CN111798411B - Concrete inner steel bar intelligent positioning method based on ground penetrating radar and deep learning - Google Patents
Concrete inner steel bar intelligent positioning method based on ground penetrating radar and deep learning Download PDFInfo
- Publication number
- CN111798411B CN111798411B CN202010493228.4A CN202010493228A CN111798411B CN 111798411 B CN111798411 B CN 111798411B CN 202010493228 A CN202010493228 A CN 202010493228A CN 111798411 B CN111798411 B CN 111798411B
- Authority
- CN
- China
- Prior art keywords
- image
- steel bar
- ground penetrating
- target
- penetrating radar
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- 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
- G01S13/885—Radar or analogous systems specially adapted for specific applications for ground probing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10044—Radar image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30132—Masonry; Concrete
-
- 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
- Y02A90/30—Assessment of water resources
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Electromagnetism (AREA)
- Quality & Reliability (AREA)
- Computer Networks & Wireless Communication (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses an intelligent positioning method for a concrete inner steel bar based on ground penetrating radar and deep learning, which comprises the following steps: collecting radar sections containing steel bar hyperbola reflecting targets on the surface of the concrete by using a ground penetrating radar, and carrying out zero-time correction and format conversion pretreatment on the data; inputting the preprocessed image into the trained SSD model for intelligent recognition; extracting all hyperbolic target areas automatically selected by the rectangular frame according to the identification result; performing Hilbert transform and diffraction superposition offset on the target frame region; determining an optimal threshold value of image binarization by using an iterative threshold value method; determining the vertex of the target area block in the binary image by using a vertex positioning method; and calculating the thickness of the protective layer of the steel bar and the horizontal distance of the steel bar according to the vertex of the steel bar. The invention can rapidly acquire the position and quantity information of the steel bars in the concrete by intelligently detecting and positioning the steel bars in the concrete by combining the ground penetrating radar and the deep learning technology.
Description
Technical Field
The invention belongs to the field of engineering nondestructive testing, and particularly relates to an intelligent positioning method for concrete inner steel bars based on ground penetrating radar and deep learning.
Background
In the completion acceptance of buildings and the safety evaluation of old construction, steel bars in reinforced concrete structures are detected according to the standard requirements, wherein the positions, the spacing, the diameters, the thickness of protective layers and the like of the steel bars are important points for steel bar detection. The traditional method is to inspect the condition of the steel bars in the concrete in a destructive manner such as cutting, drilling and the like on the surface of the concrete, and the method is time-consuming and labor-consuming, and also damages the integrity of the building components, and influences the mechanical property and the bearing capacity of the building components.
The ground penetrating radar (Ground Penetrating Radar, GPR for short) is an engineering geophysical prospecting method for transmitting and receiving ultra-wideband high-frequency pulse electromagnetic waves, has the advantages of convenience, no damage, high detection efficiency, high precision and the like, is widely applied to the fields of engineering detection and the like at present, and is also a main means for detecting the reinforcing steel bars in concrete. However, field detection of ground penetrating radar can generate a large amount of data. Interpretation of ground penetrating radar data is required to be completed by a detector with expert knowledge, which is not only labor-intensive, but also often causes the detection accuracy to be reduced by human factors. Therefore, in order to improve the working efficiency of the ground penetrating radar for detecting the steel bars in the concrete and reduce the dependence of the detection result on the expertise and experience of operators, an intelligent algorithm capable of automatically identifying radar profile steel bar targets and accurately determining the positions of the steel bar targets and the thickness of a protective layer needs to be developed.
Disclosure of Invention
The invention mainly aims to overcome the defects and shortcomings of the prior art and provides an intelligent positioning method for concrete inner steel bars based on ground penetrating radar and deep learning.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
an intelligent positioning method for a concrete inner steel bar based on ground penetrating radar and deep learning comprises the following steps:
collecting radar sections containing steel bar hyperbola reflecting targets on the surface of the concrete by using a ground penetrating radar, and carrying out zero-time correction and format conversion pretreatment on the data;
inputting the preprocessed image into a trained SSD model to identify the hyperbolic reflection of the steel bar in the radar section, and automatically selecting the area of the radar section by using a rectangular frame;
performing Hilbert transformation and offset processing on the target frame area, and reconstructing a hyperbolic target in the image into an approximate circular target for truly describing the cross-sectional shape of the steel bar;
determining an optimal threshold value of image binarization by using an iterative threshold value method, and further processing the target area image into a binary image;
determining the vertex of the target area block in the binary image by using a vertex positioning method, wherein the vertex position is the true vertex position of the steel bar;
and calculating the thickness of the protective layer of the steel bar and the horizontal distance of the steel bar according to the vertex of the steel bar.
