CN105631459B - Protective fence data reduction method and device - Google Patents
Protective fence data reduction method and device Download PDFInfo
- Publication number
- CN105631459B CN105631459B CN201511025864.XA CN201511025864A CN105631459B CN 105631459 B CN105631459 B CN 105631459B CN 201511025864 A CN201511025864 A CN 201511025864A CN 105631459 B CN105631459 B CN 105631459B
- Authority
- CN
- China
- Prior art keywords
- protective fence
- point cloud
- cloud
- point
- single frames
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a kind of protective fence data reduction method and devices.This method comprises: obtaining the shape feature of the first protective fence, first protective fence is waveform protective fence, cement crash bearer or soundproof wall；The first protective fence point cloud is extracted from single frames road waypoint cloud according to the shape feature of first protective fence.The present invention can extract the first protective fence point cloud according to the shape feature of the first protective fence after the shape feature for obtaining the first protective fence from single frames road waypoint cloud.Need manually to be extracted protective fence point cloud in the prior art by engineer, extraction efficiency is low.The present invention can extract protective fence point cloud according to the shape feature of the first protective fence by machine from road waypoint cloud, manually extract without engineer, reduce human cost and machine extraction rate is very fast, improve the recognition efficiency of protective fence.
Description
Technical field
The present embodiments relate to Point Cloud Processing technology more particularly to a kind of protective fence data reduction method and dresses
It sets.
Background technique
Three-dimensional high-precision map is known as the main direction of development of generation digital map by industry and academia, is to realize
The precondition that automatic driving and auxiliary drive be accurately positioned for autonomous driving vehicle and correct decisions provides mainly
Foundation.One important link of three-dimensional high-precision map is the mapping relations established between road image and protective fence point.
The prior art is in the mapping relations established between road image and protective fence point cloud data, first by vehicle-mounted
Point cloud obtains equipment and obtains road waypoint cloud, the corresponding single frames road waypoint cloud of each time point.Then, for each single frames road
Point cloud, by the spatial point set for finding out similar protective fence in the artificial slave road point cloud data of engineer.
However, when carrying out scatter point protraction to one section of road, it may be necessary to dozens of minutes or longer time, and every one
Time point (such as 1 millisecond) will generate a single frames road waypoint cloud, and then generate a large amount of single frames road waypoint cloud.At this point, if
Each single frames point cloud passes through engineer and manually searches protective fence point cloud, then will take a substantial amount of time, and human cost is higher, and
Protective fence point cloud recognition efficiency is low.
Summary of the invention
The present invention provides a kind of protective fence data reduction method and device, to realize through machine to anti-in road waypoint cloud
Guardrail point cloud is identified, the recognition efficiency of protective fence point cloud is improved.
In a first aspect, the embodiment of the invention provides a kind of protective fence data reduction methods, comprising:
The shape feature of the first protective fence is obtained, first protective fence is waveform protective fence, cement crash bearer or sound insulation
Wall；
The first protective fence point cloud is extracted from single frames road waypoint cloud according to the shape feature of first protective fence.
Second aspect, the embodiment of the invention also provides a kind of protective fence data reduction devices, comprising:
Shape feature acquiring unit, for obtaining the shape feature of the first protective fence, first protective fence is anti-for waveform
Guardrail, cement crash bearer or soundproof wall；
First protective fence extraction unit, first protective fence for being obtained according to the shape feature acquiring unit
Shape feature extracts the first protective fence point cloud from single frames road waypoint cloud.
The present invention can be after the shape feature for obtaining the first protective fence, according to the shape feature of the first protective fence from single frames
The first protective fence point cloud is extracted in road waypoint cloud.It needs manually to be extracted protective fence point cloud in the prior art by engineer, extracts effect
Rate is low.The present invention can extract protective fence point cloud according to the shape feature of the first protective fence by machine from road waypoint cloud, be not necessarily to
Engineer manually extracts, and reduces human cost and machine extraction rate is very fast, improve the recognition efficiency of protective fence.
