CN114612731B - Intelligent identification method and system for road flatness detection - Google Patents
Intelligent identification method and system for road flatness detection Download PDFInfo
- Publication number
- CN114612731B CN114612731B CN202210503976.5A CN202210503976A CN114612731B CN 114612731 B CN114612731 B CN 114612731B CN 202210503976 A CN202210503976 A CN 202210503976A CN 114612731 B CN114612731 B CN 114612731B
- Authority
- CN
- China
- Prior art keywords
- detection
- picture
- flatness
- pixel points
- road
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/64—Analysis of geometric attributes of convexity or concavity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- 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
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Geometry (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention relates to the technical field of image data processing, and discloses an intelligent identification method for road flatness detection, which comprises the following steps: s100, acquiring detection information of a detected vehicle running on a target road; s200, drawing and generating a corresponding detection picture according to the detection information; s300, processing the detected picture according to a preset segmentation and combination rule to obtain a standard picture; s400, the standard image sending and identifying module is used for measuring and calculating the flatness; the method has the advantages that the road flatness is rapidly evaluated and calculated, the efficiency is improved, and the calculation amount and the complexity are greatly reduced.
Description
Technical Field
The invention relates to the technical field of image data processing, in particular to an intelligent identification method for road flatness detection.
Background
As a core part of a highway maintenance management system, inspection and evaluation of road surface quality is very important. The road surface with poor flatness not only influences the driving safety of the road, reduces the driving comfort, increases the running cost of driving, but also can accelerate the damage of the road surface and shorten the maintenance period.
After the road is paved, the road side is easy to deform due to reasons of terrain settlement, deformation, rain wash, vehicle running, heavy object rolling and the like, pits or bulges and the like are formed to cause road deformation, and then the safety of road running is influenced, particularly, the road such as an airport runway, a racing runway, an experimental road and the like has extremely high requirement on the flatness of the road; therefore, it is necessary to detect the flatness of the road for a long time.
At present, a traditional 3M ruler detection method is adopted for detecting the road flatness, a ruler with the length of 3 meters is placed on a road, and the condition that the ruler is attached to the surface of the road is checked to judge the flatness of the road surface.
However, the method can only judge the problem of road surface evenness in the area within 3 meters, the area outside 3 meters cannot be detected, the whole measuring result is not standard due to no reference basis, the error is large, and the method cannot be used as the basis for subsequent reference; and a large amount of manpower and material resources are needed, and the efficiency cannot be ensured.
Disclosure of Invention
The invention aims to provide an intelligent identification method for detecting road flatness, which solves the following technical problems:
how to provide an efficient road flatness detection method.
The purpose of the invention can be realized by the following technical scheme:
an intelligent identification method for road flatness detection comprises the following steps:
s100, acquiring detection information of a detected vehicle running on a target road;
s200, drawing and generating a corresponding detection picture according to the detection information;
s300, processing the detected picture according to a preset segmentation and combination rule to obtain a standard picture;
s400, sending the standard picture to an identification module for flatness measurement and calculation;
the detection information comprises the running speed of the detection vehicle on a horizontal plane and the height change track on a vertical plane, and the detection picture comprises the motion track of the detection vehicle in the vertical direction; the recognition module is a neural network model after training is completed.
Through the technical scheme, the detection information comprising the corresponding running speed and the height change track on the vertical surface is generated when the detection vehicle runs on the target road, and the detection vehicle inevitably fluctuates in height when encountering uneven ground, so that the height change track which changes along with the change of the running distance can be shown on the detection picture generated according to the detection information, and the corresponding standard picture processed according to the preset segmentation and combination rule can be sent to the identification module for intelligent identification, so that the flatness of the road can be evaluated and calculated quickly, the efficiency is improved, and the operation amount and the complexity are reduced greatly.
As a further scheme of the invention: the step S200 includes:
establishing a rectangular coordinate system by taking the running distance of the detected vehicle on the horizontal plane as a transverse axis and the height on the vertical plane as a vertical axis;
setting the rectangular coordinate system on a blank picture, and inputting the detection information into the rectangular coordinate system to generate the detection picture;
wherein the length of the detection picture is related to the driving distance of the detection vehicle on the horizontal plane of the horizontal axis.
Through the technical scheme, the height change track on the detected picture is normalized by utilizing the rectangular coordinate system, so that the regional identification and judgment of the detected picture are favorably realized in a targeted manner, and the flexibility and the accuracy are ensured.
