distance between the insect and the insect radar; then dividing a height layer and dividing an RCS section of the insect; and counting the number of insects in each height layer and each RCS section of the insects according to the obtained insect tracks.
6. The insect density statistical method according to claim 1, wherein in the second step, the insect radar scanning volumes of different height layers in different insect RCS sections are obtained by Gaussian beam rotation integration by:
step 2.1, according to
Figure FDA0002950506120000031
Obtaining the farthest detection distance r of the kth insect RCS section0Wherein σ iskMean RCS, SNR, of the k-th insect RCS segmentminC is a constant associated with the insect radar system for the minimum detectable signal-to-noise ratio of the insect radar receiver;
step 2.2, according to the wave beam width theta of the insect radar3dbAnd the farthest detection distance r of the kth insect RCS segment0Determining the detection range of the insect radar in the kth insect RCS section by using an insect radar main lobe Gaussian beam formula;
2.3, cutting Gaussian beams according to the elevation angle of the insect radar and the correspondingly divided height layers to obtain the scanning volume of the insect radar in each height layer in the kth insect RCS section;
and 2.4, traversing all the RCS sections of the insects by adopting the methods in the steps 2.1 to 2.3 to obtain the radar scanning volumes of the insects in different RCS sections and different height layers.
7. The insect density statistic method according to claim 6, wherein said cutting the Gaussian beam is performed by approximately converting the Gaussian beam to be cut on the central axis of the Gaussian beam at the height level.
8. The insect density statistical method according to claim 1, wherein in the first step, the insect radar scans in a multi-elevation mode without overlapping elevation beams.
CN201810810986.7A 2018-07-23 2018-07-23 Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume Active CN109034356B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810810986.7A CN109034356B (en) 2018-07-23 2018-07-23 Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810810986.7A CN109034356B (en) 2018-07-23 2018-07-23 Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume

Publications (2)

Publication Number Publication Date
CN109034356A CN109034356A (en) 2018-12-18
CN109034356B true CN109034356B (en) 2021-06-01

Family

ID=64645021

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810810986.7A Active CN109034356B (en) 2018-07-23 2018-07-23 Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume

Country Status (1)

Country Link
CN (1) CN109034356B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113093179B (en) * 2021-03-08 2022-07-12 北京理工大学前沿技术研究院 Insect density monitoring method based on weather radar

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5689444A (en) * 1995-06-07 1997-11-18 The United States Of America, As Represented By The Secretary Of Commerce Statistical quality control of wind profiler data
WO2013043636A3 (en) * 2011-09-23 2014-05-15 Donald Ronning Method and system for detecting animals in three dimensional space and for inducing avoidance in an animal
CN105759252A (en) * 2016-03-24 2016-07-13 北京理工大学 Insect dimension measurement method based on multi-frequency scattering modeling
CN105928448A (en) * 2016-04-18 2016-09-07 北京理工大学 Insect dimension measurement method based on Rayleigh region analysis scattering modeling
CN106022340A (en) * 2016-05-17 2016-10-12 南京理工大学 Improved Gaussian mixed potential probability hypothesis density filtering method
CN106501802A (en) * 2016-04-18 2017-03-15 北京理工大学 High-resolution multidimensional synergistic insect is migrated Radar Measurement Instrument
CN106940734A (en) * 2017-04-24 2017-07-11 南京信息工程大学 A kind of Migrating Insects monitor recognition methods and device in the air

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7864103B2 (en) * 2007-04-27 2011-01-04 Accipiter Radar Technologies, Inc. Device and method for 3D height-finding avian radar
CN105116393A (en) * 2015-06-12 2015-12-02 中国电子科技集团公司第四十一研究所 Method for estimating flight height of high-altitude object and radar sectional area based on position fingerprint

