CN108416044A - Scene thumbnail map generalization method, apparatus, electronic equipment and storage medium - Google Patents
Scene thumbnail map generalization method, apparatus, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN108416044A CN108416044A CN201810213001.2A CN201810213001A CN108416044A CN 108416044 A CN108416044 A CN 108416044A CN 201810213001 A CN201810213001 A CN 201810213001A CN 108416044 A CN108416044 A CN 108416044A
- Authority
- CN
- China
- Prior art keywords
- point
- cluster
- dimensional coordinate
- deflection
- thumbnail
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- 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
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Remote Sensing (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention discloses a kind of scene thumbnail map generalization method, apparatus, electronic equipment and storage medium, the method includes：According to each collecting device when target scene moves acquired image data and exercise data, determine three-dimensional track point of each sample devices in the target scene；Obtain the two-dimensional coordinate point and deflection of each three-dimensional track point；According to the deflection of the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point, each two-dimensional coordinate point is clustered, each cluster point is obtained；It carries out curve fitting to each cluster point, generates the thumbnail of the target scene, for being inconvenient to take photo by plane, or the target scene that GPS is not covered with, the thumbnail of target scene can be generated, method is simple, operand is small, and quickly generating for thumbnail may be implemented.
Description
Technical field
The present embodiments relate to map acquisition technique field more particularly to a kind of scene thumbnail map generalization method, dresses
It sets, electronic equipment and storage medium.
Background technology
With the fast development of network technology, the communication technology and geographic information system technology, the thumbnail of target scene
It is real-time acquisition be particularly important.For example, whens vehicle enters parking lot, cell etc., the contracting of parking lot or cell can be passed through
Sketch map quickly grasps the planning of parking lot or cell, facilitates vehicle safe driving, destination is quickly found out convenient for user.
The acquisition methods of the prior art, target scene thumbnail are mainly, and by remote sensing or take photo by plane and obtain target scene
Image, from the extracting target from images scene thumbnail figure of shooting.Alternatively, by collecting in target scene, a large amount of moving objects
GPS (Global Positioning System, global positioning system) location data, these GPS positioning data are transported
Calculation is handled, and generates target scene thumbnail.
But the prior art, cost is higher, and is inconvenient to the place taken photo by plane or GPS coverings for underground parking etc.
Less than place, its thumbnail can not be obtained.
Invention content
A kind of scene thumbnail map generalization method, apparatus of offer of the embodiment of the present invention, electronic equipment and storage medium, with solution
The certainly prior art, cost is higher, and is inconvenient to the place taken photo by plane for underground parking etc. or place that GPS is not covered,
The problem of its thumbnail can not be obtained.
In a first aspect, the embodiment of the present invention provides a kind of scene thumbnail map generalization method, including：
According to each collecting device when target scene moves acquired image data and exercise data, determine each described adopt
Three-dimensional track point of the sample equipment in the target scene；
Obtain the two-dimensional coordinate point and deflection of each three-dimensional track point；
According to the deflection of the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point, to each two dimension
Coordinate points are clustered, and each cluster point is obtained；
It carries out curve fitting to each cluster point, generates the thumbnail of the target scene.
In a kind of possible realization method of first aspect, the two-dimensional coordinate point for obtaining each three-dimensional track point and side
To angle, specifically include：
By subpoint of the three-dimensional track point on two dimensional surface, the two-dimensional coordinate point as the three-dimensional track point；
By the line direction of the two-dimensional coordinate point and the two-dimensional coordinate point of neighbouring sampling instant, as the two-dimensional coordinate
The deflection of point.
In the alternatively possible realization method of first aspect, the two dimensional surface is apart from each three-dimensional track point
Nearest plane.
It is described according to the distance between each described two-dimensional coordinate point in the alternatively possible realization method of first aspect
With the deflection of each two-dimensional coordinate point, each two-dimensional coordinate point is clustered, each cluster point is obtained, specifically includes：
Determine adjacent the distance between two two-dimensional coordinate points；
The sum of the distance is less than default clustering distance, and the deflection be located at it is each in preset direction angle range
Adjacent two-dimensional coordinate point, as a two-dimentional point set；
Determine the deflection of each two-dimentional the point set corresponding cluster point and each cluster point.
In the alternatively possible realization method of first aspect, the corresponding cluster of each two-dimentional point set of the determination
The deflection of point and each cluster point, specifically includes：
Using the central point of the two-dimentional point set as the cluster point, by each two-dimensional coordinate point in two-dimentional point set
Deflection deflection of the average value as the cluster point, wherein the distance between adjacent two clusters point is described pre-
If clustering distance.
In the alternatively possible realization method of first aspect, the corresponding cluster of each two-dimentional point set of the determination
The deflection of point and each cluster point, specifically includes：
Using in the two-dimentional point set, first two-dimensional coordinate point is as the cluster point, by first two-dimensional coordinate
Deflection of the deflection of point as the cluster point, wherein the distance between 2 adjacent cluster points are the default cluster
Distance.
It is described to carry out curve fitting to each cluster point in the alternatively possible realization method of first aspect, it is raw
At the thumbnail of the target scene, specifically include：
Each cluster point between adjacent two turning point is classified as a fitting set, wherein the turning point is each described
In cluster point, the difference of the deflection and the deflection of adjacent cluster point is more than the cluster point of preset value；
It carries out curve fitting to each fitting set, obtains each fitting and gather corresponding matched curve；
Each matched curve is attached, the thumbnail of the target scene is generated.
It is described to carry out curve fitting to each fitting set in the alternatively possible realization method of first aspect,
It specifically includes：
Fitting a straight line is carried out to each fitting set；
Alternatively, carrying out quadratic fit to each fitting set.
It is described to be attached each matched curve in the alternatively possible realization method of first aspect, it generates
Before the thumbnail of the target scene, the method further includes：
By at a distance of the opposite matched curve of pre-determined distance and direction is less than, a matched curve is merged into.
It is described to be attached each matched curve in the alternatively possible realization method of first aspect, it generates
The thumbnail of the target scene, specifically includes：
If next cluster point of the target cluster point in the first matched curve is not in first matched curve, really
The second matched curve where fixed next cluster point, wherein first matched curve is in each matched curve
Any curve；
Determine that the target cluster point, next cluster point and first matched curve are intended with described second
Close the triangle that the intersection point of curve is constituted；
If being carried out to first matched curve and second matched curve without matched curve in the triangle
Fitting connection, generates the thumbnail of the target scene.