As a further scheme of the invention: the step S300 includes:
segmenting the detection picture according to a preset segmentation and combination rule to obtain n segmented pictures with equal length;
and merging the n segmented pictures according to a preset segmentation merging rule.
Through the technical scheme, the standard picture obtained after the segmented pictures are combined has the length which is reduced by n times compared with the previously obtained detection picture, so that the calculation amount of the identification module is reduced, and the complexity is reduced.
As a further scheme of the invention: the preset segmentation and combination rule comprises the following steps:
when the detection picture is segmented, establishing a new origin on the segmented picture after segmentation, wherein the origin is the midpoint of the left edge of the segmented picture;
when the segmented pictures are combined, combining all pixel points representing a new original point, a rectangular coordinate system of the original point and a height change track;
all pixel points except the height change track are defined as standard color pixel points, and all pixel points forming the height change track are colored pixel points;
and changing the pixel brightness of the colored pixel points according to a preset value every time the colored pixel points are combined.
Through above-mentioned technical scheme, establish new initial point can be with the uniformity of convenient back merging process, and the rethread distinguishes standard color pixel and coloured pixel, can promote the degree of discernment of picture, and the abscissa of the coloured pixel of combination represents the road surface position that is in uniform height, and its pixel luminance changes the back, can reflect the road surface position quantity condition that is in this height through the luminance value, is favorable to identification module's accurate discernment.
As a further scheme of the invention: the preset segmentation and combination rule comprises the following steps:
when the colored pixel point at the merging position is located in a first position area, setting the preset value as a first change value;
when the colored pixel point at the merging position is located in the second position area, the preset value is set as a second change value;
the first position area is an area which is separated from the normal direction of the transverse axis by m pixel points, and the second position area is an area except the first position area;
the first variation value is negative and the second variation value is positive.
Through the technical scheme, general newly-built road surface, its roughness is not easily perceived, consequently, the position that is close to the cross axle more can have the condition of more coloured pixel coincidence, the coincidence quantity in normal region is more better, and the quantity of the coloured pixel coincidence of keeping away from the cross axle position then is more better less, consequently, for the discernment degree of further increase standard picture, can set up the region that is close to the cross axle into first position region, the coloured pixel quantity of coincidence in first position region is more, the more luminance that the coincidence quantity is lower more, the coloured pixel quantity of coincidence in second position region is less, the more luminance that the coincidence quantity is higher, from this can increase in the standard picture between the coloured pixel and the luminance difference between the different position regions, thereby promote the discernment degree, indirectly be favorable to promoting the precision of identification module discernment.
As a further scheme of the invention: and changing the colored pixel points with the brightness values lower than the preset reduction threshold value into colorless pixel points.
As a further scheme of the invention: the step S400 includes:
the flatness measurement result comprises the following steps: flat, uneven and partially repaired.
As a further scheme of the invention: further comprising:
if the flatness measurement result is flatness, the flatness of the target road is qualified;
if the flatness measurement result is uneven, the flatness of the target road is unqualified;
and if the flatness measurement result indicates that partial repair is needed, entering a repair road section confirmation procedure.
As a further scheme of the invention: the repair link confirmation program includes:
acquiring a horizontal axis coordinate of a colored pixel point with a brightness value higher than a preset repairing threshold;
and determining the position of the target road needing to be repaired according to the horizontal axis coordinate.
An intelligent recognition system for road flatness detection, comprising:
the sampling module is used for acquiring detection information of a detection vehicle running on a target road;
the drawing module is connected with the sampling module and used for drawing and generating a corresponding detection picture according to the detection information;
the processing module is connected with the drawing module and used for processing the detection picture according to a preset segmentation and combination rule to obtain a standard picture;
the recognition module is a trained neural network model;
and the communication module is connected with the processing module and used for sending the standard picture to the identification module for flatness measurement and calculation.
The invention has the beneficial effects that:
according to the invention, the detection information comprising the corresponding running speed and the height change track on the vertical surface is generated when the detection vehicle runs on the target road, and the detection vehicle inevitably fluctuates in height when encountering uneven ground, so that the height change track changing along with the change of the running distance can be shown on the detection picture generated according to the detection information, and the corresponding standard picture processed according to the preset segmentation and combination rule can be sent to the identification module for intelligent identification, so that the road flatness can be rapidly evaluated and calculated, the operation amount and the complexity are greatly reduced, and the efficiency is improved.