As an optimal technical scheme, the SSD model is an SSD model capable of accurately identifying a reinforcement hyperbola target, and the output result of the SSD model is that a hyperbola in an image is automatically selected by a rectangular frame.
As an optimal technical scheme, the Hilbert transform is to convert radar signals into analysis signals and then take envelopes to enhance the contrast of radar images, so that the recognition of targets is facilitated.
As an preferable technical scheme, the offset processing adopts an offset algorithm, the offset algorithm is a diffraction superposition offset algorithm, and the calculation formula is as follows:
wherein E is in The amplitude value of the gray level map of the ground penetrating radar before offset is shown as E, the amplitude value of the image after offset is shown as t i Is the sampling time of the ith reflected signal, z represents the vertical position of the pixel in the image, x i For the horizontal position of the ground penetrating radar transmitting antenna, x 0 For the distance between the transmitting and receiving antennas, x represents the horizontal position corresponding to the imaging point, and v is the propagation speed of electromagnetic waves in the concrete.
As a preferable technical scheme, the method for determining the optimal threshold value of the image binarization by using the iterative threshold value method specifically comprises the following steps:
for a frame selection radar image E containing a steel bar target, firstly, utilizing the maximum amplitude E of pixels in the E max And minimum amplitude E min An initial threshold t=0.6× (E max +E min ) And dividing the image into two regions P and Q, wherein the pixel value E of the P region i Pixel value E of < T, Q area i ＞T；
Calculating the average gray value of two areas
And->
Updating threshold values
Calculate Δ= |t-T h I, when delta < 0.001, then T h Is the optimal threshold value; let t=t no h Then, the image is segmented again, and iterative circulation is carried out;
after the binary processing is completed, white area blocks in the image represent targets, and the remaining black parts are the backgrounds.
As a preferable technical scheme, the vertex positioning method is to position the vertex of a target white area block in a binary image, the origin position of the image is at the upper left corner, the abscissa axis is positive to the right, the ordinate axis is positive downwards, firstly, the barycenter of the target area block is determined, and the coordinates (X c ，Y c ) Is calculated as follows:
where m is the total number of pixels in the region, x j Is the abscissa of pixel j, y j Is the ordinate of pixel j; then utilize the minimum ordinate y of the region block min And a maximum ordinate y max Calculating the height h=y of the region block max -y min The method comprises the steps of carrying out a first treatment on the surface of the Then through the region blockCentroid upward translation
The vertex position, vertex coordinates (X apex ，Y apex ) The following is indicated:
as a preferable technical scheme, after determining the fixed point position of the reinforcing steel bar, when returning to the initial image of the ground penetrating radar, position calculation is carried out by combining the coordinates of the rectangular frame and the top point of the target, wherein the upper left corner of the rectangular frame is the minimum coordinate value point (X min ,Y min ) The method comprises the steps of carrying out a first treatment on the surface of the In the ground penetrating radar image, the horizontal position X of the target hyperbola i i And vertical position Y i The calculation is as follows:
X i ＝X min +X apex ,Y i ＝Y min +Y apex
horizontal spacing of steel bars in concrete passes through horizontal position X of target i Calculating, i.e. the horizontal spacing S of two adjacent reinforcing bars A and B AB Is that
S AB ＝|X A -X B |×Δs
Wherein deltas is the horizontal distance between each signal when the ground penetrating radar collects data. The thickness of the protective layer of the reinforcing steel bar passes through the vertical position Y of the target i And (3) calculating:
Z＝Y i ×v×Δt
wherein deltat is the time interval of adjacent sampling points set during data acquisition of the ground penetrating radar, and v is the propagation speed of electromagnetic waves in concrete.
Compared with the prior art, the invention has the following advantages and beneficial effects:
according to the invention, intelligent detection and positioning are carried out on the steel bars in the concrete by combining the ground penetrating radar and the deep learning technology, so that the position and quantity information of the steel bars in the concrete can be rapidly obtained, the construction quality of the reinforced concrete structure can be timely evaluated on site by combining the design and specification requirements, the detection efficiency is improved, the dependence of data interpretation on the professional knowledge and engineering experience of operators is reduced, the detection period can be shortened, the detection precision is improved, and important economic and social values are realized.