Second aspect, the embodiment of the present invention provide a kind of scene thumbnail map generalization device, including：
Determining module, for according to each collecting device when target scene move acquired image data and move number
According to determining three-dimensional track point of each sample devices in the target scene；
Acquisition module, two-dimensional coordinate point and deflection for obtaining each three-dimensional track point；
Cluster module, for the direction according to the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point
Angle clusters each two-dimensional coordinate point, obtains each cluster point；
Generation module generates the thumbnail of the target scene for carrying out curve fitting to each cluster point.
In a kind of possible realization method of second aspect, the acquisition module is specifically used for the three-dimensional track
Subpoint of the point on two dimensional surface, the two-dimensional coordinate point as the three-dimensional track point；By the two-dimensional coordinate point with close on
The line direction of the two-dimensional coordinate point of sampling instant, the deflection as the two-dimensional coordinate point.
In the alternatively possible realization method of second aspect, the two dimensional surface is apart from each three-dimensional track point
Nearest plane.
In the alternatively possible realization method of second aspect, the cluster module includes：
Determination unit, for determining adjacent the distance between two two-dimensional coordinate points；
Cluster cell, for the sum of the distance to be less than default clustering distance, and the deflection is located at preset side
Each adjacent two-dimensional coordinate point into angular region, as a two-dimentional point set；
The determination unit is additionally operable to determine the direction of each two-dimentional the point set corresponding cluster point and each cluster point
Angle.
In the alternatively possible realization method of second aspect, the determination unit is specifically used for the two-dimensional points
The central point of set as the cluster point, using the average value of the deflection of each two-dimensional coordinate point in two-dimentional point set as
The deflection of the cluster point, wherein the distance between 2 adjacent cluster points are the default clustering distance.
In the alternatively possible realization method of second aspect, the determination unit, also particularly useful for by the two dimension
First two-dimensional coordinate point is as the cluster point in point set, using the deflection of first two-dimensional coordinate point as described in
Cluster the deflection of point, wherein the distance between 2 adjacent cluster points are the default clustering distance.
In the alternatively possible realization method of second aspect, the generation unit includes：
Sort out unit, for each cluster point between adjacent two turning point to be classified as a fitting set, wherein described turn
It is in each cluster point to point, the difference of the deflection and the deflection of adjacent cluster point is more than the cluster of preset value
Point；
It is corresponding quasi- to obtain each fitting set for carrying out curve fitting to each fitting set for fitting unit
Close curve；
Connection unit generates the thumbnail of the target scene for each matched curve to be attached.
In the alternatively possible realization method of second aspect, fitting unit is specifically used for gathering each fitting
Carry out fitting a straight line；
Alternatively, carrying out quadratic fit to each fitting set.
In the alternatively possible realization method of second aspect, the fitting unit, being additionally operable to will be default at a distance of being less than
The opposite matched curve of distance and direction, merges into a matched curve.
In the alternatively possible realization method of second aspect, connection unit, if specifically in the first matched curve
Target cluster point next cluster point not in first matched curve, it is determined that it is described it is next cluster point where
Second matched curve, and determine target cluster point, next cluster point and first matched curve with it is described
The triangle that the intersection point of second matched curve is constituted；If without matched curve in the triangle, to first fitting
Curve and second matched curve are fitted connection, generate the thumbnail of the target scene, wherein first fitting
Curve is any curve in each matched curve.
The third aspect, the embodiment of the present invention provide a kind of electronic equipment, including：
Memory, for storing computer program；
Processor, for executing the computer program, to realize the scene thumbnail map generalization side described in first aspect
Method.
Fourth aspect, the embodiment of the present invention provide a kind of computer storage media, computer are stored in the storage medium
Program, the computer program realize scene thumbnail map generalization method described in first aspect when being executed.
The embodiment of the present invention has the beneficial effect that：
According to each collecting device when target scene moves acquired image data and exercise data, determine each described adopt
Three-dimensional track point of the sample equipment in the target scene obtains the two-dimensional coordinate point and deflection of each three-dimensional track point,
According to the deflection of the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point, each two-dimensional coordinate is clicked through
Row cluster, obtains each cluster point, carries out curve fitting to each cluster point, generate the thumbnail of the target scene.In this way,
For the target scene for being inconvenient to take photo by plane or GPS is not covered with, its thumbnail can be generated using the method for the present embodiment,
And the method for the present embodiment is simple, operand is small, and quickly generating for thumbnail may be implemented.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is this hair
Some bright embodiments for those of ordinary skill in the art without having to pay creative labor, can be with
Obtain other attached drawings according to these attached drawings.
Fig. 1 is the flow chart for the scene thumbnail map generalization method that the embodiment of the present invention one provides；
Fig. 2 is the two-dimensional coordinate point distribution map that the embodiment of the present invention one is related to；
Fig. 3 is the cluster point distribution map that the embodiment of the present invention one is related to；
Fig. 4 is the thumbnail for the target scene that the embodiment of the present invention one is related to；
Fig. 5 is the flow chart of scene thumbnail map generalization method provided by Embodiment 2 of the present invention；
Fig. 6 is the cluster point distribution map that the embodiment of the present invention two is related to；
Fig. 7 is the flow chart for the scene thumbnail map generalization method that the embodiment of the present invention three provides；
Fig. 8 is the matched curve schematic diagram that the embodiment of the present invention three is related to；
Fig. 9 is another schematic diagram of matched curve that the embodiment of the present invention three is related to；
Figure 10 is the thumbnail for the target scene that the embodiment of the present invention three is related to；
Figure 11 is the structural schematic diagram for the scene thumbnail map generalization device that the embodiment of the present invention one provides；
Figure 12 is the structural schematic diagram of scene thumbnail map generalization device provided by Embodiment 2 of the present invention；
Figure 13 is the structural schematic diagram for the scene thumbnail map generalization device that the embodiment of the present invention three provides；
Figure 14 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
The every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
The method of the present embodiment is suitable for map planning, road topology, contour of object simplification etc. and needs to obtain thumbnail
Field.