Drawings
The invention will be further described with reference to the accompanying drawings.
FIG. 1 is a basic flow diagram of the intelligent recognition method of the present invention;
FIG. 2 is a schematic diagram of processing a detected picture according to a preset segmentation and merging rule in the present invention;
fig. 3 is a schematic diagram showing a standard picture after the standard picture is synthesized from segmented pictures in the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, the present invention is an intelligent identification method for road flatness detection, including the following steps:
s100, acquiring detection information of a detected vehicle running on a target road;
s200, drawing and generating a corresponding detection picture according to the detection information;
s300, processing the detected picture according to a preset segmentation and combination rule to obtain a standard picture;
s400, sending the standard picture to an identification module for flatness measurement and calculation;
the detection information comprises the running speed of the detected vehicle on the horizontal plane and the height change track on the vertical plane, and the detection picture comprises the motion track of the detected vehicle in the vertical direction; the recognition module is a neural network model after training is completed.
According to the invention, the detection information comprising the corresponding running speed and the height change track on the vertical surface is generated when the detection vehicle runs on the target road, and the detection vehicle inevitably fluctuates in height when encountering uneven ground, so that the height change track changing along with the change of the running distance can be shown on the detection picture generated according to the detection information, and the corresponding standard picture processed according to the preset segmentation and combination rule can be sent to the identification module for intelligent identification, so that the road flatness can be rapidly evaluated and calculated, the operation amount and the complexity are greatly reduced, and the efficiency is improved.
In step S100 of this embodiment of the present invention, a projection background long plate may be disposed along one side of the target road, a laser unit for irradiating laser onto the projection background long plate is fixed on the detection vehicle, the projection of the laser unit onto the projection background long plate is in a dot shape, the dot projection may be recorded by using an image pickup device, and the height change trajectory may be obtained by processing a dot object in the image pickup record.
For example, the projection background long plate is a long rectangle which can be selected to be white, the point-shaped laser projection can be captured by a camera, the width of the long rectangle is known, the distance from the point-shaped projection to the upper edge of the long rectangle can be calculated from the image, and therefore, the height change track reflecting the road leveling condition can be obtained through the change of the distance from the point-shaped projection to the upper edge of the long rectangle.
As a further scheme of the invention: the step S200 includes:
establishing a rectangular coordinate system by taking the running distance of the detected vehicle on the horizontal plane as a transverse axis and the height on the vertical plane as a vertical axis;
setting the rectangular coordinate system on a blank picture, and inputting detection information into the rectangular coordinate system to generate a detection picture;
wherein, the length of the detection picture is related to the driving distance of the detection vehicle on the horizontal plane of the horizontal axis.
Through the technical scheme, the height change track on the detected picture is normalized by utilizing the rectangular coordinate system, so that the regional identification and judgment of the detected picture are favorably realized in a targeted manner, and the flexibility and the accuracy are ensured. Therefore, an origin point can be set at the left middle point of the blank picture, and then the trend of the transverse axis is determined in the direction that the transverse axis of the rectangular coordinate system is parallel to the length direction side of the blank picture, so that the fluctuation of the height change track floats up and down around the transverse axis.
As a further scheme of the invention: step S300 includes:
segmenting the detection picture according to a preset segmentation and combination rule to obtain n segmented pictures with equal length;
and merging the n segmented pictures according to a preset segmentation merging rule.
Through the technical scheme, the standard picture obtained after the segmented pictures are combined has the length which is reduced by n times compared with the previously obtained detection picture, so that the calculation amount of the identification module is reduced, and the complexity is reduced.
For example, when the flatness detection is performed on a road segment of which the target road is 100 meters, a detection picture reflecting the distance of 100 meters in length can be obtained, and when n is 100, each segmented picture includes a height change track of 1 meter in length, and the total number of pixel points after the 100 segmented pictures are combined is greatly reduced, so that the calculation amount of an identification module is fully reduced, and the complexity is reduced.
As shown in fig. 2, the detected picture is divided into 9 segments of segmented pictures and then synthesized to obtain a corresponding standard picture.
As a further scheme of the invention: the preset segmentation and combination rule comprises the following steps:
when the detection picture is segmented, establishing a new origin on the segmented picture after segmentation, wherein the origin is the midpoint of the left edge of the segmented picture;
when the segmented pictures are combined, combining all pixel points representing a new original point, a rectangular coordinate system of the original point and a height change track;
all pixel points except the height change track are defined as standard color pixel points, and all pixel points forming the height change track are colored pixel points;
and changing the pixel brightness of the colored pixel points according to a preset value every time the colored pixel points are combined.