Drawings
Fig. 1 is a flowchart of a method for intelligent positioning of a steel bar based on ground penetrating radar and deep learning in an embodiment of the invention;
FIG. 2 is a graph of the results of hyperbolas in a ground penetrating radar image selected by rectangular boxes;
FIG. 3 is a flow chart of an iterative thresholding method;
FIG. 4 is a schematic diagram of white area block vertex positioning for a binary image;
fig. 5 is an effect diagram of the positioning of the reinforcing steel bar vertex in the embodiment of the invention.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but embodiments of the present invention are not limited thereto.
Examples
The invention discloses an intelligent positioning method for a concrete inner steel bar based on ground penetrating radar and deep learning, which is shown in fig. 1 and comprises the following steps:
s1, acquiring a radar section containing a steel bar hyperbola reflecting target on the surface of concrete by using a ground penetrating radar, and performing zero-time correction and format conversion pretreatment on data;
s2, inputting the preprocessed image into a trained SSD model to identify the hyperbolic reflection of the steel bar in the radar section, and automatically selecting the area of the radar section by using a rectangular frame;
s3, hilbert transformation and offset processing are carried out on the target frame area, and a hyperbolic target in the image is reconstructed into an approximate circular target for truly describing the cross-sectional shape of the reinforcing steel bar;
s4, determining an optimal threshold value of image binarization by using an iterative threshold value method, and further processing the target area image into a binary image;
s5, determining the vertex of the target area block in the binary image by using a vertex positioning method, wherein the vertex position is the true vertex position of the steel bar;
and S6, calculating the thickness of the protective layer of the reinforcing steel bar and the horizontal distance of the reinforcing steel bar according to the vertex of the reinforcing steel bar.
Compared with the prior art, the intelligent detection and positioning method for the steel bars in the concrete based on the ground penetrating radar and the deep learning can rapidly acquire the position and quantity information of the steel bars in the concrete, timely evaluate the construction quality of the reinforced concrete structure on site by combining design and specification requirements, improve the detection efficiency, reduce the dependence of data interpretation on professional knowledge and engineering experience of operators, shorten the detection period and improve the detection precision, and have important economic and social values.
Specifically, in step S1, a radar section including a steel bar hyperbola reflecting target is acquired on the concrete surface by using a ground penetrating radar, and zero-time correction and format conversion preprocessing are performed on the data;
the time corresponding to the first positive peak of the direct wave signal in the data is set as the zero time because the zero time is not the start time corresponding to the ground surface due to the influence of factors such as the delay of the transmission line.
In step S2, inputting the preprocessed image into a trained SSD model to identify the hyperbolic reflection of the steel bars in the radar section, and automatically selecting the area of the hyperbolic reflection by using a rectangular frame;
and inputting the preprocessed data image into an SSD model with the capability of identifying the hyperbola of the steel bar, selecting the hyperbola in the image by using a rectangular frame (shown in fig. 2) according to the output result of the model, and adding the category and the confidence value of the target. And extracting the target area according to the coordinate information of the rectangular frame, and further analyzing.
In step S3, hilbert transform and offset processing are carried out on the target frame area, and a hyperbolic target in the image is reconstructed into an approximate circular target for truly describing the cross-sectional shape of the reinforcing steel bar;
after the target frame region is extracted, hilbert transformation and diffraction superposition offset processing are required to be performed on the image in order to obtain the thickness value of the protective layer of the reinforcing steel bar. The Hilbert transform is used for converting radar signals into analysis signals and then taking envelopes to enhance the contrast of radar images, so that the identification of targets is facilitated. The diffraction superposition offset is used for superposing pixel amplitude values of a region representing the characteristics of the steel bar target in the image at the position of the steel bar target in an offset mode, so that the near-circular shape of the steel bar target is restored on the image, and the calculation formula is as follows:
wherein E is in The amplitude value of the gray level map of the ground penetrating radar before offset is shown as E, the amplitude value of the image after offset is shown as t i Is the sampling time of the ith reflected signal, z represents the vertical position of the pixel in the image, x i For the horizontal position of the ground penetrating radar transmitting antenna, x 0 For the distance between the transmitting and receiving antennas, x represents the horizontal position corresponding to the imaging point, and v is the propagation speed of electromagnetic waves in the concrete.