The technical solution of the present embodiment acquires image data and fortune using the camera and sensor of collecting device itself
Dynamic data, and according to the image data and exercise data of acquisition, three-dimensional track point of each sample devices in target scene is generated,
These three-dimensional track points are converted into two-dimensional coordinate point, and cluster and curve matching are carried out to transformed two-dimensional coordinate point, it is raw
At the thumbnail of target scene.In this way, for the target scene for being inconvenient to take photo by plane or GPS is not covered with, this reality can be used
The method for applying example generates its thumbnail, and the method for the present embodiment is simple, and operand is small, and the fast fast-growing of thumbnail may be implemented
At.
Technical scheme of the present invention is described in detail with specifically embodiment below.These specific implementations below
Example can be combined with each other, and same or analogous concept or process may be repeated no more in some embodiments.
Fig. 1 is the flow chart for the scene thumbnail map generalization method that the embodiment of the present invention one provides, as shown in Figure 1, this reality
The method for applying example may include：
S101, according to each collecting device when target scene moves acquired image data, determine it is each it is described sampling set
The standby three-dimensional track point in the target scene.
The executive agent of the present embodiment is any electronic equipment with data-handling capacity, which includes at least
Memory and processor.The electronic equipment can be individual equipment, for example, thumbnail generating apparatus, optionally, the electronics
Equipment can also be integrated in other products, such as be integrated on vehicle, as a part for vehicle, for example, on vehicle
Car-mounted device.
Optionally, which can also have display function, can show the breviary of the target scene ultimately produced
Figure.For example, electronic equipment is the car-mounted device with display function, in this way, the user in target scene can be by this
Car-mounted device knows the thumbnail of target scene, drives vehicle convenient for user, and then improve the driving experience of user.Meanwhile
The method of the present embodiment is also convenient for realizing unmanned.
Optionally, the electronic equipment of the present embodiment can also be network-side processor, such as cloud processor.
The collecting device of the present embodiment can be vehicle, robot etc., be equipped on the vehicle or robot camera and
Sensor etc., wherein sensor include mainly inertial navigation device, and odometer etc., camera can be monocular cam, binocular
Camera or more mesh cameras.
The collecting device of the present embodiment can be the vehicle or robot for signing crowdsourcing agreement, which has reward
Mechanism can obtain corresponding remuneration when collecting device completes acquisition tasks.More collecting devices can be encouraged to add in this way
Enter into acquisition tasks.
In the same target area, multiple collecting devices can be dominated and carry out Image Acquisition, can formed so a large amount of
Image data and exercise data.
During actual acquisition, collecting device shoots target scene using the camera that itself is installed, and generates
The movement number of the image data of target scene, the velocity sensor installed using itself and inertial navigation sensors to collecting device
According to being acquired.Wherein, exercise data includes mileage information, the velocity information of collecting device and the direction information of collecting device
Deng optionally, exercise data also includes the acceleration information, angular velocity information and the compass information that are provided by inertial navigation device
Deng.
Wherein, above-mentioned image data and exercise data use identical timestamp, or are ensured using hardware trigger mechanism
The time synchronization of image data and exercise data.
The collecting device of the present embodiment can be connect with electronic equipment wireless telecommunications, for example, by 3G/4G cellular communications,
WIFI (Wireless Fidelity, Wireless Fidelity) wireless telecommunications such as equipment or bluetooth equipment connect.Optionally, collecting device
Wired connection can be carried out with electronic equipment, for example, electronic equipment is car-mounted device, which passes through vehicle bus and vehicle
Camera on is connected with sensor.
In this way, the image data of acquisition and exercise data can be sent to electronic equipment by collecting device, for example, acquisition is set
The standby image data that will be acquired and exercise data real-time delivery electron equipment, alternatively, collecting device is by the image data of acquisition
Local cache is first carried out with exercise data, after to be collected, then the image data of caching and exercise data is sent into electron
Equipment.
Then, using existing map planing method, for example, SLAM (Simultaneous LocationAnd
Mapping, immediately positioning and map structuring), (Structure from Motion, restore three dimensional field to SfM from movable information
Scape structure) method or multi-sensor fusion technology, the image data and exercise data of the acquisition of each collecting device are handled, it is raw
Location point (i.e. three-dimensional coordinate point) at each sample devices in different moments in target scene, using these location points as respectively adopting
Collect three-dimensional track point of the equipment in target scene.
The method of the present embodiment acquires the picture number of target scene using the camera and sensor of collecting device itself
According to, it is compared compared with prior art using remote sensing or take photo by plane to acquire image data, it is at low cost.Meanwhile for inconvenient remote sensing
Or the place taken photo by plane, such as underground parking or the unlapped places GPS, the method for the present embodiment can effectively obtain
The thumbnail in these above-mentioned places, i.e. the method use scope of the present embodiment is wide, and process is simple.
S102, the two-dimensional coordinate point and deflection for obtaining each three-dimensional track point.
The thumbnail of the present embodiment is the two-dimensional projection of target scene, as such, it is desirable to each three-dimensional that above-mentioned steps are obtained
Tracing point is converted into two-dimensional coordinate point, for example, three-dimensional track point A (x1, y1, z1), about the two-dimensional coordinate point on X/Y plane
It is a2 (x1, z1) about the two-dimensional coordinate point in XZ planes, about the two-dimensional coordinate point in YZ planes for a1 (x1, y1)
For a3 (y1, z1).
Wherein, three-dimensional track point has sequencing, i.e. three-dimensional track point includes not only location information, also includes direction
Information.For example, in vehicle-mounted preceding viewing system, each three-dimensional track point being calculated represents vehicle in different moments in target field
The sequencing of position in scape, three-dimensional track point represents the direction of motion of the vehicle in target scene.
In a kind of example, by subpoint of each three-dimensional track point on two dimensional surface, as the three-dimensional track
The two-dimensional coordinate point of point；
By the line direction of the two-dimensional coordinate point and the two-dimensional coordinate point for closing on sampling instant, as the two-dimensional coordinate
The deflection of point.
The two dimensional surface of the present embodiment can be any plane.
Optionally, which is horizontal plane, such as ground.