In the above technical solution, the standard color pixel points are colors of the hollow white picture or the projection background long plate in the above technical solution, and the colors of the colored pixel points are all the same, but are different from the colors of the standard color pixel points, and the combined colored pixel points have a change in brightness, which is related to a preset value.
Therefore, the new original point is established to conveniently promote the uniformity of subsequent combination, the standard color pixel points and the colored pixel points are distinguished, the identification degree of the picture can be improved, the abscissa of the combined colored pixel points represents the position of the road surface at the same height, and after the pixel brightness is changed, the condition of the quantity of the road surface positions at the height can be reflected through the brightness value, so that the accurate identification of the identification module is facilitated.
As a further scheme of the invention: referring to fig. 3, the preset segmentation merging rule includes:
when the colored pixel point of the merging position is located in the first position area, the preset value is set as a first change value;
when the colored pixel point of the merging position is located in the second position area, the preset value is set as a second change value;
the first position area is an area which is separated from the transverse axis by m pixel points in the normal direction, and the second position area is an area except the first position area;
the first variation value is negative and the second variation value is positive.
The invention is suitable for the condition that the flatness of the road surface is not easy to detect, so that the situation that colored pixel points coincide occurs at the position closer to a transverse shaft in a standard picture, because the flatness of most road surfaces is calculated to be normal, the more the coincidence quantity in a normal area is natural, the better the coincidence quantity is, and the less the coincidence quantity of the colored pixel points far away from the transverse shaft is, the better the coincidence quantity is.
Therefore, as shown in fig. 3, for the discernment degree of further increase standard picture, can set up the region that is close to the cross axle into first position region, the colored pixel number of coincidence that is in first position region is more, the more luminance that the coincidence quantity is lower, the colored pixel number of coincidence that is in second position region is less, the more luminance that the coincidence quantity is higher, from this, can increase in the standard picture between the colored pixel and the luminance difference between the different position regions, thereby promote the discernment degree, indirectly be favorable to promoting the precision of identification module discernment.
Therefore, in the standard picture, not only the shape of the height change track influences the judgment result of the identification module, but also the number and the brightness of the overlapped colored pixel points and the brightness difference of different areas can reflect the condition of the flatness.
For example, if the overlapping number of the colored pixels in the first position region is more, it indicates that the flatness is better.
If the brightness of the colored pixel points in the first position area is higher, the flatness is poor.
If the brightness difference between the colored pixel points in the first position area and the second position area is larger, the flatness is poorer.
Of course, in the invention, a third position area can be continuously added, and the third position area occupies a position of the second position area far away from the first position area, so that the judgment kernel rule of the flatness is further subdivided, and the detection accuracy of the identification module is improved.
As a further scheme of the invention: and changing the colored pixel points with the brightness values lower than the preset reduction threshold value into colorless pixel points.
As a further scheme of the invention: step S400 includes:
the result of the flatness measurement comprises the following steps: flat, uneven and partially repaired.
As a further scheme of the invention: further comprising:
if the flatness measurement result is flatness, the flatness of the target road is qualified;
if the flatness measurement result is uneven, the flatness of the target road is unqualified;
and if the flatness measurement result indicates that partial repair is needed, entering a repair road section confirmation procedure.
As a further scheme of the invention: the repair link confirmation program includes:
acquiring a horizontal axis coordinate of a colored pixel point with a brightness value higher than a preset repairing threshold;
and determining the position of the target road to be repaired according to the horizontal axis coordinate.
In the embodiment, when the detection picture is segmented, the segmented picture is labeled according to the time sequence, so that the later pavement position is convenient to determine.
An intelligent recognition system for road flatness detection, comprising:
the sampling module is used for acquiring detection information of a detection vehicle running on a target road;
the drawing module is connected with the sampling module and used for drawing and generating a corresponding detection picture according to the detection information;
the processing module is connected with the drawing module and used for processing the detected picture according to a preset segmentation and combination rule to obtain a standard picture;
the recognition module is a trained neural network model;
and the communication module is connected with the processing module and is used for sending the standard picture to the identification module for flatness measurement and calculation.
