CN111198352B - Edge extraction method for target detection radar imaging - Google Patents

Edge extraction method for target detection radar imaging Download PDF

Info

Publication number
CN111198352B
CN111198352B CN201911406672.1A CN201911406672A CN111198352B CN 111198352 B CN111198352 B CN 111198352B CN 201911406672 A CN201911406672 A CN 201911406672A CN 111198352 B CN111198352 B CN 111198352B
Authority
CN
China
Prior art keywords
edge
detected
point
radar
area
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
CN201911406672.1A
Other languages
Chinese (zh)
Other versions
CN111198352A (en
Inventor
王宏宇
肖楠
江志远
彭璐
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.)
Suzhou Science And Technology Leike Sensing Technology Co ltd
Original Assignee
Suzhou Science And Technology Leike Sensing Technology Co ltd
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 Suzhou Science And Technology Leike Sensing Technology Co ltd filed Critical Suzhou Science And Technology Leike Sensing Technology Co ltd
Priority to CN201911406672.1A priority Critical patent/CN111198352B/en
Publication of CN111198352A publication Critical patent/CN111198352A/en
Application granted granted Critical
Publication of CN111198352B publication Critical patent/CN111198352B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
    • G01S7/021—Auxiliary means for detecting or identifying radar signals or the like, e.g. radar jamming signals
    • 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
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/06—Systems determining position data of a target
    • G01S13/42—Simultaneous measurement of distance and other co-ordinates
    • 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
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
    • G01S7/40—Means for monitoring or calibrating

Abstract

The invention provides an edge extraction method for target detection radar imaging, which can improve the edge extraction efficiency without performing a large amount of measurement and calibration work in advance. After the erection position of a millimeter wave detection radar in a region to be detected is fixed, the radar is started to perform reconnaissance scanning, a scene one-dimensional distance image is fed back and scanned, the one-dimensional distance image is subjected to processing such as a universal gravitation edge detection algorithm, a Gaussian low-pass filtering denoising algorithm, a fitting reconstruction method and the like, finally, edge information of the region to be detected is extracted, the edge information of the region to be detected is stored in a FLASH memory of a radar processor, target detection calculation is only performed in the region to be detected after each scanning of the radar, and the problem that the efficiency of an edge extraction method is low is solved.