Optionally, which is to be distributed most planes apart from each three-dimensional track point, i.e., apart from each three-dimensional track point
It is distributed nearest plane.
The three-dimensional track point being distributed on the horizontal planes of usual ground or desktop etc. is most, it is preferred, therefore, that the two of the present embodiment
Dimensional plane is the horizontal plane nearest apart from each three-dimensional track point.
The present embodiment is not limited the position of two dimensional surface, is determined with specific reference to actual needs.
After determining two dimensional surface according to the above method, each three-dimensional track point is projected on the two dimensional surface, each in this way three
Dimension tracing point forms two-dimensional projection on the two dimensional surface, using projection of each three-dimensional track point on the two dimensional surface as each three
The two-dimensional coordinate point of tracing point is tieed up, it is specific as shown in Figure 2.
In the present embodiment, electronic equipment can obtain the sampling time of each three-dimensional track point, in this way, can be by current two dimension
The line direction of coordinate points two-dimensional coordinate point corresponding with sampling instant is closed on, the deflection as current two-dimensional coordinate point.
For example, current two-dimensional coordinate point is b1, then the two-dimensional coordinate of the corresponding three-dimensional track point of next sampling instant is
B2, in this way by the line direction of b1 and b2, i.e.,Deflection as two-dimensional coordinate point b1.
S103, according to the deflection of the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point, to each institute
It states two-dimensional coordinate point to be clustered, obtains each cluster point.
Specifically, each two-dimensional coordinate point generated to above-mentioned steps clusters, cluster point as shown in Figure 3 is generated.
In a kind of example, according to the distance between each two-dimensional coordinate point and direction, each two-dimensional coordinate point is ranked up,
Then, by the adjacent two-dimensional coordinate o'clock of preset quantity as a cluster point.
Optionally, it can also be according to preset distance (or preset interval), by each two-dimensional coordinate point of sequence, phase
Some two-dimensional coordinate o'clock away from pre-determined distance (or preset interval) is as a cluster point.
It is pre-determined distance by mutual distance in another example, and each two-dimensional coordinate that direction is identical or 180 degree is opposite
Point is as a cluster point.
In another example, each two-dimensional coordinate point is divided using preset length of window, a group will be divided into
Interior each two-dimensional coordinate o'clock is as a cluster point.
Optionally, the present embodiment can also be adopted with other methods, clustered to each two-dimensional coordinate point, obtain each cluster
Point.
In the present embodiment, since the data volume of the three-dimensional track point of acquisition is more, the data of the two-dimensional coordinate point of projection
Amount is also larger, in this way, in order to reduce computation burden, improves data processing speed, is then clustered to two-dimensional coordinate point, will be multiple
Two-dimensional coordinate point is reduced to a cluster point, greatly reduces data volume, is convenient for subsequent curve matching.
S104, it carries out curve fitting to each cluster point, generates the thumbnail of the target scene.
Specifically, after obtaining cluster point shown in Fig. 3 according to above-mentioned steps, carry out curve fitting, generates to each cluster point
The thumbnail of target scene as shown in Figure 4.
Optionally, line fitting method, such as one-variable linear regression algorithm or least square method etc., fitting can be used each
Point is clustered, the matched curve of each cluster point is obtained, these matched curves constitute the thumbnail of target scene.
Optionally, curve-fitting method, such as least square method, Moving Least, cubic spline can also be used
Function method or Lagrange's interpolation etc. are fitted each cluster point, obtain the matched curve of each cluster point, these matched curves are constituted
The thumbnail of target scene.
For example, the target scene of the present embodiment is parking lot, then the thumbnail that the present embodiment is formed can be parking lot
Road and parking stall etc..
The method of the present embodiment carries out curve fitting to each cluster point, generates the thumbnail of target scene, whole process
Simply, processing speed is fast, and result of calculation is accurate.
Scene thumbnail map generalization method provided in an embodiment of the present invention, according to each collecting device when target scene moves
Acquired image data and exercise data determine three-dimensional track point of each sample devices in the target scene, obtain
The two-dimensional coordinate point and deflection for taking each three-dimensional track point, according to the distance between each described two-dimensional coordinate point and each described
The deflection of two-dimensional coordinate point clusters each two-dimensional coordinate point, obtains each cluster point, is carried out to each cluster point
Curve matching generates the thumbnail of the target scene.In this way, for the target field for being inconvenient to take photo by plane or GPS is not covered with
Scape can generate its thumbnail using the method for the present embodiment, and the method for the present embodiment is simple, and operand is small, may be implemented
Thumbnail quickly generates.
Fig. 5 is the flow chart of scene thumbnail map generalization method provided by Embodiment 2 of the present invention.In above-described embodiment
On the basis of, the present embodiment refers to the direction according to the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point
Angle clusters each two-dimensional coordinate point, obtains the detailed process of each cluster point.As shown in figure 5, the present embodiment can wrap
It includes：
S201, adjacent the distance between two two-dimensional coordinate points are determined.
The present embodiment can calculate the distance between two neighboring two-dimensional coordinate point by Euclidean distance.
For example, adjacent two two-dimensional coordinate point b1=(x1, y1), b2=(x2, y2), then the distance between b1 and b2
In this way, adjacent the distance between each two-dimensional coordinate point can be obtained according to the above method.
S202, the sum of the distance is less than default clustering distance, and the deflection error is located at preset deflection
Each adjacent two-dimensional coordinate point in error range, as a two-dimentional point set.
S203, the deflection for determining each two-dimentional the point set corresponding cluster point and each cluster point.
It illustrates, it is assumed that default clustering distance is dk, preset deflection error range ± 1 °, each two-dimensional coordinate point root
It is arranged as according to the sequencing of sampling instant：b1、b2、b3…..bn.Since first two-dimensional coordinate point b1, b1 and second
The distance between two-dimentional sampled point b2 is d1, d1<Dk, deflection are 80 °；Between the two-dimentional sampled point b3 of b2 and third away from
From for d2, (d1+d2)<Dk, deflection are 80.5 °, 80.5 ° -80 °=0.5 °<1°；B3 and the 4th two-dimentional sampled point b4 it
Between distance be d3, (d1+d2+d3)>Dk, deflection are 80.4 °, 80.4 ° -80.5 °=- 0.1 °<1°.