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5689444A (en) * 1995-06-07 1997-11-18 The United States Of America, As Represented By The Secretary Of Commerce Statistical quality control of wind profiler data
WO2013043636A3 (en) * 2011-09-23 2014-05-15 Donald Ronning Method and system for detecting animals in three dimensional space and for inducing avoidance in an animal
CN105759252A (en) * 2016-03-24 2016-07-13 北京理工大学 Insect dimension measurement method based on multi-frequency scattering modeling
CN105928448A (en) * 2016-04-18 2016-09-07 北京理工大学 Insect dimension measurement method based on Rayleigh region analysis scattering modeling
CN106501802A (en) * 2016-04-18 2017-03-15 北京理工大学 High-resolution multidimensional synergistic insect is migrated Radar Measurement Instrument
CN106022340A (en) * 2016-05-17 2016-10-12 南京理工大学 Improved Gaussian mixed potential probability hypothesis density filtering method
CN106940734A (en) * 2017-04-24 2017-07-11 南京信息工程大学 A kind of Migrating Insects monitor recognition methods and device in the air

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
Development of vertical-looking radar technology for monitoring insect migration;Jintao Zheng,等;《Computers and Electronics in Agriculture》;20020831;第35卷(第2-3期);第95-110页 *
Estimation of unbiased insect densities and density profiles with vertically pointing entomological radars;Drake, V.A.等;《International Journal of Remote Sensing》;20140731;第35卷;第4630-4654页 *
Identification of Migratory Insects from their Physical Features using a Decision-Tree Support Vector Machine and its Application to Radar Entomology;Cheng Hu等;《Scientific Reports》;20180403;第1-11页 *
一种生命探测雷达微多普勒测量灵敏度分析新方法;胡程等;《雷达学报》;20161005;第5卷(第5期);第455-461页 *

Also Published As

Publication number Publication date
CN109034356A (en) 2018-12-18

Similar Documents

Publication Publication Date Title
CN110161474B (en) Low-slow small target detection method based on digital beam forming technology
US6750805B1 (en) Full polarization synthetic aperture radar automatic target detection algorithm
CN110426690B (en) Automatic calibration method for airborne weather radar beam pointing
CN107561518B (en) Three-coordinate radar point trace condensation method based on two-dimensional sliding window local extreme value
US6756935B1 (en) Full polarization ground moving target indicator radar automatic target detection algorithm
EP1980873B1 (en) History or image based methods for altitude determination in a radar altimeter
US7646329B2 (en) Method for detecting a target
CN108957419B (en) Asynchronous interference suppression method based on notch filtering processing
CN110988839B (en) Quick identification method for wall health condition based on one-dimensional convolutional neural network
Long et al. Entomological radar overview: System and signal processing
CN112051568B (en) Pitching angle measurement method of two-coordinate radar
CN109212500A (en) A kind of miscellaneous covariance matrix high-precision estimation method of making an uproar of KA-STAP based on sparse reconstruct
CN106383339B (en) A kind of mirror target suppressing method of multi-site radar signal grade joint-detection
CN108318865B (en) Multichannel SAR deception jamming identification and self-adaptive suppression method
CN110632569B (en) Radar fixed-wing unmanned aerial vehicle using dual-polarization characteristics and clutter recognition method
CN110146873B (en) Target position and speed estimation method of distributed non-coherent radar
CN113504525B (en) Fog region visibility inversion method and system
CN109034356B (en) Insect density statistical method based on nearest neighbor method correlation and Gaussian beam volume
CN112327265A (en) Division and treatment detection method based on semantic segmentation network
KR101714198B1 (en) Target detection method and apparatus using radar
CN114966579A (en) Method and device for acquiring calibration parameters of radar system
CN113009473B (en) Multi-beam staring radar low elevation target height measurement method, device and medium
CN109581366B (en) Discrete sidelobe clutter identification method based on target steering vector mismatch
Wilson Probability of detecting aircraft targets
CN115546526B (en) Three-dimensional point cloud clustering method, device and storage 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