Description

Edge extraction method for target detection radar imaging
Technical Field
The invention belongs to the technical field of edge extraction, and particularly relates to an edge extraction method for target detection radar imaging.
Background
Currently, a millimeter wave target detection radar for an area to be detected reports distance and orientation information of all detected targets in a scanning detection range to main control software, the main control software converts the target distance and orientation information reported by the radar into coordinate information of the area to be detected according to previous measurement and calibration, filters out detection results outside the area to be detected, displays the detection results of the targets in the edge of the area to be detected on an interface and gives an alarm.
In the prior art, a target detection radar of a millimeter wave to-be-detected area performs target detection in all scanning ranges, and foreign matters outside the to-be-detected area do not need to be detected, so that a large amount of unnecessary calculation is caused, and the detection efficiency is low; in addition, in order to convert the target distance and the target direction information reported by the radar into the coordinate information of the area to be detected and screen out the target in the area to be detected, the main control software needs to do a large amount of measurement and calibration work in advance, and the efficiency is low.
Disclosure of Invention
In view of this, the invention provides an edge extraction method for target detection radar imaging, which can improve the edge extraction efficiency without performing a large amount of measurement and calibration work in advance.
In order to achieve the above object, the edge extraction method for target detection radar imaging of the present invention includes the following steps:
step 1, performing field-prospecting scanning on a target in an area to be detected by using a millimeter wave detection radar to obtain a scene one-dimensional range profile;
step 2, processing the scene one-dimensional distance image by adopting a universal gravitation edge detection algorithm, which comprises the following steps:
step 21, calculating the horizontal mass M (i, j) of each power point P (i, j) in the one-dimensional distance imagexAnd vertical direction mass M (i, j)yWhere P (i, j) represents the power value at the point in the ith row and jth column in the one-dimensional range profile, M (i, j)xAnd M (i, j)yThe expression of (a) is as follows:
Figure BDA0002348826570000021
Figure BDA0002348826570000022
in the formula, T (i, j)xAnd T (i, j)yRespectively representing a horizontal direction gradient and a vertical direction gradient, and the specific expression is as follows:
T(i,j)xf (P (i, j) -P (i-1, j)) -F (-P (i +1, j) -P (i, j)) (formula 2-1)
T(i,j)yF (P (i, j) -P (i, j-1)) -F (-P (i, j +1) -P (i, j)) (formula 2-2)
F in the formula is a non-linear operator,
Figure BDA0002348826570000023
step 22, calculating the horizontal component force of the power point P (i, j) at the point P (m, N) in the NxN neighborhood
Figure BDA0002348826570000024
And component force in the vertical direction
Figure BDA0002348826570000025
Figure BDA0002348826570000026
Figure BDA0002348826570000027
Wherein i-N is not less than m not more than i + N, j-N is not less than N not more than j + N, and r represents the distance between the point P (i, j) and the point P (m, N);
step 23, calculating the resultant force of horizontal universal gravitation of point P (m, n) to point P (i, j) in the neighborhood
Figure BDA0002348826570000028
Resultant force of vertical universal gravitation
Figure BDA0002348826570000029
Figure BDA0002348826570000031
Figure BDA0002348826570000032
Step 24, calculating the resultant force of the universal gravitation borne by the point P (i, j):
Figure BDA0002348826570000033
when the resultant force is less than or equal to a set threshold value, judging the point as an edge point, otherwise, not judging the point as the edge point;
step 3, denoising the universal gravitation edge detection result to obtain a denoised result;
and 4, fitting and interpolating the denoised result, removing the fuzzy phenomenon, ensuring that the region to be detected is closed, longitudinally intercepting according to the radar detection distance requirement, and extracting the final edge of the region to be detected.
And 5, storing the edge information of the area to be detected to the millimeter wave target detection radar of the area to be detected, calling the edge information of the area to be detected and detecting only the area in the area to be detected during each subsequent detection and scanning.
And sending the edge information of the area to be detected to a millimeter wave area target detection radar to be detected through upper computer software, and storing the information in a FLASH memory by the radar.
In the step 3, denoising processing is performed on the universal gravitation edge detection result by adopting a Gaussian low-pass filtering denoising algorithm.
The transfer function of the gaussian low-pass filter comprises a one-dimensional gaussian function and a two-dimensional gaussian function, and the expressions are respectively as follows:
Figure BDA0002348826570000034
Figure BDA0002348826570000035
wherein δ is the standard deviation.
Wherein the edges are edges of railways, rivers, highways, and airport runways.
Has the advantages that:
according to the method, after the erection position of a millimeter wave detection radar in a region to be detected is fixed, the radar is started to perform field surveying scanning, a one-dimensional range image of a scanned scene is returned, the one-dimensional range image is processed by a universal gravitation edge detection algorithm, a Gaussian low-pass filtering denoising algorithm, a fitting reconstruction method and the like, finally edge information of the region to be detected is extracted, the edge information of the region to be detected is stored in a FLASH memory of a radar processor, target detection calculation is only performed in the region to be detected after each scanning of the radar, and the problem that the efficiency of an edge extraction method is low is solved.
Drawings
FIG. 1 is a one-dimensional range profile of a scene according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a universal gravitation edge detection result according to an embodiment of the invention.
FIG. 3 is a schematic diagram of a Gaussian low-pass filtering denoising result according to an embodiment of the present invention.
Fig. 4 is a schematic diagram of a road surface edge extraction result according to an embodiment of the present invention.
Fig. 4(a) is a schematic diagram of an individual road surface edge, and fig. 4(b) is a schematic diagram of a road surface edge in the whole scene.
FIG. 5 is a flow chart of the present invention.
Detailed Description
The invention is described in detail below by way of example with reference to the accompanying drawings.
The edge extraction method for target detection radar imaging, as shown in fig. 5, includes the following steps:
step 1, performing reconnaissance scanning on a target in an area to be detected by using a millimeter wave detection radar, specifically as follows:
after the radar is erected and fixed, the scanning scene is also fixed, and at the moment, the reconnaissance scanning is carried out, namely, the whole scene is scanned, and the scene one-dimensional distance image is transmitted back to the upper computer software for the extraction processing of the edge of the area to be detected. The one-dimensional range image of the scene of the present embodiment is shown in fig. 1, in which the x-axis represents the azimuth angle (unit: °), the y-axis represents the distance (unit: m), and the z-axis represents the scene echo power (unit: dBm).
Step 2, processing the scene one-dimensional distance image by adopting a universal gravitation edge detection algorithm, which comprises the following steps:
step 21, calculating the horizontal mass M (i, j) of each power point P (i, j) in the one-dimensional distance imagexAnd vertical direction mass M (i, j)yWhere P (i, j) represents the power value at the point in the ith row and jth column in the one-dimensional range profile, M (i, j)xAnd M (i, j)yThe expression of (a) is as follows:
Figure BDA0002348826570000051
Figure BDA0002348826570000052
in the formula, T (i, j)xAnd T (i, j)yRespectively representing a horizontal direction gradient and a vertical direction gradient, and the specific expression is as follows:
T(i,j)xf (P (i, j) -P (i-1, j)) -F (-P (i +1, j) -P (i, j)) (formula 2-1)
T(i,j)yF (P (i, j) -P (i, j-1)) -F (-P (i, j +1) -P (i, j)) (formula 2-2)
F in the formula is a nonlinear operator, and the specific expression is as follows:
Figure BDA0002348826570000053
step 22, calculating the horizontal component force of the inner point P (m, n) in the 3 × 3 neighborhood (not limited to) of the power point P (i, j)
Figure BDA0002348826570000054
And component force in the vertical direction
Figure BDA0002348826570000055
i-3. ltoreq. m.ltoreq.i +3, j-3. ltoreq. n.ltoreq.j +3, wherein r represents the distance between the point P (i, j) and the point P (m, n):
Figure BDA0002348826570000056
Figure BDA0002348826570000057
step 23, calculating the resultant force of horizontal universal gravitation of point P (m, n) to point P (i, j) in the neighborhood
Figure BDA0002348826570000061
Resultant force of vertical universal gravitation
Figure BDA0002348826570000062
The specific expression is as follows:
Figure BDA0002348826570000063
Figure BDA0002348826570000064
step 24, calculating the resultant force of the universal gravitation borne by the point P (i, j), wherein the specific expression is as follows:
Figure BDA0002348826570000065
when the resultant force is less than or equal to a set threshold (empirical value), the point is determined to be an edge point, otherwise, the point is not an edge point. The universal gravitation edge detection result of the embodiment is shown in fig. 2.
And 3, carrying out Gaussian low-pass filtering denoising algorithm on the universal gravitation edge detection result to carry out denoising treatment.
Gaussian Low Pass filters (Gaussian Low Pass filters) are a type of linear smoothing Filter whose transfer function is a Gaussian function. The expressions of the one-dimensional gaussian function and the two-dimensional gaussian function (transfer function of gaussian low-pass filter) are respectively as follows:
Figure BDA0002348826570000066
Figure BDA0002348826570000067
in the formula, δ is a standard deviation, and the larger the standard deviation of the gaussian function is, the smoother the gaussian curve is, the stronger the denoising capability is, but the more blurred the image is. After the universal gravitation edge detection result is processed by the gaussian low-pass filter, the denoising result is shown in fig. 3.
And 4, fitting, interpolating and intercepting the denoised result to obtain the final edge of the area to be detected. As shown in fig. 4, fig. 4(a) is a schematic diagram of an individual road surface edge, and fig. 4(b) is a schematic diagram of a road surface edge in the whole scene.
And (3) carrying out fitting and interpolation by using simulation software to remove the fuzzy phenomenon, ensuring that the region to be detected is closed (each unit angle at an endpoint corresponds to one distance value, and each unit angle at other points corresponds to two distance values), and then carrying out longitudinal interception according to the radar detection distance requirement, thus extracting the final region edge to be detected.
And 5, storing the edge information of the area to be detected to the millimeter wave target detection radar of the area to be detected.
In the embodiment, the upper computer software sends the edge information of the area to be detected to the millimeter wave area target detection radar to be detected, and the radar stores the information in the FLASH memory to ensure that the power failure is not lost. And during each subsequent detection scanning, the edge information of the area to be detected is called, only the area in the area to be detected is detected, so that the calculated amount is reduced, the calculation efficiency is improved, and good detection can be realized for small targets (less than 5 cm).
The method is suitable for edge extraction of railways, rivers, highways, airport runways and the like
Taking an airport pavement as an example, the method has the following significant advantages compared with the existing airport pavement target detection edge extraction method:
1. the work of large-scale coordinate measurement on the airport pavement is omitted.
The existing airport pavement foreign object detection needs to carry out coordinate measurement on all pavement detection areas in advance so as to screen out foreign object information in the pavement; by the edge extraction method, a large amount of measurement is not needed, and the method can greatly reduce the workload and reduce the cost considering that the measurement can only be carried out when the airport stops navigating.
2. The detection calculation amount of the radar for detecting the pavement target is obviously reduced.
The detection of foreign objects on the existing airport pavement needs radar to detect all targets in a detection range, and subsequent main control can filter the information of the target detection result outside the pavement; by the edge extraction method, the detection range is reduced before the radar starts to perform detection calculation, the calculation amount is obviously reduced, and the calculation efficiency is improved.
3. The early warning efficiency of the main control software is obviously improved.
After the foreign objects on the prior airport pavement detect all target information in the detection range reported by a radar, the target information needs to be judged and calculated, and targets in the pavement are screened out; by the edge extraction method, screening calculation is not required by master control software, target information reported by the radar is directly displayed and early-warned, and early-warning efficiency is improved.
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 (6)

