CN111091049A - Road surface obstacle detection method based on reverse feature matching - Google Patents
Road surface obstacle detection method based on reverse feature matching Download PDFInfo
- Publication number
- CN111091049A CN111091049A CN201911061150.2A CN201911061150A CN111091049A CN 111091049 A CN111091049 A CN 111091049A CN 201911061150 A CN201911061150 A CN 201911061150A CN 111091049 A CN111091049 A CN 111091049A
- Authority
- CN
- China
- Prior art keywords
- road
- point
- points
- obstacle
- feature
- 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.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 40
- 230000008859 change Effects 0.000 claims abstract description 33
- 238000012216 screening Methods 0.000 claims abstract description 16
- 230000004888 barrier function Effects 0.000 claims abstract description 13
- 238000012163 sequencing technique Methods 0.000 claims description 5
- 102100029469 WD repeat and HMG-box DNA-binding protein 1 Human genes 0.000 claims description 3
- 101710097421 WD repeat and HMG-box DNA-binding protein 1 Proteins 0.000 claims description 3
- 238000007670 refining Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 6
- 238000012545 processing Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 2
- 239000004575 stone Substances 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000000746 purification Methods 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000004347 surface barrier Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/211—Selection of the most significant subset of features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- 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)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (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)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a road surface obstacle detection method based on reverse feature matching, which belongs to the field of remote sensing change detection, and is characterized in that based on space-sky high-resolution remote sensing images, change information in the images is extracted by using an SIFT feature point reverse matching method, and area and position information of road surface obstacles in the images is obtained by processing the change information. The method has the characteristics or key links that: respectively calculating SIFT feature point sets of the remote sensing images before and after the path loss, implementing feature matching, and negating matching results to obtain a change information feature point set; screening a road buffer area to obtain a change information characteristic point set in a road area, and screening out false change points and tiny change points by a sliding window method; and finally, grouping and combining the characteristic points to obtain position information of the obstacle, and then combining a region growing method to obtain the approximate range of the obstacle and estimating the area. The key technology of the invention belongs to the field of remote sensing change detection, but the position and area information of the road barrier can be obtained without image registration, and the degree of automation is high.
Description
Technical Field
The invention relates to the detection problem of road surface obstacles, in particular to the detection problem of road passing obstacles generated by disasters such as debris flow, landslide, collapse and the like by using a remote sensing technology, belonging to the technical field of remote sensing.
Background
By the end of 2018, the total road mileage in China reaches 484.65 kilometers, and the road density reaches 50.48 kilometers per hundred square kilometers. The mountain area of China accounts for more than two thirds of the total area of the whole country, and mountain roads are easily damaged by disasters such as debris flow, landslide, collapse, rolling stones and the like. And the geological disasters which cause road damage in 2018 have 501, and huge economic losses are caused. The method has the advantages that the barrier information in the road passing direction is found out in time, the rush-to-pass time is judged according to factors such as the property and the size of the barrier, the advancing scheme is adjusted in real time, and the method has great practical significance to military and civil traffic organizations.
The traditional road obstacle information acquisition mode mainly depends on manual on-site investigation, and has the defects of large consumption of manpower and material resources, difficulty in data organization, low informatization degree, poor instantaneity and the like. With the development of remote sensing technology, road damage detection is carried out by more remote sensing technologies at present. The main methods are of three types: human-machine interaction interpretation, machine interpretation based on image classification, and machine interpretation based on change detection. The damage assessment is performed in a man-machine interaction interpretation mode, the reliability is high, the automation degree is low, and the workload is large. The method based on image classification can perform qualitative analysis on the road obstacle extraction result, but the accuracy of the method depends on the accuracy of classification. The accuracy of the change detection-based method depends on the registration accuracy, and the detection result needs further interpretation to obtain more accurate obstacle information.
Therefore, the existing road damage remote sensing detection technology has the defects of low automation degree, difficult registration, requirement of personnel with professional knowledge for operation and the like.