Detailed description of the invention
Fig. 1 is the flow chart of the protective fence data reduction method in the embodiment of the present invention one；
Fig. 2 is the flow chart of the protective fence data reduction method in the embodiment of the present invention two；
Fig. 3 is the flow chart of the protective fence data reduction method in the embodiment of the present invention three；
Fig. 4 is the flow chart of the protective fence data reduction method in the embodiment of the present invention four；
Fig. 5 is the structural schematic diagram of the protective fence data reduction device in the embodiment of the present invention five.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just
Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
Embodiment one
Fig. 1 is the flow chart for the protective fence data reduction method that the embodiment of the present invention one provides, and the present embodiment is applicable to
The case where protective fence point cloud is extracted from road waypoint cloud, this method can be held by the terminal for carrying out protective fence data reduction
Row, which can be server, PC, laptop, tablet computer, automobile center console, smart phone or intelligence
Wearable device etc., this method comprises:
S110, the shape feature for obtaining the first protective fence.
Wherein, the first protective fence is waveform protective fence, cement crash bearer or soundproof wall.
Waveform protective fence is also known as corrugated guardrail, and corrugated guardrail is highway crash facility, primarily to preventing errant vehicle
Go out road.Corrugated guardrail is a kind of mutually spliced with corrugated steel guard rail plate and by the continuous structure of upright supports, waveform shield
Column can be double wave corrugated guardrail or three wave corrugated guardrails.Therefore usually the point cloud sequence in wave shape in road waypoint cloud is determined
For waveform crash barrier, the point cloud sequence of the waveform can be that double wave waveform can also be three wave waveforms.
Cement crash bearer is also known as cement crash bearer, for being placed on turning, the entrance, charging aperture of highway and urban road
And on the road surface of other hazardous locations for needing isolation or anticollision, security isolation, warning, pre- anticollision can simultaneously collide
When work as a buffer, absorb and reduce the facility of collision impact.The trapezoidal top of the usual pedestal of cement crash barrier is rectangle.
Since cement crash barrier is directly placed on road surface, block raised on road surface can be determined as cement crash bearer.
Soundproof wall (acoustic) is mainly used for the sound insulation and noise reducing of highway, overhead composite road and other noise sources,
Soundproof wall is located at both sides of the road, and in order to reach higher sound-absorbing effect, soundproof wall is usually in the form of sheets.It therefore can be by road two
The tablet of side and road surface disposition is determined as soundproof wall.
The shape feature of first protective fence can be the given shape feature of engineer's input.The shape of first protective fence
Feature can also be to be trained by the known protective fence point cloud to a large amount of (such as larger than 500), obtains waveform protective fence, water
The shape feature of mud crash bearer or soundproof wall.
S120, the first protective fence point cloud is extracted from single frames road waypoint cloud according to the shape feature of the first protective fence.
Different protective fences has corresponding style characteristic, can extract one or more the according to use demand at the extraction
One protective fence.In one implementation, more comprehensively to identify the protective fence point cloud in waypoint cloud, respectively according to wave
The row shape feature of shape protective fence, cement crash bearer and soundproof wall is searched in road waypoint cloud, and extracts the point cloud found
Sequence.In another implementation, according to different surface conditions, can targetedly select one or two kinds of protective fences into
Row extracts.For example, usually running at a low speed section in charge station and intersection of roads etc. on a highway places cement anticollision
Pier carries out water conservancy diversion, therefore waveform protective fence and cement crash bearer are extracted from road waypoint cloud running at a low speed section.It is public in high speed
Waveform protective fence and soundproof wall is mainly arranged in other sections on road, therefore waveform crash barrier is extracted when entering fast running section
And soundproof wall.
The present invention can be after the shape feature for obtaining the first protective fence, according to the shape feature of the first protective fence from single frames
The first protective fence point cloud is extracted in road waypoint cloud.It needs manually to be extracted protective fence point cloud in the prior art by engineer, extracts effect
Rate is low.The present invention can extract protective fence point cloud according to the shape feature of the first protective fence by machine from road waypoint cloud, be not necessarily to
Engineer manually extracts, and reduces human cost and machine extraction rate is very fast, improve the recognition efficiency of protective fence.