1. An edge extraction method for target detection radar imaging is characterized by comprising the following steps:
step 1, performing field-prospecting scanning on a target in an area to be detected by using a millimeter wave detection radar to obtain a scene one-dimensional range profile;
step 2, processing the scene one-dimensional distance image by adopting a universal gravitation edge detection algorithm, which comprises the following steps:
step 21, calculating the horizontal mass M (i, j) of each power point P (i, j) in the one-dimensional distance imagexAnd vertical direction mass M (i, j)yWhere P (i, j) represents the power value at the point in the ith row and jth column in the one-dimensional range profile, M (i, j)xAnd M (i, j)yThe expression of (a) is as follows:
Figure RE-FDA0002452528940000011
Figure RE-FDA0002452528940000012
in the formula, T (i, j)xAnd T (i, j)yRespectively representing a horizontal direction gradient and a vertical direction gradient, and the specific expression is as follows:
T(i,j)xf (P (i, j) -P (i-1, j)) -F (-P (i +1, j) -P (i, j)) (formula 2-1)
T(i,j)yF (P (i, j) -P (i, j-1)) -F (-P (i, j +1) -P (i, j)) (formula 2-2)
F in the formula is a non-linear operator,
Figure RE-FDA0002452528940000013
step 22, calculating the horizontal component force of the power point P (i, j) at the point P (m, N) in the NxN neighborhood
Figure RE-FDA0002452528940000014
And component force in the vertical direction
Figure RE-FDA0002452528940000015
Figure RE-FDA0002452528940000016
Figure RE-FDA0002452528940000017
Wherein i-N is not less than m not more than i + N, j-N is not less than N not more than j + N, and r represents the distance between the point P (i, j) and the point P (m, N);
step 23, calculating the resultant force of horizontal universal gravitation of point P (m, n) to point P (i, j) in the neighborhood
Figure RE-FDA0002452528940000018
Resultant force of vertical universal gravitation
Figure RE-FDA0002452528940000021
Figure RE-FDA0002452528940000022
Figure RE-FDA0002452528940000023
Step 24, calculating the resultant force of the universal gravitation borne by the point P (i, j):
Figure RE-FDA0002452528940000024
when the resultant force is less than or equal to a set threshold value, judging the point as an edge point, otherwise, not judging the point as the edge point;
step 3, denoising the universal gravitation edge detection result to obtain a denoised result;
and 4, fitting and interpolating the denoised result, removing the fuzzy phenomenon, ensuring that the region to be detected is closed, longitudinally intercepting according to the radar detection distance requirement, and extracting the final edge of the region to be detected.
2. The edge extraction method for target detection radar imaging according to claim 1, further comprising a step 5 of storing edge information of the area to be detected in the millimeter wave target detection radar of the area to be detected, and calling the edge information of the area to be detected each time of subsequent detection scanning, so as to detect only the area in the area to be detected.
3. The edge extraction method for target detection radar imaging according to claim 1 or 2, characterized in that the upper computer software issues the edge information of the area to be detected to the millimeter wave target detection radar of the area to be detected, and the radar stores the information in the FLASH memory.
4. The method for extracting edges of radar imaging for object detection according to claim 1, wherein in the step 3, denoising processing is performed on the detection result of the gravitation edges by using a gaussian low-pass filtering denoising algorithm.
5. The method of edge extraction for target detection radar imaging as claimed in claim 4, wherein the transfer function of the Gaussian low pass filter includes a one-dimensional Gaussian function and a two-dimensional Gaussian function, and the expressions are respectively as follows:
Figure RE-FDA0002452528940000031
Figure RE-FDA0002452528940000032
wherein δ is the standard deviation.
6. The method of object detection radar imaged edge extraction of claim 1, wherein the edge is an edge of a railway, river, highway, and airport runway.
CN201911406672.1A 2019-12-31 2019-12-31 Edge extraction method for target detection radar imaging Active CN111198352B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911406672.1A CN111198352B (en) 2019-12-31 2019-12-31 Edge extraction method for target detection radar imaging

