Summary of the invention
For the deficiencies in the prior art, an object of the present invention is to solve present in above-mentioned prior art
One or more problems.Such as, an object of the present invention be for original earthquake data amount very big in the case of propose one
Kind new curve reestablishing method, completes the process of mass data effectively utilizing limited internal memory.
To achieve these goals, the invention provides the curve reestablishing method of a kind of extensive geological data.Described side
First method carries out the terrain blocks of extensive geological data, specifically includes and layer bit data and geological constraining polygon data is divided
Enter in different terrain block and preserve to external memory;Then curve reestablishing is carried out with each terrain block for ultimate unit;Wherein, described
Terrain blocks step uses the mode dynamically adding terrain block, specifically: given initial landform block's attribute, described terrain block attribute
Including length, X-direction terrain block number and the Y-direction terrain block number of terrain block, wherein, X-direction terrain block number and Y-direction
The minima of terrain block number is 1;If the geological data read in is in initial given terrain block, then geological data is distributed
In terrain block;If the geological data read in is outside initial given terrain block, then initial given terrain block is carried out
Expand with the geological data comprising reading.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method reads extensive
Predetermined row in geological data, and the geological data of described predetermined row is carried out terrain blocks process, then repeat described
Read and terrain blocks operates, until extensive geological data is read and has processed.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described to initial given ground
Shape block ensures when expanding that whole region is rectangular.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, in described preservation step,
One id is set for each terrain block, and uses the mode of two-dimensional array to carry out the id index of terrain block, determine a two-dimemsional number
Group a [M] [N], the X-direction of the one-dimensional representation terrain block of wherein said two-dimensional array, the two-dimensional representation landform of two-dimensional array
The Y-direction of block, the value in two-dimensional array is the id of terrain block.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described curve reestablishing step can
To include: the external memory from respective path reads the layer bit data within terrain block and geological constraining polygon data, then
In terrain block, build initial mesh, in each initial mesh, carry out Constrained Delaunay triangulation and carry out three dimensions
The triangulation network recovers, and obtains three-dimensional topology, finally the point on three-dimensional topology is carried out surface interpolation, obtain three dimensional geological curved surface.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, to geology constrained polygon
When data carry out piecemeal process, when the polygon form point of only one of which geological constraining in a terrain block, obtain two adjacent geology
The line segment that constrained polygon point is linked to be and the intersection point on terrain block border, and add the geological constraining polygon in the terrain block of correspondence
In data.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method also includes retaining
A overall closed polygon data, and, when each terrain block is carried out curved surface reconstruction operation, closing will be called many
Limit graphic data carries out attributive judgment.Specifically:
Definition closed polygon data acquisition system P:
P={P1, P2..., PmEquation (1)
Wherein
Pi={ upi, lpi, idiEquation (2)
In equation (1) and equation (2), P represents the set of closed polygon, PiRepresent in closed polygon set P
I-th closed polygon, 1≤i≤m, m are total number of closed polygon, upiRepresent that the upper dish of i-th closed polygon is many
Limit form point set, lpiRepresent the lower wall polygon point set of i-th closed polygon, idiRepresent the root of i-th closed polygon
Polygon id.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method can also include
Introduce the index attributes of the polygon form point of geological constraining, to identify the polygon form point of geological constraining in whole geology constrained polygon
Position, and it is used for judging between the polygon form point of geological constraining the most adjacent, only index adjacent two geological constraining polygon
Put and could connect formation binding side, specifically, definition geological constraining polygon point set:
Vp={ V1, V2..., VnEquation (3)
Wherein
Vj={ Xj, Yj, IDj, IndexjEquation (4)
In equation (3) and equation (4), VpRepresent the polygon form point of all geological constraining on a geological constraining polygon
Set, VjRepresenting the polygon form point of jth geological constraining on geological constraining polygon, 1≤j≤n, n are that a geological constraining is polygon
Total number of geology constrained polygon point, X in shapejAnd YjRepresent the position of the polygon form point of geological constraining, IDjRepresent jth geology
The geological constraining polygon at constrained polygon point place ID, Index in geological constraining polygon setjRepresent jth ground
Matter constrained polygon point location index in geological constraining polygon.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method can also include
Opening up buffer area at internal memory, in curve reestablishing operates, the mutual and internal memory performing external memory interval with caching is interval with caching
Mutual, specifically, when carrying out the scheduling of buffer area terrain block, first determine whether currently processed terrain block and neighbouring terrain block thereof
Whether read in buffer area, if finding currently processed terrain block in buffer area and closing on terrain block, then directly from buffer area
In choose corresponding terrain block;If not finding currently processed terrain block in buffer area and closing on terrain block, then judge slow
Deposit district the fullest, if buffer area less than and buffer area remaining space enough, from external memory, the data block of needs is added to slow
Deposit in district;If buffer area is full or buffer area remaining space is less than the space processed required for current block, then use terrain block
Replacement Strategy, by the terrain block removal buffer area farthest apart from currently processed terrain block, by facing of currently processed terrain block
Nearly block moves into buffer area.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, it is right that described interpolation procedure includes
The non-boundary point of terrain block and boundary point carry out degree of depth interpolation, wherein, non-boundary point carry out the step of degree of depth interpolation for net
Lattice point interpolation, then tomography polygon is clicked on row interpolation, finally compensate interpolation;The boundary point of terrain block carries out the degree of depth insert
During value, all points with current border point Topology connection will participate in the degree of depth interpolation of current border point as seed points
Calculate;Further, boundary point is added interpolation whether attribute, the attribute of the boundary point after interpolation is set to interpolation, to
The point of interpolation does not carry out quadratic interpolation.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method can also include
Before building initial mesh step in described terrain block, the layer bit data in terrain block and geological constraining polygon data are entered
Row pretreatment.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method is to different ground
Shape block, determines different resolution according to the distribution of landform, specifically, uses energy function method to calculate the resolution that terrain block is initial
Rate, is then finely adjusted the tomography distribution situation in the resolution value Combining with terrain block calculated, if a terrain block
The tomography constrained polygon of interior existence is the most, and resolution is the highest.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method uses discrete
LOD model carries out curve reestablishing.