Disclosure of Invention
The invention provides a road surface obstacle detection method based on reverse feature matching aiming at the defects of the prior art, the position and area information of a road obstacle is obtained by processing feature points which are not matched with an image before disaster in an image after disaster, the method belongs to the field of remote sensing change detection, image registration is not needed, and automation is basically realized.
In order to solve the technical problem, the technical scheme of the invention is as follows, the method for detecting the road surface obstacle based on reverse feature matching comprises the following steps:
step one, calculating SIFT feature point sets of remote sensing images before and after the road loss respectively, performing feature matching, and negating matching results to obtain a change information feature point set;
step two, screening by using a road buffer area to obtain a change information characteristic point set in a road area;
thirdly, screening out false change characteristic points and micro change characteristic points according to the distribution concentration of the characteristic points;
step four, preliminarily grouping the characteristic points according to the point intervals, and distinguishing the characteristic points belonging to different obstacles;
step five, detecting the primary grouping result obtained in the last step, and obtaining a final grouping result through similarity judgment;
and step six, obtaining the position information of each obstacle according to the grouping result, and obtaining the distribution range of the obstacles by combining a multi-seed point region growing method and carrying out area estimation.
In a further improvement of the present invention, in the step one, the inverse feature matching formula is:
Qu=CQQm(1)
point set QuIs a point set QmComplement in Q, i.e.Wherein Q is a characteristic point set of the post-disaster image, QmIs a set of feature points, Q, in the post-disaster image that can be matched with the pre-disaster image feature pointsuThe image change information feature point set is obtained.
In the second step, the existing road vector data is used to generate a road buffer, and a change feature point set Q in the buffer is screenedurThe formula is as follows:
where Poly is the road buffer polygon, i is 0,1, …, k, k is the set of points QuTotal number of feature points.
In the third step, the specific implementation steps of screening the feature points according to the distribution concentration degree are as follows:
and 3.1, selecting a circular detection window with the radius r equal to 15 pixels, and continuously moving the detection window by taking each unmatched feature point as the center of a circle.
And 3.2, counting the number N1 of unmatched feature points and the number N2 of all feature points in the window, and calculating the proportion of the unmatched feature points to the total number of the feature points, namely the value of N1/N2, wherein if N1 is larger than or equal to N and N1/N2 is larger than m, the feature points are reserved.
The reference formula of the step is as follows:
the obstacle feature point set Q can be obtained through the stepsobs。
In the fourth step, the preliminary grouping of the feature points according to the point spacing mainly includes the following steps:
and 4.1, determining the road direction. The inclination angle theta (theta epsilon [0, pi)) of the road section midline can be calculated according to the road midline data. If it isThe link direction is specified as the vertical direction. If it isThe link direction is specified to be the horizontal direction.
And 4.2, sequencing the feature points. The feature points of the road sections in the vertical direction are sorted according to the coordinate y, and the feature points of the road sections in the horizontal direction are sorted according to the coordinate x.
And 4.3, searching backwards from the first point, wherein the distance between two adjacent points does not exceed R, the two points belong to the same group until a certain point q is found, and if no other characteristic point exists in the distance range of R after the point q, the first group of search is completed. Taking the median of all the characteristic point coordinates in the group as the candidate coordinate point O of the obstaclei. And the next feature point is taken as the first point.
Step 4.4, repeating step 4.3 until Q is checkedobsAnd obtaining a candidate coordinate point set O of the obstacle by using all the characteristic points.
In the fifth step, the specific steps of refining the grouping result through similarity judgment are as follows:
and 5.1, acquiring a standard road sample, intercepting a road center line by using the obstacle candidate coordinate point set O obtained in the previous step, and intercepting the distance between a road section and a positioning point R/2 to avoid getting a road obstacle pixel point. Taking each point on the extracted road center line as the center, and taking its eight neighborhoods as a road sample, as shown in fig. 2. And (4) counting the mean value and the variance of all road samples, sequencing the samples according to the two statistics values, and taking the samples corresponding to the statistics mode as standard road surface samples.