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911406672.1A CN111198352B (en) 2019-12-31 2019-12-31 Edge extraction method for target detection radar imaging

Publications (2)

Publication Number Publication Date
CN111198352A CN111198352A (en) 2020-05-26
CN111198352B true CN111198352B (en) 2022-04-05

Family

ID=70746632

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911406672.1A Active CN111198352B (en) 2019-12-31 2019-12-31 Edge extraction method for target detection radar imaging

Country Status (1)

Country Link
CN (1) CN111198352B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101256625A (en) * 2007-12-19 2008-09-03 重庆大学 Method for extracting human ear image edge combining multiple methods
KR101341288B1 (en) * 2013-10-24 2013-12-12 사회복지법인 삼성생명공익재단 Quality assurance system and the method for radiotherapy apparatus
CN108828549A (en) * 2018-07-31 2018-11-16 安徽四创电子股份有限公司 A kind of target extraction method based on airport surface detection radar system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101256625A (en) * 2007-12-19 2008-09-03 重庆大学 Method for extracting human ear image edge combining multiple methods
KR101341288B1 (en) * 2013-10-24 2013-12-12 사회복지법인 삼성생명공익재단 Quality assurance system and the method for radiotherapy apparatus
CN108828549A (en) * 2018-07-31 2018-11-16 安徽四创电子股份有限公司 A kind of target extraction method based on airport surface detection radar system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
不确定性边缘表示与提取的认知物理学方法;吴涛等;《物理学报》;20130331;第064201-1-064201-13页 *
基于光谱空间数据场模型的遥感影像边缘提取;刘楠;《中国优秀博硕士学位论文全文数据库（硕士）基础科学辑》;20060515;第23-71页 *

Also Published As

Publication number Publication date
CN111198352A (en) 2020-05-26

Similar Documents

Publication Publication Date Title
Wu et al. Road pothole extraction and safety evaluation by integration of point cloud and images derived from mobile mapping sensors
Kang et al. Pothole detection system using 2D LiDAR and camera
Lei et al. A novel tunnel-lining crack recognition system based on digital image technology
CN107154040B (en) Tunnel lining surface image crack detection method
CN107590470B (en) Lane line detection method and device
CN103714538B (en) road edge detection method, device and vehicle
CN107942329B (en) Method for detecting sea surface ship target by maneuvering platform single-channel SAR
CN104408932A (en) Drunk driving vehicle detection system based on video monitoring
CN104574393A (en) Three-dimensional pavement crack image generation system and method
CN103679167A (en) Method for processing CCD images
CN107392139A (en) A kind of method for detecting lane lines and terminal device based on Hough transformation
CN202134079U (en) Unmanned vehicle lane marker line identification and alarm device
CN105574894A (en) Method and system for screening moving object feature point tracking results
KR101207220B1 (en) Accurate extraction method of boundary line in high-resolution sar amplitude images
CN103914829B (en) Method for detecting edge of noisy image
CN102663385A (en) Detection method for spot target on satellite
CN109541601A (en) Differentiating obstacle and its detection method based on millimeter wave
CN106803262A (en) The method that car speed is independently resolved using binocular vision
CN104330048A (en) Image-based railway snow depth measurement device and method
Kumar et al. Vehicle speed detection using corner detection
CN103455985A (en) Road crack enhancement method based on Hessian structural analysis
CN110992398A (en) Image fusion tracking and anti-interference method for laser radar and infrared detection system
Liu et al. The target detection for GPR images based on curve fitting
CN111198352B (en) Edge extraction method for target detection radar imaging
Gooda et al. Automatic detection of road cracks using EfficientNet with residual U-net-based segmentation and YOLOv5-based detection

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