In step S4, an optimal threshold value of image binarization is determined by using an iterative threshold value method, and then the target area image is processed into a binary image;
after the offset process, the image is divided into two parts by a binary method, wherein the pixels of the target area are assigned 255 and the rest are assigned 0, so that the target is distinguished from the background. As shown in fig. 3, the threshold of the binary method is calculated by an iterative method, and the steps of obtaining the iterative threshold can be summarized as follows:
a. for a frame selection radar image E containing a steel bar target, firstly, utilizing the maximum amplitude E of pixels in the E max And minimum amplitude E min An initial threshold t=0.6× (E max +E min ) And dividing the image into two regions P and Q, wherein the pixel value E of the P region i Pixel value E of < T, Q area i ＞T；
b. Calculating the average gray value of two areas
And->
c. Updating threshold values
d. Calculate Δ= |t-T h I, when delta < 0.001, then T h Is the optimal threshold value; let t=t no h Then, the image is segmented again, and an iterative loop is performed.
e. After the binary processing is completed, white area blocks in the image represent targets, and the remaining black parts are the backgrounds.
In step S5, using a vertex positioning method to determine the vertex of the target area block in the binary image, wherein the vertex position is the true vertex position of the steel bar;
the vertex positioning method is to position the vertex of the target white area block in the binary image (the origin position of the image is at the upper left corner, the abscissa axis is positive to the right, and the ordinate axis is positive to the down), firstly determine the mass center of the target area block, and the coordinates (X c ，Y c ) Is calculated as follows:
where m is the total number of pixels in the region, x j Is the abscissa of pixel j, y j Is the ordinate of pixel j; as shown in fig. 4, the minimum ordinate y of the region block is then utilized min And a maximum ordinate y max Calculating the height h=y of the region block max -y min The method comprises the steps of carrying out a first treatment on the surface of the Then by upwards translating the mass center of the regional block
The vertex position, vertex coordinates (X apex ,Y apex ) The following is indicated:
after the regional block vertexes are determined, the vertexes of the reinforcing steel bars are approximately determined.
Further, in the method for positioning the steel bar, after determining the vertex of the steel bar, when returning to the initial image of the ground penetrating radar, position calculation needs to be performed by combining the coordinates of the rectangular frame and the vertex of the target, wherein the upper left corner of the rectangular frame is the minimum coordinate value point (X min ,Y min ) As shown in fig. 5. In the ground penetrating radar image, the horizontal position X of the target hyperbola i i And vertical position Y i The calculation is as follows:
X i ＝X min +X apex ,Y i ＝Y min +Y apex
horizontal spacing of steel bars in concrete passes through horizontal position X of target i Calculating, i.e. the horizontal spacing S of two adjacent reinforcing bars A and B AB Is that
S AB ＝|X A -X B |×Δs
Wherein deltas is the horizontal distance between each signal when the ground penetrating radar collects data. The thickness of the protective layer of the reinforcing steel bar passes through the vertical position Y of the target i And (3) calculating:
Z＝Y i ×v×Δt
wherein deltat is the time interval of adjacent sampling points set during data acquisition of the ground penetrating radar, and v is the propagation speed of electromagnetic waves in concrete.
The above examples are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above examples, and any other changes, modifications, substitutions, combinations, and simplifications that do not depart from the spirit and principle of the present invention should be made in the equivalent manner, and the embodiments are included in the protection scope of the present invention.