It can be seen from the above, the sum of the distance between adjacent two-dimensional coordinate point b1, b2, b3 (d1+d2+d3) are less than default gather
Class distance dk, and the error of the deflection of b1, b2 and b3 is in preset direction angle error ± 1 °, in this way can sit two dimension
Punctuate b1, b2, b3 are as a two-dimentional point set B.
It is then determined the cluster point of the two dimension point set, and each deflection for clustering point.
Then, it using above-mentioned cluster point as starting point, using default clustering distance dk as distance, is inserted in subsequent two-dimensional coordinate point
Enter another cluster point, in this way so that the distance of adjacent two clusters point is identical, is equal to default clustering distance dk.
Wherein, above-mentioned S203 can also be realized according to following manner：
In a kind of example, using any one two-dimensional coordinate point in two-dimentional point set as the cluster of the two dimension point set
Point, meanwhile, using the corresponding deflection of two-dimensional coordinate point as the deflection of the cluster point.
For example, with continued reference to above-mentioned example, any one two-dimensional coordinate point b3 in two-dimentional point set B (b1, b2, b3) is made
For the cluster point of two dimension point set B, meanwhile, the deflection by the corresponding deflections of the b3 (i.e. 80.4 °) as the cluster point.
It, will be in two-dimentional point set using the central point of the two-dimentional point set as the cluster point in another example
Deflection of the average value of the deflection of each two-dimensional coordinate point as the cluster point, wherein 2 adjacent cluster points
Between distance be the default clustering distance.
For example, with continued reference to above-mentioned example, using the coordinate midpoint bm of three points in two-dimentional point set B (b1, b2, b3) as
The cluster point of two dimension point set B, meanwhile, using the average value of the deflection of tri- points of b1, b2, b3 as the direction of the cluster point
Angle.
Alternatively, using the central point b2 in two-dimentional point set B (b1, b2, b3) as the cluster point of two dimension point set B, together
When, the deflection by the corresponding deflections of the b2 (i.e. 80.5 °) as the cluster point.
In another example, using in the two-dimentional point set, first two-dimensional coordinate point is as the cluster point, by institute
State the deflection of the deflection of first two-dimensional coordinate point as the cluster point, wherein between 2 adjacent cluster points away from
From for the default clustering distance.
For example, with continued reference to above-mentioned example, first two-dimensional coordinate point b1 in two-dimentional point set B (b1, b2, b3) is made
For the cluster point of two dimension point set B, meanwhile, the deflection by the corresponding deflections of the b1 (i.e. 80 °) as the cluster point.
I.e. the present embodiment can obtain cluster point distribution map as shown in FIG. 6 according to the above method.
Scene thumbnail map generalization method provided in an embodiment of the present invention, by two adjacent two-dimensional coordinate points of determination it
Between distance, the sum of the distance is less than default clustering distance, and the deflection is located in preset direction angle range
Each adjacent two-dimensional coordinate point, as a two-dimentional point set；Determine each two-dimentional corresponding cluster point of point set and each cluster
The deflection of point, and then realize the accurate cluster to each two-dimensional coordinate point.
Fig. 7 is the flow chart for the scene thumbnail map generalization method that the embodiment of the present invention three provides.In above-described embodiment
On the basis of, the present embodiment refers to carry out curve fitting to each cluster point, generates the thumbnail of the target scene
Detailed process.As shown in fig. 7, the present embodiment may include：
S301, each cluster point between adjacent two turning point is classified as to a fitting set, gathered as fitting, wherein
The turning point is in each cluster point, and the difference of deflection and the deflection of adjacent cluster point is more than the cluster of preset value
Point.
Specifically, as shown in fig. 6, the deflection difference of the deflection and adjacent cluster point positioned at the cluster point of corner
It is larger, each cluster point can be traversed in this way, compares the deflection of cluster point, if adjacent thereto next of certain cluster point deflection
A cluster point is compared to big variation is had occurred, for example, the side of the deflection and adjacent next cluster point bk+1 of cluster point bk
It has been more than preset value c to the difference at angle, then has can determine that cluster point bk is turning point.
According to the above method, the turning point in each cluster point can be obtained, each steering can be clicked through in order to facilitate identification
Line flag.
At this point, the direction of each cluster point between two adjacent turning points is almost the same, it can be by two neighboring steering
Each cluster point between point is gathered as a fitting.
In this way, according to the above method, each cluster point is divided into each fitting set, obtains multiple fitting set.
S302, it carries out curve fitting to each fitting set, obtains each fitting and gather corresponding matched curve.
Specifically, as unit of a fitting set, carry out curve fitting to each cluster point in each fitting set, example
Such as, fitting a straight line is carried out, or is carried out curve fitting, each fitting is generated and gathers corresponding matched curve.
Optionally, it if the curve error after fitting is excessive, is fitted respectively again after being segmented.
Fig. 8 is to carry out fitting a straight line, multiple matched curves of generation to cluster point shown in fig. 6.
In a kind of possible realization method of the present embodiment, by a distance of the opposite matched curve of pre-determined distance and direction,
Merge into a matched curve.
Specifically, as shown in figure 8, due to the direction angle range of point is [0,360), so in curve matching, deflection a
(a<180) the cluster point with the deflection of a+180 or so is considered as both direction, is fitted respectively.And in real road,
A and a+180 can be considered two different travel directions of same path, so 180 or so can will be differed apart from close and direction
Curve merge, be merged into a curve.
As shown in figure 8, the distance between matched curve 1 and matched curve 2 are less than pre-determined distance, and matched curve 1 with it is quasi-
The deflection of curve 2 is closed on the contrary, in this way, matched curve 1 and matched curve 2 can be merged, generates fitting shown in Fig. 9
Curve.
Wherein, matched curve 1 and matched curve 2 merge, specifically, by each cluster point in matched curve 1 and quasi-
Each cluster point closed on curve 2 is fitted again, generates a matched curve.
That is the method for the present embodiment can be used for detecting two-way lane information.
S303, each matched curve is attached, generates the thumbnail of the target scene.
Specifically, according to above-mentioned steps, matched curve as shown in Figure 9 can be obtained, then, by these matched curves into
Row connection, generates the thumbnail of target scene as shown in Figure 10.