The training process of the recognition module is the same as that of the prior art, a large number of training samples are needed, the acquisition mode of the training samples is the same as that of the standard picture in the invention, and a correct flatness measurement result label is attached.
While one embodiment of the present invention has been described in detail, the description is only a preferred embodiment of the present invention and should not be taken as limiting the scope of the invention. All equivalent changes and modifications made within the scope of the present invention shall fall within the scope of the present invention.
Claims (7)
1. An intelligent identification method for road flatness detection is characterized by comprising the following steps:
s100, acquiring detection information of a detected vehicle running on a target road;
s200, drawing and generating a corresponding detection picture according to the detection information;
s300, processing the detected picture according to a preset segmentation and combination rule to obtain a standard picture;
s400, sending the standard picture to an identification module for flatness measurement and calculation;
the detection information comprises the running speed of the detection vehicle on a horizontal plane and the height change track on a vertical plane, and the detection picture comprises the motion track of the detection vehicle in the vertical direction; the recognition module is a trained neural network model;
the step S300 includes:
segmenting the detection picture according to a preset segmentation and combination rule to obtain n segmented pictures with equal length;
merging the n segmented pictures according to a preset segmentation merging rule;
the preset segmentation and combination rule comprises the following steps:
when the detection picture is segmented, establishing a new origin on the segmented picture after segmentation, wherein the origin is the midpoint of the left edge of the segmented picture;
when the segmented pictures are combined, combining all pixel points representing a new original point, a rectangular coordinate system of the original point and a height change track;
all pixel points except the height change track are defined as standard color pixel points, and all pixel points forming the height change track are colored pixel points;
changing the pixel brightness of the colored pixel points according to a preset value every time the colored pixel points are combined;
the preset segmentation and combination rule further comprises:
when the colored pixel point at the merging position is located in a first position area, setting the preset value as a first change value;
when the colored pixel point at the merging position is located in the second position area, the preset value is set as a second change value;
the first position area is an area which is separated from the normal direction of the transverse axis by m pixel points, and the second position area is an area except the first position area;
the first variation value is negative and the second variation value is positive.
2. The intelligent recognition method for road flatness detection according to claim 1, wherein the step S200 includes:
establishing a rectangular coordinate system by taking the running distance of the detected vehicle on the horizontal plane as a transverse axis and the height on the vertical plane as a vertical axis;
setting the rectangular coordinate system on a blank picture, and inputting the detection information into the rectangular coordinate system to generate the detection picture;
wherein the length of the detection picture is related to the driving distance of the detection vehicle on the horizontal plane of the horizontal axis.
3. The intelligent recognition method for road flatness detection according to claim 2, wherein colored pixels with brightness values lower than a preset subtraction threshold are changed to colorless pixels.
4. The intelligent recognition method for road flatness detection according to claim 2, wherein said step S400 includes:
the flatness measurement result comprises the following steps: flat, uneven and partially repaired.
5. The intelligent recognition method for road flatness detection according to claim 4, further comprising:
if the flatness measurement result is flatness, the flatness of the target road is qualified;
if the flatness measurement result is uneven, the flatness of the target road is unqualified;
and if the flatness measurement result indicates that partial repair is needed, entering a repair road section confirmation procedure.
6. The intelligent recognition method for road flatness detection according to claim 5, wherein the repaired section confirmation procedure includes:
acquiring a horizontal axis coordinate of a colored pixel point with a brightness value higher than a preset repairing threshold;
and determining the position of the target road needing to be repaired according to the horizontal axis coordinate.