Claims (4)
1. The intelligent concrete inner steel bar positioning method based on ground penetrating radar and deep learning is characterized by comprising the following steps of:
collecting radar sections containing steel bar hyperbola reflecting targets on the surface of the concrete by using a ground penetrating radar, and carrying out zero-time correction and format conversion pretreatment on the data;
inputting the preprocessed image into a trained SSD model to identify the hyperbolic reflection of the steel bar in the radar section, and automatically selecting the area of the radar section by using a rectangular frame;
performing Hilbert transformation and offset processing on the target frame area, and reconstructing a hyperbolic target in the image into an approximate circular target for truly describing the cross-sectional shape of the steel bar; the offset processing adopts an offset algorithm, the offset algorithm is a diffraction superposition offset algorithm, and the calculation formula is as follows:
wherein E is in The amplitude value of the gray level map of the ground penetrating radar before offset is shown as E, the amplitude value of the image after offset is shown as t i Is the sampling time of the ith reflected signal, z represents the vertical position of the pixel in the image, x i For the horizontal position of the ground penetrating radar transmitting antenna, x o For the distance between the receiving and transmitting antennas, x represents the horizontal position corresponding to the imaging point, and v is the propagation speed of electromagnetic waves in the concrete;
determining an optimal threshold value of image binarization by using an iterative threshold value method, and further processing the target area image into a binary image;
determining the vertex of the target area block in the binary image by using a vertex positioning method, wherein the vertex position is the true vertex position of the steel bar; the vertex positioning method is to position the vertex of the target white area block in the binary image, the origin position of the image is at the upper left corner, the abscissa axis is positive to the right, the ordinate axis is positive downwards, firstly, the mass center of the target area block is determined, and the coordinates (X c ,Y c ) Is calculated as follows:
where m is the total number of pixels in the region, x j Is the abscissa of pixel j, y j For the longitudinal seating of pixel jMarking; then utilize the minimum ordinate y of the region block min And a maximum ordinate y max Calculating the height h=y of the region block max -y min The method comprises the steps of carrying out a first treatment on the surface of the Then by upwards translating the mass center of the regional block
The vertex position, vertex coordinates (X apex ,Y apex ) The following is indicated:
X apex ＝X c ，
according to the top of the steel bar, calculating the thickness of the protective layer of the steel bar and the horizontal distance of the steel bar; after the fixed point position of the steel bar is determined, when returning to the initial image of the ground penetrating radar, position calculation is carried out by combining the coordinates of the rectangular frame and the vertex of the target, wherein the upper left corner of the rectangular frame is the minimum coordinate value point (X min ,Y min ) The method comprises the steps of carrying out a first treatment on the surface of the In the ground penetrating radar image, the horizontal position X of the target hyperbola i i And vertical position Y i The calculation is as follows:
X i ＝X min +X apex ,Y i ＝Y min +Y apex
horizontal spacing of steel bars in concrete passes through horizontal position X of target i Calculating, i.e. the horizontal spacing S of two adjacent reinforcing bars A and B AB Is S AB ＝|X A -X B |×Δs
Wherein deltas is the horizontal distance between each signal when the ground penetrating radar collects data, and the thickness of the protective layer of the steel bar passes through the vertical position Y of the target i And (3) calculating:
Z＝Y i ×v×Δt
wherein deltat is the time interval of adjacent sampling points set during data acquisition of the ground penetrating radar, and v is the propagation speed of electromagnetic waves in concrete.
2. The intelligent positioning method for the steel bars in the concrete based on the ground penetrating radar and the deep learning according to claim 1, wherein the SSD model is an SSD model capable of accurately identifying the hyperbolic target of the steel bars, and the output result of the SSD model is that the hyperbolic curve in the image is automatically selected by a rectangular frame.
3. The intelligent concrete reinforcement positioning method based on ground penetrating radar and deep learning according to claim 1, wherein the Hilbert transform is to convert radar signals into analysis signals and then take envelopes to enhance contrast of radar images, so that recognition of targets is facilitated.
4. The intelligent positioning method for the steel bars in the concrete based on the ground penetrating radar and the deep learning according to claim 1, wherein the optimal threshold value for image binarization is determined by using an iterative threshold value method, and the method is specifically as follows:
for a frame selection radar image E containing a steel bar target, firstly, utilizing the maximum amplitude E of pixels in the E max And minimum amplitude E min An initial threshold t=0.6× (E max +E min ) And dividing the image into two regions P and Q, wherein the pixel value E of the P region i <Pixel value E of T, Q region i >T；
Calculating the average gray value of two areas
And->
Updating threshold values
Calculate Δ= |t-T h |, when delta<At 0.001, then T h Is the optimal threshold value; let t=t no h Then, the image is segmented again, and iterative circulation is carried out;