Embodiment two
Fig. 2 is the flow chart of protective fence data reduction method provided by Embodiment 2 of the present invention, optionally, S110, basis
The shape feature of first protective fence extracts the first protective fence point cloud from single frames road waypoint cloud, can be carried out by following manner
Implement:
S110a, by the curve form point cloud sequential extraction procedures in single frames road waypoint cloud be waveform protective fence point cloud.
Waveform protective fence usually passes through fixed bracket and is fixed on both sides of the road, and there are pre-determined distance (such as 0.8-1.2 with road surface
Rice).The shape of waveform protective fence is wavy, is specifically as follows double wave waveform or three wave waveforms.Optionally, by the curve of S type
Form point cloud sequence is extracted as waveform protective fence point cloud.S type curve can be unicast, double wave or three waves.Point cloud sequence is used
In the combined sequence for the spatial point for indicating to be made of spatial point.
Specifically, obtaining any one first spatial point, judge whether the space curve by first spatial point is S
Type.If it is S type, S type curve is extracted.If not S type, then obtain another second space point, judge by this second
Whether the space curve of spatial point is S type.And so on obtain the waveform protective fence point cloud in waypoint cloud.
According to the curve shape of road waypoint cloud midpoint cloud Sequence composition, it may be determined that protected in road waypoint cloud with the presence or absence of waveform
Column improves the recognition efficiency of waveform protective fence.
Optionally, S110, the first protection is extracted according to the shape feature of first protective fence from single frames road waypoint cloud
Column point cloud, can be implemented by following manner:
S110b, there will be the point cloud sequential extraction procedures of height trip point or gradient trip point for cement in single frames road waypoint cloud
Crash bearer point cloud.
Since cement crash bearer is generally positioned on road surface, when along road surface, point cloud is scanned to cement crash bearer, can go out
The now trip point of height and the gradient.Trip point is the point on the intersection of road surface and cement crash bearer.Road surface after trip point, due to
From road surface, jump is cement crash bearer, therefore after trip point, and the height of point cloud sequence is increased sharply；Simultaneously as cement crash bearer
Mop can be to be trapezoidal, therefore after trip point, the gradient of point cloud sequence is consequently increased, not such as road surface gently.
In order to avoid because of erroneous judgement caused by the big rise and fall of road surface itself itself, can increase sharply to height or ramp degree into
Professional etiquette model.Such as: judge after height trip point, whether the height of road waypoint cloud sequence is greater than preset height threshold value (such as
40cm).Another example is: judging after height trip point, whether the gradient of road waypoint cloud sequence is greater than default gradient threshold value (such as 30
Degree).
Further, judge after height trip point, put cloud sequence height whether cement crash bearer normal range (NR)
It is interior, such as be highly greater than 40cm and be less than 70cm.Alternatively, judging after gradient trip point, whether the gradient of cloud sequence is put in cement
In the normal range (NR) of crash bearer bottom slope, as the gradient is greater than 30 degree less than 60 degree.
The bottom profile that cement crash bearer can be found from the point cloud of road surface by height trip point or gradient trip point, is based on
The bottom profile of cement crash bearer can get the point Yun Xulie of cement crash bearer.
Optionally, S110, the first protection is extracted according to the shape feature of first protective fence from single frames road waypoint cloud
Column point cloud, can be implemented by following manner:
S110c, there will be the point cloud sequential extraction procedures of plane characteristic for soundproof wall point cloud in single frames road waypoint cloud.
Three-dimensional Hough transformation (Hough Transform) can be used to extract the plane characteristic of point cloud sequence.Three-dimensional Huffman
Transformation may recognize that the point Yun Xulie with flat shape feature in waypoint cloud.The normal vector of spatial point A is obtained first, so
Afterwards in the multiple spatial point Bs adjacent with spatial point A with lookup in the vertical plane of the normal vector, if the multiple spaces found
Point B is in same plane, it is determined that spatial point A and multiple spatial point B constitutes plane.
It further, can be according to the plane of cloud sequence composition since the normal vector of soundproof wall is vertical with road surface
Normal vector is screened, and is soundproof wall point cloud by the normal vector planar point cloud sequential extraction procedures vertical with road surface.
By judging that spatial point cloud with the presence or absence of plane characteristic, can extract soundproof wall point cloud from road waypoint cloud, improve
The extraction efficiency of soundproof wall point cloud.