7. An intelligent recognition system for road flatness detection, comprising:
the sampling module is used for acquiring detection information of a detection vehicle running on a target road;
the drawing module is connected with the sampling module and used for drawing and generating a corresponding detection picture according to the detection information;
the processing module is connected with the drawing module and used for processing the detection picture according to a preset segmentation and combination rule to obtain a standard picture;
the recognition module is a trained neural network model;
the communication module is connected with the processing module and used for sending the standard picture to the identification module for flatness measurement and calculation;
wherein, the processing the detection picture according to the preset segmentation and combination rule to obtain the standard picture comprises:
segmenting the detection picture according to a preset segmentation and combination rule to obtain n segmented pictures with equal length;
merging the n segmented pictures according to a preset segmentation merging rule;
wherein the preset segmentation and merging rules comprise:
when the detection picture is segmented, establishing a new origin on the segmented picture after segmentation, wherein the origin is the midpoint of the left edge of the segmented picture;
when the segmented pictures are combined, combining all pixel points representing a new original point, a rectangular coordinate system of the original point and a height change track;
all pixel points except the height change track are defined as standard color pixel points, and all pixel points forming the height change track are colored pixel points;
changing the pixel brightness of the colored pixel points according to a preset value every time the colored pixel points are combined;
the preset segmentation merging rule further comprises:
when the colored pixel point at the merging position is located in the first position area, the preset value is set as a first change value;
when the colored pixel point at the merging position is located in the second position area, the preset value is set as a second change value;
the first position area is an area which is separated from the normal direction of the transverse axis by m pixel points, and the second position area is an area except the first position area;
the first variation value is negative and the second variation value is positive.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210503976.5A CN114612731B (en) | 2022-05-10 | 2022-05-10 | Intelligent identification method and system for road flatness detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210503976.5A CN114612731B (en) | 2022-05-10 | 2022-05-10 | Intelligent identification method and system for road flatness detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114612731A CN114612731A (en) | 2022-06-10 |
| CN114612731B true CN114612731B (en) | 2022-08-09 |
Family
ID=81870219
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210503976.5A Active CN114612731B (en) | 2022-05-10 | 2022-05-10 | Intelligent identification method and system for road flatness detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114612731B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114780907B (en) * | 2022-06-22 | 2022-09-02 | 中国铁路设计集团有限公司 | Method for determining surface settlement limit value of underpass road of subway interval tunnel |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101126638A (en) * | 2007-09-29 | 2008-02-20 | 张小虎 | Pick-up measuring method for checking road surface planeness |
| CN101246000A (en) * | 2008-03-21 | 2008-08-20 | 哈尔滨工业大学 | Pavement planeness automatic detection device and detecting method |
| CN101319485A (en) * | 2008-07-16 | 2008-12-10 | 张小虎 | Photographing measurement method for detecting highroad pavement planeness |
| CN103874615A (en) * | 2011-10-06 | 2014-06-18 | Lg伊诺特有限公司 | Apparatus and method for measuring road flatness |
| CN204023380U (en) * | 2014-08-26 | 2014-12-17 | 长安大学 | Based on the pavement flatness checking device of multisensor |
| CN109167818A (en) * | 2018-08-06 | 2019-01-08 | 长安大学 | Road evenness detection system based on smart phone crowdsourcing acquisition |
| CN109468929A (en) * | 2018-12-25 | 2019-03-15 | 河南城建学院 | Flatness detection trolley based on CPLD |
| CN110318327A (en) * | 2019-06-10 | 2019-10-11 | 长安大学 | A kind of surface evenness prediction technique based on random forest |
| CN111127459A (en) * | 2019-12-30 | 2020-05-08 | 武汉理工大学 | Real-time image processing system for road track detection |
| CN112050756A (en) * | 2020-09-04 | 2020-12-08 | 南通大学 | Rock ore slice and resin target surface flatness recognition processing method |
| CN113062182A (en) * | 2021-03-30 | 2021-07-02 | 山东建筑大学 | Fine road surface smoothness detection system and method based on multi-sensor fusion |
| CN113353085A (en) * | 2021-07-03 | 2021-09-07 | 西北工业大学 | Road surface unevenness identification method based on Kalman filtering theory |
| GB202112265D0 (en) * | 2020-08-28 | 2021-10-13 | Nvidia Corp | Object image completion |
| CN113932758A (en) * | 2021-09-15 | 2022-01-14 | 同济大学 | Road surface flatness prediction method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110856112B (en) * | 2019-11-14 | 2021-06-18 | 深圳先进技术研究院 | Crowd-sourcing perception multi-source information fusion indoor positioning method and system |