Step 5.2, marking the property of the coordinate point, defining the advancing direction of the road as the coordinate axis forward direction, taking R/2 as the step length along the advancing direction of the road, and marking the property of the coordinate point at each coordinate point OiThe front side and the rear side of the road surface are respectively provided with a point, the similarity between the front side point and the rear side point and the standard road surface sample and the similarity between the front side point and the rear side point and the obstacle point are respectively compared, and different labels are given to the positioning points under four different conditions as shown in table 1.
TABLE 1 Properties of obstacle candidate points
And 5.3, merging, if the properties of two continuous coordinate points are +1 and-1 respectively along the advancing direction of the road, judging that the two points are points positioned at two sides of the same barrier, and taking the average value of the two coordinates as the positioning point of the barrier. Obtaining a final barrier positioning point set Robs:
In the formula ci,ci+1Respectively represent the coordinate points O of the obstaclei,Oi+1The nature of (c).
According to the method, in the sixth step, a multi-seed point area growing method is adopted, wherein all the positioning points and the neighborhood points with the difference smaller than a threshold value α from the standard road sample mean value are used as seed points, and the difference between the growing points and any seed points is smaller than a threshold value α, so that the growing condition is met.
In the step 3.2, the value of n is generally between 3 and 6; m is generally 0.5.
In step 4.3, the threshold R is set according to the project requirement and the image resolution, and the higher the precision requirement is, the smaller R is, otherwise, the larger R is.
In a further improvement of the present invention, in the seventh step, the value of the threshold α is generally between 30 and 70.
According to the scheme, the space remote sensing image and the sky remote sensing image are utilized, the feature point set of the road surface change area is obtained through a reverse feature point matching technology, the change feature point set is clustered according to the feature point distribution density and the adjacent point distance, the obstacles at different positions can be segmented, and then each obstacle is positioned and the area of each obstacle is calculated. The scheme is improved in aspects of obstacle segmentation, positioning, uniqueness judgment of positioning points of various obstacles and the like. The obstacle identification and positioning method has the advantages of being wide in retrieval range, free of image registration, high in automation degree, high in detection accuracy rate and the like. As the case shows: the missing detection rate of the road surface barrier is 0, the false detection rate is lower than 10%, the area estimation error is less than 10%, and the method can provide a basis for emergency and disaster relief route planning and road traffic engineering quantity estimation.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention;
FIG. 2 is a schematic diagram of the extraction of a standard road sample in step 4.1;
FIG. 3 is a schematic representation of the nature of the candidate points for the obstacle as identified in step 4.2;
FIG. 4 is a schematic diagram of the distribution of obstacles in a post-disaster image used in an embodiment of the present invention;
FIG. 5 is a diagram illustrating road buffer screening results according to an embodiment of the present invention;
FIG. 6 is a graph showing the results of the sliding window screening according to the embodiment of the present invention;
FIG. 7 is a graph of the grouping results of the neighboring dot spacing method according to an embodiment of the present invention;
fig. 8 is a graph showing the result of obstacle detection in the embodiment of the present invention.
Detailed Description
The invention will be described in further detail with reference to the following figures and detailed description for the purpose of promoting an understanding and appreciation of the invention, but the invention is not limited to the embodiments given.
Example 1: referring to fig. 1, a road surface obstacle detection method based on reverse feature matching, the method comprising the steps of:
step one, calculating SIFT feature point sets of remote sensing images before and after the road loss respectively, performing feature matching, and negating matching results to obtain a change information feature point set;
step two, screening by using a road buffer area to obtain a change information characteristic point set in a road area;
thirdly, screening out false change characteristic points and micro change characteristic points according to the distribution concentration of the characteristic points;
step four, preliminarily grouping the characteristic points according to the point intervals, and distinguishing the characteristic points belonging to different obstacles;
step five, detecting the primary grouping result obtained in the last step, and obtaining a final grouping result through similarity judgment;
and step six, obtaining the position information of each obstacle according to the grouping result, and obtaining the distribution range of the obstacles by combining a multi-seed point region growing method and carrying out area estimation.