One embodiment of the curve reestablishing method according to the extensive geological data of the present invention, described method uses
Constrained Delaunay retrains triangulation, such as, circulates method of changing sides.
Compared with prior art, the beneficial effect comprise that efficiently solving separate unit PC internal memory is difficult to
The problem that the survey data in whole extensive geology work area is carried out overall curve reestablishing.
Detailed description of the invention
Hereinafter, accompanying drawing will be combined and exemplary embodiment describes the extensive geological data according to the present invention in detail
Curve reestablishing method.
First the curve reestablishing method of extensive according to an exemplary embodiment of the present invention geological data is carried out on a large scale
The terrain blocks of shake data, is then that ultimate unit carries out curve reestablishing with each terrain block (referred to as block).Main innovation point
It is that the method using dynamic terrain blocks and the three dimensions triangulation network to recover solves the curve reestablishing difficulty of extensive geological data
Topic.
The problem that present invention mainly solves is: has first had to the terrain blocks of extensive geological data, has been directed to
The piecemeal of layer bit data and the piecemeal of constrained polygon data.Secondly, for graphic data, the many resolutions of introducing the most simplifiedly
Concept, the data resolution of different terrain block is relevant to the landform in concrete block, and adjacent terrain block is likely to be due to resolution
Difference, and produce crack phenomenon, it should take certain measure to be avoided.Again, whole extensive geology work area is drawn
It is divided into terrain block, although solve extensive geologic data lower surface camber and rebuild the problem of low memory, but terrain block quantity is false
As too much, will certainly cause between internal memory and hard disk is the most mutual, accordingly, it is determined that a suitable scheduling memory algorithm, with reality
The curve reestablishing process of existing relative efficiency is also the most necessary.Finally, the interpolation problem on border between multiple terrain block will be solved.
For ease of understanding, applicant simply introduces with regard to the theory involved by the application.
1, Delaunay Triangulation technology
Delaunay Triangulation is defined as follows: assuming that Plane-point collection is combined into V, limit collection is combined into E, for appointing in E
Anticipate a limit e, if there is circle C, only comprises two end points of e, and does not comprise other arbitrfary points, then limit e is one
Delaunay limit.Assume the triangulation that T is planar point set set, if all limits of T are Delaunay limit, then T
For Delaunay Triangulation.
Delaunay Triangulation is the triangulation methodology of a kind of optimization, and the triangulation network of generation has many good spies
Property, empty circle property, Minimum Internal Angle maximality, uniqueness, locality, the convex closure etc. of such as circumscribed circle.
The several common methods of Delaunay Triangulation are briefly described below:
(1) incremental algorithm
The flow process of incremental algorithm is: first, is determined to contain a convex closure of all discrete points data as border
Polygon.Secondly, successively original point data are inserted in boundary polygon, during each insertion point, the position of newly inserted point is added
To judge, connect newly inserted point and triangular apex about.Again, to the new limit connected, draw according to circumscribed circle criterion
Carry out LOP optimization.Finally, it is inserted into a little, completes the structure of the triangulation network.Some existing frequently-used Delaunay triangles cut open
Point, the method generating triangle gridding all can classify as pixel-level generation method, simply also has initial border many in some data structures
Limit shape structure aspect has been distinguished.
(2) segmentation conflation algorithm (divide and conquer)
Complicated problem is first divided into fairly simple subproblem solve, then by all subproblems with certain rule
Then merging, the thought solving former problem is referred to as thought of dividing and ruling.Use segmentation conflation algorithm to carry out Delaunay Triangulation mainly to have
Following step: first, is ranked up discrete some collective data, general order be X-direction by left-to-right, Y-direction
From top to bottom, first again Y-direction is sorted after sequence drained to X-direction.Secondly, scatterplot collective data is carried out recurrence division, directly
No more than 3 are counted in all divisions subclass out.Again, the point in all subclass is connected as triangle.?
Eventually, merge the subset triangulation network, when merging, carry out local optimum according to circumscribed circle criterion every time, be still Delaunay after making merging
The triangulation network, finally gives the Delaunay triangulation network of entirety.It is illustrated in figure 1 and 6 original scatterplot are used segmentation conflation algorithm
Carrying out the process schematic of triangulation, wherein (a) shows the initial data after sequence, and (b) represents respectively to subset
Dividing, subset is merged by (c) expression.
(3) triangulation network growth method
The main flow of triangulation network growth method is as follows: first, chooses 1 A, and find Europe in all discrete point set V
Family name, apart from nearest 1 B of upper distance A, obtains limit AB.Secondly, based on AB limit, find and meet Delaunay Triangulation
The thirdly C of circumscribed circle decision criteria, obtains △ ABC.Again, based on three limits of △ ABC, find eligible successively
Triangle, finally give Delaunay Triangulation produce the triangulation network.
2. constraint Delaunay triangulation network technology
During actual Geologic modeling, geological prospecting data are frequently not independent, but have one each other
Fixed restriction relation, relatively common restriction relation has the geosutures on stratum, the tomography upper lower burrs strike line of reversed fault, geology to survey
Terrain trend constrained line that spy personnel artificially draw etc..The existence retrained due to these, uses common Delaunay subdivision
Although the triangulation network that method produces can obtain certain digital terrain model, but cannot present the actual complexity in earth's surface about
Bundle relation, the most just contributes less in the demand to actual Geologic modeling.Then, add during Delaunay Triangulation
Certain restriction relation, becomes the most crucial technology and research direction.That commonly use in realization is Constrained
Delaunay retrains triangulation.
Binding side is carried out whole piece embedding by Constrained Delaunay constraint triangulation, does not add some points
Operation, does not splits, so the impact on original topology is less than Conforming Delaunay method.Also it is the present invention
The method carrying out Delaunay Triangulation use under Complex Constraints.Circulation method of changing sides is more representational as one
Constrained Delaunay retrains triangulation methodology, and its step is as follows:
(1) two end points of binding side are inserted the original unconfinement triangulation network.
(2) calculate the non-end points intersection edges intersected with binding side in the triangulation network, obtain intersection edges set E={e0,e1,...,
en}。
(3) the intersection edges set E found is traveled through, to each edge ei, find with eiFor cornerwise two trianglees, it is judged that
The concavity and convexity of the tetragon that two trianglees are formed.If tetragon is Strict Convex polygon, then exchange tetragon diagonal,
Otherwise skip eiOther limits in E are judged.As determined tetragon diagonal to be exchanged, then by limit eiRemove from E, sentence
The new limit end points the most non-with binding side produced after changing diagonal of breaking off a friendship intersects.Intersection edges set E is updated.
(4) one take turns after traversal terminates, and the limit intersected if there is no non-end points illustrates that binding side is completely embedded.
If still suffered from, then repeat step (3), until binding side intersects with the end points the most nothing but of the limit in every triangle gridding.
The flow chart of Constrained Delaunay constraint triangulation is as shown in (a) in Fig. 2 to (f).
3, the theoretical basis that extensive geological data lower surface camber is rebuild
Large-scale work area geological prospecting data, feature is that data area scope is big, and data point is many.Work at some super large
Under district, the initial data of geological prospecting has been even up to G or T rank, and the memory range of existing ordinary PC only has
4G-32G.Owing to being restricted by objective hardware condition, build to solve to carry out three-dimensional geological under extensive geological data
The problem of mould, a lot of experts and scholars are studied and have been explored, and main method is to process original earthquake data, reduce and need
Total amount of data to be processed and once read in the data volume of internal memory.Terrain blocks technology and level-of-detail technology are also aimed at greatly
Scale data carries out data organization technique during Geologic modeling, is next introduced it.
3.1 terrain blocks technology
Terrain blocks technology, as the term suggests, it is simply that by whole geology work area according to the geographical coordinate at place, it is divided into area
The same some pieces.In a stand-alone mode, by rational segment partition scheme, the extensive geological data of arbitrary size can be solved
Curve reestablishing with draw problem.The extensive geological data leaving external memory in is carried out piecemeal based on spatial distribution operation,
Make the terrain block data of each generation all can disposable graftabl.Terrain blocks step has the following steps:
(1) determine the size of each piece, conventional sized by identical square block.
(2) geological data is put in block, if a certain piece does not exist data, then add sky for overall integrity
In vain.
(3) if there is grid in block, ensure that the Grid dimension of every piece is (2 the most as far as possiblek+1)×(2k+1)。
(4) generated terrain block data file is preserved in a hard disk, facilitate real-time calling.
During actual terrain blocks, having Railway Project is to need to consider: be first exactly the size of piecemeal, if ground
Shape piecemeal is excessive, and possible internal memory cannot deposit single block file, affects memory efficient.If piecemeal is too small, although internal memory aspect
There is no problem, but it is mutual to may result in too much interior external memory I/O, affects whole efficiency, and meanwhile, block file is too much, also
Extra external memory storage administration overhead can be increased.Secondly, the data point on each block boundary, incessantly affected by current block, week
Enclose block also can it be had an impact, during so entering internal memory process, making several adjacent block enter internal memory often simultaneously, the most just relating to
And the problem to caching design, the most how to set up caching mechanism to the frequent interaction problems processing between internal memory and external memory.
3.2 level-of-detail technology
Level of detail LOD (Level of Detail) technology is usually used in showing in real time, in actual curve reestablishing and display
During, some areal geology exploratory spots are than comparatively dense, and there is the geologic rule constraint such as tomography or fold, in these districts
More original point is often applied in territory, builds the grid data than comparatively dense.And in other regions, landform is relatively flat,
Mainly based on layer position, there is not constraint, build simple mesh by small part original point and can complete regional reconstruction.So, exist
Different levels and different range, use the grid of different resolution to rebuild initial data, both ensure that the efficiency of operation,
In turn ensure that the curved surface effect of reconstruction.
At present conventional LOD model has a following two:
(1) discrete LOD model
Discrete LOD model refers to, in advance by calculating, obtain multiple copies in whole work area, and each copy correspond to one
Weight resolution, this results in a pyramid shape data model based on different resolution.When curve reestablishing is with display, root
Corresponding resolution is chosen according to actual requirement.The method realizes relatively simple, but when switching in the region of different resolution, can produce
Obvious saltus step sense.And the resolution grid point of redundancy can be stored, cause the waste of storage.
(2) continuous print LOD (CLOD) model
Because discrete LOD model is the most not enough in a lot of places, the concept of LOD Model is suggested.LOD mould continuously
The concept of type is similar with discrete LOD, and difference is that LOD Model has not been previously stored the data of different resolution, but
Calculate concrete resolution in real time.
Discrete LOD model realization is relatively simple, and amount of calculation is little, once generates and is used for multiple times, the least to the burden of CPU.
Shortcoming is to store the grid data under multiple resolution, also hopping phenomenon.LOD Model effect in display is excellent
In discrete LOD model, but calculate more, complicated and time-consuming.Owing to the present invention is not related to the dynamic change of resolution, at curved surface
During reconstruction, data only need to preserve once, so considering the thought using discrete LOD model.
In one exemplary embodiment, method based on terrain blocks is used to carry out the curved surface under extensive geological data
Rebuild, specifically include following operation:
A, the terrain blocks overall flow of extensive geological data
First, original extensive geological data and constraint polygon data are subdivided into different terrain block according to landform
In, and preserve to external memory (such as, hard disk).Secondly, curve reestablishing is carried out with each terrain block for ultimate unit, first to landform
Layer bit data and layer data in block carry out pretreatment, then the data after processing are carried out two-dimensional grid structure, according to net
Lattice data construct two dimension Delaunay Triangulation also carries out three dimensions triangulation network recovery, finally carries out surface interpolation and preserves
Curved surface topological data after reconstruction, display time the curved surface topology in all terrain block is shown together, thus realize whole greatly
The curve reestablishing in scale work area and display.Overall flow such as Fig. 3 institute of extensive geology work area geological data curve reestablishing method
Show.Here, the constrained polygon and the geology polygon that occur in the application are the polygonal abbreviations of geological constraining, and the present invention says
Geological constraining polygon refers in particular to tomography polygon.
Owing to separate unit PC internal memory is difficult to the survey data in whole extensive geology work area carries out overall curved surface weight
Building, so using terrain blocks technology, overall region being divided into some terrain block, as shown in Figure 4.Due to complete in overall data
Before portion has read in, it is impossible to determine the max min of entirety, also cannot determine the number of terrain block, so landform to be used
The method that block dynamically adds, dynamically landform and attribute to terrain block judge.
Terrain blocks basic procedure under a, extensive geological data
The step of the terrain blocks of extensive geological data mainly has the following steps:
(1) given initial landform block's attribute
Mainly there is a following items: the length of terrain block, X-direction terrain block number, Y-direction terrain block number.These several genus
Property determined by the actual size in work area, can be specified by outside, the minima of the terrain block of X-direction and Y-direction is 1, i.e.
There is a terrain block for extensive earthquake work area data at least, now this block is in the range of whole geology work area.
(2) initial data is read
Read geological prospecting initial data, including layer bit data and tomography polygon data.Under the data of extensive work area
The data volume of original geology survey data is relatively big, if carrying out terrain blocks behaviour again after all data all read from file
Making, data manipulation amount is relatively big, spends the time longer.And if often one data of reading just carry out piecemeal operation, operation can be relatively
Frequently, efficiency is affected.Therefore the thought using similar piecemeal carries out data reading, if read in seismic data acquisition the most every time
These several rows data are carried out terrain blocks operation by dry row, are further continued for carrying out reading in operation after having operated, until geological data
File is read and has processed.
In common terrain blocks operation, initialize and whole initial datas are read in, obtain the maximum of data
According to the concrete scope of data and the single block length of formulation, little value, determines that terrain block is in X-direction and the number of Y-direction.And
During the curve reestablishing of extensive work area geological data, data volume is relatively big, if all read initial data in advance
Entering, can consume the regular hour, in order to improve treatment effeciency, the present invention uses a kind of method of dynamic interpolation terrain block, i.e. at the beginning of
Begin given a range of terrain block, if the geological data read in is in initial given terrain block, is then divided by geological data
It is fitted in terrain block.If the geological data read in is outside initial given terrain block, then given terrain block is expanded
Fill, be allowed to comprise reading geological data.Preferably, curve reestablishing operation and curved-surface display for convenience, when expanding terrain block
Ensure the rectangular shape in whole region.
As it is shown in figure 5, initial setting up acquiescence X-direction and Y-direction are respectively arranged with two terrain block, when reading in data, V1、
V2、V3、V4Deng point through judging in the terrain block existed, just it is added directly into inside corresponding terrain block.Point V5Not
Inside terrain block through existing, just existing terrain block is dynamically added, add new terrain block to comprise a V5, it is simultaneously
Ensure the integrity of overall region, with the addition of the terrain block of blank.
(4) data are distributed in terrain block
Original geology survey data is mainly layer bit data and layer data, according to the two-dimensional coordinate of geological data, finds
Corresponding terrain block, puts into the terrain block of correspondence by geological data.
(5) data in terrain block are preserved
Geological data in terrain block is preserved to hard disk, determines save location for convenience, each terrain block is given
A fixed block id, the present invention uses the mode of two-dimensional array to carry out the id index of block, determines a two-dimensional array a [M] [N], its
The X-direction of the one-dimensional representation terrain block of middle array, the Y-direction of the two-dimensional representation terrain block of array, the value in array is ground
The id of shape block.Such as, the value of a [0] [0] is 0, illustrates the 0th of X-direction, block id is 0 in the 0th terrain block of Y-direction.With
The dynamic interpolation of terrain block, two-dimensional array a [M] [N] also can dynamically change, terrain block id of time changing correspondence position.With
File in one terrain block is finally saved under the path that same terrain block id is corresponding.
After above operation, completing the reading to all original earthquake datas, original extensive geological data turns
Turn to terrain block data and preserve in a hard disk, preparing for next step operation, terrain block data such as Fig. 6 of distribution in a hard disk
Shown in, the file of the entitled block of suffix is terrain block file, and terrain block file is with the rule of " terrain block id_ terrain block resolution "
Then name.
Under b, extensive geological data, Complex Constraints piecemeal processes.
In the process of extensive geological data, in addition to original layer bit data, there is also Various Complex constraint information,
Presented in geological constraining polygon.Wherein, tomography polygon data is the polygon that geological prospecting personnel manually draw,
In the present invention, geological constraining polygon refers to tomography polygon, it addition, closed polygon also refers in particular to tomography polygon, namely
Say that all of polygon of the present invention is all referring to tomography polygon.When carrying out terrain blocks process under extensive geological data, ground
The process of matter constrained polygon data has a following difficult point:
First have to it is considered that the distribution problem of the polygon form point of geological constraining.Geology polygon all shows on two dimensional surface
For the form of broken line, adjacent two the point often spacing of geological constraining polygon data farther out, are distributed at actual geologic data
In, it is more likely that occur that two adjacent polygon form points cross over the situation of terrain block, now may only have in a terrain block
One polygon form point of geology.Due to the present invention use terrain block as the ultimate unit of curve reestablishing, in each terrain block only
In the case of having a polygon form point of geology, it is impossible to build binding side and realize follow-up Delaunay constraint triangulation computing,
So respective handling to be made.In one exemplary embodiment, the processing method taked is that to obtain the adjacent geology of the two many
The line segment that limit form point is linked to be and the intersection point on terrain block border, and add in the geology polygon data in the terrain block of correspondence.As
Shown in Fig. 7, V1、V2、V3For three junction points on geology polygon, first find the friendship of geological constraining polygon and terrain block
Point V4、V5, by V4、V5Point is respectively put into the terrain block of correspondence, now it should be noted that owing to being boundary point, so adjacent
Terrain block will be put into.After operating, the polygon form point of geology in terrain block 1 is V3、V4, the polygon form point in block 2 is
V4、V1, the polygon form point of geology in block 3 is V5、V2。
It is in place of another complexity that Complex Constraints information under extensive geological data is divided in terrain block, geology
Constrained polygon exists the vertical fault polygon do not closed, there is also reversed fault polygon, normal fault polygon, reversing many
The polygon that limit shape etc. are closed, as shown in Figure 8.Polygonal data boundary pair can be applied during carrying out Based Model for Reconstructing Geological Surface
Tuple and the attribute of point judge.And the piecemeal operation of closed polygon is it is possible that situation about not conforming to the actual conditions, such as Fig. 8
In block 1 internal, some A is the polygon form point of geological constraining, and some B, C are the intersection point of geological constraining polygon and terrain block, in block 1
Constrained polygon is judged as the closed polygon that some A, B, C are formed by connecting, hence it is evident that do not conform to the actual conditions.So on the one hand will be by ground
The point data of matter constrained polygon is divided in each terrain block, on the other hand to retain a overall closing geology polygon
Data (abbreviation closed polygon).
Definition closed polygon set P:
P={P1, P2..., PmEquation (1)
Polygon data is as follows, comprises the some set of dish polygon, the some set of lower wall polygon and root polygon id
Pi={ upi, lpi, idiEquation (2)
That is, in equation (2), upiThe set of dish polygon point, lp in expressioniRepresent the some set of lower wall polygon, idiTable
Show root polygon id, on two dimensional surface, all give an id to all of point, this Rigen polygon idiRefer to current i-th
This id concentrated in two-dimensional points in individual polygon.
It is to say, in equation (1) and equation (2), P represents the set of all closed polygon data, each Pi
It is all a closed polygon, collects PiRepresenting the i-th closed polygon in closed polygon set P, 1≤i≤m, m are for closing
Polygonal total number, equation (2) is for equation (1) some polygon PiData structure, each polygonized structure
It is just as, upiRepresent the upper dish polygon point set of i-th closed polygon, lpiRepresent the lower wall of i-th closed polygon
Polygon point set, idiRepresent the root polygon id of i-th closed polygon.
Closed polygon data exist as an entirety, when each terrain block is carried out curved surface reconstruction operation, all
Closed polygon data to be called carry out attributive judgment.
Also have and a little need consideration, it is simply that geology polygon line segment iterates through the situation of same terrain block.Such as Fig. 9
Shown in, geology polygon repeats to enter terrain block 1 and terrain block 2, now, if according to division methods mentioned above, then block 1
Internal polygon form point is followed successively by: V1、V2、V5、V6、V7、V8, owing to terrain block is by the ultimate unit of curve reestablishing, at block
With polygonal sets of line segments cooperation for retraining when inside carrying out Constrained Delaunay triangulation, i.e. V2、V5The line segment of composition is also
Can become constraint, and actually the two point being not attached to, the most there is not constraint.Only index adjacent two and could connect shape
Become binding side.In order to solve this problem, introduce an attribute: the index of the polygon form point of geology.A most current geology is many
What kind of position limit form point is in whole polygon.Definition polygon point set at this:
Vp={ V1, V2..., VnEquation (3)
Wherein ViFor i-th point, ViIn data include that the position of polygon point, polygon id that point is corresponding, point are polygon
Location index in shape.
Vi={ Xi, Yi, IDi, IndexiEquation (4)
That is, in equation (4), XiAnd YiRepresent the position of the polygon form point of geological constraining, IDiRepresent that i-th point is corresponding
Geological constraining polygon ID, IndexiRepresent i-th point location index in geological constraining polygon;1≤i≤n, n are ground
Total number of matter constrained polygon point.
It should be added that, the V in equation (3)pRefer on some geological constraining polygon institute a little (i.e.
Matter constrained polygon point) set, V in equation (4)iFor i-th point, X on this geological constraining polygoniAnd YiRepresent geological constraining
The position of polygon form point.Owing to each polygon has an ID, so IDiRepresent be this place geology polygon it
Polygon ID, Index at geology polygon setiJust represent this location index on this geological constraining polygon, be
This unique mark on this geological constraining polygon.
As a example by Fig. 9, after introducing the index attributes of polygon form point, understand V through judgement2、V5On original geology polygon
Positional value is the most non-conterminous, will not be connected to become binding side, it is ensured that the correctness of result.
B, the multiresolution of extensive geological data are arranged
In order to simplify large-scale geological data further, improving the efficiency of curve reestablishing, the present invention is exemplary
Embodiment uses the Level Of Details differentiated more, to different terrain block, determines different resolution according to the distribution of landform.
In one exemplary embodiment, use energy function method, if (x y) is point (x, y) degree of depth at place to z=H
Value, for the shaped area of one piece of n data point, defines fcFor this region energy function a little.
Wherein PiRelative depth for this region is distributed:
In equation (6), Z0For minimum depth value in this region, ε is the least peace in order to prevent denominator from being zero interpolation
Full number.
Use energy function method, resolution can be good at embodying the fluctuating situation of landform, due to the most all
Take part in computing, it is therefore prevented that occur that the overstocked situation of partial points causes resolution deviation.The present invention uses energy function method to calculate
Initial resolution, is then finely adjusted the tomography distribution situation in the resolution value combined block calculated, if one
The tomography or the reversing constraint that exist in terrain block are more, and resolution will be of a relatively high.
In exemplary embodiments of the present invention, the method for actual curved surface topology rebuilding is to build initial four limits in terrain block
Shape grid, carries out Constrained Delaunay triangulation in quadrilateral mesh and obtains two-dimensional topology, finally two-dimensional topology is recovered
For three-dimensional topology.Need in each terrain block to build the quadrilateral mesh necessarily crossing number, the resolution in terrain block and grid
Density is directly related, it is assumed that in a terrain block, and resolution is P, then the initial quadrilateral mesh built in this terrain block
Number is 2P×2P。
Terrain block scheduling under C, extensive geological data
Due to data scale and the problem of appointment block length, it is possible to produce multiple landform after carrying out terrain blocks operation
Block, if the most one by one terrain block being carried out curved surface reconstruction operation, may relate to repeatedly the mutual of internal memory and hard disk, impact
Curve reestablishing efficiency.In view of above problem, exemplary embodiment of the present opens up a part of memory headroom conduct further
Buffer area, in curve reestablishing operates, performs the mutual of the interval mutual and internal memory of hard disk-caching-caching interval, and such one
Aspect avoid terrain block number too much time produce frequently mutual, on the other hand can ensure that the degree of depth of topology point in terrain block
The correctness of value.
For each terrain block, in curved surface three-dimensional topology process of reconstruction, the borderline mesh point of terrain block deep
Angle value only can not be operated acquisition by the row interpolation that clicks within current terrain block, also to be considered by the depth value of the point in adjacent blocks
Coming in, problem is sewed up on the border having related to terrain block, and the setting of buffer size and terrain block are just being cached by this point
Scheduling in district proposes requirement, and the terrain block under extensive geological data is dispatched mainly the following aspects:
(1) buffer area is arranged
The size of buffer area is relevant to the size of the number of terrain block and each terrain block, if single landform blocks of data amount
Relatively big, whole internal memory all cannot put into multiple terrain block, and nature can limit the terrain block number in buffer area.If terrain block
Number is fewer and data volume little, all terrain block can be put into buffer area completely.For general extensive geological data,
Terrain block is more, and overall data is relatively big, not only to use the data in current terrain block, also use when operating each terrain block
To the data (except borderline region terrain block) of 8 terrain block about, the minima that therefore buffer area is arranged is at least 9 ground
Shape block.
(2) terrain block is added in buffer area
When carrying out the scheduling of buffer area terrain block, in buffer area, first find currently processed terrain block and it closes on landform
Block.During terrain blocks above, have been set up represent block index relative two-dimensional array a [M] [N], by this two
Dimension group, when processing some terrain block, it is possible to is quickly obtained the id of its adjacent terrain block.I.e. in two-dimensional array
The neighbouring relations of reality between terrain block that neighbouring relations are the most corresponding, as the position of current terrain block corresponds in two-dimensional array as a
[m] [n], then its adjacent terrain block be a [m-1] [n-1], a [m-1] [n], a [m-1] [n+1], a [m] [n+1], a [m+1] [n
+ 1] terrain block, corresponding to a [m+1] [n], a [m+1] [n-1], a [m] [n-1], and then obtain terrain block id and correspondence
Blocks of data file.If finding currently processed terrain block in buffer area and closing on terrain block, directly choose from buffer area
Corresponding terrain block operates.Otherwise judge that buffer area is the fullest, if buffer area less than and buffer area remaining space enough,
From fixed disk file, the data block file of needs is added in buffer area.As shown in Figure 10, dotted line region is currently processed
Terrain block 1, block 2,3,4 closes on terrain block for it, when processing block 1, finds block 1,2,3,4, do not look in buffer area
To corresponding terrain block and buffer area less than, then from hard disk, add the data of block 1,2,3,4 in buffer zone.
(3) terrain block displacement is carried out in buffer area
If buffer area is full or buffer area remaining space is less than the space processed required for current block, landform to be used
Block Replacement Strategy, replaces a part of terrain block in buffer area by new terrain block data.Conventional buffer area dispatching method
Time and the access times decision of buffer area is mainly entered according to terrain block.For the present invention, terrain block to be considered it
Between close on relation, it is ensured that when certain terrain block is processed, this terrain block close on terrain block all in buffer area.At this
On the basis of the present invention use a kind of terrain block method of replacing based on distance, when current terrain block is processed, if caching
District's insufficient space, by the terrain block removal buffer area farthest apart from current terrain block, moves into slow by the adjacent blocks of current terrain block
Deposit district.As shown in figure 11, the terrain block that original state processes is block 12, and block 12 and adjacent blocks thereof are all in buffer area.Process
After block 12 processing block 13, need the adjacent blocks of block 13 is moved into buffer area, now buffer area is the fullest, sentences through position
Disconnected, in the adjacent blocks of block 13, block 7, block 14, block 17 be not in buffer area, three pieces by farthest for distance block 13 in buffer area: block 10,
Block 11, block 20 remove buffer area, and block 7, block 14, block 17 are moved into buffer area.
Use buffer area that terrain block is carried out the overall flow schematic diagram of data dispatch as shown in figure 12.
Terrain block interpolation under D, extensive geological data
The operation performed in terrain block, is in a terrain block, and first reading in the fixed disk file from respective path should
Layer bit data file within terrain block and constraint polygon data file, the resolution further according to this terrain block sets up certain step
Long initial mesh, carries out Constrained Delaunay triangulation structure inside each grid, obtains network of triangle LF Topology, then
The two-dimensional topology built is carried out concatenation, obtains three-dimensional topology, finally the point in topology is carried out degree of depth interpolation, obtain
Real three dimensional geological curved surface.Point degree of depth interpolation mainly has following two steps in terrain block:
(1) non-boundary point interpolation
For non-terrain block boundary point, degree of depth interpolation procedure is first to mesh point interpolation, then carries out tomography polygon point
Interpolation, finally compensates interpolation.
First mesh point is carried out interpolation, the curved surface topology connected, mesh point occupies the overwhelming majority a little, thus the most right
Mesh point carries out interpolation, during the foundation of curved surface topology, original layer site is assigned to correspondence as seed points
Grid in, the value of mesh point is carried out interpolation by the seed points of corresponding grid and obtains.Secondly to clicking on row interpolation on polygon.
Point quantity for mesh point on polygon is less, and is generally individually linked to be topology limit with mesh point.Complete mesh point
Interpolation Process after, obtained the actual grade value of mesh point, it is possible to use gridden data pair is connected many with mesh point
Limit form point carries out interpolation.
Finally compensating interpolation, through gridding interpolation and polygon point interpolation two step, only small part point may be still
There is no actual grade value, interpolation operation to be compensated, use in the topology obtained by two steps above and had reality
The point of border depth value, clicks on row interpolation to still not having actual grade value.Compensate interpolation not stipulated number, until all topologys
Point in connection is respectively provided with actual grade value.
(2) boundary point interpolation
During the boundary point of terrain block is carried out degree of depth interpolation, not only to consider the impact of point in current terrain block,
It is also contemplated that the impact that it is carried out by the point within adjacent block.As shown in figure 13, at the boundary point A of upper left corner terrain block, to an A
When carrying out depth value interpolation, utilization is other points being connected topology with an A, at topological midpoint A and the point of this terrain block inside
B, C, F, G are connected, and put A in it closes on terrain block and be connected with some C, D, E, F, all above and some A Topology connection
Point will participate in the degree of depth interpolation calculation of some A as seed points.
The terrain block boundary point interpolation problem that another should be noted that is the repetition interpolation problem between different terrain block,
The most multiple terrain block are all to the same situation clicking on row interpolation.As a example by some A in Figure 13, lower-left terrain block is being carried out a little
During interpolation, find the point-to-point A being connected with an A in each terrain block and carried out interpolation, entered in upper left terrain block
Also an A can be carried out an interpolation during operation of row point interpolation, total algorithm will certainly be imitated by the repetition interpolation operation of identical point
Rate produces impact.In order to avoid repeating interpolation, boundary point is added interpolation whether attribute, every time to this point after the point interpolation of border
Attribute be set to interpolation, the point of interpolation is not carried out quadratic interpolation.
In sum, the present invention combines the practical situation of extensive geology work area geological data, it is proposed that divide based on landform
The method of the extensive geological data Based Model for Reconstructing Geological Surface of block, is divided into overall region some terrain block, efficiently solves
The problem that separate unit PC internal memory is difficult to the survey data in whole extensive geology work area carries out overall curve reestablishing.This
The bright process to mass data has very important meaning, has a extensive future in physical prospecting field.
Although the most describing the present invention by combining exemplary embodiment, but those skilled in the art should be clear
Chu, in the case of the spirit and scope limited without departing from claim, can be carried out respectively the exemplary embodiment of the present invention
Plant modifications and changes.