For example, each matched curve in Fig. 9 is carried out extension intersection, thumbnail shown in Fig. 10 is formed.
In a kind of example, curve connection can be carried out according to following manner：
If next cluster point of the target cluster point on S3031, the first matched curve is not in first matched curve
On, it is determined that the second matched curve where next cluster point, wherein first matched curve is each fitting
Any curve in curve.
S3032, determine target cluster point, next cluster point and first matched curve with it is described
The triangle that the intersection point of second matched curve is constituted.
If bent to first matched curve and second fitting without matched curve in S3033, the triangle
Line is fitted connection, generates the thumbnail of the target scene.
For example, in each matched curve shown in Fig. 9, conduct the first matched curve L1 is chosen, in L1
Each cluster presses order traversal, if some cluster point p1 (the cluster point is denoted as target and clusters point) thereon, next
For a cluster point p2 not on first matched curve L1, then the matched curve where inquiring cluster point p2 is the second matched curve
L2。
It is then determined the explanation p of the first matched curve L1 and the second matched curve L2, explanation p are close to p1's and p2
Intersection point.
It determines the triangle Δ p1p2p that p1, p2 and p are constituted, and judges bent with the presence or absence of other fittings in Δ p1p2p
Line will be based on the direction of p1, p2 and p1 and p2, L1 is carried out curve fitting with L2 and is connect if not having.If in Δ p1p2p
There are other matched curves, illustrate p1, connected curve is had existed between p2, need not be again coupled to.
Scene thumbnail map generalization method provided in an embodiment of the present invention will respectively cluster a little between adjacent two turning point, make
For a fitting set, carries out curve fitting to each fitting set, obtains each fitting and gather corresponding matched curve,
Each matched curve is attached, generates the thumbnail of the target scene, and then realize the thumbnail to target scene
Accurately generate.
Figure 11 is the structural schematic diagram for the scene thumbnail map generalization device that the embodiment of the present invention one provides, such as Figure 11 institutes
Show, the scene thumbnail map generalization device 100 of the present embodiment may include：
Determining module 110 is used for acquired image data and movement when target scene moves according to each collecting device
Data determine three-dimensional track point of each sample devices in the target scene；
Acquisition module 120, two-dimensional coordinate point and deflection for obtaining each three-dimensional track point；
Cluster module 130, for the side according to the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point
To angle, each two-dimensional coordinate point is clustered, each cluster point is obtained；
Generation module 140 generates the thumbnail of the target scene for carrying out curve fitting to each cluster point.
In a kind of possible realization method of the present embodiment, the two-dimensional coordinate point for obtaining each three-dimensional track point and side
To angle, specifically include：
The acquisition module 120 is specifically used for the subpoint by the three-dimensional track point on two dimensional surface, as described
The two-dimensional coordinate point of three-dimensional track point；By the line direction of the two-dimensional coordinate point and the two-dimensional coordinate point for closing on sampling instant,
Deflection as the two-dimensional coordinate point.
In the alternatively possible realization method of the present embodiment, the two dimensional surface is apart from each three-dimensional track point
Nearest plane.
Figure 12 is the structural schematic diagram of scene thumbnail map generalization device provided by Embodiment 2 of the present invention, such as Figure 12 institutes
Show, on the basis of the above embodiments, above-mentioned cluster module 130 includes：
Determination unit 131, for determining adjacent the distance between two two-dimensional coordinate points；
Cluster cell 132, for the sum of the distance to be less than default clustering distance, and the deflection is positioned at preset
Each adjacent two-dimensional coordinate point in direction angle range, as a two-dimentional point set；
The determination unit 132 is additionally operable to determine the side of each two-dimentional the point set corresponding cluster point and each cluster point
To angle.
In a kind of possible realization method of the present embodiment, the determination unit 131 is specifically used for the two-dimensional points
The central point of set as the cluster point, using the average value of the deflection of each two-dimensional coordinate point in two-dimentional point set as
The deflection of the cluster point, wherein the distance between 2 adjacent cluster points are the default clustering distance.
In the alternatively possible realization method of the present embodiment, the determination unit 131, also particularly useful for by described two
First two-dimensional coordinate point is as the cluster point in dimension point set, using the deflection of first two-dimensional coordinate point as institute
State the deflection of cluster point, wherein the distance between 2 adjacent cluster points are the default clustering distance.
Figure 13 is the structural schematic diagram for the scene thumbnail map generalization device that the embodiment of the present invention three provides, such as Figure 13 institutes
Show, on the basis of the above embodiments, above-mentioned generation module 140 includes：
Sort out unit 141, for each cluster point between adjacent two turning point to be classified as a fitting set, wherein institute
It is in each cluster point to state turning point, and the difference of the deflection and the deflection of adjacent cluster point is more than the poly- of preset value
Class point；
It is corresponding to obtain each fitting set for carrying out curve fitting to each fitting set for fitting unit 142
Matched curve；
Connection unit 143 generates the thumbnail of the target scene for each matched curve to be attached.
In a kind of possible realization method of the present embodiment, fitting unit 142 is specifically used for gathering each fitting
Carry out fitting a straight line；Alternatively, carrying out quadratic fit to each fitting set.
In the alternatively possible realization method of the present embodiment, above-mentioned fitting unit 142, being additionally operable to will be pre- at a distance of being less than
If the opposite matched curve of distance and direction, merges into a matched curve.
In the alternatively possible realization method of the present embodiment, connection unit 143, if being specifically used for the first matched curve
On target cluster point next cluster point not in first matched curve, it is determined that the next cluster point place
The second matched curve, and determine target cluster point, next cluster point and first matched curve and institute
State the triangle that the intersection point of the second matched curve is constituted；If quasi- to described first without matched curve in the triangle
It closes curve and second matched curve is fitted connection, generate the thumbnail of the target scene, wherein described first is quasi-
It is any curve in each matched curve to close curve.
It should be noted that：The scene thumbnail map generalization device that above-described embodiment provides is in the life for carrying out scene thumbnail figure
At processing when, only the example of the division of the above functional modules, in practical application, can as needed and will be above-mentioned
Function distribution is completed by different function module, i.e., the internal structure of device is divided into different function modules, with complete with
The all or part of function of upper description.In addition, scene thumbnail map generalization device and the above-mentioned scene of above-described embodiment offer
The generating means embodiment of thumbnail belongs to same design, and specific implementation process refers to embodiment of the method, and which is not described herein again.