In a further improvement of the present invention, in the step one, the inverse feature matching formula is:
Qu=CQQm(1)
point set QuIs a point set QmComplement in Q, i.e.Wherein Q is a characteristic point set of the post-disaster image, QmIs a set of feature points, Q, in the post-disaster image that can be matched with the pre-disaster image feature pointsuThe image change information feature point set is obtained.
In the second step, the existing road vector data is used for generating a road buffer area, and a change feature point set Q in the buffer area is screenedurThe formula is as follows:
where Poly is the road buffer polygon, i is 0,1, …, k, k is the set of points QuTotal number of feature points.
In the third step, the specific implementation steps for screening the feature points according to the distribution concentration degree are as follows:
and 3.1, selecting a circular detection window with the radius r equal to 15 pixels, and continuously moving the detection window by taking each unmatched feature point as the center of a circle.
And 3.2, counting the number N1 of unmatched feature points and the number N2 of all feature points in the window, and calculating the proportion of the unmatched feature points to the total number of the feature points, namely the value of N1/N2, wherein if N1 is larger than or equal to N and N1/N2 is larger than m, the feature points are reserved.
The reference formula of the step is as follows:
the obstacle feature point set Q can be obtained through the stepsobs。
In the fourth step, the primary grouping of the feature points according to the point spacing mainly comprises the following steps:
and 4.1, determining the road direction. The inclination angle theta (theta epsilon to 0 and pi) of the road section central line can be calculated according to the road central line data. If it isThe link direction is specified as the vertical direction. If it isThe link direction is specified to be the horizontal direction.
And 4.2, sequencing the feature points. The feature points of the road sections in the vertical direction are sorted according to the coordinate y, and the feature points of the road sections in the horizontal direction are sorted according to the coordinate x.
And 4.3, searching backwards from the first point, wherein the distance between two adjacent points does not exceed R, the two points belong to the same group until a certain point q is found, and if no other characteristic point exists in the distance range of R after the point q, the first group of search is completed. Taking the median of all the characteristic point coordinates in the group as the candidate coordinate point O of the obstaclei. And the next feature point is taken as the first point.
Step 4.4, repeating step 4.3 until Q is checkedobsAnd obtaining a candidate coordinate point set O of the obstacle by using all the characteristic points.
In the fifth step, the specific steps of judging the refined grouping result through the similarity are as follows:
and 5.1, acquiring a standard road sample, intercepting a road center line by using the obstacle candidate coordinate point set O obtained in the previous step, and intercepting the distance between a road section and a positioning point R/2 to avoid getting a road obstacle pixel point. Taking each point on the extracted road center line as the center, and taking its eight neighborhoods as a road sample, as shown in fig. 2. And (4) counting the mean value and the variance of all road samples, sequencing the samples according to the two statistics values, and taking the samples corresponding to the statistics mode as standard road surface samples.
Step 5.2, marking the property of the coordinate point, defining the advancing direction of the road as the coordinate axis forward direction, taking R/2 as the step length along the advancing direction of the road, and marking the property of the coordinate point at each coordinate point OiOne point is taken from each of the front and rear sides of the road surface, and the similarities between the front side point and the rear side point and the standard road surface sample and the obstacle point are compared respectively, as shown in fig. 3. Assigning localization points to four different situationsThe various labels are shown in table 1.
TABLE 1 Properties of obstacle candidate points
And 5.3, merging, if the properties of two continuous coordinate points are +1 and-1 respectively along the advancing direction of the road, judging that the two points are points positioned at two sides of the same barrier, and taking the average value of the two coordinates as the positioning point of the barrier. Obtaining a final barrier positioning point set Robs:
In the formula ci,ci+1Respectively represent the coordinate points O of the obstaclei,Oi+1The nature of (c).