Embodiment three
Fig. 3 is the flow chart for the protective fence data reduction method that the embodiment of the present invention three provides, with protective fence technology
It continues to develop, other than waveform protective fence, cement crash bearer and soundproof wall, more and more Novel protective handrails come out, at this time
According only to waveform protective fence, cement crash bearer or soundproof wall shape feature carry out an extraction for cloud sequence can not be to novel protective
Column point cloud extracts.Based on this, first is extracted from single frames road waypoint cloud in S120, according to the shape feature of the first protective fence
After protective fence point cloud, further includes:
S130, the first protective fence point cloud, road surface point cloud and body of rod point cloud are filtered out from single frames road waypoint cloud, obtain left point
Cloud.
S120 can determine the first protective fence point cloud, road surface point cloud and body of rod point cloud in single frames road waypoint cloud.From single frames road
The first protective fence point cloud, road surface point cloud and body of rod point cloud are deleted in waypoint cloud, obtain left point cloud.
S140, left point cloud corresponding K arest neighbors relational graph (K Nearest Neighbor Graph, KNN are calculated
Graph)。
K nearest neighbor algorithm is a kind of Data Mining Classification technology.K arest neighbors indicates K nearest neighbours, i.e., each sample
(spatial point) can be represented with its immediate K neighbour.The core concept of KNN algorithm is if a sample is in feature
Most of in K in space most adjacent samples belong to some classification, then the sample also belongs to this classification.Due to new
The shape feature of type protective fence is uncertain, therefore the point cloud sequence in left point cloud can be classified by KNN algorithm.
S150, left point cloud is split according to K arest neighbors relational graph, obtains at least one graphical dots cloud.
After the K arest neighbors relational graph for obtaining being made of multiple cloud sequences, KNN CUT technology, that is, K arest neighbors can be used
Relational graph cutting technique cuts K arest neighbors relational graph.When cutting, the spatial point with same point cloud sequence is cut
It cuts, obtains at least one graphical dots cloud.
S160, the rotation image and bounding box for obtaining any graphical dots cloud.
Rotation image (Spin Image) is generated according to the space coordinate of spatial point each in picture point cloud.According to picture point
The coordinate threshold value of the spatial point of cloud determines that bounding box, bounding box are the circumscribed rectangular body of picture point cloud.
S170, extraction rotation image and bounding box meet default protective fence point cloud model from least one graphical dots cloud
Second protective fence point cloud.
Default protective fence point cloud model can be by support vector machines (Support Vector Machine, SVM) or random
Forest (Random forest) obtains.Second protective fence point cloud known to multiple (such as larger than 500) is updated to support vector machines
Or be trained in random forest, the shape feature of the second protective fence point cloud is obtained, image and bounding box are such as rotated.Rotate image
For describing the shape of the second protective fence point cloud, bounding box is used to describe the size of the second protective fence point cloud.
Technical solution provided in this embodiment, trained support vector machines or random forest can be to the figures of input
The rotation image and bounding box of point cloud are identified, determine whether the graphical dots cloud of input meets the shape of the second protective fence point cloud
Feature is known into the shape feature for realizing customized second protective fence, and then to Novel protective handrail (the second protective fence) point cloud
Not, the availability of protective fence point cloud identification is improved.
Example IV
Fig. 4 is the flow chart of protective fence data reduction method that the embodiment of the present invention four provides, in S120, according to described the
The shape feature of one protective fence is after extracting the first protective fence point cloud in single frames road waypoint cloud, further includes:
S180, the first protective fence point cloud is tracked along garage track, obtains the continuous feature of the first protective fence point cloud.
From the spatial point α found in certain single frames road waypoint cloud A in the first protective fence point cloud a.When being obtained according to road waypoint cloud
Point cloud obtains the motion profile (garage track) of equipment, determines the deviation post of the first protective fence point cloud.Judge single frames road waypoint
Corresponding to the deviation post in the adjacent single frames road waypoint cloud B of cloud A whether there is the first protective fence point cloud a.If it is present really
It is continuous that the first protective fence point cloud a is found in order frame road waypoint cloud A.
If S190a, the first protective fence point Yun Lianxu, retain the first protective fence point cloud.