Figure 14 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention, as shown in figure 14, the electricity of the present embodiment
Sub- equipment 200 includes：
Memory 210, for storing computer program；
Processor 220, for executing the computer program, to realize above-mentioned scene thumbnail map generalization method, in fact
Existing principle is similar with technique effect, and details are not described herein again.
Further, when at least part function of the generation method of Scene thumbnail of the embodiment of the present invention passes through software
When realization, the embodiment of the present invention also provides a kind of computer storage media, and computer storage media is above-mentioned to field for being stored as
The computer software instructions of scape breviary map generalization, when run on a computer so that computer can execute above-mentioned side
Various possible scene thumbnail map generalization methods in method embodiment.The computer execution is loaded and executes on computers to refer to
When enabling, can entirely or partly it generate according to the flow or function described in the embodiment of the present invention.The computer instruction can be deposited
Storage is transmitted in computer storage media, or from a computer storage media to another computer storage media, described
Transmission can be by wireless (such as cellular communication, infrared, short-distance wireless, microwave etc.) mode to another web-site, meter
Calculation machine, server or data center are transmitted.The computer storage media can be that computer can access it is any can
With medium either comprising data storage devices such as one or more usable mediums integrated server, data centers.It is described can
Can be magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (example with medium
Such as SSD).
Finally it should be noted that：The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Present invention has been described in detail with reference to the aforementioned embodiments for pipe, it will be understood by those of ordinary skill in the art that：Its according to
So can with technical scheme described in the above embodiments is modified, either to which part or all technical features into
Row equivalent replacement；And these modifications or replacements, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (12)
1. a kind of scene thumbnail map generalization method, which is characterized in that including：
According to each collecting device when target scene moves acquired image data and exercise data, determine it is each it is described sampling set
The standby three-dimensional track point in the target scene；
Obtain the two-dimensional coordinate point and deflection of each three-dimensional track point；
According to the deflection of the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point, to each two-dimensional coordinate
Point is clustered, and each cluster point is obtained；
It carries out curve fitting to each cluster point, generates the thumbnail of the target scene.
2. according to the method described in claim 1, it is characterized in that, the two-dimensional coordinate point for obtaining each three-dimensional track point and side
To angle, specifically include：
By subpoint of the three-dimensional track point on two dimensional surface, the two-dimensional coordinate point as the three-dimensional track point；
By the line direction of the two-dimensional coordinate point and the two-dimensional coordinate point of neighbouring sampling instant, as the two-dimensional coordinate point
Deflection.
3. according to the method described in claim 2, it is characterized in that, the two dimensional surface be apart from each three-dimensional track point most
Close plane.
4. according to the method described in claim 3, it is characterized in that, it is described according to the distance between each described two-dimensional coordinate point and
The deflection of each two-dimensional coordinate point clusters each two-dimensional coordinate point, obtains each cluster point, specifically include：
Determine adjacent the distance between two two-dimensional coordinate points；
The sum of the distance is less than default clustering distance, and deflection is located at each adjacent two dimension in preset direction angle range
Coordinate points, as a two-dimentional point set；
Determine the deflection of each two-dimentional point set corresponding cluster point and each cluster point.
5. according to the method described in claim 4, it is characterized in that, the corresponding cluster point of each two-dimentional point set of the determination with respectively
The deflection for clustering point, specifically includes：
Using the central point of the two-dimentional point set as the cluster point, by the side of each two-dimensional coordinate point in the two-dimentional point set
To the deflection of the average value as the cluster point at angle, wherein the distance between adjacent two clusters point is described default poly-
Class distance.
6. according to the method described in claim 3, it is characterized in that, described carry out curve fitting to each cluster point, generate
The thumbnail of the target scene, specifically includes：
Each cluster point between adjacent two turning point is classified as a fitting set, wherein the turning point is each cluster
In point, the difference of deflection and the deflection of adjacent cluster point is more than the cluster point of preset value；
It carries out curve fitting to each fitting set, obtains each fitting and gather corresponding matched curve；
Each matched curve is attached, the thumbnail of the target scene is generated.
7. according to the method described in claim 6, it is characterized in that, described carry out curve fitting to each fitting set, have
Body includes：
Fitting a straight line is carried out to each fitting set；
Alternatively, carrying out quadratic fit to each fitting set.
8. the method according to the description of claim 7 is characterized in that described be attached each matched curve, institute is generated
Before the thumbnail for stating target scene, the method further includes：
By at a distance of the opposite matched curve of pre-determined distance and direction is less than, a matched curve is merged into.
9. according to claim 6 to 8 any one of them method, which is characterized in that described to connect each matched curve
It connects, generates the thumbnail of the target scene, specifically include：
If next cluster point of the target cluster point in the first matched curve is not in first matched curve, it is determined that institute
The second matched curve where next cluster point is stated, wherein first matched curve is any in each matched curve
Curve；
Determine that the target cluster point, next cluster point and first matched curve and second fitting are bent
The triangle that the intersection point of line is constituted；
If being fitted to first matched curve and second matched curve without matched curve in the triangle
Connection, generates the thumbnail of the target scene.
10. a kind of scene thumbnail map generalization device, which is characterized in that including：
Determining module, for according to each collecting device when target scene moves acquired image data and exercise data, really
Fixed three-dimensional track point of each sample devices in the target scene；
Acquisition module, two-dimensional coordinate point and deflection for obtaining each three-dimensional track point；
Cluster module, it is right for the deflection according to the distance between each described two-dimensional coordinate point and each two-dimensional coordinate point
Each two-dimensional coordinate point is clustered, and each cluster point is obtained；
Generation module generates the thumbnail of the target scene for carrying out curve fitting to each cluster point.
11. a kind of electronic equipment, which is characterized in that including：
Memory, for storing computer program；
Processor, for executing the computer program, to realize scene thumbnail figure as claimed in any one of claims 1-9 wherein
Generation method.