And in the sixth step, a multi-seed point region growing method is adopted, wherein positioning points and neighborhood points with the difference between the standard road sample average value and the standard road sample average value smaller than a threshold value α are all used as seed points, the difference between the growing points and any seed points is smaller than a threshold value α, the growing condition is met, the distribution range of the obstacles can be obtained after the growing is finished, the number of pixels of each obstacle in the road region is counted, and the actual occupied area of each obstacle can be roughly estimated by combining the image resolution.
In the step 3.2, the value of n is generally between 3 and 6; m is generally 0.5.
In the step 4.3, the threshold R is set according to the project requirement and the image resolution, and the higher the precision requirement is, the smaller R is, otherwise, the larger R is.
In the seventh step, the value of the threshold α is generally between 30 and 70.
Application example 1: a method for detecting a road surface obstacle based on reverse feature matching, the method comprising the steps of: the technical flow chart of the invention is shown in figure 1. The key steps of the flow chart, fig. 1, will be described separately below.
Step one, using SIFT algorithm to detect two stages before and after disasterAnd (3) carrying out feature point matching on the image feature point sets (respectively marked as P and Q), reducing mismatching by using a ratio purification method and eliminating outlier matching by using an RANSAC algorithm to further purify a matching result in order to ensure the accuracy of subsequent obstacle feature point extraction. Obtaining a change information feature point set Q of the post-disaster image after reverse feature matchingu;
Step two, extracting a change information characteristic point set Q in the road area after screening the road buffer areaur. FIG. 4 is a result of extracting a road region change feature point set using a road ROI;
step three, according to the image resolution and the texture characteristics of the road area, the parameters of the sliding window screening method are r-15 pixels, n-5 (the number of unmatched points in the window is more than or equal to 5), m-0.5 (the number of unmatched points is more than half of the total number of the feature points), and fig. 6(a) and (b) are examples of the screening results of the sliding window, so that a plurality of feature point clusters with high aggregation degree are obtained;
step four, the case sets road obstacle points within 10m of the actual distance without being distinguished, and the threshold value R is 20 since the resolution of the experimental image is 0.5 m. Calculating to obtain the horizontal trend of the road from the road vector data, so as to collect the characteristic point set Q of the barrierobsSorting by coordinate x, and sorting Q by RobsThe image is divided into 15 groups, and 15 obstacle candidate points are obtained by calculating the median value in each group. 6(a) and (b) are examples of candidate points of the obstacle obtained by using an adjacent point interval grouping method;
step five, as can be seen from fig. 6(b), the detected feature points all belong to the same obstacle, but there are two candidate points after being grouped and positioned as shown in fig. 6(b), which need to be merged, and the results of property calibration of 15 candidate points according to the implementation steps of the similarity comparison method are shown in table 2:
TABLE 2 calibration results of candidate Point Properties
Through comparative statistics, 4 groups of coordinate points need to be merged. And taking the combined new coordinate value as the positioning point of the 4 obstacles, keeping the other 7 coordinate points unchanged, and finally obtaining the positioning point coordinates of the 11 obstacles in the table 3.
And sixthly, after the accurate barrier positioning points are obtained, the shape information of each barrier can be obtained through a multi-subregion growing method, and the white part in the graph 8 is the distribution range of each barrier. As can be seen from FIG. 8, the firstThe obstacle has no growth result because ofThe similarity between the neighborhood of the positioning point of the obstacle and the standard pavement sample is high, and no proper seed point is found, so that the area of the obstacle is set to be 0. The area of the other 10 obstacles is estimated by counting the number of pixels of the area growth result of the obstacles in the road area, and the calculation result is listed in table 3; the reference area in table 3 is an area calculated by using a boundary range of the obstacle defined by a human visual interpretation. The area is the area constructed by the boundary of the road buffer zone and the extracted boundary of the obstacle, and can be directly used as the reference of the amount of the earth and the stone for emergency rescue.