If the first protective fence point cloud is continuous in the continuous single frames road waypoint cloud for presetting continuous quantity, it is determined that the
One protective fence point Yun Lianxu, wherein presetting continuous quantity greater than 10, preferably 20.
If S190b, the first protective fence point cloud are discontinuous, the first protective fence point cloud is filtered out.
It is filtered, filters it should be noted that scheme provided in this embodiment can also be used for the second protective fence point cloud
Except the insufficient protective fence of continuity.
Technical solution provided in this embodiment passes through the continuous spy to the first protective fence point cloud (or the second protective fence point cloud)
Sign extracts and judges, determines whether the first protective fence point cloud (or the second protective fence point cloud) has continuity, and then filters out
Intermittent first protective fence point cloud (or the second protective fence point cloud), improves the recognition accuracy of protective fence.
Embodiment five
Fig. 5 is the structural schematic diagram for the protective fence data reduction device that the embodiment of the present invention five provides, and described device is located at
In terminal, comprising:
Shape feature acquiring unit 11, for obtaining the shape feature of the first protective fence, first protective fence is waveform
Protective fence, cement crash bearer or soundproof wall；
First protective fence extraction unit 12, first protection for being obtained according to the shape feature acquiring unit 11
The shape feature on column extracts the first protective fence point cloud from single frames road waypoint cloud.
Further, the first protective fence extraction unit 12 is specifically used for, by the curve form point in single frames road waypoint cloud
Cloud sequential extraction procedures are waveform protective fence point cloud.
Further, the first protective fence extraction unit 12 is specifically used for, and in single frames road waypoint cloud will there is height to jump
The point cloud sequential extraction procedures of height or gradient trip point are cement crash bearer point cloud.
Further, the first protective fence extraction unit 12 is specifically used for, and will have plane special in single frames road waypoint cloud
The point cloud sequential extraction procedures of sign are soundproof wall point cloud.
Further, described device further includes the second protective fence data reduction unit 13, and the second protective fence point cloud mentions
Unit 13 is taken to be used for:
The first protective fence point cloud, the road surface that the first protective fence extraction unit 12 extracts are filtered out from single frames road waypoint cloud
Point cloud and body of rod point cloud, obtain left point cloud；
Calculate the corresponding K arest neighbors relational graph of the left point cloud；
The left point cloud is split according to the K arest neighbors relational graph, obtains at least one graphical dots cloud；
Obtain the rotation image and bounding box of any graphical dots cloud；
Rotation image is extracted from least one described graphical dots cloud and bounding box meets default protective fence point cloud model
Second protective fence point cloud.
Further, described device further includes continuity judging unit 14, and the continuity judging unit 14 is used for:
The the first protective fence point cloud extracted to the first protective fence extraction unit 12 is tracked along garage track, is obtained
The continuous feature of the first protective fence point cloud；
If the first protective fence point Yun Lianxu retains the first protective fence point cloud.
The embodiment of the present invention one can be performed to method provided by example IV in above-mentioned apparatus, has and executes above method phase
The functional module and beneficial effect answered.The not technical detail of detailed description in the present embodiment, reference can be made to the embodiment of the present invention one
With method provided by example IV.
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that
The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention
It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also
It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.
Claims (10)
1. a kind of protective fence data reduction method characterized by comprising
The shape feature of the first protective fence is obtained, first protective fence is waveform protective fence, cement crash bearer or soundproof wall；
The first protective fence point cloud is extracted from single frames road waypoint cloud according to the shape feature of first protective fence；
It after extracting the first protective fence point cloud in single frames road waypoint cloud, is also wrapped according to the shape feature of first protective fence
It includes:
The first protective fence point cloud, road surface point cloud and body of rod point cloud are filtered out from single frames road waypoint cloud, obtain left point cloud；
Calculate the corresponding K arest neighbors relational graph of the left point cloud；
The left point cloud is split according to the K arest neighbors relational graph, obtains at least one graphical dots cloud；
The rotation image and bounding box of any graphical dots cloud are obtained, the bounding box is the coordinate of the spatial point of the graphical dots cloud
The shape feature that threshold value determines；
Rotation image is extracted from least one described graphical dots cloud and bounding box meets the second of default protective fence point cloud model
Protective fence point cloud.