12. a kind of computer storage media, which is characterized in that store computer program, the computer in the storage medium
Program realizes any scene thumbnail map generalization method in claim 1-9 when being executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810213001.2A CN108416044B (en) | 2018-03-15 | 2018-03-15 | Scene thumbnail generation method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810213001.2A CN108416044B (en) | 2018-03-15 | 2018-03-15 | Scene thumbnail generation method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108416044A true CN108416044A (en) | 2018-08-17 |
| CN108416044B CN108416044B (en) | 2021-11-09 |
Family
ID=63131609
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810213001.2A Active CN108416044B (en) | 2018-03-15 | 2018-03-15 | Scene thumbnail generation method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108416044B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110060343A (en) * | 2019-04-24 | 2019-07-26 | 百度在线网络技术（北京）有限公司 | Map constructing method and system, server, computer-readable medium |
| CN114972828A (en) * | 2022-05-14 | 2022-08-30 | 上海贝特威自动化科技有限公司 | Clustering-based two-dimensional coordinate row and column sorting method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070172105A1 (en) * | 2006-01-25 | 2007-07-26 | Claus Bahlmann | System and method for local pulmonary structure classification for computer-aided nodule detection |
| CN105404844A (en) * | 2014-09-12 | 2016-03-16 | 广州汽车集团股份有限公司 | Road boundary detection method based on multi-line laser radar |
| CN105741355A (en) * | 2016-02-01 | 2016-07-06 | 华侨大学 | Block segmentation method for triangular grid model |
| CN105825510A (en) * | 2016-03-17 | 2016-08-03 | 中南大学 | Automatic matching method between point of interest and road network |
| CN105993034A (en) * | 2014-02-13 | 2016-10-05 | 微软技术许可有限责任公司 | Contour completion for augmenting surface reconstructions |
| US20170109373A1 (en) * | 2015-10-15 | 2017-04-20 | Telogis, Inc. | Systems and methods for database geocoding |
-
2018
- 2018-03-15 CN CN201810213001.2A patent/CN108416044B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070172105A1 (en) * | 2006-01-25 | 2007-07-26 | Claus Bahlmann | System and method for local pulmonary structure classification for computer-aided nodule detection |
| CN105993034A (en) * | 2014-02-13 | 2016-10-05 | 微软技术许可有限责任公司 | Contour completion for augmenting surface reconstructions |
| CN105404844A (en) * | 2014-09-12 | 2016-03-16 | 广州汽车集团股份有限公司 | Road boundary detection method based on multi-line laser radar |
| US20170109373A1 (en) * | 2015-10-15 | 2017-04-20 | Telogis, Inc. | Systems and methods for database geocoding |
| CN105741355A (en) * | 2016-02-01 | 2016-07-06 | 华侨大学 | Block segmentation method for triangular grid model |
| CN105825510A (en) * | 2016-03-17 | 2016-08-03 | 中南大学 | Automatic matching method between point of interest and road network |
Non-Patent Citations (3)
| Title |
|---|
| KIM, JE, SEOK, ET AL: "Boundary detection with a road model for occupancy grids in the curvilinear coordinate system using a downward-looking lidar sensor", 《PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS, PART D. JOURNAL OF AUTOMOBILE ENGINEERING》 * |
| 李辉等: "架空输电线路中设备轮廓提取方法的研究", 《电子测量技术》 * |
| 柯欣琦等: "郊外低对比度图像线特征提取算法研究", 《激光与红外》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110060343A (en) * | 2019-04-24 | 2019-07-26 | 百度在线网络技术（北京）有限公司 | Map constructing method and system, server, computer-readable medium |
| CN110060343B (en) * | 2019-04-24 | 2023-06-20 | 阿波罗智能技术(北京)有限公司 | Map construction method and system, server and computer readable medium |
| CN114972828A (en) * | 2022-05-14 | 2022-08-30 | 上海贝特威自动化科技有限公司 | Clustering-based two-dimensional coordinate row and column sorting method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108416044B (en) | 2021-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jeong et al. | Complex urban dataset with multi-level sensors from highly diverse urban environments | |
| US20210172756A1 (en) | Lane line creation for high definition maps for autonomous vehicles | |
| US11353589B2 (en) | Iterative closest point process based on lidar with integrated motion estimation for high definition maps | |
| CN112105893B (en) | Real-time map generation system for an autonomous vehicle | |
| CN110497901B (en) | Parking space automatic searching method and system based on robot VSLAM technology | |
| US11313684B2 (en) | Collaborative navigation and mapping | |
| CN111771141B (en) | LIDAR positioning for solution inference using 3D CNN network in autonomous vehicles | |
| US11364931B2 (en) | Lidar localization using RNN and LSTM for temporal smoothness in autonomous driving vehicles | |
| CN113406682B (en) | Positioning method, positioning device, electronic equipment and storage medium | |
| JP2021111385A (en) | System and method enabling collaborative 3d map data fusion platform and virtual world system thereof | |
| JP2019518222A (en) | Laser scanner with real-time on-line egomotion estimation | |
| JP2018084573A (en) | Robust and efficient algorithm for vehicle positioning and infrastructure | |
| CN107888828A (en) | Space-location method and device, electronic equipment and storage medium | |
| CN109461208A (en) | Three-dimensional map processing method, device, medium and calculating equipment | |
| WO2024027350A1 (en) | Vehicle positioning method and apparatus, computer device and storage medium | |
| CN111768489B (en) | Indoor navigation map construction method and system | |
| CN111652072A (en) | Track acquisition method, track acquisition device, storage medium and electronic equipment | |
| CN110411457A (en) | Localization method, system, terminal and the storage medium merged with vision is perceived based on stroke | |
| CN114387319A (en) | Point cloud registration method, device, equipment and storage medium | |
| CN112950710A (en) | Pose determination method and device, electronic equipment and computer readable storage medium | |
| CN110458885A (en) | Localization method, system, terminal and the storage medium merged with vision is perceived based on stroke | |
| CN108416044A (en) | Scene thumbnail map generalization method, apparatus, electronic equipment and storage medium | |
| Cui et al. | Mcsfm: multi-camera-based incremental structure-from-motion | |
| CN116858269A (en) | Tobacco industry finished product warehouse flat warehouse inventory robot path optimization method based on laser SLAM | |
| Xu et al. | Smartphone-based crowdsourcing for panoramic virtual tour construction |
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 |