Table 3 obstacle extraction results
The embodiments described above are only preferred embodiments of the present invention, and are not intended to limit the present invention in any other way, and modifications or additions made to the technical spirit of the invention or the like can be made thereto without departing from the scope of the invention as claimed.
Claims (10)
1. A road surface obstacle detection method based on reverse feature matching is characterized by specifically comprising the following steps:
step one, calculating SIFT feature point sets of remote sensing images before and after the road loss respectively, performing feature matching, and negating matching results to obtain a change information feature point set;
step two, screening by using a road buffer area to obtain a change information characteristic point set in a road area;
thirdly, screening out false change characteristic points and micro change characteristic points according to the distribution concentration of the characteristic points;
step four, preliminarily grouping the characteristic points according to the point intervals, and distinguishing the characteristic points belonging to different obstacles;
step five, detecting the primary grouping result obtained in the last step, and obtaining a final grouping result through similarity judgment;
and step six, obtaining the position information of each obstacle according to the grouping result, and obtaining the distribution range of the obstacles by combining a multi-seed point region growing method and carrying out area estimation.
2. The method for detecting a road surface obstacle based on reverse feature matching according to claim 1, characterized in that: in step one, the inverse feature matching formula is Qu=CQQmSet of points QuIs a point set QmComplement in Q, i.e.Wherein Q is a feature point set of the post-disaster image, QmIs a set of feature points, Q, in the post-disaster image that can be matched with the pre-disaster image feature pointsuIs a feature point set of image change information.
3. The method for detecting road surface obstacles based on reverse feature matching according to claim 1, wherein in step two, the used road buffer is generated from the existing road vector data to obtain the changed feature point set Q in the road bufferurThe reference formula is:
where Poly is a road buffer polygon, i is 0,1, k, k is a pointQ collectionuTotal number of feature points.
4. The method for detecting the road surface obstacle based on the reverse feature matching as claimed in claim 1, wherein in the third step, the specific implementation steps of screening the feature points according to the distribution concentration degree are as follows:
step 3.1, selecting a circular detection window with the radius r equal to 15 pixels, and continuously moving the detection window by taking each unmatched feature point as the center of a circle;
step 3.2, counting the number N1 of unmatched feature points and the number N2 of all feature points in the window, calculating the proportion of the unmatched feature points in the total number of the feature points, namely the value of N1/N2, and if N1 is not less than N and N1/N2 is greater than m, reserving the feature points;
the reference formula of the step is as follows:
the obstacle feature point set Q can be obtained through the stepsobs。
5. The method for detecting road surface obstacles based on reverse feature matching according to claim 1, wherein in the fourth step, the preliminary grouping of the feature points according to the point spacing mainly comprises the following steps:
step 4.1, determining the road direction, calculating the inclination angle theta (theta belongs to [0, pi ]) of the road section central line according to the road central line data, and if the inclination angle theta belongs to [0, pi ])The direction of the link is defined as the vertical direction ifDefining the direction of the road section as a horizontal direction;
step 4.2, sorting the feature points, wherein the feature points of the road section in the vertical direction are sorted according to a coordinate y, and the feature points of the road section in the horizontal direction are sorted according to a coordinate x;
4.3, searching backwards from the first point, if the distance between two adjacent points does not exceed R, the two points belong to the same group until a certain point q is found, if no other characteristic point exists in the distance range of R after the point, the first group is searched, and the median of the coordinates of all the characteristic points in the group is taken as a candidate coordinate point O of the obstacleiAnd the next feature point is taken as the first point.
Step 4.4, repeating step 4.3 until Q is checkedobsAnd obtaining the obstacle candidate coordinate point set O by using all the characteristic points in the image.