2. protective fence data reduction method according to claim 1, which is characterized in that described according to first protective fence
Shape feature the first protective fence point cloud is extracted from single frames road waypoint cloud, comprising:
It is waveform protective fence point cloud by the curve form point cloud sequential extraction procedures in single frames road waypoint cloud.
3. protective fence data reduction method according to claim 1, which is characterized in that described according to first protective fence
Shape feature the first protective fence point cloud is extracted from single frames road waypoint cloud, comprising:
It is cement crash bearer point cloud by the point cloud sequential extraction procedures with height trip point or gradient trip point in single frames road waypoint cloud.
4. protective fence data reduction method according to claim 1, which is characterized in that described according to first protective fence
Shape feature the first protective fence point cloud is extracted from single frames road waypoint cloud, comprising:
It is soundproof wall point cloud by the point cloud sequential extraction procedures in single frames road waypoint cloud with plane characteristic.
5. protective fence data reduction method according to any one of claim 1 to 4, which is characterized in that according to
The shape feature of first protective fence is after extracting the first protective fence point cloud in single frames road waypoint cloud, further includes:
First protective fence point cloud is tracked along garage track, obtains the continuous feature of the first protective fence point cloud；
If the first protective fence point Yun Lianxu retains the first protective fence point cloud.
6. a kind of protective fence data reduction device characterized by comprising
Shape feature acquiring unit, for obtaining the shape feature of the first protective fence, first protective fence be waveform protective fence,
Cement crash bearer or soundproof wall；
First protective fence extraction unit, the shape of first protective fence for being obtained according to the shape feature acquiring unit
Feature extracts the first protective fence point cloud from single frames road waypoint cloud；
Second protective fence data reduction unit, the second protective fence data reduction unit are used for:
Filtered out from single frames road waypoint cloud the first protective fence point cloud that the first protective fence extraction unit extracts, road surface point cloud and
Body of rod point cloud, obtains left point cloud；
Calculate the corresponding K arest neighbors relational graph of the left point cloud；
The left point cloud is split according to the K arest neighbors relational graph, obtains at least one graphical dots cloud；
The rotation image and bounding box of any graphical dots cloud are obtained, the bounding box is the coordinate of the spatial point of the graphical dots cloud
The shape feature that threshold value determines；
Rotation image is extracted from least one described graphical dots cloud and bounding box meets the second of default protective fence point cloud model
Protective fence point cloud.
7. protective fence data reduction device according to claim 6, which is characterized in that the first protective fence extraction unit
It is specifically used for, is waveform protective fence point cloud by the curve form point cloud sequential extraction procedures in single frames road waypoint cloud.
8. protective fence data reduction device according to claim 6, which is characterized in that the first protective fence extraction unit
It is specifically used for, is cement crash bearer by the point cloud sequential extraction procedures with height trip point or gradient trip point in single frames road waypoint cloud
Point cloud.
9. protective fence data reduction device according to claim 6, which is characterized in that the first protective fence extraction unit
It is specifically used for, is soundproof wall point cloud by the point cloud sequential extraction procedures in single frames road waypoint cloud with plane characteristic.