-
2022
- 2022-05-10 CN CN202210503976.5A patent/CN114612731B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101126638A (en) * | 2007-09-29 | 2008-02-20 | 张小虎 | Pick-up measuring method for checking road surface planeness |
| CN101246000A (en) * | 2008-03-21 | 2008-08-20 | 哈尔滨工业大学 | Pavement planeness automatic detection device and detecting method |
| CN101319485A (en) * | 2008-07-16 | 2008-12-10 | 张小虎 | Photographing measurement method for detecting highroad pavement planeness |
| CN103874615A (en) * | 2011-10-06 | 2014-06-18 | Lg伊诺特有限公司 | Apparatus and method for measuring road flatness |
| CN204023380U (en) * | 2014-08-26 | 2014-12-17 | 长安大学 | Based on the pavement flatness checking device of multisensor |
| CN109167818A (en) * | 2018-08-06 | 2019-01-08 | 长安大学 | Road evenness detection system based on smart phone crowdsourcing acquisition |
| CN109468929A (en) * | 2018-12-25 | 2019-03-15 | 河南城建学院 | Flatness detection trolley based on CPLD |
| CN110318327A (en) * | 2019-06-10 | 2019-10-11 | 长安大学 | A kind of surface evenness prediction technique based on random forest |
| CN111127459A (en) * | 2019-12-30 | 2020-05-08 | 武汉理工大学 | Real-time image processing system for road track detection |
| GB202112265D0 (en) * | 2020-08-28 | 2021-10-13 | Nvidia Corp | Object image completion |
| CN112050756A (en) * | 2020-09-04 | 2020-12-08 | 南通大学 | Rock ore slice and resin target surface flatness recognition processing method |
| CN113062182A (en) * | 2021-03-30 | 2021-07-02 | 山东建筑大学 | Fine road surface smoothness detection system and method based on multi-sensor fusion |
| CN113353085A (en) * | 2021-07-03 | 2021-09-07 | 西北工业大学 | Road surface unevenness identification method based on Kalman filtering theory |
| CN113932758A (en) * | 2021-09-15 | 2022-01-14 | 同济大学 | Road surface flatness prediction method and device |
Non-Patent Citations (5)
| Title |
|---|
| A deep learning approach to automatic road surface monitoring and pothole detection;Braian Varona 等;《Springer-Verlag London Ltd., part of Springer Nature 2019》;20190527;1-16 * |
| 基于多传感器数据融合的路面平整度检测系统;邵志超;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20160215(第02期);C034-29 * |
| 基于路面测量值的沥青路面平整度数学模型研究;李伟;《公路工程》;20170630;第42卷(第3期);314-319 * |
| 基于车载激光的道路平整度检测系统研究;郭姣;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20140215(第02期);C034-23 * |
| 路面检测技术综述;马键 等;《交通运输工程学报》;20171031;第17卷(第5期);121-137 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114612731A (en) | 2022-06-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105260699B (en) | A kind of processing method and processing device of lane line data | |
| CN103824452B (en) | A kind of peccancy parking detector based on panoramic vision of lightweight | |
| EP1890264B1 (en) | Vehicle, image processing system, image processing method, image processing program, image processing system configuration method, and server | |
| CN107180416B (en) | Train wheel tread image deformity correction method and system | |
| CN111784657A (en) | Digital image-based system and method for automatically identifying cement pavement diseases | |
| CN109635737B (en) | Auxiliary vehicle navigation positioning method based on road marking line visual identification | |
| CN110379168B (en) | Traffic vehicle information acquisition method based on Mask R-CNN | |
| CN109782364B (en) | Traffic sign board missing detection method based on machine vision | |
| CN107798293A (en) | A kind of crack on road detection means | |
| CN103077387B (en) | Carriage of freight train automatic testing method in video | |
| CN114612731B (en) | Intelligent identification method and system for road flatness detection | |
| CN113011283B (en) | Non-contact type rail sleeper relative displacement real-time measurement method based on video | |
| CN117173661B (en) | Asphalt road quality detection method based on computer vision | |
| Kargah-Ostadi et al. | Evaluation framework for automated pavement distress identification and quantification applications | |
| CN110826412A (en) | Highway visibility detection system and method | |
| Chua et al. | Simple procedure for identifying pavement distresses from video images | |
| CN113312987B (en) | Recognition method based on unmanned aerial vehicle road surface crack image | |
| CN114663859A (en) | Sensitive and accurate complex road condition lane deviation real-time early warning system | |
| CN113343782A (en) | Expressway sign plate detection method based on unmanned aerial vehicle remote sensing | |
| KR102562757B1 (en) | Prediction and recognition method of road marking information and road maintenance method | |
| CN116631187B (en) | Intelligent acquisition and analysis system for case on-site investigation information | |
| CN116433527A (en) | Road surface laser line extraction device and method | |
| CN116935157A (en) | Pavement disease detection method based on PConv-YOLOv7 | |
| CN115100173B (en) | Road pavement image crack geometric property detection and crack identification method based on deep convolutional neural network | |
| CN114693722B (en) | Vehicle driving behavior detection method, detection device and detection equipment |
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 |