after the binary processing is completed, white area blocks in the image represent targets, and the remaining black parts are the backgrounds.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010493228.4A CN111798411B (en) | 2020-06-03 | 2020-06-03 | Concrete inner steel bar intelligent positioning method based on ground penetrating radar and deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010493228.4A CN111798411B (en) | 2020-06-03 | 2020-06-03 | Concrete inner steel bar intelligent positioning method based on ground penetrating radar and deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111798411A CN111798411A (en) | 2020-10-20 |
| CN111798411B true CN111798411B (en) | 2023-07-07 |
Family
ID=72806197
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010493228.4A Active CN111798411B (en) | 2020-06-03 | 2020-06-03 | Concrete inner steel bar intelligent positioning method based on ground penetrating radar and deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111798411B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111854666A (en) * | 2020-07-09 | 2020-10-30 | 广东雄炜建筑工程检测有限公司 | Method for scanning steel bars and detecting thickness of concrete protection layer |
| CN112666545A (en) * | 2020-12-11 | 2021-04-16 | 南通勘察设计有限公司 | Concrete reinforcement protective layer thickness detection system and detection method thereof |
| CN112731381A (en) * | 2020-12-16 | 2021-04-30 | 华南农业大学 | Method for intelligently detecting hard foreign matters in soil by utilizing android debugging bridge and vehicle-mounted radar |
| CN112800664B (en) * | 2021-01-08 | 2024-03-08 | 江南大学 | Method for estimating tree root diameter based on ground penetrating radar A-scan data |
| CN114359369B (en) * | 2021-12-28 | 2024-05-03 | 华南农业大学 | Fruit tree root system identification and positioning method based on deep learning and ground penetrating radar |
| CN114777642B (en) * | 2022-03-16 | 2023-07-28 | 交通运输部公路科学研究所 | Three-dimensional space position detection system for load transmission component of in-service concrete pavement slab |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108981557A (en) * | 2018-09-05 | 2018-12-11 | 广州大学 | Detection method that is a kind of while measuring reinforcement in concrete diameter and its protective layer thickness |
| CN109923264A (en) * | 2016-09-14 | 2019-06-21 | 爱马特伦系统有限责任公司 | Reinforce the method for cement works by high speed extrusion printing and using the equipment of this method |
-
2020
- 2020-06-03 CN CN202010493228.4A patent/CN111798411B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109923264A (en) * | 2016-09-14 | 2019-06-21 | 爱马特伦系统有限责任公司 | Reinforce the method for cement works by high speed extrusion printing and using the equipment of this method |
| CN108981557A (en) * | 2018-09-05 | 2018-12-11 | 广州大学 | Detection method that is a kind of while measuring reinforcement in concrete diameter and its protective layer thickness |
Non-Patent Citations (1)
| Title |
|---|
| 雷达技术在混凝土结构无损检测中的应用;王茹;邵珍奇;;核电子学与探测技术(第02期);第459-462页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111798411A (en) | 2020-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111798411B (en) | Concrete inner steel bar intelligent positioning method based on ground penetrating radar and deep learning | |
| CN112130132B (en) | Underground pipeline detection method and system based on ground penetrating radar and deep learning | |
| Liu et al. | Detection and localization of rebar in concrete by deep learning using ground penetrating radar | |
| Hou et al. | Improved Mask R-CNN with distance guided intersection over union for GPR signature detection and segmentation | |
| CN109685011A (en) | A kind of underground utilities detection recognition method based on deep learning | |
| WO2021082904A1 (en) | Ground penetrating radar map adaptive selection method for detecting water damage | |
| CN107621626A (en) | Radar signal Railway Roadbed detection method based on depth convolutional neural networks | |
| CN111751392B (en) | Steel bar corrosion detection method based on dual-polarization ground penetrating radar | |
| CN111445515B (en) | Underground cylinder target radius estimation method and system based on feature fusion network | |
| WO2021083394A1 (en) | Method for detection of water damage to asphalt road surface based on adaptive selection of plot grayscales | |
| Ma et al. | Automatic detection of steel rebar in bridge decks from ground penetrating radar data | |
| Hu et al. | A study of automatic recognition and localization of pipeline for ground penetrating radar based on deep learning | |
| Xiong et al. | Automatic detection and location of pavement internal distresses from ground penetrating radar images based on deep learning | |
| CN113468804B (en) | Underground pipeline identification method based on matrix bundles and deep neural network | |
| Hu et al. | Defects identification and location of underground space for ground penetrating radar based on deep learning | |
| Liu et al. | Advances in automatic identification of road subsurface distress using ground penetrating radar: State of the art and future trends | |
| Huang et al. | Automated detection of through-cracks in pavement using three-instantaneous attributes fusion and Swin Transformer network | |
| Xiong et al. | Deep learning-based detection of tie bars in concrete pavement using ground penetrating radar | |
| CN112257735A (en) | Internal defect radar detection system based on artificial intelligence | |
| CN105844644A (en) | Morphological median derivative-based sea-land clutter scene segmentation method | |
| CN115859081A (en) | Visual detection method and device for pipeline | |
| CN115630492A (en) | Intelligent inversion method, system and storage medium for tunnel lining disease change characteristics | |
| Guo et al. | Research on tunnel lining image target recognition method based on YOLOv3 | |
| CN114693776A (en) | Cable position information determining method, device, equipment and storage medium | |
| Song et al. | Underground diseases identification of airport runway using GPR |
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 |