10. protective fence data reduction device according to any one of claims 6 to 9, which is characterized in that further include continuous
Property judging unit, the continuity judging unit are used for:
The the first protective fence point cloud extracted to the first protective fence extraction unit is tracked along garage track, obtains described the
The continuous feature of one protective fence point cloud；
If the first protective fence point Yun Lianxu retains the first protective fence point cloud.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511025864.XA CN105631459B (en) | 2015-12-31 | 2015-12-31 | Protective fence data reduction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511025864.XA CN105631459B (en) | 2015-12-31 | 2015-12-31 | Protective fence data reduction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105631459A CN105631459A (en) | 2016-06-01 |
| CN105631459B true CN105631459B (en) | 2019-11-26 |
Family
ID=56046371
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511025864.XA Active CN105631459B (en) | 2015-12-31 | 2015-12-31 | Protective fence data reduction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105631459B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107918753B (en) | 2016-10-10 | 2019-02-22 | 腾讯科技（深圳）有限公司 | Processing Method of Point-clouds and device |
| CN108345822B (en) * | 2017-01-22 | 2022-02-01 | 腾讯科技（深圳）有限公司 | Point cloud data processing method and device |
| CN114325755B (en) * | 2021-11-26 | 2023-08-01 | 江苏徐工工程机械研究院有限公司 | Retaining wall detection method and system suitable for automatic driving vehicle |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103711050A (en) * | 2013-12-31 | 2014-04-09 | 中交第二公路勘察设计研究院有限公司 | Laser radar road reconstruction and expansion exploratory survey design method |
| CN104091321A (en) * | 2014-04-14 | 2014-10-08 | 北京师范大学 | Multi-level-point-set characteristic extraction method applicable to ground laser radar point cloud classification |
| CN105184852A (en) * | 2015-08-04 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Laser-point-cloud-based urban road identification method and apparatus |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8078396B2 (en) * | 2004-08-31 | 2011-12-13 | Meadow William D | Methods for and apparatus for generating a continuum of three dimensional image data |
-
2015
- 2015-12-31 CN CN201511025864.XA patent/CN105631459B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103711050A (en) * | 2013-12-31 | 2014-04-09 | 中交第二公路勘察设计研究院有限公司 | Laser radar road reconstruction and expansion exploratory survey design method |
| CN104091321A (en) * | 2014-04-14 | 2014-10-08 | 北京师范大学 | Multi-level-point-set characteristic extraction method applicable to ground laser radar point cloud classification |
| CN105184852A (en) * | 2015-08-04 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Laser-point-cloud-based urban road identification method and apparatus |
Non-Patent Citations (1)
| Title |
|---|
| 车载LiDAR点云中建筑物的自动识别与立面几何重建方法;魏征;《中国博士学位论文全文数据库 信息科技辑》;20130615;第54至57页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105631459A (en) | 2016-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hegde et al. | Yet another deep learning approach for road damage detection using ensemble learning | |
| CN108267747B (en) | Road feature extraction method and device based on laser point cloud | |
| KR102062680B1 (en) | Laser point cloud based urban road recognition method, device, storage medium and device | |
| WO2018068653A1 (en) | Point cloud data processing method and apparatus, and storage medium | |
| JP6024229B2 (en) | Monitoring device, monitoring method, and program | |
| CN102509457B (en) | Vehicle tracking method and device | |
| CN108021844A (en) | A kind of road edge recognition methods and device | |
| CN105631459B (en) | Protective fence data reduction method and device | |
| CN105005771A (en) | Method for detecting full line of lane based on optical flow point locus statistics | |
| US20160202355A1 (en) | Object detecting device, radar device, and object detection method | |
| CN102997900A (en) | Vehicle systems, devices, and methods for recognizing external worlds | |
| CN102862574A (en) | Method for realizing active safety of vehicle on the basis of smart phone | |
| CN103605967A (en) | Subway fare evasion prevention system and working method thereof based on image recognition | |
| Chong et al. | Integrated real-time vision-based preceding vehicle detection in urban roads | |
| CN105627938A (en) | Pavement asphalt thickness detection method based on vehicle-mounted laser scanning spot cloud | |
| CN103280052B (en) | Be applied to the intrusion detection method of long distance track circuit intelligent video monitoring | |
| RU2008109755A (en) | SURROUNDING SURVEY DATA FORMATION SYSTEM, SURROUNDING SURVEY SYSTEM, SURROUNDING SURVEILLANCE SYSTEM FORMATION | |
| JP6465421B1 (en) | Structural deformation detector | |
| Yadav et al. | Pole-shaped object detection using mobile LiDAR data in rural road environments | |
| KR102564856B1 (en) | Method and apparatus of detecting road line | |
| CN116739183B (en) | Mine safety risk early warning prediction system | |
| CN110660211A (en) | Parking area map improvement using occupancy behavior anomaly detector | |
| Janda et al. | Road boundary detection for run-off road prevention based on the fusion of video and radar | |
| CN114241443A (en) | Blind area early warning method and device, electronic equipment and storage medium | |
| Ma et al. | Virtual analysis of urban road visibility using mobile laser scanning data and deep learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |