CN106033620B - A kind of point cloud model restorative procedure, device and computing device - Google Patents
A kind of point cloud model restorative procedure, device and computing device Download PDFInfo
- Publication number
- CN106033620B CN106033620B CN201510111559.6A CN201510111559A CN106033620B CN 106033620 B CN106033620 B CN 106033620B CN 201510111559 A CN201510111559 A CN 201510111559A CN 106033620 B CN106033620 B CN 106033620B
- Authority
- CN
- China
- Prior art keywords
- cloud model
- point cloud
- point
- neighborhood
- scattered points
- 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
Abstract
A kind of point cloud model restorative procedure of offer of the embodiment of the present invention, device and computing device, wherein method include：Determine point cloud model to be repaired；Each scattered points of the point cloud model are traversed, and determine the corresponding neighborhood of each scattered points traversed；If in the corresponding neighborhood of each scattered points traversed, there are the pending neighborhoods that dot density at random is less than density threshold, then in pending neighborhood without determining insertion point in scattered points region；The point cloud model is repaired according to identified insertion point.The present invention may make the cavity of point cloud model to reduce, and improve the validity of point cloud model.
Description
Technical field
The present invention relates to threedimensional model processing technology fields, and in particular to a kind of point cloud model restorative procedure, device and meter
Calculate equipment.
Background technology
Point cloud model is after carrying out three-dimensional modeling to an object (including arbitrary entity, such as street flooring, trees etc.)
The scattered points of threedimensional model, the object mainly detected by cloud detecting devices are constituted；At present in the building process of point cloud model
In, mainly use LiDar (Light Detection And Ranging, optical detection and measurement) equipment to carry out the at random of object
Point detection, LiDar equipment are one kind of point cloud detecting devices, and the data measured by LiDar equipment are the DSM (Digital of object
Surface Model, digital surface model) discrete point indicate, the space three-dimensional information containing object in measured data
With laser intensity information etc..
The present inventor has found in the course of the research, is detected to the scattered points of object in cloud detecting devices
In the process, due to other entities block or the influence of shadow crevice projection angle, finally by the scattered points of the object detected
There is a large amount of cavity in constructed object point cloud model, these are empty by the object point cloud model constructed by strong influence
Validity；By taking streetscape point cloud model as an example, during vehicle-mounted LiDar equipment detects the scattered points of building, due to vehicle-mounted
Blocking for laser penetration glass metope that LiDar equipment uses and other objects, will cause the building object point cloud finally built
There is a large amount of cavity in model.Fig. 1 shows the schematic diagram of the building point cloud model in the presence of cavity, it can be seen that constructed
Building point cloud model window, glass metope, there is a large amount of cavity in the part such as roof, building point cloud model it is true
It spends relatively low.
Therefore, how the point cloud model that there is cavity is repaired, solves the point cloud for the existing structure that inventor has found
There is a large amount of cavity in model, the relatively low problem of validity, the problem of becoming those skilled in the art's urgent need to resolve.
Invention content
In view of this, a kind of point cloud model restorative procedure of offer of the embodiment of the present invention, device and computing device, with to existing
The object point cloud model in cavity is repaired, and the object point cloud model for solving existing structure has an a large amount of cavity, validity compared with
Low problem.
To achieve the above object, the embodiment of the present invention provides the following technical solutions：
A kind of point cloud model restorative procedure, including：
Determine point cloud model to be repaired；
Each scattered points of the point cloud model are traversed, and determine the corresponding neighborhood of each scattered points traversed；
If in the corresponding neighborhood of each scattered points traversed, it is pending less than density threshold that there are dot densities at random
Neighborhood, then in pending neighborhood without determining insertion point in scattered points region；
The point cloud model is repaired according to identified insertion point.
The embodiment of the present invention also provides a kind of point cloud model prosthetic device, including：
Model determining module, for determining point cloud model to be repaired；
Neighborhood determining module, each scattered points for traversing the point cloud model, and determine each traversed
The corresponding neighborhood of scattered points；
Insertion point determining module, if in the corresponding neighborhood of each scattered points for being traversed, there are dot densities at random
Less than the pending neighborhood of density threshold, then in pending neighborhood without determining insertion point in scattered points region；
Repair module, for repairing the point cloud model according to identified insertion point.
The embodiment of the present invention also provides a kind of computing device, including point cloud model prosthetic device described above.
Based on the above-mentioned technical proposal, point cloud model restorative procedure provided in an embodiment of the present invention is determining complex point cloud to be repaired
After model, each scattered points of the point cloud model can be traversed, and determine the corresponding neighborhood of each scattered points traversed；
If in the corresponding neighborhood of each scattered points traversed, there are the pending neighborhoods that dot density at random is less than density threshold, then
In pending neighborhood without determining insertion point in scattered points region；The point cloud model is repaired according to identified insertion point.This
It is low to find dot density at random in the corresponding neighborhood of each scattered points by traversing each scattered points of point cloud model for inventive embodiments
In the pending neighborhood of density threshold, to, without insertion point is determined in scattered points region, obtain multiple insert in pending neighborhood
Access point repairs the cavity of point cloud model by obtained multiple insertion points, may make the cavity of point cloud model to reduce, improves point
The validity of cloud model.
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 only this
The embodiment of invention for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is the schematic diagram in the presence of the building point cloud model in cavity；
Fig. 2 is the flow chart of point cloud model restorative procedure provided in an embodiment of the present invention；
Fig. 3 is another flow chart of point cloud model restorative procedure provided in an embodiment of the present invention；
Fig. 4 is the method flow diagram of determining insertion point provided in an embodiment of the present invention；
Fig. 5 is that insertion point provided in an embodiment of the present invention determines schematic diagram；
Fig. 6 is the another method flow chart of determining insertion point provided in an embodiment of the present invention；
Fig. 7 is that another insertion point provided in an embodiment of the present invention determines schematic diagram；
Fig. 8 is the method flow diagram that point cloud model is repaired in insertion point determined by basis provided in an embodiment of the present invention；
Fig. 9 is another flow chart of point cloud model restorative procedure provided in an embodiment of the present invention；
Figure 10 is the another flow chart of point cloud model restorative procedure provided in an embodiment of the present invention；
Figure 11 is the structure diagram of point cloud model prosthetic device provided in an embodiment of the present invention；
Figure 12 is the structure diagram of insertion point determining module provided in an embodiment of the present invention；
Figure 13 is the structure diagram that insertion point provided in an embodiment of the present invention determines execution unit；
Figure 14 is another structure diagram that insertion point provided in an embodiment of the present invention determines execution unit；
Figure 15 is the structure diagram of repair module provided in an embodiment of the present invention；
Figure 16 is another structure diagram of point cloud model prosthetic device provided in an embodiment of the present invention；
Figure 17 is another structure diagram of point cloud model prosthetic device provided in an embodiment of the present invention；
Figure 18 is the hardware block diagram of computing device provided in an embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Fig. 2 is the flow chart of point cloud model restorative procedure provided in an embodiment of the present invention, and this method can be applied to have number
According to the computing device that can carry out data processing to point cloud model of processing capacity, such as smart mobile phone, PC (personal computer) use
The network side equipments such as family equipment or server；It, can be to arbitrary using point cloud model restorative procedure provided in an embodiment of the present invention
Cavity is repaired existing for the point cloud model of standalone object；With reference to Fig. 2, this method may include：
Step S100, point cloud model to be repaired is determined；
Point cloud model to be repaired can be the point cloud model for arbitrarily existing cavity, be not limited to the initial point cloud of standalone object
Model, it is also possible to be to carry out empty repair process, but there are still the point cloud models in cavity.
Step S110, each scattered points of the point cloud model are traversed, and determine each scattered points pair traversed
The neighborhood answered；
Point cloud model has multiple scattered points, and in primary traversal, the embodiment of the present invention can traverse each of point cloud model
A scattered points determine the corresponding neighborhood of each scattered points traversed；
Optionally, the embodiment of the present invention can set contiguous range, for each scattered points of point cloud model, it may be determined that go out
Including the scattered points traversed, and size is to set the neighborhood of contiguous range；Preferably, during the scattered points that can be traversed are
The heart determines that size is to set the neighborhood of contiguous range；
Further, the shape in identified field can be regular, such as spherical, cuboid, the shapes such as cube；Neighborhood
Interior scattered points (the corresponding neighborhood point of scattered points traversed) can be calculated by KNN (k-NearestNeighbor, nearest neighbor algorithm)
Method obtains.
After determining the corresponding neighborhood of each scattered points traversed, the points at random with the point cloud model can be obtained
Corresponding neighborhood is measured, a scattered points correspond to a neighborhood.
If in the corresponding neighborhood of each scattered points step S120, traversed, there are dot densities at random to be less than density threshold
The pending neighborhood of value, then in pending neighborhood without determining insertion point in scattered points region；
For each neighborhood, whether the embodiment of the present invention can determine whether the dot density at random in neighborhood less than the density threshold set
Value, if the dot density at random in neighborhood is less than density threshold, which is pending neighborhood, can be from the pending neighborhood really
The insertion point in the multiple cavity of periodical repair；The dot density at random of neighborhood can be obtained by scattered points number in neighborhood divided by neighborhood area；
Setting density threshold may be considered point cloud model in the case that there is no the dot density at random in cavity, concrete numerical values
Visual practical situations setting；
No scattered points region is the zonule in pending neighborhood, and the area in no scattered points region is less than pending neighborhood
Area；No scattered points region may be considered scattered points in pending neighborhood and be distributed sparse region；Optionally, no scattered points area
The size in domain may be considered area of the point cloud model occupied by there is no cavity, a scattered points are average, without scattered
Disorderly the size in point region can regard practical situations and set；
Optionally, insertion point quantity determined by a pending neighborhood is chosen as 1, it is clear that is also not necessarily limited to this value.
Step S130, the point cloud model is repaired according to identified insertion point.
Optionally, correct in identified insertion point, insertion point as defined and the point cloud model
Morphological feature matches, then identified insertion point can be used to repair the cavity of point cloud model.
Point cloud model restorative procedure provided in an embodiment of the present invention can traverse described after determining point cloud model to be repaired
Each scattered points of point cloud model, and determine the corresponding neighborhood of each scattered points traversed；If each traversed
In the corresponding neighborhood of scattered points, there are the pending neighborhoods that dot density at random is less than density threshold, then in the nothing of pending neighborhood
Insertion point is determined in scattered points region；The point cloud model is repaired according to identified insertion point.The embodiment of the present invention by time
Each scattered points for going through point cloud model find dot density at random and wait locating less than density threshold in the corresponding neighborhood of each scattered points
Neighborhood is managed, so that, without insertion point is determined in scattered points region, multiple insertion points are obtained, by obtained in pending neighborhood
The cavity of point cloud model is repaired in multiple insertion points, may make the cavity of point cloud model to reduce, improves the validity of point cloud model.
As can be seen that each scattered points by once traversing point cloud model, it may be determined that go out multiple insertion points and repair point
The cavity of cloud model so that the cavity of point cloud model reduces, and improves the validity of point cloud model；Therefore optional, the present invention is real
Each scattered points of point cloud model can repeatedly be traversed by applying example, and the sky of point cloud model is repaired by way of the multiple insertion point of determination
Hole so that point cloud model tends to no cavity, further promotes the validity of point cloud model, wherein the point cloud mould traversed next time
Type is the point cloud model that insertion point is repaired determined by each scattered points of the point cloud model traversed according to the last time；Traversal time
Number can be set according to practical situations.
Fig. 3 is another flow chart of point cloud model restorative procedure provided in an embodiment of the present invention, and with reference to Fig. 3, this method can
To include：
Step S200, judge the number of each scattered points of current traversal point cloud model, if it is more than setting number, if
It is no, step S210 is executed, if so, executing step S250；
It step S210, will be according to insertion point reparation determined by each scattered points of the point cloud model of last time traversal
Point cloud model, as point cloud model to be repaired；
Optionally, point cloud model is repaired according to insertion point determined by each scattered points of last traversal point cloud model
Process be referred to shown in Fig. 2, even in the corresponding neighborhood of each scattered points of the upper point cloud model once traversed, deposit
When dot density at random is less than the pending neighborhood of density threshold, then it is inserted into without determining in scattered points region in pending neighborhood
Point carries out point cloud model reparation according to identified insertion point；Its substantive processing procedure is consistent with Fig. 2；
Optionally, if current point cloud model is initial traverse, there is no each of last traversal point cloud model
The case where scattered points, can execute follow-up process by current point cloud model directly as point cloud model to be repaired.
Step S220, each scattered points of point cloud model to be repaired are traversed, and determine each scattered points traversed
Corresponding neighborhood；
If in the corresponding neighborhood of each scattered points step S230, traversed, there are dot densities at random to be less than density threshold
The pending neighborhood of value, then in pending neighborhood without determining insertion point in scattered points region；
Step S240, point cloud model to be repaired is repaired according to identified insertion point；
Step S250, terminate flow.
It is worth noting that, the embodiment of the present invention carry out point cloud model repair process in, not by find point
The hole region of cloud model repairs cavity, but by each neighborhood of a point density at random of point cloud model carries out insertion point really
It is fixed, to be distributed sparse region interpolation access point in the scattered points of point cloud model, pass through neighborhood of a point density domination point cloud mould at random
It repairs in the cavity of type；The mode in cavity is repaired compared to the hole region for finding point cloud model, the present invention need not find a cloud
The hole region of model, calculation amount substantially reduce, and repair accuracy higher (the searching calculation amount of the hole region of point cloud model
Greatly, and the decision logic that is related to is extremely complex, and confined condition is more, the case where hole region misjudgment easily occurs).
Optionally, exactness is fixed really without insertion point in scattered points region for the pending neighborhood of raising, the present invention is implemented
Example can set the size in no scattered points region, can specifically set size of the area as no scattered points region；
Corresponding, no scattered points region is the region of the setting area there is no scattered points in pending neighborhood, and setting area, which is less than, to be waited for
Handle the area of neighborhood；Can be in the process without determining insertion point in scattered points region of pending neighborhood, in pending neighbour
There is no determine insertion point in the region of the setting area of scattered points in domain.
Preferably, setting area can choose point cloud model in the case that there is no cavity, and a scattered points are average shared
According to area, concrete numerical value visual practical situations setting.
Fig. 4 shows the region of the setting area provided in an embodiment of the present invention that scattered points are not present in pending neighborhood
The method flow diagram of interior determining insertion point, this method is mainly according to each scattered points in neighborhood, with corresponding traversed scattered points
Position relationship, constrain the position of insertion point so that identified insertion point is in no scattered points region；With reference to Fig. 4, the party
Method may include：
Step S300, each scattered points for determining pending neighborhood, to the midpoint gather of corresponding traversed scattered points；
Step S310, for each point of midpoint gather, choose the point of midpoint gather to pending neighborhood each scattered points
The distance for meeting first condition in distance, obtains the distance set for meeting first condition；
The embodiment of the present invention can select symbol from midpoint gather a little in the distance of each scattered points of pending neighborhood
A distance of first condition is closed, each point of centering point set makees this processing, then can obtain the distance set for meeting first condition；
Optionally, the distance for meeting first condition can be minimum range, specifically, the embodiment of the present invention can choose midpoint
The point of set to pending neighborhood each scattered points apart from minimum value, obtain minimum range set；Obviously, first condition also may be used
To be the second minimum range or third minimum range etc., mainly using the setting sequence of minimum range as the determination of first condition
Strategy.
Step S320, the one of second condition will be met in the distance set for meeting first condition apart from corresponding Point Set
The point of conjunction is determined as insertion point.
Optionally, the distance for meeting second condition can be maximum distance, specifically, the embodiment of the present invention can will meet the
The point of the midpoint gather corresponding to the maximum value in the distance set of one condition is determined as insertion point；Obviously, second condition
It may be the second maximum distance or third maximum distance etc., mainly sorted using the setting of maximum distance as second condition
Determine strategy.
Optionally, there is no determine that the mode of insertion point can in the region of the setting area of scattered points in pending neighborhood
Think：Each scattered points for determining pending neighborhood, to the midpoint gather of corresponding traversed scattered points；For midpoint gather
Each point, determine the point of midpoint gather to pending neighborhood each scattered points apart from minimum value, obtain minimum range set；It will most
The point of the midpoint gather corresponding to the maximum value in small distance set is determined as insertion point.As example, Fig. 5 is shown pair
The insertion point answered determines schematic diagram.
With reference to Fig. 5, by taking neighborhood shape is spherical as an example, for ease of example, Fig. 5 indicates spherical situation with circular flat；It is real
Point v is the scattered points that are traversed, has multiple scattered points N (v) in the spheric neighbo(u)rhood of the setting radius size centered on point v,
Other real points i.e. in figure in addition to point v；It can obtain in the scattered points N (v) in spheric neighbo(u)rhood, the midpoint of each point-to-point v at random
Set Mid (v, N (v)), as shown in hollow dots in figure；For each hollow dots, calculate hollow dots to spheric neighbo(u)rhood in it is at random
Point N (v) apart from minimum value, obtain minimum range set；It will be hollow corresponding to the maximum value in minimum range set
It is the insertion point finally determined that point, which is used as insertion point, i.e. hollow dots m,.
It optionally, can be by pending neighbour after in determining pending neighborhood there is no the region of the setting area of scattered points
There is no the central points in the region of the setting area of scattered points in domain as insertion point, can also be random there is no at random
Insertion point is determined in the region of the setting area of point.
Preferably, the mode that insertion point is determined in the region of the setting area of scattered points is not present in pending neighborhood,
It can be realized by carrying out the region division of setting area to pending neighborhood, search out the division region there is no scattered points
Afterwards, then insertion point is determined in the division region there is no scattered points.
Fig. 6 shows the region of the setting area provided in an embodiment of the present invention that scattered points are not present in pending neighborhood
The another method flow chart of interior determining insertion point, with reference to Fig. 6, this method may include：
Step S400, pending neighborhood is divided with setting area, obtains multiple dividing regions with setting area
Domain；
With reference to Fig. 7, by taking neighborhood shape is spherical as an example, for ease of example, Fig. 7 indicates spherical situation with circular flat；It is real
Point v is the scattered points that are traversed, has multiple scattered points N (v) in the spheric neighbo(u)rhood of the setting radius size centered on point v；
The embodiment of the present invention can set area as the single area size divided, carry out region division to neighborhood, such as Fig. 7 is obtained more
A division region, each size for dividing region are setting area.
If step S410, there is the division region there is no scattered points, in the division region there is no scattered points really
Determine insertion point.
For there are the division regions of scattered points, then the embodiment of the present invention is negligible, does not make the determination of insertion point；For not
There are the division regions of scattered points, can be determined as the center in the division region there is no scattered points, or the point randomly selected
Insertion point；Optionally, may have multiple there is no the division region of scattered points, the embodiment of the present invention can be only at identified one
There is no insertion point is determined in the division region of scattered points, an insertion point is obtained；Obviously, also scattered points can be not present multiple
Division region in respectively determine insertion point, obtain multiple insertion points.
With reference to Fig. 7, after carrying out region division to neighborhood, dash area is then may be used there is no the division region of scattered points
Insertion point, Diamond spot as shown in Figure 7 are determined in the division region there is no scattered points.
Foregoing illustrates a variety of in pending neighborhood without the mode for determining insertion point in scattered points region, noticeable
It is that mode illustrated above is only optional mode, for others by neighborhood density domination strategy, is dissipated in the nothing of pending neighborhood
The mode that insertion point is disorderly determined in point region, is within the scope of the invention.
Optionally, after pending neighborhood in scattered points region without insertion point is determined, the embodiment of the present invention can be according to institute
Point cloud model is repaired in determining insertion point, and benefit point is carried out to point cloud model so that the scattered points quantity of point cloud model increases, and is promoted
The validity of point cloud model.
Further, the morphological feature for insertion point determined by guarantee and point cloud model matches, and the embodiment of the present invention can
Judge whether to repair point cloud model using identified insertion point by the normal orientation of insertion point.Fig. 8 shows this hair
The method flow diagram of point cloud model is repaired in insertion point determined by the basis that bright embodiment provides, and with reference to Fig. 8, this method can wrap
It includes：
Step S500, for identified each insertion point, identified insertion neighborhood of a point is determined；
Optionally, the mode for being inserted into neighborhood of a point is determined, it can be with the mode of the determining neighborhood of a point at random traversed above
It is similar, it can refer to；Flow to simplify the calculation, the embodiment of the present invention can directly be dissipated with corresponding traverse in identified insertion point
Random neighborhood of a point, as the insertion neighborhood of a point.
Step S510, the normal direction of insertion point determined by judgement, with it is identified be inserted into neighborhood of a point normal direction whether phase
Together, if so, executing step S520, if it is not, executing step S530；
Step S520, insertion point determined by reservation, using identified insertion point as new in the point cloud model plus
Enter scattered points；
By regarding identified insertion point as the scattered points in point cloud model, benefit point can be carried out to point cloud model, promoted
The scattered points quantity of point cloud model.
Step S530, insertion point determined by discarding.
A kind of preferred point cloud model restorative procedure flow is provided below, Fig. 9 is provided in an embodiment of the present invention cloud mould
Another flow chart of type restorative procedure, with reference to Fig. 9, this method may include：
Step S600, judge the number of each scattered points of current traversal point cloud model, if it is more than setting number, if
It is no, step S610 is executed, if so, executing step S690；
It step S610, will be according to the point that insertion point is repaired determined by each scattered points of last traversal point cloud model
Cloud model, as point cloud model to be repaired；
If current point cloud model is initial traverse, there is no each scattered points of last traversal point cloud model
Situation, can by the current point cloud model initial point cloud model of object to be repaired (may be) directly as point cloud model to be repaired,
Execute follow-up process.
Step S620, each scattered points of point cloud model to be repaired are traversed, and determine each scattered points traversed
Corresponding neighborhood；
Step 630 judges whether that dot density at random is less than the pending neighborhood of density threshold, if so, executing step
S640, if it is not, executing step S690；
Step S640, for each pending neighborhood, each scattered points of pending neighborhood are determined, are dissipated to corresponding traversed
The midpoint gather disorderly put, for each point of midpoint gather, determine the point of midpoint gather to pending neighborhood each scattered points away from
From minimum value, minimum range set is obtained, the point of the midpoint gather in minimum range set corresponding to the maximum value is true
It is set to insertion point；
Wherein, a pending neighborhood can determine that an insertion point.
Step S650, for identified each insertion point, identified insertion neighborhood of a point is determined；
Step S660, the normal direction of insertion point determined by judgement, with it is identified be inserted into neighborhood of a point normal direction whether phase
Together, if so, executing step S670, if it is not, executing step S680；
Step S670, insertion point determined by reservation, using identified insertion point as new in the point cloud model plus
Enter scattered points；
Step S680, insertion point determined by discarding；
Step S690, terminate flow.
Above-described point cloud model restorative procedure is applicable to arbitrarily carry out mending the stage of point to point cloud model.
Further, when carrying out mending point to point cloud model, the embodiment of the present invention can be divided into filling-up hole stage and encrypting stage, mend
The hole stage is similar with the processing logic of point cloud model restorative procedure that the processing logical AND of encrypting stage is described above, difference
It is the contiguous range selected by the filling-up hole stage, is more than the contiguous range selected by encrypting stage.
Figure 10 is the another flow chart of point cloud model restorative procedure provided in an embodiment of the present invention, referring to Fig.1 0, this method
May include：
Step S700, point cloud model to be repaired is determined；
Step S710, each scattered points of point cloud model to be repaired are traversed, and determine each scattered points traversed
Corresponding filling-up hole neighborhood；
If in the corresponding filling-up hole neighborhood of each scattered points step S720, traversed, there are dot densities at random less than close
The pending filling-up hole neighborhood of threshold value is spent, then in pending filling-up hole neighborhood without determining insertion point in scattered points region；
Step S730, filling-up hole is carried out to point cloud model to be repaired according to identified insertion point；
It is worth noting that, step S700~step S730 is the filling-up hole stage, the filling-up hole stage determines point cloud model to be repaired
Mode can be：Initial point cloud model to be repaired is determined as the point cloud model；If alternatively, having traversed complex point cloud to be repaired
The number of each scattered points of model not up to sets number, then will be dissipated according to each of the point cloud model of last time traversal
The point cloud model that disorderly insertion point determined by point is repaired, as current point cloud model to be repaired；Wherein, the point traversed next time
Cloud model is the point cloud model that the insertion point determined according to each scattered points of the point cloud model of last time traversal is repaired.
Step S740, point cloud model to be encrypted is determined；
Identified point cloud model to be encrypted can be the filling-up hole stage, each for having traversed point cloud model to be repaired is at random
When the number of point reaches setting number, corresponding point cloud model；Or, the filling-up hole stage, each scattered points traversed correspond to
Filling-up hole neighborhood dot density at random be not less than density threshold when, corresponding point cloud model.
Step S750, each scattered points of point cloud model to be encrypted are traversed, and determine each scattered points traversed
Corresponding encryption neighborhood, wherein the range for encrypting neighborhood is less than the range of filling-up hole neighborhood；
Optionally, if neighborhood is regular shape, such as spherical shape, then the half of filling-up hole neighborhood need to be less than by encrypting the radius of neighborhood
Diameter.
If in the corresponding encryption neighborhood of each scattered points step S760, traversed, there are dot densities at random less than close
The pending encryption neighborhood of threshold value is spent, then in pending encryption neighborhood without determining insertion point in scattered points region；
Step S770, the point cloud model to be encrypted is repaired according to identified insertion point.
As can be seen that step S740~step S760 is encrypting stage, it is to after step S700~step S730 filling-up holes
Point cloud model is encrypted, and the processing logic of filling-up hole stage and encrypting stage is almost the same, and the result of the two is regarded as pair
Point cloud model carries out benefit point；It can refer to the description of the parts Fig. 2~Fig. 9 above.Filling-up hole stage and encrypting stage difference lies in：It mends
Contiguous range selected by the stage of hole is larger, first insertion point is determined by neighborhood density domination in wide range, to be mended
Point；And the contiguous range selected by encrypting stage is smaller, it is further close by neighborhood in smaller range after the filling-up hole stage
Degree control determines insertion point, to promote the quantity for mending point.
One application of point cloud model restorative procedure provided in an embodiment of the present invention is, to the building in streetscape point cloud model
Object point cloud model carries out empty reparation.Streetscape point cloud model is mainly made of building point cloud model and near-earth point cloud model, this
Inventive embodiments can be after being partitioned into building point cloud model, to the detached building point cloud being partitioned into streetscape point cloud model
Model carries out empty reparation；Repair mode can refer to the description of the parts Fig. 2~Fig. 9 above, it is clear that filling-up hole and encryption can also be used
In conjunction with mode carry out empty reparation.
Below in a manner of being combined using filling-up hole and encryption, the repair process of building point cloud model is introduced：
1, building point cloud model is partitioned into from streetscape point cloud model, the building point cloud model being partitioned into should be only
Vertical；Specifically, can determine the building point cloud model in streetscape point cloud model, by the building point cloud model of different height into
Row segmentation, mutually level building point cloud model is mutually clustered, at least one independent building point cloud model is obtained；
2, after being partitioned into independent building point cloud model, each scattered points of building point cloud model can be traversed,
And determine the corresponding filling-up hole neighborhood of each scattered points traversed；Optionally, the first radius can be set and determine that each is dissipated
Disorderly the corresponding spherical filling-up hole neighborhood of point, the first radius of setting can choose the half of building story height, it is clear that concrete numerical value
It also can be depending on actually using situation；
If in the corresponding filling-up hole neighborhood of each scattered points 3, traversed, there are dot densities at random less than density threshold
Pending filling-up hole neighborhood, then in pending filling-up hole neighborhood without determining insertion point in scattered points region；Optional method of determination can be such as
Shown in Fig. 4 or Fig. 6；
4, the normal direction of insertion point is compared with its neighborhood normal direction, is retained and the identified normal direction for being inserted into neighborhood of a point
Identical insertion point carries out filling-up hole by the insertion point retained；
5, cycle executes above-mentioned 2~4 step, until the number of each scattered points of traversal building point cloud model reaches
When setting number, alternatively, the dot density at random of the corresponding filling-up hole neighborhood of each scattered points traversed is not less than density threshold
When value, 6 are entered step；It should be noted that the building point cloud model traversed is, according to the building of last time traversal next time
The insertion point that each scattered points of object point cloud model determine carries out the building point cloud model after filling-up hole；
6, by the building point cloud model after the execution obtained filling-up hole of step 5, as building point cloud model to be encrypted；
7, each scattered points of point cloud model to be encrypted are traversed, and determines that each scattered points traversed are corresponding and adds
Close neighborhood, wherein the range for encrypting neighborhood is less than the range of filling-up hole neighborhood；Optionally, can set the second radius determine it is each
The corresponding spherical encryption neighborhood of a scattered points, the second radius of setting are less than the first radius of setting, and the second radius of setting can be chosen
0.5 meter, it is clear that concrete numerical value also can be depending on actually using situation；
If in the corresponding encryption neighborhood of each scattered points 8, traversed, there are dot densities at random less than density threshold
Pending encryption neighborhood, then in pending encryption neighborhood without determining insertion point in scattered points region；Optional method of determination can be such as
Shown in Fig. 4 or Fig. 6；
9, the normal direction of insertion point is compared with its neighborhood normal direction, is retained and the identified normal direction for being inserted into neighborhood of a point
Identical insertion point is encrypted by the insertion point retained；
10, cycle executes above-mentioned 7~9 step, until the number of each scattered points of traversal building point cloud model reaches
When to setting number, alternatively, the dot density at random of the corresponding encryption neighborhood of each scattered points traversed is not less than density
When threshold value, the building point cloud model after being repaired；It should be noted that the building point cloud model traversed next time is,
Building object point after being encrypted according to the insertion point of each scattered points determination of the building point cloud model of last time traversal
Cloud model；
11, due to the object of near-earth point cloud model be distributed it is especially complex, mainly have trees, pedestrian, vehicle, billboard and
The objects such as various traffic signboards；This partial data is most of all than more complete, but due to the speed of collecting vehicle operation is uneven
The even scattered points that can cause near-earth point cloud model are unevenly distributed；Therefore near-earth point cloud model can be added after step 10
Close processing improves the quality of streetscape point cloud model；To the process that near-earth point cloud model is encrypted, and to building point cloud model
The process being encrypted is similar, can phase reference.
The embodiment of the present invention is found and is dissipated in the corresponding neighborhood of each scattered points by traversing each scattered points of point cloud model
Random dot density be less than density threshold pending neighborhood, to pending neighborhood without in scattered points region determine insertion point,
Multiple insertion points are obtained, the cavity of point cloud model is repaired by obtained multiple insertion points, may make the cavity of point cloud model
Reduce, improves the validity of point cloud model；Meanwhile the present invention passes through the cavity of neighborhood of a point density domination point cloud model at random
It repaiies, the hole region of point cloud model need not be found, calculation amount substantially reduces, and repairs accuracy higher.
Point cloud model prosthetic device provided in an embodiment of the present invention is introduced below, point cloud model described below is repaiied
Apparatus for coating can correspond reference with above-described point cloud model restorative procedure.
Figure 11 is the structure diagram of point cloud model prosthetic device provided in an embodiment of the present invention, which can be applied to have
The computing device that data processing can be carried out to point cloud model of data-handling capacity；Referring to Fig.1 1, which may include：
Model determining module 100, for determining point cloud model to be repaired；
Neighborhood determining module 200, each scattered points for traversing the point cloud model, and determination are traversed each
The corresponding neighborhood of a scattered points；
Insertion point determining module 300, if in the corresponding neighborhood of each scattered points for being traversed, it is close that there are scattered points
Degree is less than the pending neighborhood of density threshold, then in pending neighborhood without determining insertion point in scattered points region；
Repair module 400, for repairing the point cloud model according to identified insertion point.
Optionally, no scattered points region can be the region of the setting area there is no scattered points in pending neighborhood；Institute
State the area that setting area is less than pending neighborhood；Figure 12 shows insertion point determining module 300 provided in an embodiment of the present invention
A kind of alternative construction, referring to Fig.1 2, insertion point determining module 300 may include：
Insertion point determines execution unit 310, for there is no the regions of the setting area of scattered points in pending neighborhood
Interior determining insertion point.
Optionally, Figure 13 shows that insertion point provided in an embodiment of the present invention determines a kind of optional knot of execution unit 310
Structure, referring to Fig.1 3, insertion point determines that execution unit 310 may include：
Midpoint gather determination subelement 3101, each scattered points for determining pending neighborhood, to corresponding traversed
The midpoint gather of scattered points；
Subelement 3102 is chosen, for each point for midpoint gather, chooses the point of midpoint gather to pending neighborhood
The distance for meeting first condition in the distance of each scattered points, obtains the distance set for meeting first condition；
Optionally, the distance for meeting first condition can be minimum range.
Subelement 3103 is chosen in insertion point, for will meet in the distance set for meeting first condition the one of second condition away from
Point from corresponding midpoint gather is determined as insertion point；
Optionally, the distance for meeting second condition can be maximum distance.
Optionally, Figure 14 shows that insertion point provided in an embodiment of the present invention determines that the another kind of execution unit 310 is optional
Structure, referring to Fig.1 4, insertion point determines that execution unit 310 may include：
Subelement 3111 is divided, for being divided to pending neighborhood with setting area, is obtained multiple with setting face
Long-pending division region；
Divide region in determination subelement 3112, if for exist there is no scattered points division region, there is no
Insertion point is determined in the division region of scattered points.
Optionally, Figure 15 shows a kind of alternative construction of repair module 400 provided in an embodiment of the present invention, with reference to figure
15, repair module 400 may include：
It is inserted into vertex neighborhood determination unit 410, for for identified each insertion point, determining the neighbour of identified insertion point
Domain；
Normal direction judging unit 420, the normal direction for judging identified insertion point, with identified insertion neighborhood of a point
Whether normal direction is identical；
Stick unit 430, if the normal direction for identified insertion point, with the identified normal direction phase for being inserted into neighborhood of a point
Together, then insertion point determined by retaining, using identified insertion point as the new addition scattered points in the point cloud model；
Discarding unit 440, if the normal direction for identified insertion point, not with the identified normal direction for being inserted into neighborhood of a point
Together, then insertion point determined by abandoning.
Optionally, model determining module 100 determines that the process of point cloud model to be repaired can be specific：If currently traversing a little
The number of each scattered points of cloud model not up to sets number, then will be dissipated according to each of last traversal point cloud model
The point cloud model that disorderly insertion point determined by point is repaired, as point cloud model to be repaired.Obviously, if current point cloud model is initial
The case where traversing, then each scattered points of last traversal point cloud model are not present, can directly make current point cloud model
For point cloud model to be repaired.
Optionally, the mode of filling-up hole combining encryption can be used in the embodiment of the present invention, is repaired to point cloud model.Figure 11 institutes
Show that function module can execute the filling-up hole stage；Specifically, repair module 400 can be according to identified insertion point to described cloud mould
Type carries out filling-up hole, corresponding, and neighborhood is filling-up hole neighborhood；Point cloud model to be repaired determined by model determining module 100 can be,
Initial point cloud model to be repaired, alternatively, the number for having traversed each scattered points of point cloud model to be repaired is not up to set
When number, according to the point cloud model that insertion point is repaired determined by each scattered points of the point cloud model of last time traversal；Its
In, the point cloud model traversed next time is the insertion point determined according to each scattered points of the point cloud model of last time traversal
The point cloud model repaired.On this basis, Figure 16 shows the another of point cloud model prosthetic device provided in an embodiment of the present invention
One structure diagram, referring to Fig.1 6, which can also include：
Point cloud model determining module 500 to be encrypted, for determining point cloud model to be encrypted；
Optionally, point cloud model to be encrypted can be to have been traversed to be repaired in the stage for carrying out filling-up hole to the point cloud model
When the number of each scattered points of complex point cloud model reaches setting number, corresponding point cloud model, or, to described cloud
Model carries out the stage of filling-up hole, and the dot density at random of the corresponding filling-up hole neighborhood of each scattered points traversed is not less than density
When threshold value, corresponding point cloud model.
Neighborhood determining module 600, each scattered points for traversing point cloud model to be encrypted are encrypted, and determination is traversed
The corresponding encryption neighborhood of each scattered points, wherein encrypt neighborhood range be less than filling-up hole neighborhood range
Insertion point determining module 700 is encrypted, if in the corresponding encryption neighborhood of each scattered points for being traversed, is existed
Dot density at random is less than the pending encryption neighborhood of density threshold, then in pending encryption neighborhood without determining in scattered points region
Insertion point；
Repair module 800 is encrypted, for repairing the point cloud model to be encrypted according to identified insertion point.
Optionally, the structure of encryption insertion point determining module 700 can be similar with the structure of insertion point determining module 300, can
Phase reference；The structure for encrypting repair module 800 can be similar with the structure of repair module 400 shown in Figure 15, can phase reference.
Optionally, point cloud model prosthetic device provided in an embodiment of the present invention can be applied to building in streetscape point cloud model
It builds object point cloud model and carries out empty reparation；The embodiment of the present invention can be partitioned into building point cloud model is by streetscape point cloud model
Afterwards, empty reparation is carried out to independent building point cloud model.Figure 17 shows point cloud models provided in an embodiment of the present invention to repair
Another structure diagram of apparatus for coating, in conjunction with shown in Figure 11 and Figure 17, which can also include：
Divide module 900, for determining the building point cloud model in streetscape point cloud model, by the building of different height
Point cloud model is split, and mutually level building point cloud model is mutually clustered, and obtains at least one building point cloud model；
Obtained building point cloud model can carry out empty reparation by point cloud model prosthetic device shown in Figure 11, also may be used
By point cloud model prosthetic device shown in Figure 16, the mode of filling-up hole combining encryption is used.
Further, after being repaired to building point cloud model, the embodiment of the present invention can also to near-earth point cloud model into
Row encryption.
The embodiment of the present invention also provides a kind of computing device, which may include that point cloud model described above is repaiied
Apparatus for coating；The computing device, can be by each scattered points of traversal point cloud model, in each scattered points when carrying out point cloud model reparation
In corresponding neighborhood, the pending neighborhood that dot density at random is less than density threshold is found, thus in pending neighborhood without at random
Point determines insertion point in region, obtains multiple insertion points, and the cavity of point cloud model is repaired by obtained multiple insertion points, can
So that the cavity of point cloud model reduces, the validity of point cloud model is improved；Meanwhile the computing device is close by neighborhood of a point at random
The cavity of degree control point cloud model is repaiied, and the hole region of point cloud model need not be found, and calculation amount substantially reduces, and repairs accurately
Spend higher.
Figure 18 is the hardware block diagram of computing device provided in an embodiment of the present invention, and computing device can be such as intelligent hand
The network side equipments such as user equipmenies or server such as machine, PC (personal computer)；Referring to Fig.1 8, calculating may include：Processor
1, communication interface 2, memory 3 and communication bus 4；
Wherein processor 1, communication interface 2, memory 3 complete mutual communication by communication bus 4；
Optionally, communication interface 2 can be the interface of communication module, such as the interface of gsm module；
Processor 1, for executing program；
Memory 3, for storing program；
Program may include program code, and said program code includes computer-managed instruction.
Processor 1 may be a central processor CPU or specific integrated circuit ASIC (Application
Specific Integrated Circuit), or be arranged to implement the integrated electricity of one or more of the embodiment of the present invention
Road.
Memory 3 may include high-speed RAM memory, it is also possible to further include nonvolatile memory (non-volatile
Memory), a for example, at least magnetic disk storage.
Wherein, program can be specifically used for：
Determine point cloud model to be repaired；
Each scattered points of the point cloud model are traversed, and determine the corresponding neighborhood of each scattered points traversed；
If in the corresponding neighborhood of each scattered points traversed, it is pending less than density threshold that there are dot densities at random
Neighborhood, then in pending neighborhood without determining insertion point in scattered points region；
The point cloud model is repaired according to identified insertion point.
Each embodiment is described by the way of progressive in this specification, the highlights of each of the examples are with other
The difference of embodiment, just to refer each other for identical similar portion between each embodiment.For device disclosed in embodiment
For, since it is corresponded to the methods disclosed in the examples, so description is fairly simple, related place is said referring to method part
It is bright.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is implemented in hardware or software actually, depends on the specific application and design constraint of technical solution.Profession
Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered
Think beyond the scope of this invention.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest range caused.
Claims (15)
1. a kind of point cloud model restorative procedure, which is characterized in that including：
Determine point cloud model to be repaired；
Each scattered points of the point cloud model are traversed, and determine the corresponding neighborhood of each scattered points traversed；
If in the corresponding neighborhood of each scattered points traversed, there are the pending neighbours that dot density at random is less than density threshold
Domain, then in pending neighborhood without determining insertion point in scattered points region；
The point cloud model is repaired according to identified insertion point.
2. point cloud model restorative procedure according to claim 1, which is characterized in that the no scattered points region is pending
There is no the regions of the setting area of scattered points in neighborhood；The setting area is less than the area of pending neighborhood；
It is described to include without determining insertion point in scattered points region in pending neighborhood：
There is no determine insertion point in the region of the setting area of scattered points in pending neighborhood.
3. point cloud model restorative procedure according to claim 2, which is characterized in that described to be not present in pending neighborhood
Determine that insertion point includes in the region of the setting area of scattered points：
Each scattered points for determining pending neighborhood, to the midpoint gather of corresponding traversed scattered points；
For each point of midpoint gather, chooses in the point to the distance of each scattered points of pending neighborhood of midpoint gather and meet first
One distance of condition, obtains the distance set for meeting first condition；
The point apart from corresponding midpoint gather for meeting second condition in the distance set for meeting first condition is determined as
Insertion point.
4. point cloud model restorative procedure according to claim 2, which is characterized in that described to be not present in pending neighborhood
Determine that insertion point includes in the region of the setting area of scattered points：
Pending neighborhood is divided with setting area, obtains multiple division regions with setting area；
If there is the division region there is no scattered points, insertion point is determined in the division region there is no scattered points.
5. according to claim 1-4 any one of them point cloud model restorative procedures, which is characterized in that determined by the basis
Repair the point cloud model in insertion point：
For identified each insertion point, identified insertion neighborhood of a point is determined；
Whether the normal direction of insertion point determined by judging is identical as the identified normal direction for being inserted into neighborhood of a point；
It is identical as the identified insertion normal direction of neighborhood of a point if the normal direction of identified insertion point, then retain identified insert
Access point, using identified insertion point as the new addition scattered points in the point cloud model；
It is different from the identified insertion normal direction of neighborhood of a point if the normal direction of identified insertion point, then abandon identified insert
Access point.
6. point cloud model restorative procedure according to claim 1, which is characterized in that the determination point cloud model packet to be repaired
It includes：
If currently the number of each scattered points of traversal point cloud model not up to sets number, point will be traversed according to the last time
The point cloud model that insertion point determined by each scattered points of cloud model is repaired, as point cloud model to be repaired.
7. point cloud model restorative procedure according to claim 1, which is characterized in that insertion point is repaiied determined by the basis
The point cloud model includes again：
Filling-up hole is carried out to the point cloud model according to identified insertion point, wherein the neighborhood is filling-up hole neighborhood；
The determination point cloud model to be repaired includes：
Initial point cloud model to be repaired is determined as to current point cloud model to be repaired；
If or, traversed each scattered points of point cloud model to be repaired number not up to set number, will be according to upper one
The point cloud model that insertion point determined by each scattered points of the point cloud model of secondary traversal is repaired, as current complex point to be repaired
Cloud model；Wherein, the point cloud model traversed next time is to be determined according to each scattered points of the point cloud model of last time traversal
The point cloud model repaired of insertion point；
The method further includes：
Determine point cloud model to be encrypted；Wherein, the point cloud model to be encrypted is in the rank for carrying out filling-up hole to the point cloud model
Section, traversed each scattered points of point cloud model to be repaired number reach setting number when, corresponding point cloud model,
Or, in the stage for carrying out filling-up hole to the point cloud model, the scattered points of the corresponding filling-up hole neighborhood of each scattered points traversed
When density is not less than the density threshold, corresponding point cloud model；
Each scattered points of point cloud model to be encrypted are traversed, and determine that the corresponding encryption of each scattered points traversed is adjacent
Domain, wherein the range for encrypting neighborhood is less than the range of filling-up hole neighborhood；
If in the corresponding encryption neighborhood of each scattered points traversed, there are dot density at random waiting for less than the density threshold
Processing encryption neighborhood, then in pending encryption neighborhood without determining insertion point in scattered points region；
The point cloud model to be encrypted is repaired according to identified insertion point.
8. point cloud model restorative procedure according to claim 1, which is characterized in that the point cloud model is building object point cloud
Model；By being partitioned into streetscape point cloud model, the streetscape point cloud model includes the building point cloud model：Build object point cloud
Model and near-earth point cloud model；
The process that building point cloud model is partitioned into the point cloud model by streetscape includes：
Determine the building point cloud model in streetscape point cloud model；
The building point cloud model of different height is split, mutually level building point cloud model is mutually clustered, is obtained
At least one building point cloud model.
9. a kind of point cloud model prosthetic device, which is characterized in that including：
Model determining module, for determining point cloud model to be repaired；
Neighborhood determining module, each scattered points for traversing the point cloud model, and determine that each traversed is at random
The corresponding neighborhood of point；
Insertion point determining module, if in the corresponding neighborhood of each scattered points for being traversed, there are dot densities at random to be less than
The pending neighborhood of density threshold, then in pending neighborhood without determining insertion point in scattered points region；
Repair module, for repairing the point cloud model according to identified insertion point.
10. point cloud model prosthetic device according to claim 9, which is characterized in that the no scattered points region is to wait locating
Manage the region of the setting area in neighborhood there is no scattered points；The setting area is less than the area of pending neighborhood；It is described to insert
Access point determining module includes：
Insertion point determines execution unit, for there is no determine to insert in the region of the setting area of scattered points in pending neighborhood
Access point.
11. point cloud model prosthetic device according to claim 10, which is characterized in that the insertion point determines execution unit
Including：
Midpoint gather determination subelement, each scattered points for determining pending neighborhood, to corresponding traversed scattered points
Midpoint gather；
Subelement is chosen, for each point for midpoint gather, chooses the point of midpoint gather to each scattered points of pending neighborhood
Distance in meet a distance of first condition, obtain the distance set for meeting first condition；
Subelement is chosen in insertion point, for will meet in the distance set for meeting first condition corresponding to a distance of second condition
The point of midpoint gather be determined as insertion point.
12. point cloud model prosthetic device according to claim 10, which is characterized in that the insertion point determines execution unit
Including：
Subelement is divided, for being divided to pending neighborhood to set area, obtains multiple divisions with setting area
Region；
Determination subelement in region is divided, if for there is the division region there is no scattered points, there is no scattered points
It divides and determines insertion point in region.
13. according to claim 9-12 any one of them point cloud model prosthetic devices, which is characterized in that the repair module packet
It includes：
It is inserted into vertex neighborhood determination unit, for for identified each insertion point, determining identified insertion neighborhood of a point；
Normal direction judging unit, the normal direction for judging identified insertion point are with the identified normal direction for being inserted into neighborhood of a point
It is no identical；
Stick unit is then protected if the normal direction for identified insertion point, identical as the identified insertion normal direction of neighborhood of a point
Identified insertion point is stayed, using identified insertion point as the new addition scattered points in the point cloud model；
Discarding unit is then lost if the normal direction for identified insertion point, different from the identified insertion normal direction of neighborhood of a point
Abandon identified insertion point.
14. point cloud model prosthetic device according to claim 9, which is characterized in that the repair module is specifically used for root
Filling-up hole is carried out to the point cloud model according to identified insertion point；The neighborhood is filling-up hole neighborhood；
The model determining module is specifically used for for initial point cloud model to be repaired being determined as current point cloud model to be repaired；
If, will be according to last time or, the number for having traversed each scattered points of point cloud model to be repaired not up to sets number
The point cloud model that insertion point determined by each scattered points for the point cloud model gone through is repaired, as current complex point cloud mould to be repaired
Type；Wherein, the point cloud model traversed next time is to be inserted according to what each scattered points of the point cloud model of last time traversal determined
The point cloud model that access point is repaired；
The point cloud model prosthetic device further includes：
Point cloud model determining module to be encrypted, for determining point cloud model to be encrypted；Wherein, the point cloud model to be encrypted be
The stage of filling-up hole is carried out to the point cloud model, the number for having traversed each scattered points of point cloud model to be repaired reaches setting
When number, corresponding point cloud model, or, in the stage for carrying out filling-up hole to the point cloud model, each traversed is at random
When the dot density at random of the corresponding filling-up hole neighborhood of point is not less than the density threshold, corresponding point cloud model；
Encryption neighborhood determining module, each scattered points for traversing point cloud model to be encrypted, and determination are traversed each
The corresponding encryption neighborhood of a scattered points, wherein the range for encrypting neighborhood is less than the range of filling-up hole neighborhood
Insertion point determining module is encrypted, if in the corresponding encryption neighborhood of each scattered points for being traversed, there are scattered points
Density is less than the pending encryption neighborhood of the density threshold, then slotting without being determined in scattered points region in pending encryption neighborhood
Access point；
Repair module is encrypted, for repairing the point cloud model to be encrypted according to identified insertion point.
15. a kind of computing device, which is characterized in that including claim 9-14 any one of them point cloud model prosthetic devices.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510111559.6A CN106033620B (en) | 2015-03-13 | 2015-03-13 | A kind of point cloud model restorative procedure, device and computing device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510111559.6A CN106033620B (en) | 2015-03-13 | 2015-03-13 | A kind of point cloud model restorative procedure, device and computing device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106033620A CN106033620A (en) | 2016-10-19 |
| CN106033620B true CN106033620B (en) | 2018-10-19 |
Family
ID=57150602
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510111559.6A Active CN106033620B (en) | 2015-03-13 | 2015-03-13 | A kind of point cloud model restorative procedure, device and computing device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106033620B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112287907B (en) * | 2020-12-24 | 2021-03-19 | 湖南联智科技股份有限公司 | Hole identification method based on point cloud density gradient |
| CN112767554B (en) * | 2021-04-12 | 2021-07-16 | 腾讯科技（深圳）有限公司 | Point cloud completion method, device, equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101127123A (en) * | 2007-09-11 | 2008-02-20 | 东南大学 | Sign point hole filling method based on neural network in tri-D scanning point cloud |
| CN104282039A (en) * | 2014-09-29 | 2015-01-14 | 樊晓莉 | Skeleton orthosis brace shaping method based on 3D scanning |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9767598B2 (en) * | 2012-05-31 | 2017-09-19 | Microsoft Technology Licensing, Llc | Smoothing and robust normal estimation for 3D point clouds |
| US8874454B2 (en) * | 2013-03-15 | 2014-10-28 | State Farm Mutual Automobile Insurance Company | Systems and methods for assessing a roof |
-
2015
- 2015-03-13 CN CN201510111559.6A patent/CN106033620B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101127123A (en) * | 2007-09-11 | 2008-02-20 | 东南大学 | Sign point hole filling method based on neural network in tri-D scanning point cloud |
| CN104282039A (en) * | 2014-09-29 | 2015-01-14 | 樊晓莉 | Skeleton orthosis brace shaping method based on 3D scanning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106033620A (en) | 2016-10-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zheng et al. | Traditional soil particle sphericity, roundness and surface roughness by computational geometry | |
| CN112465948B (en) | Vehicle-mounted laser pavement point cloud rarefying method capable of retaining spatial features | |
| JP5053705B2 (en) | Optimal route determination using estimation function | |
| CN102890828B (en) | Point cloud data compacting method based on normal included angle | |
| CN106548520A (en) | A kind of method and system of cloud data denoising | |
| CN106340061B (en) | A kind of mountain area point cloud filtering method | |
| CN110109142A (en) | Point cloud filtering method, device, computer equipment and storage medium | |
| JP2008186440A (en) | Topology evolution optimization computing method for structural design | |
| CN110906940B (en) | Lane sideline aggregation method based on track direction | |
| CN108919954B (en) | Dynamic change scene virtual and real object collision interaction method | |
| CN102750730B (en) | Characteristic-maintained point cloud data compacting method | |
| CN114529633B (en) | Method for supporting continuous LOD (level of detail) drawing of GIS (geographic information system) line object and surface object | |
| CN106033620B (en) | A kind of point cloud model restorative procedure, device and computing device | |
| CN109506672A (en) | A kind of acquisition methods and device of pavement markers laser point cloud | |
| CN115578536A (en) | Node merging method and device for layered and partitioned three-dimensional model and electronic device | |
| CN102800114B (en) | Data point cloud downsizing method based on Poisson-disk sampling | |
| CN109087344B (en) | Image selection method and device in three-dimensional reconstruction | |
| Wu et al. | Building reconstruction from high-resolution multiview aerial imagery | |
| US20050107992A1 (en) | Method and program of converting three-dimensional shape data into cell internal data | |
| CN113191311A (en) | Filling boundary identification method, device and equipment of vector PDF drawing and storage medium | |
| CN116764225A (en) | Efficient path-finding processing method, device, equipment and medium | |
| CN108446343B (en) | Method and device for area aggregation and electronic equipment | |
| CN116168384A (en) | Point cloud target detection method and device, electronic equipment and storage medium | |
| CN115457493A (en) | Target detection method, target detection device, computer equipment and storage medium | |
| CN114926590A (en) | Mass point cloud data visualization method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230706 Address after: 518000 Tencent Building, No. 1 High-tech Zone, Nanshan District, Shenzhen City, Guangdong Province, 35 Floors Patentee after: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. Patentee after: TENCENT CLOUD COMPUTING (BEIJING) Co.,Ltd. Address before: 2, 518000, East 403 room, SEG science and Technology Park, Zhenxing Road, Shenzhen, Guangdong, Futian District Patentee before: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. |