6. The method for detecting the road surface obstacle based on the reverse feature matching as claimed in claim 1, wherein in the fifth step, the concrete steps of refining the grouping result through the similarity judgment are as follows:
step 5.1, acquiring a standard road sample, intercepting a road center line by using the obstacle candidate coordinate point set O obtained in the previous step, and intercepting the distance between a road section and a positioning point R/2 to avoid acquiring a road obstacle pixel point; taking each point on the taken road center line as a center, taking eight neighborhoods of the point as a road sample, counting the mean value and variance of all road samples, sequencing the samples according to the two statistics, and taking the sample corresponding to the statistics mode as a standard road surface sample;
step 5.2, marking the property of the coordinate point, defining the advancing direction of the road as the coordinate axis forward direction, taking R/2 as the step length along the advancing direction of the road, and marking the property of the coordinate point at each coordinate point OiRespectively taking one point from the front side and the rear side, respectively comparing the similarity between the front side point and the rear side point and the standard road surface sample and the similarity between the front side point and the rear side point and the obstacle point, and giving different labels to the positioning points aiming at four different conditions as shown in a table 1:
TABLE 1 Properties of obstacle candidate points
And 5.3, merging, namely judging that two continuous coordinate points are positioned if the properties of the two continuous coordinate points are +1 and-1 respectively along the advancing direction of the roadTaking the average value of the two coordinates as the positioning point of the obstacle at the point on the two sides of the same obstacle; obtaining a final barrier positioning point set Robs:
In the formula ci,ci+1Respectively represent the coordinate points O of the obstaclei,Oi+1The nature of (c).
7. The method for detecting the road surface obstacles based on the reverse feature matching as claimed in claim 1, wherein in the sixth step, a multi-seed point area growing method is adopted, wherein all the positioning points and the neighborhood points with the difference with the standard road sample mean value smaller than a threshold value α are used as seed points, the difference between the to-be-grown points and any seed points is smaller than a threshold value α, so that the growing condition is met, the distribution range of the obstacles can be obtained after the growing is finished, the number of pixels of each obstacle in the road area is counted, and the actual floor area of each obstacle can be approximately estimated by combining the image resolution.
8. The method for detecting the road surface obstacle based on the reverse feature matching as claimed in claim 4, wherein in the step 3.2, the value of n is between 3 and 6; m is 0.5.
9. The method for detecting the road surface obstacle based on the reverse feature matching as claimed in claim 5, wherein in step 4.3, the threshold R is set according to project requirements and image resolution, and the higher the precision requirement is, the smaller R is, otherwise, the larger R is.
10. The method for detecting the road surface obstacle based on the reverse feature matching as claimed in claim 7, wherein the threshold value α is between 30 and 70.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911061150.2A CN111091049B (en) | 2019-11-01 | 2019-11-01 | Road surface obstacle detection method based on reverse feature matching |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911061150.2A CN111091049B (en) | 2019-11-01 | 2019-11-01 | Road surface obstacle detection method based on reverse feature matching |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111091049A true CN111091049A (en) | 2020-05-01 |
CN111091049B CN111091049B (en) | 2024-02-09 |
Family
ID=70393483
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911061150.2A Active CN111091049B (en) | 2019-11-01 | 2019-11-01 | Road surface obstacle detection method based on reverse feature matching |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111091049B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111783700A (en) * | 2020-07-06 | 2020-10-16 | 中国交通通信信息中心 | Automatic recognition early warning method and system for road foreign matters |
CN111858342A (en) * | 2020-07-23 | 2020-10-30 | 深圳慕智科技有限公司 | Fuzzy test data generation method based on intelligent traffic image input feature recognition |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060013438A1 (en) * | 2004-07-13 | 2006-01-19 | Susumu Kubota | Obstacle detection apparatus and a method therefor |
CN103914688A (en) * | 2014-03-27 | 2014-07-09 | 北京科技大学 | Urban road obstacle recognition system |
WO2016076449A1 (en) * | 2014-11-11 | 2016-05-19 | Movon Corporation | Method and system for detecting an approaching obstacle based on image recognition |
CN109785307A (en) * | 2019-01-09 | 2019-05-21 | 武汉大学 | A kind of unmanned plane image road Damage assessment method based on vector guidance |
CN109815859A (en) * | 2019-01-10 | 2019-05-28 | 中交信息技术国家工程实验室有限公司 | A kind of inferior grade road automatically extracts and mutation analysis method |
-
2019
- 2019-11-01 CN CN201911061150.2A patent/CN111091049B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060013438A1 (en) * | 2004-07-13 | 2006-01-19 | Susumu Kubota | Obstacle detection apparatus and a method therefor |
CN103914688A (en) * | 2014-03-27 | 2014-07-09 | 北京科技大学 | Urban road obstacle recognition system |
WO2016076449A1 (en) * | 2014-11-11 | 2016-05-19 | Movon Corporation | Method and system for detecting an approaching obstacle based on image recognition |
CN109785307A (en) * | 2019-01-09 | 2019-05-21 | 武汉大学 | A kind of unmanned plane image road Damage assessment method based on vector guidance |
CN109815859A (en) * | 2019-01-10 | 2019-05-28 | 中交信息技术国家工程实验室有限公司 | A kind of inferior grade road automatically extracts and mutation analysis method |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111783700A (en) * | 2020-07-06 | 2020-10-16 | 中国交通通信信息中心 | Automatic recognition early warning method and system for road foreign matters |
CN111783700B (en) * | 2020-07-06 | 2023-11-24 | 中国交通通信信息中心 | Automatic recognition and early warning method and system for pavement foreign matters |
CN111858342A (en) * | 2020-07-23 | 2020-10-30 | 深圳慕智科技有限公司 | Fuzzy test data generation method based on intelligent traffic image input feature recognition |
Also Published As
Publication number | Publication date |
---|---|
CN111091049B (en) | 2024-02-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109949326B (en) | Building contour line extraction method based on knapsack type three-dimensional laser point cloud data | |
CN110378879A (en) | A kind of Bridge Crack detection method | |
CN112287807B (en) | Remote sensing image road extraction method based on multi-branch pyramid neural network | |
CN110008854B (en) | Unmanned aerial vehicle image highway geological disaster identification method based on pre-training DCNN | |
CN108197583A (en) | The building change detecting method of optimization and image structure feature is cut based on figure | |
CN108921076B (en) | Pavement crack disease self-adaptive constant false alarm detection method based on image | |
Bremer et al. | Eigenvalue and graph-based object extraction from mobile laser scanning point clouds | |
CN112801075B (en) | Automatic rural road boundary line extraction method based on aerial image | |
CN105069395B (en) | Roadmarking automatic identifying method based on Three Dimensional Ground laser scanner technique | |
CN111091049A (en) | Road surface obstacle detection method based on reverse feature matching | |
CN116563279B (en) | Measuring switch detection method based on computer vision | |
CN109187548B (en) | Rock mass fracture identification method | |
CN104636750A (en) | Pavement crack recognition algorithm method and system based on dual-scale clustering algorithm | |
CN115115627B (en) | Soil saline-alkali soil monitoring method based on data processing | |
CN117710399B (en) | Crack contour extraction method in geological survey based on vision | |
CN111091071A (en) | Underground target detection method and system based on ground penetrating radar hyperbolic wave fitting | |
CN113033386B (en) | High-resolution remote sensing image-based transmission line channel hidden danger identification method and system | |
CN107564015B (en) | A kind of segmentation and labeling method and device based on polychrome grade image connectivity domain | |
CN116718599A (en) | Apparent crack length measurement method based on three-dimensional point cloud data | |
CN112906627B (en) | Green pricklyash peel identification method based on semantic segmentation | |
CN115797649A (en) | Crack extraction method under complex background | |
CN104504702A (en) | Cement notch pavement crack recognition method based on square lattice searching method | |
Wijaya et al. | Building crack due to lombok earthquake classification based on glcm features and svm classifier | |
CN114139099A (en) | Road intersection information extraction method based on track density homogenization and hierarchical segmentation | |
Zhang et al. | Research on pavement marking recognition and extraction method |
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 |