CN103903061A - Information comprehensive processing device and method in three-dimensional mineral resource prediction evaluation - Google Patents
Information comprehensive processing device and method in three-dimensional mineral resource prediction evaluation Download PDFInfo
- Publication number
- CN103903061A CN103903061A CN201410077136.2A CN201410077136A CN103903061A CN 103903061 A CN103903061 A CN 103903061A CN 201410077136 A CN201410077136 A CN 201410077136A CN 103903061 A CN103903061 A CN 103903061A
- Authority
- CN
- China
- Prior art keywords
- module
- initial predicted
- dimensional
- predicted body
- index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses an information comprehensive processing device and method in three-dimensional mineral resource prediction evaluation, wherein the method comprises the following steps: step one, constructing an initial prediction body of three-dimensional mineralization prediction; secondly, quickly extracting prediction elements of the space discrete points from the initial predictor; rasterizing the prediction elements expressed by the surface model, and assigning the rasterization result to an attribute table of the initial prediction body; and step four, screening and optimizing the attribute table of the initial prediction object to obtain a prediction object. The method is used for assisting geologists to solve the problem of the deep and peripheral mine exploration work of known mines, and meets the requirements of realizing the automatic information simulation, comprehensive analysis, visualization and the like of the multivariate mine formation information based on geological, geophysical, geochemical and other data.
Description
Technical field
The present invention relates to geomathematics and Geographic Information System map making field, particularly relate to technology and the device of the three-dimensional metallogenic prognosis multiple information of a kind of large scale unified model rapid build, information automatic Synthesis and visual analyzing.
Background technology
Along with ore deposit, earth's surface, the minimizing day by day in superficial part ore deposit and ore deposit easy to identify, look for mine disaster subsist benefit increase, prospecting result reduces day by day, since entering eighties 21 century, emphasis is looked for Deep-concealed Ore, newtype ore deposit and frontier ore deposit have become the range of reconnaissance that countries in the world are paid close attention to, wherein use three-dimensional stereo model to carry out deep concealed complex geologic body (, second looks for space, ore deposit) searching, become many countries and regions and looked for the main object in ore deposit, therefore the effect of the metallogenic prognosis of large scale (being conventionally greater than 1:50000) is more outstanding, it has become important component part (Zhao Peng great etc. of Mineral exploration work, 1992).
At present, also do not develop the application software that is specifically designed to ore deposit, deep tune, the three-dimensional metallogenic prognosis work of large scale of a set of maturation both at home and abroad.First a large amount of three-dimensional Deeo-Space Metallogenic Predication Study complete the work such as metallogenic information extraction and three-dimensional visible fractional analysis by mine exploration software, secondly, in other softwares, complete information comprehensive analysis.Mining area exploration three-dimensional modeling and visual both at home and abroad, the applications such as geochemistry, the fractional analysis of geophysical data three-dimensional visible have been developed many ripe practical software of dimensionally.From software application angle analysis, domestic three-dimensional software market is mainly monopolized by external software vendor.Therefore more need the Software tool of the actual requirements such as large scale Deeo-Space Metallogenic Predication Study and application.
In sum, form a set of collection rapid modeling, high-precision three-dimensional visual analyzing, location, quantitative and determine the three-dimensional metallogenic prognosis method of the specialty with independent intellecture property that probability forecasting method is integrated, become look for miner do in the technical support means of urgent need.For reaching this purpose, need to be applied to following technology:
1) pinpoint target figure layer informix model in mineral resource assessment
In mineral resource assessment field, separate figure layer refers between ore-controlling evidence layer it is all independently.Depend on another control ore deposit factor if a control ore deposit exists, and using these two all as evidence layer, will produce too high or too low one-tenth ore deposit probability, will cause final predicting the outcome to be affected.Therefore, in actual applications, must use mathematical model to carry out condition independence test to the figure layer that enters metallogenic prognosis, reject dependent figure layer; Target figure layer refers to from original variable, to select with predicting mineral the ore-controlling evidence layer compared with substantial connection, and this process of selecting realizes based on mathematical model; Informix refers at ore-controlling evidence figure layer to have on independent and favourable with becoming ore deposit basis, the information of the different spaces same position in same ratio chi, same Fundamentals of Mathematics is superimposed, attribute using each figure layer as an input feature vector, by certain mathematical model, above-mentioned characteristic attribute is combined for describing new feature, i.e. MINERAL PREDICTION target area.
2) two-dimentional surface sweeping line algorithm
The method that two-dimensional space is set up index structure has a lot, for example: Adelson-Velskii-Landis tree, RBTree, BSP tree, evenly band cut apart, etc. graticule mesh cut apart, the method such as Quadtree Partition and self-adaptation piecemeal.In the time that data volume is very large, in order to complete fast query and search, consider time and space complexity issue simultaneously, generally use grid index technology.Two-dimensional scan line is to be applied to two dimensional image process field at first.Its main thought is the state by carry out certain conditional search and determine respectively this pixel to 4 adjacent directions of pixel to be determined.In order to realize the constructing technology of high-precision three-dimensional multiple information unified model quickly and accurately, above-mentioned two kinds of algorithm ideas are applied to three dimensions, solve high precision cube lattice model (the being piece segment model) rapid modeling of complex geologic body.
3) three dimensions K rank neighbor search technology
K-NN Query is a space-like searching algorithm, mainly solves the search problem for spatial discrete points, is one of rudimentary algorithm of Geographic Information System.Along with the technology of obtaining of 3 d-dem point is constantly perfect, constantly increase for demands such as classification, interpolation and the curve reestablishings of a large amount of 3 d-dem points, realizing fast and efficiently K-NN Query under specified criteria has become the hot issue of current three-dimensional geographic information system research.K-NN Query is in two dimension or 3 d-dem data set S, to search K the point nearest with Euclidean distance to be made an inventory of.
In large scale Deeo-Space Metallogenic Predication Study, there is a large amount of geology, geochemistry, geophysics original point sampled data.The method of extracting for these raw data is the spatial interpolation algorithm based on predicting unit lattice central point, for example: the point take unit center of a lattice as interpolation, the attribute information of the method acquiring unit lattice such as use Ke Lige, anti-distance weighted method.Therefore, K rank neighbour's Fast search technique is one of important step realizing component of forecast rapid extraction.
4) the latent target selection technology based on evidence weight algorithm and three-dimensional visualization technique
In recent years, the common way of the application of the GIS technology in mineral resources exploration and evaluation is to adopt so-called evidence-right-weight " method.Weights-of-evidence method is the method for quantitative test and comprehensive multi-source special topic earth science data layer, generating quantification containing ore deposit potentiality figure.Adopting in the mineral exploration of classic method, usually need the thematic maps such as comparative analysis geology, geophysics and geochemistry, iris out the target area that is worth further work, in fact weights-of-evidence method is exactly the digitizing " version " of traditional exploration method.
" evidence " in evidence-right-weight " method is made up of a series of data sets (learning thematic maps) of reconnoitring, and " weighted value " estimates (Basab Mukhopadhyay etc., 2000) according to known mineral deposit or the specific district's exploration model of reconnoitring.Its basic ideas are under GIS environment, to carry out space stack and comprehensive analysis by composing respectively with the thematic map data (evidence) of different weights, generation has the different grid thematic maps containing ore deposit probable value, this quantitative analysis results has reflected different from the distribution spatially of ore deposit potentiality subregion, thus for further in detail preliminry basic research the leading evaluation result on region is provided.
5) three-dimensional visualization technique based on OpenGL
The three-dimensional visualization technique of mentioning in the present invention realizes the exploitation that mainly refers to the three-dimensional visualization program based on OpenGL technology.OpenGL is the abbreviation of OpenGraphicsLib, is that a set of three-dimensional picture is processed storehouse, is also the industrial standard in this field.Computerized three-dimensional figure refers to the technology that the three dimensions by data description is converted to two dimensional image and shown or print by calculating.OpenGL is designed to be independent of hardware, be independent of window system, on the various computing machines of the various operating systems of operation, all can use, and can, at net environment with Client/Server work, be the test pattern storehouse of the high-end application such as professional graphics process, science calculating.
Summary of the invention
The object of the present invention is to provide a kind of three-dimensional metallogenic prognosis multiple information Integrated Processing Unit and method thereof, solve known deep and surrounding of mine for auxiliary geology expert and look for miner to do, meet and realize the information automatic imitation of the multiple ore forming information based on data such as geology, geophysics, geochemistry, comprehensive analysis and the demand of the aspect such as visual.
To achieve these goals, the invention provides informix treating apparatus in a kind of three-dimensional mineral resource prediction evaluation, it is characterized in that, comprising:
Predictor builds module, for building the initial predicted body of three-dimensional metallogenic prognosis;
Space search module, connects described predictor and builds module, for the component of forecast to initial predicted body rapid extraction spatial discrete points;
Variable extraction module, connects described space search module, for the component of forecast of being expressed by surface model is carried out to rasterizing, and by rasterizing result assignment in the attribute list of initial predicted body;
Target selection module, connects described predictor and builds module, described variable extraction module, for the attribute list of initial predicted body being screened with preferred, obtains target of prediction body.
Described device, wherein, described predictor builds module and further comprises:
Outer bounding box extraction module, for extracting the outer bounding box of surface or curved surface data;
Outer bounding box gridding module, for being decomposed into hexahedron according to the grid granularity of setting by outer bounding box;
Face body is asked friendship module, asks friendship to calculate for carry out tri patch and hexahedron according to outer bounding box;
Hexahedron sort module, based on se ed filling algorithm and ask friendship result of calculation, carries out rapid screening to the hexahedron that forms three-dimensional metallogenic prognosis body, generates the attribute list of initial predicted body and hexahedron model.
Described device, wherein, described space search module further comprises:
Module set up in the first index, for setting up the Octree Spatial Index of 3 d-dem point;
Module set up in the second index, for setting up the three-dimensional R-Tree tree index of Octree Spatial Index;
Hunting zone acquisition module, for obtaining hunting zone to be made an inventory of;
Space interpolation is realized module, for setting up module according to hunting zone and the first index, the result of module set up in the second index, and apply Ke Lige, anti-distance weighted formula carries out space interpolation;
Search assignment module, for by space interpolation result assignment to initial predicted body, and be retained in the attribute list of initial predicted body.
Described device, wherein, described variable rapid extraction module further comprises:
Rasterizing processing module, for rasterizing surface model body or curved surface data;
Result assignment module, for by rasterizing result according to locus assignment the attribute list to initial predicted body.
Described device, wherein, described target selection module further comprises:
Becoming ore deposit probability calculation module, is primary data for the attribute list take initial predicted body, calculates each hexahedral one-tenth ore deposit probability in three-dimensional prediction body based on evidence-right-weight " method;
Preferred body module, for take initial predicted body Model as basis, and is combined into ore deposit probability calculation result, is extracted into ore deposit target of prediction body.
To achieve these goals, the invention provides information comprehensive processing method in a kind of three-dimensional mineral resource prediction evaluation, it is characterized in that, comprising:
Step 2, to the component of forecast of initial predicted body rapid extraction spatial discrete points;
Step 3, carries out rasterizing to the component of forecast of being expressed by surface model, and by rasterizing result assignment in the attribute list of initial predicted body;
Step 4, screens with preferred the attribute list of initial predicted body, obtains target of prediction body.
Described method, wherein, described step 1, further comprises:
Extract the outer bounding box of surface or curved surface data;
According to the grid granularity of setting, outer bounding box is decomposed into hexahedron;
Carrying out tri patch and hexahedron according to outer bounding box asks to hand over and calculates;
Based on se ed filling algorithm and ask friendship result of calculation, the hexahedron that forms three-dimensional metallogenic prognosis body is carried out to rapid screening, generate the attribute list of initial predicted body and hexahedron model.
Described method, wherein, described step 2, further comprises:
Set up the Octree Spatial Index of 3 d-dem point;
Set up the three-dimensional R-Tree tree index of Octree Spatial Index;
Obtain hunting zone to be made an inventory of;
According to hunting zone and the first index set up module, the result of module set up in the second index, and apply Ke Lige, anti-distance weighted formula carries out space interpolation;
Space interpolation result assignment, to initial predicted body, and is retained in the attribute list of initial predicted body.
Described method, wherein, described step 3, further comprises:
Rasterizing surface model body or curved surface data;
By rasterizing result according to locus assignment in the attribute list of initial predicted body.
Described method, wherein, described step 4, further comprises:
Take the attribute list of initial predicted body as primary data, calculate each hexahedral one-tenth ore deposit probability in three-dimensional prediction body based on evidence-right-weight " method;
Take initial predicted body Model as basis, and be combined into ore deposit probability calculation result, be extracted into ore deposit target of prediction body.
The method that the present invention proposes, be applicable to component of forecast simulation and the polynary fields such as ore deposit information comprehensive analysis model rapid build and visual analyzing thereof of looking in the three-dimensional metallogenic prognosis of large scale, be mainly used in the various data that obtain in large scale mineral exploration process, such as sampling number certificate, geology terrain data etc., utilize the three-dimensional computations machine technologies such as rapid space interpolation, curved surface automatic Fitting, Raster Data Model rapid build, realize the three-dimensional information automatic imitation of the multivariate datas such as geology, geophysics, geochemistry, comprehensive analysis and visual demonstration.
Accompanying drawing explanation
Fig. 1 is three-dimensional mineral resource prediction multiple information Integrated Processing Unit structural drawing;
Fig. 2 is three-dimensional mineral resource prediction multiple information integrated conduct method process flow diagram;
Fig. 3 A, 3B are R-Tree spatial index model schematic diagram;
Fig. 4 is the rapid space interpolation method flow diagram of the polynary sampled point of Large scale prosecing assessments of mineral resources of the present invention;
Fig. 5 A, 5B, 5C are basis---the neighbouring relations between voxel of the grid rapid screening algorithm based on Flood-Fill thought;
Fig. 6 be the present invention is based on the constraint of complicated geological body Model and earth's surface surface model look for ore deposit information comprehensive analysis model fast construction method process flow diagram;
Fig. 7 is that Flood-Fill algorithm two-dimensional space of the present invention is realized schematic diagram;
Fig. 8 is that Flood-Fill algorithm three dimensions of the present invention is realized schematic diagram;
Fig. 9 is the schematic diagram of variable extraction module in the present invention;
Figure 10 is the schematic diagram of variable extraction module in the present invention.
Embodiment
Describe the present invention below in conjunction with the drawings and specific embodiments, but not as a limitation of the invention.
As shown in Figure 1, be three-dimensional mineral resource prediction multiple information Integrated Processing Unit structural drawing.This device 100 comprises: predictor builds module 10, space search module 20, variable extraction module 30, target selection module 40.
Predictor structure module 10, space search module 20, variable extraction module 30, target selection module 40 are to be mutually related.
Carry out the three-dimensional metallogenic prognosis of large scale, first need to build high-precision three-dimensional multiple information Comprehensive Model body (grid body).This model is made up of a large amount of hexahedrons, for representing initial predicted body, can use afterwards based on hexahedral quantitative and localization method, calculates the position of target of prediction body.Target of prediction body is take initial predicted body as basis, search for and variable abstraction function obtains by cell.In the time carrying out three-dimensional metallogenic prognosis, first need difference figure layer, dissimilar data to extract in the attribute list of three-dimensional prediction body, therefore, need 30 two modules of space search module 20 and variable extraction module to realize the screening to initial predicted body.
Predictor builds module 10, for building the initial predicted body of three-dimensional metallogenic prognosis.This initial predicted body can be changed by algorithm by surface model body, also can directly calculate structure.This surface model body is made up of a large amount of hexahedrons, for representing initial predicted body.
Space search module 20, for take three dimensions discrete point K rank neighbour's fast search algorithm as basis, realizes the component of forecast of rapid extraction spatial discrete points etc.
Variable extraction module 30, is mainly that the component of forecast to reaching based on surface model body surface carries out improved quick rasterizing, realizes the component of forecast variable of rapid extraction curved surface and predictor etc.
Target selection module 40, for according to evidence-right-weight " method, the initializaing variable attribute list of three-dimensional metallogenic prognosis body being screened with preferred, obtains target of prediction body.
Three-dimensional visualization is among whole large scale metallogenic prognosis process, for predictor structure module 10, space search module 20, rapid extraction module 30, target selection module 40 provide Computerized three-dimensional Visualization true to nature.
Further, predictor structure module 10 comprises:
Outer bounding box extraction module, for extracting the outer bounding box of surface model body or curved surface data;
Outer bounding box gridding module, for the grid granularity according to certain, is decomposed into one group of hexahedron by outer bounding box;
Face body is asked friendship module, for carrying out tri patch according to outer bounding box and hexahedron is asked friendship; Tri patch herein refers to be configured to the element figure of surface, i.e. triangle.
Hexahedron sort module, based on Flood-Fill algorithm and ask friendship result of calculation, carries out rapid screening to the hexahedron that forms three-dimensional metallogenic prognosis body, generates initial predicted body, generates the attribute list of hexahedron model, i.e. the attribute list of initial predicted body simultaneously.
Further, space search module 20 comprises:
Module set up in the first index, and for setting up the Octree Spatial Index of 3 d-dem point, this index is the first index;
Module set up in the second index, and for setting up the three-dimensional R-Tree tree index of Octree Spatial Index, this index is the second index;
Hunting zone acquisition module, for by input with automatically calculate hunting zone to be made an inventory of;
Space interpolation is realized module, for setting up module according to hunting zone and the first index, the result of module set up in the second index, and apply ordinary kriging, anti-distance weighted formula carries out rapid space interpolation;
The second index is the index being based upon on the first index., set up the first index based on raw data.Afterwards, on the basis of the first index, set up the second index.In the time using Ke Lige or anti-distance weighted calculating, first use the second index; Afterwards, on the basis of the second index calculation, use the first index, obtain final result of calculation.
Screening assignment module, for by space interpolation result assignment to initial predicted body, and be retained in the attribute list of initial predicted body.
Further, variable rapid extraction module 30 comprises:
Rasterizing processing module, for building the algorithm translation function of module 10 according to predictor, rasterizing surface model body or curved surface data, and consistent with predictor structure module 10 of the parameter such as the length of grid;
Result assignment module, for by rasterizing result according to locus indirect assignment the attribute list to initial predicted body.
The assignment that screening assignment module realizes refers in assignment procedure and needs advanced line search to calculate, then carries out assignment.So-called search is calculated and is referred to that the hexahedron in interpolation result and initial predicted body is not one to one, therefore, need to determine how which or which interpolation result is remained in initial predicted body according to filtering algorithm.And in result assignment module, this assignment procedure is one to one.
The attribute list of initial predicted body refers to the attribute library of forecast model, and its existence is through whole device.Its fundamental purpose is the result of preserving screening assignment module and the calculating of result assignment module.
Further, target selection module 40 comprises:
Becoming ore deposit probability calculation module, is primary data for the attribute list take initial predicted body, calculates each hexahedral one-tenth ore deposit probability in three-dimensional metallogenic prognosis body based on evidence-right-weight " method;
Evidence-right-weight " method (Bonham-Carter, 1989) is prior art, is the method through being usually used in mineral resource assessment work.
Preferred body module, for take initial predicted body Model as basis, realizes three-dimensional interactive editting function setting space screening scope based on OpenGL, and is combined into ore deposit probability calculation result, is extracted into ore deposit target of prediction body.
As shown in Figure 2, be three-dimensional mineral resource prediction multiple information integrated conduct method process flow diagram.
The method is applicable to the field such as integration of multivariate information analysis model rapid build and visual analyzing of deep, mine and Peripheral Prospecting, belong to geomathematics and Geographic Information System map making category, main application is for information such as the geology obtaining in large scale mineral exploration process, geochemistry, geophysicses, utilize Computerized three-dimensional technology, realize automatic imitation based on multiple ore forming information, comprehensive analysis with visual, auxiliary geology expert solves known deep and surrounding of mine and looks for model and the device thereof in ore deposit.
The key step of the method is as follows:
Step 201, builds the initial predicted body of three-dimensional metallogenic prognosis.
Further, step 201 is following step 202 bases to step 204.Step 203 and step 204 are all the steps that the initial predicted body for being generated by step 201 extracts initializaing variable.The core content of step 201 is based on Flood-Fill thought, realizes the quick conversion of vector body to grid body; The latent target selection based on evidence weight algorithm and three-dimensional visualization technique in step 204 is to carry out cutting apart of initial predicted body on the basis of step 201.
The Raster Data Model of body is called again piece segment model, and this model is relatively applicable to representing the space distribution of heterogeneous body data, is used widely in fields such as the metallogenic prognosis of the dark limit of large scale portion, 3-d inversion, mine reserve estimations.In large scale Deeo-Space Metallogenic Predication Study, first need to set up initial predicted grid body.Initial predicted grid body refers on the basis of the regularity of ore formation and the research of study area geologic background, a scope for metallogenic prognosis of preliminary delineation.For the computer expression of this scope, typically use an individuality and represent.The source of volume data, is to utilize newly-built one of volume modeling function, and more general method is to form an individuality by earth's surface and a downward extended distance.Set up high-precision three-dimensional metallogenic prognosis body and need to solve the quick transfer problem to Raster Data Model by the surface model of body.
Step 202, searches between geological sampling point data Quick air.
Further, step 202, take the result of step 201 as basis, is extracted geological sampling point data.The core content of step 202 is the K-NN Query algorithms based on Octree and R-Tree tree, can be used as the basis of any spatial interpolation algorithm; In step 204, by using the results model that is extracted the step 201 of attribute by step 202, divide Bu Zhaokuang target area, dark limit.
While using the Multivariate Discrete sampling point informations such as geology, geophysics, geochemistry, generally first need to carry out space interpolation.It is more that this step mainly solves in three dimensions interpolation point, exists under the prerequisite of any direction search need, fast the solution of implementation space interpolation simultaneously.The present invention mainly uses ordinary kriging and anti-distance weighted method to realize rapid space interpolation;
Step 203, discrete point diagram layer is to the extraction of the one-tenth ore deposit information of pinpoint target figure layer.
On the basis of step 201 and step 202, step 203 is selected and wants vegetarian refreshments according to the one-tenth ore deposit searching selecting take hexahedral element center of a lattice as space interpolation, calculates the corresponding one-tenth of each hexahedral element lattice ore deposit key element value.Become the calculating of ore deposit key element value mainly to comprise 6 kinds of methods.That is: the method such as maximal value, minimum value, arithmetic mean value, algorithm summation, ordinary kriging interpolation and anti-distance weighted method interpolation obtains.It is the K-NN Query of discrete point that discrete point becomes the core of the extraction of ore deposit information.
Step 204, complicated body or surface chart layer are to the metallogenic information extraction of pinpoint target figure layer.
In the present invention, body and curved surface are all represented by surface model.Complicated body refers to and comprises multiple parts but there is no the crossing triangulation network of tri patch.Wherein, the position relationship of multiple parts can be both from, also can mutually comprise.
The core of this step is the algorithm in step 201, and complicated body or curved surface are converted to grid body.Unique difference is: the hexahedral element lattice that comprise in grid body are not set by the user, and must automatically decide, parameter is identical with the parameter that user in step 201 arranges, be that its standard is the cell size of prediction grid body, synkaingenesis becomes the outer bounding box of cell and the outer bounding box of initial predicted body to carry out " asking also " calculating, certain overlapping with the cell of some prediction grid bodies on locus to guarantee newly-generated cell, as shown in Fig. 9,10.Meanwhile, property value corresponding to prediction grid body unit lattice overlapping with newly-generated cell is set to 1, otherwise is set to 0.
In Fig. 9, the schematic diagram of variable extraction module in the present invention has been described, its fundamental purpose is in order to illustrate the process of extracting body or curved surface attribute.First by body or curved surface rasterizing, secondly itself and the Raster Data Model of initial predicted body are asked and handed over calculating; Finally preserve the attribute that exists of body or curved surface.
In Figure 10, the schematic diagram of variable extraction module in the present invention is described.Mainly to improve for the method in Fig. 9.Adopt " alignment " strategy.Mainly solve the second step described in Fig. 9, " ask to hand over the Raster Data Model of initial predicted body and calculate ", this step will cause a large amount of calculating.When adopting after " alignment " strategy, do not need to ask to hand over and calculate, only need whether to exist in comparison same position cube lattice of two rasterizings simultaneously.
Step 205, is filtered into target area, ore deposit based on evidence weight algorithm.
First, each Geologic Indicators figure layer is all represented with binary-state variable, represent that with 1 Geologic Indicators exists, 0 represents that Geologic Indicators does not exist; Secondly, each Geologic Indicators calculates a pair of weight coefficient, the power while representing that this mark exists, power when another represents that this mark does not exist.In the time cannot determining that whether this mark exists, making weight coefficient is 0; The 3rd, the logarithm value of the posterior probability ratio (odds) of prediction mineral mineral deposit output equals the logarithm value of prior probability ratio and the weight coefficient sum of various Geologic Indicators.The method comes from the Bayes's relational expression in theory of probability, in the time being calculated to be ore deposit posterior probability, need to use the probability multiplication formule of limited independent random event, therefore, each control ore deposit geologic agent is with respect to this probability event of mineral deposit output, must be all condition independently.
Step 206, on the basis of geologic background and research on mineralization, applying three-dimensional is visual, carries out interactive editing, is extracted into target area, ore deposit.
According to step 205, sand smeller can classify to initial predicted grid body according to posterior probability, for example, posterior probability be more than or equal to 0.95 be category-A target area, interval [0.8,0.95) on be category-B target area, interval [0.6,0.8) on be C class target area.But in the ordinary course of things, the regularity of distribution of each class target area may not be inconsistent with sand smeller in geologic background and the research on mineralization to study area, now needs to carry out certain interactive editing work., some geology experts are thought to the independent delineation in important region out.At this, need the three-dimensional editting function of application based on OpenGL, draw a regular screening hexahedron, friendship is asked in the A that this screening hexahedron and hope are screened or B or C class target area, extracts dark limit, final mining area portion ore body.
Adopt above 6 steps, can meet to a certain extent the demand that the aspect such as ore deposit information extraction and Comprehensive Analysis Model of Unit rapid build and three-dimensional visualization is looked in the three-dimensional one-tenth of large scale ore deposit.Below in conjunction with accompanying drawing, as above method flow is specifically described.
As shown in Fig. 3 A-3B, it is R-Tree spatial index model schematic diagram; Fig. 4 is the rapid space interpolation method flow diagram of the polynary sampled point of Large scale prosecing assessments of mineral resources of the present invention.
The K-NN Query of three dimensions discrete point is the major issue of three-dimensional geographic information system research.Compared with the K-NN Query of two-dimensional space discrete point, three dimensions discrete point is higher to the rate request of K-NN Query, and its realization is also more complicated.At present, be mostly based upon on the basis of a certain space index structure about the research of K-NN Query in three dimensions.But, because each spatial index all exists deficiency, therefore, use single space index structure to carry out the effect of K-NN Query of three dimensions discrete point not very good.For the problems referred to above, the present invention proposes a kind of the improving one's methods of K-NN Query of three dimensions discrete point.The method, by Octree and two kinds of space index structure compound uses of R-Tree tree, completes the K-NN Query of three dimensions discrete point, has improved the efficiency of K-NN Query.Experiment showed, that the method can carry out the K-NN Query of three dimensions discrete point quickly and accurately.In three dimensions, along with the quantity of point rolls up, K-NN Query is very consuming time.For example, in three dimensions, 12000 points are carried out to anti-distance weighted space interpolation, establish interpolation point and reach 1,000 ten thousand, if use linear method to carry out interpolation, its time complexity will reach O (12000 × 10000000).The time complexity that can find out said method is that user is unacceptable.
At present, carried out a series of fast algorithm research for this problem.These algorithms mainly can be divided into two large classes.The first kind, utilizes Voronoi Diagram of Point-sets to carry out the search of K closest approach, but the calculated amount of the Voronoi figure of point set is very large.Equations of The Second Kind, utilizes auxiliary space index structure to carry out the search of K closest approach.The people such as Gumhold propose space separating strategy, but the method can not guarantee that its space separating has the best or close to best search speed, can not guarantee that each data point can find K field recently.Xiong Bangshu etc. improve these class methods, but still need to estimate the size of piecemeal.The three-dimensional R-Tree spatial index of the application build such as Liu Yu, improves seek rate, but when initial sampled point a lot (>=10
4) time, R-Tree sets the time and space expense spending and can not ignore.Meanwhile, in the time carrying out K-NN Query, its search radius domain of influence rectangle or ellipse that not necessarily level is put.Particularly, in mineral resource assessment field, interpolation point in space is subject to the constraint of known three-dimensional model spatial shape, for example: ore body occurrence information etc.Therefore,, in the time carrying out space interpolation, also need to consider the form of the search radius domain of influence in space.
Some relevant concepts of paper and term are to hereinafter discuss and use.
Definition 1.R-Tree space index structure
R tree is a kind of height balanced tree, is the natural expansion of B tree on K dimension (K>=2) space.R tree is made up of intermediate node and leaf node, and the minimum boundary rectangle of real data object is stored in leaf node, and intermediate node forms by the boundary rectangle of assembling its low-level nodes, comprises all these boundary rectangles.Meanwhile, R tree is a kind of dynamic indexing structure, that is: its inquiry can or be deleted and carry out simultaneously with insertion, and does not need termly tree construction to be reorganized.As shown in Fig. 3 A, 3B.
Fig. 3 A represents that R-Tree is an index tree, can use this result to carry out space querying; The locus of the raw data of the R-Tree that Fig. 3 B presentation graphs 3A explains distributes.Particularly, rectangular block R8～R19 does not comprise any other rectangular block, and therefore, these rectangular blocks are leaf node in Fig. 3 A.By that analogy, can obtain Fig. 3 A by Fig. 3 B.
Define 2. Octrees
By whole area of space according to 2
n× 2
n× 2
nmode recurrence cut zone in the quadrant of eight, space, be progressively decomposed into the cube region being included by single type area.
Variable n is a determined value, and is unique value in concrete a calculating.Particularly, n is calculated that value of 2 logarithm maximum in the length in three dimensions by initial predicted body.
In sum, the overall procedure of summing up this part algorithm is as follows, as shown in Figure 4, has described the rapid space interpolation method of the polynary sampled point of Large scale prosecing assessments of mineral resources.
In this step, first, the outer bounding box of calculation level cloud; Secondly, by outer bounding box according to 2
npower mode is cut apart, until reach acquiescence or user-specific criteria.
Variable n is a determined value, and is unique value in concrete a calculating.Particularly, n is calculated that value of 2 logarithm maximum in the length in three dimensions by initial predicted body.
Step 402, the three-dimensional R-Tree that sets up Octree Spatial Index sets index;
In this step, be that the subregion being obtained by Octree decomposition space is set to modeling rule according to R-Tree, build the R-Tree search tree of space subregion.
Space Octree by whole area of space according to 2
n× 2
n× 2
nmode recurrence cut zone in the quadrant of eight, space, obtain one group of subregion.First, these subregions are regular, but the subregion that comprises data is not rule.Therefore, use R-Tree to manage the subregion that these comprise data.
Step 403, obtains the hunting zone of interpolation point by matrix computations;
In this step, in space interpolation computation process, first need to determine the hunting zone of interpolation point.This scope, is generally used hexahedron, spheroid or spheroid to represent.Evaluate field at mineral resource prediction, generally use hexahedron or spheroid.In three dimensions, three axles determining hexahedron and spheroid not necessarily with the X of cartesian coordinate system, Y, Z axis is parallel;
Matrix computations herein refers to three dimensional space coordinate conversion.From one section above, the search of point is used hexahedron and spheroid.Its search has direction, i.e. the spatial shape of hexahedron or spheroid, is determined by tendency, inclination angle and pitch.In the time carrying out spatial data search, first carry out space coordinate transformation according to the spatial shape of search hexahedron or spheroid.Secondly, judge that interpolation point is whether in hexahedron or spheroid.Concrete conversion comprises translation and rotation.
Step 404, application ordinary kriging or anti-distance weighted method are carried out space interpolation;
In this step, the fundamental purpose of carrying out this step is the property value that obtains interpolation point according to cloud data.
According to the Search Results in step 403, use ordinary kriging or anti-distance weighted method to calculate the value at interpolation point place.Particularly, when there being a space interpolation point, first use the searching algorithm in step 403 to search in cloud data, secondly, use ordinary kriging or anti-distance weighted method to estimate the value of point to be inserted.Here ordinary kriging is simply introduced.Ordinary Kriging Interpolation (Ordinary Kriging) is the Linear Estimation of regionalized variable, and its tentation data is varied to normal distribution, thinks that the expectation value of regionalized variable Z is unknown.Interpolation Process is similar to weighting running mean, definite spatial data analysis that comes from of weighted value.
As shown in Fig. 5 A, 5B, 5C, be the neighbouring relations between voxel of the present invention; Fig. 6 be the present invention is based on the constraint of complicated geological body Model and earth's surface surface model look for ore deposit information comprehensive analysis model fast construction method process flow diagram; Fig. 7 is that Flood-Fill algorithm two-dimensional space of the present invention is realized schematic diagram; Fig. 8 is that Flood-Fill algorithm three dimensions of the present invention is realized schematic diagram.
At present, for said three-dimensional body graphic data model and modeling method thereof, existing large quantity research both at home and abroad.Its modeling method mainly comprises: piece section structure modulus method, wire frame structure modulus method, surperficial structure modulus method, entity structure modulus method and section structure modulus method.The principal feature of piece segment model is that the unevenness in complex geologic body is had to stronger ability to express, and be easy to realize quantitative calculating and the spatial analysis of complex geologic body, be therefore widely used in the mine actual productions such as 3-d inversion, mine reserve estimation and the three-dimensional metallogenic prognosis of large scale.
Set up the piece segment model of complex geologic body, first, need to complete wire frame or the surperficial structure mould of complex geologic body.Secondly, complete on this basis the piece section structure mould of complex geologic body by space querying computing.Piece segment model is relatively suitable for the application of voxel given in advance (the present invention mainly refers to rectangular parallelepiped) size in the application of mine, as reserve estimate and three-dimensional latent complex geologic body prediction work.Nonetheless,, because needs guarantee the precision of quantitatively calculating, above-mentioned application still can produce a large amount of voxel cell (>=10
6).In order to solve above-mentioned transfer problem, Chinese scholars conducts in-depth research.Its core concept is to set up auxiliary space index, accelerates the speed of rectangular parallelepiped and complicated geological surface polygon space query count.Tomaminen proposes to use rays method to realize the conversion of surface model to piece segment model, because the ray that need to draw each voxel is processed, is therefore difficult to meet the demand that a large amount of cubes of lattice calculate fast, and the method exists precision problem simultaneously.On the basis of Tomaminen work, Amanatides and Jiang propose further to improve transfer algorithm based on BSP tree space index structure.For reaching 10
6the voxel of the order of magnitude, uses BSP tree construction, can cause the degree of depth of tree excessively deeply to cause degradation problem under search efficiency.Bi Lin etc. and Jing Yongbin etc. propose to use the piece segment model of Linear Octree structure rapid build complex geologic body, and the method can meet quick generation complex geologic body surface voxel.But in the time judging complex geologic body voxel of object, still continuity has been used the thought of algorithm Tomaminen, in the time that data volume is very large, whole efficiency of algorithm still can obviously reduce.In addition, researchist finds that in actual development octree model is applicable to the piece segment model of having set up to manage more, and setting up in piece segment model process, uses octree structure will increase undoubtedly system overhead.
In sum, the key that realizes complex geologic body piece section structure mould is fast, accurately to complete by wire frame or surface model to change to piece segment model.The subject matter existing in research is at present, when need rectangular parallelepiped to be processed Cuboid(>=10
6) quantity is when larger, above-mentioned algorithm does not all provide gratifying modeling method, therefore needs to propose solution.
Some relevant concepts of paper and term are to hereinafter use.
Define 1. complex geologic bodies
Body is three-dimensional geometry element, the consistent useful space of dimension that uses confining surface to surround.The body of general satisfaction foregoing description is called canonical body., any must be a surperficial part, can not hang face; Every there are and only have two proximal surfaces on limit, can not be unsettled; Each summit is at least adjacent with three limits, can not isolate.On this basis, complex geologic body is made up of one group of body, and these bodies mutually disjoint, but can mutually comprise or separate.
Number=1.5 × leg-of-mutton the number on theorem 1 complicated geological body Model triangle gridding limit
Prove: the limit that known triangular mesh comprises and triangle exist following relation:
3T=2E
i+E
b
Wherein, T is leg-of-mutton number, E
ifor the number of all common edge, E
bfor the number of non-common edge.I, b in this formula can be any given.
The known complicated geological body Model in space with topological structure is sealing triangle gridding, E again
b=0, there is following relation: 3T=2E
i, i.e. E
i=1.5 × T
Prove complete.
Define the neighbouring relations between 2 voxels
Voxel can be understood as the expansion of two-dimensional pixel in three dimensions, is one group of cubic units that is distributed in orthogonal grid center.Between two voxels, be divided into 26-in abutting connection with, 18-adjacency and 6-adjacency, as shown in Fig. 5 A, 5B, 5C according to concurrent, altogether limit and the qualifications such as coplanar respectively.
The a certain cube of Fig. 5 A-5C representation space and other cubical position relationship.In Fig. 5 A, represent 26-adjacency; Fig. 5 B represents 18-adjacency; Fig. 5 C represents 6-adjacency.
Define 3. complex geologic body border rectangular parallelepipeds
If there is rectangular parallelepiped Cell, the 6-of Cell has at least 1 rectangular parallelepiped not exist in rectangular parallelepiped, claims that Cell is the border rectangular parallelepiped of complex geologic body.
The present invention proposes to apply uniform grid and cuts apart with Flood-fill thought and realize the conversion of complex geologic body surface model to piece segment model.The medium mesh segmentation algorithm of algorithm proposed be mainly used in the outer bounding box of subdivision complex geologic body in the present invention.The benefit of doing is like this that the time complexity of subdivision process is O (1), and rectangular parallelepiped after subdivision is orderly, and its locating query time complexity is also O (1).By waiting rectangular parallelepiped of obtaining of mesh segmentation can be divided into 3 classes according to the relation of itself and complex geologic body.The first kind, in the outside of complex geologic body; Equations of The Second Kind, in the inside of complex geologic body; The 3rd class, on the surface of complex geologic body.Known according to the definition of bodily form piece segment model, piece segment model is made up of Equations of The Second Kind and the 3rd class rectangular parallelepiped.The overwhelming majority in the 3rd class rectangular parallelepiped can obtain fast by asking to hand over to calculate on the basis of waiting mesh segmentation.The 3rd part for class and the rectangular parallelepiped of Equations of The Second Kind need to apply Flood-fill algorithm and calculate acquisition.
In sum, summing up the present invention, to propose the overall procedure of algorithm as follows, as shown in Figure 6:
Step 601, sets up the outer bounding box of complex geologic body;
In this step, get original geologic body at X, Y, the scope [(xmin, ymin, zmin), (xmax, ymax, zmax)] of tri-coordinate axis of Z, builds a hexahedron.
Step 602, the rectangular parallelepiped size uniform of inputting according to user is cut apart the outer bounding box of complex geologic body;
In this step, adopt man-machine interactive operation pattern, specified the precision of cutting apart outer bounding box by user, at X, Y, the step-length in three directions of Z axis.
Step 603, judges that according to the 6-syntople between rectangular parallelepiped after even partition whether bin (triangle) is crossing with rectangular parallelepiped;
Bin is the base unit of describing geometric objects in graphics." bin refers to triangle " is described, avoids ambiguity.In other different algorithms, likely bin is rectangle.
In this step, 6-syntople as shown in Figure 5 C.Hand over and calculate about asking of spatial line segment, can in computational geometry book, find.
Step 604, on the basis of known complex geologic body surface cuboid, application scanning line thought detects residue rectangular parallelepiped whether in the inside of complex geologic body;
In this step, as shown in following algorithm 3.
Step 605, retains complicated geological surface and inner rectangular parallelepiped and is distinguished, and obtains final piece segment model.
In this step, as shown in following algorithm 3.
In sum, according to algorithm main thought and overall procedure, sum up the flow process of algorithm as shown in Figure 6.
According to as above describing, algorithm of the present invention mainly comprises two large divisions.Part I is to search the lip-deep rectangular parallelepiped of complex geologic body, and Part II is the rectangular parallelepiped of searching complex geologic body inside.
Part I mainly comprises two steps: the first step is the rectangular parallelepiped of searching on complex geologic body surface-boundary limit; Second step is the rectangular parallelepiped of searching on the tri patch of complicated geological surface.
The present invention will use following algorithm 1 and algorithm 2 to complete searching of Part I rectangular parallelepiped.
If being one group, BM obtains rectangular parallelepiped by mesh segmentation such as grade;
it is the limit on complex geologic body.
Input:
and BM
Output:
one group of rectangular parallelepiped CubeArray of process
Step (1) calculation level A and the rectangular parallelepiped CellA and the CellB that put in the BM of B place;
Step (2) judges whether rectangular parallelepiped CellA and CellB equate.If unequal, calculate
with the crossing rectangle P of rectangular parallelepiped CellA and forward step (3) to, otherwise, record rectangular parallelepiped CellA and forward step (4) to;
Step (3) in abutting connection with definition and crossing rectangle P, obtains the rectangular parallelepiped Cell ' adjacent with rectangular parallelepiped CellA, the attribute m_bCellStatus=1 of mark rectangular parallelepiped Cell ' according to 6-.If Cell ' unequal to CellB returns to step (2), otherwise forward step (4) to.
Step (4) finishes.
Rectangular parallelepiped Cell ' is a rectangular parallelepiped newly obtaining; It is according to obtaining with 6-adjacency and the rectangle P of CellA.
Algorithm 2 judges that whether tri patch is crossing with rectangular parallelepiped
Input: CubeArray and tri patch t
Output: one group of rectangular parallelepiped cTArray on tri patch
Step 2, in abutting connection with search rule, searches a rectangular parallelepiped c ' according to 6-, c ' meet at least exist a limit e ' (e ' ≠ e) crossing with tri patch t.If there is such rectangular parallelepiped c ', continue to search calculating.Otherwise forward step 3 to.
Step 3, finishes.
After acquisition and the lip-deep rectangular parallelepiped voxel of complex geologic body, also need to obtain other rectangular parallelepipeds on piece segment model.In to the search procedure of Part II rectangular parallelepiped, the present invention proposes Flood-fill algorithm application to three dimensions, the 6-by searching rectangular parallelepiped to be judged, in abutting connection with the inside and outside state of rectangular parallelepiped in direction, determines that whether current rectangular parallelepiped is in complex geologic body inside.Because this deterministic process is the result based on waiting mesh segmentation, therefore the required expense of algorithm is relatively low.The thought of algorithm 3 as shown in FIG. 7 and 8.In two-dimensional space, as Fig. 7, application Flood-fill algorithm determines that current grid cell is in the time of the state of Polygon, needs to search for to 4 directions of grid cell lattice respectively, determines the inside and outside state of grid cell to be judged according to Search Results.In like manner, in three dimensions, as shown in Figure 8, rectangular parallelepiped to be determined can be by its 6-the inside and outside state in abutting connection with rectangular parallelepiped in direction, determine current its whether in complex geologic body inside.
Finally, obtain the residue rectangular parallelepiped of piece segment model based on Flood-fill algorithm.Use the outside rectangular parallelepiped of scanning line method mark part.Follow the tracks of the rectangular parallelepiped Cell of its process along directions of rays, if Cell is the rectangular parallelepiped of complicated geological surface, stopping tracking and mark Cell is definedCube; Otherwise mark Cell is outside rectangular parallelepiped and continues to follow the tracks of until run into another border rectangular parallelepiped; Travel through all unlabelled rectangular parallelepipeds.If wherein any one unmarked rectangular parallelepiped undefinedCube, does ray according to its 6-Adjacent rule along 6 directions, the direction of getting the ray r that runs into recently outside rectangular parallelepiped judges direction as undefinedCube; Calculate ray r the quantity n judgement cube lattice of rectangular parallelepiped of process inside and outside geologic body.As described in algorithm 3 below.
Algorithm 3 obtains the residue rectangular parallelepiped of piece segment model based on scan-line algorithm
Input: border rectangular parallelepiped S set
Output: piece segment model BlockModel
Step 2 is used the outside rectangular parallelepiped of scanning line method mark part.Follow the tracks of the rectangular parallelepiped Cell of its process along directions of rays, if Cell is the rectangular parallelepiped on ore body surface, stop following the tracks of, and mark Cell is definedCube; Otherwise mark Cell is outside rectangular parallelepiped and continues to follow the tracks of until run into another border rectangular parallelepiped;
Step 3 travels through all unlabelled rectangular parallelepipeds.If wherein any one unmarked rectangular parallelepiped undefinedCube, does ray according to its 6-Adjacent rule along 6 directions, the direction of getting the ray r that runs into recently outside rectangular parallelepiped judges direction as undefinedCube;
Step 4 calculate ray r the quantity n of rectangular parallelepiped of process, if n is odd number, undefineCube is inner rectangular parallelepiped; Otherwise UndefinedCell is outside rectangular parallelepiped;
Step 5 finishes.
The method that the present invention proposes, be applicable to component of forecast simulation and the polynary fields such as ore deposit information comprehensive analysis model rapid build and visual analyzing thereof of looking in the three-dimensional metallogenic prognosis of large scale, be mainly used in the various data that obtain in large scale mineral exploration process, such as sampling number certificate, geology terrain data etc., utilize the three-dimensional computations machine technologies such as rapid space interpolation, curved surface automatic Fitting, Raster Data Model rapid build, realize the three-dimensional information automatic imitation of the multivariate datas such as geology, geophysics, geochemistry, comprehensive analysis and visual demonstration.
Claims (10)
1. an informix treating apparatus in three-dimensional mineral resource prediction evaluation, is characterized in that, comprising:
Predictor builds module, for building the initial predicted body of three-dimensional metallogenic prognosis;
Space search module, connects described predictor and builds module, for the component of forecast to initial predicted body rapid extraction spatial discrete points, by result assignment to initial predicted body and be retained in the attribute list of initial predicted body;
Variable extraction module, connects described predictor and builds module, for the component of forecast of being expressed by surface model is carried out to rasterizing, and by rasterizing result assignment in the attribute list of initial predicted body;
Target selection module, connects described space search module, described variable extraction module, for the attribute list of initial predicted body being screened with preferred, obtains target of prediction body.
2. device according to claim 1, is characterized in that, described predictor builds module and further comprises:
Outer bounding box extraction module, for extracting the outer bounding box of surface or curved surface data;
Outer bounding box gridding module, for being decomposed into hexahedron according to the grid granularity of setting by outer bounding box;
Face body is asked friendship module, asks friendship to calculate for carry out tri patch and hexahedron according to outer bounding box;
Hexahedron sort module, for based on se ed filling algorithm and ask friendship result of calculation, carries out rapid screening to the hexahedron that forms three-dimensional metallogenic prognosis body, generates the attribute list of initial predicted body and initial predicted body.
3. device according to claim 1, is characterized in that, described space search module further comprises:
Module set up in the first index, for setting up the Octree Spatial Index of 3 d-dem point;
Module set up in the second index, for setting up the three-dimensional R-Tree tree index of Octree Spatial Index;
Hunting zone acquisition module, for obtaining hunting zone to be made an inventory of;
Space interpolation is realized module, for setting up module according to hunting zone and the first index, the result of module set up in the second index, and apply Ke Lige, anti-distance weighted formula carries out space interpolation;
Search assignment module, for by space interpolation result assignment to initial predicted body, and be retained in the attribute list of initial predicted body.
4. according to the device described in claim 1,2 or 3, it is characterized in that, described variable rapid extraction module further comprises:
Rasterizing processing module, for rasterizing surface model body or curved surface data;
Result assignment module, for by rasterizing result according to locus assignment the attribute list to initial predicted body.
5. according to the device described in claim 1,2 or 3, it is characterized in that, described target selection module further comprises:
Becoming ore deposit probability calculation module, is primary data for the attribute list take initial predicted body, calculates each hexahedral one-tenth ore deposit probability in three-dimensional prediction body based on evidence-right-weight " method;
Preferred body module, for take initial predicted body Model as basis, and is combined into ore deposit probability calculation result, is extracted into ore deposit target of prediction body.
6. an information comprehensive processing method in three-dimensional mineral resource prediction evaluation, is characterized in that, comprising:
Step 1, builds the initial predicted body of three-dimensional metallogenic prognosis;
Step 2, to the component of forecast of initial predicted body rapid extraction spatial discrete points, by result assignment to initial predicted body and be retained in the attribute list of initial predicted body;
Step 3, carries out rasterizing to the component of forecast of being expressed by surface model, and by rasterizing result assignment in the attribute list of initial predicted body;
Step 4, screens with preferred the attribute list of initial predicted body, obtains target of prediction body.
7. method according to claim 6, is characterized in that, described step 1, further comprises:
Extract the outer bounding box of surface or curved surface data;
According to the grid granularity of setting, outer bounding box is decomposed into hexahedron;
Carrying out tri patch and hexahedron according to outer bounding box asks to hand over and calculates;
According to se ed filling algorithm and ask friendship result of calculation, the hexahedron that forms three-dimensional metallogenic prognosis body is carried out to rapid screening, generate the attribute list of initial predicted body and initial predicted body.
8. method according to claim 6, is characterized in that, described step 2, further comprises:
Set up the Octree Spatial Index of 3 d-dem point;
Set up the three-dimensional R-Tree tree index of Octree Spatial Index;
Obtain hunting zone to be made an inventory of;
According to hunting zone and the first index set up module, the result of module set up in the second index, and apply Ke Lige, anti-distance weighted formula carries out space interpolation;
Space interpolation result assignment, to initial predicted body, and is retained in the attribute list of initial predicted body.
9. according to the method described in claim 6,7 or 8, it is characterized in that, described step 3, further comprises:
Rasterizing surface model body or curved surface data;
By rasterizing result according to locus assignment in the attribute list of initial predicted body.
10. according to the method described in claim 6,7 or 8, it is characterized in that, described step 4, further comprises:
Take the attribute list of initial predicted body as primary data, calculate each hexahedral one-tenth ore deposit probability in three-dimensional prediction body based on evidence-right-weight " method;
Take initial predicted body Model as basis, and be combined into ore deposit probability calculation result, be extracted into ore deposit target of prediction body.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410077136.2A CN103903061B (en) | 2014-03-04 | 2014-03-04 | Information comprehensive processing device and method in three-dimensional mineral resource prediction evaluation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410077136.2A CN103903061B (en) | 2014-03-04 | 2014-03-04 | Information comprehensive processing device and method in three-dimensional mineral resource prediction evaluation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103903061A true CN103903061A (en) | 2014-07-02 |
| CN103903061B CN103903061B (en) | 2016-08-31 |
Family
ID=50994371
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410077136.2A Active CN103903061B (en) | 2014-03-04 | 2014-03-04 | Information comprehensive processing device and method in three-dimensional mineral resource prediction evaluation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103903061B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104167021A (en) * | 2014-07-31 | 2014-11-26 | 北京大学 | Space target indexing method based on terrestrial space three-dimensional partition |
| CN105678399A (en) * | 2015-12-29 | 2016-06-15 | 中国地质科学院矿产资源研究所 | Regional mineral resource quantity estimation analysis method and system |
| CN106295848A (en) * | 2015-06-26 | 2017-01-04 | 中国石油化工股份有限公司 | A kind of method and system determining that geology risk domain is distributed |
| CN108241593A (en) * | 2018-01-22 | 2018-07-03 | 西南林业大学 | A kind of Spatial sampling object localization method and system |
| CN110826122A (en) * | 2019-10-12 | 2020-02-21 | 中广核工程有限公司 | Voxel method and system for nuclear power three-dimensional layout design model |
| CN111583423A (en) * | 2020-05-19 | 2020-08-25 | 北京数字绿土科技有限公司 | Method and device for extracting cross section line of point cloud data |
| CN111739164A (en) * | 2019-03-19 | 2020-10-02 | 北京京东尚科信息技术有限公司 | Box body modeling method and device, robot picking system, electronic device and medium |
| US10839598B2 (en) | 2016-07-26 | 2020-11-17 | Hewlett-Packard Development Company, L.P. | Indexing voxels for 3D printing |
| CN113703068A (en) * | 2021-10-26 | 2021-11-26 | 核工业北京地质研究院 | Method for determining uranium ore distribution |
| CN114332369A (en) * | 2021-12-28 | 2022-04-12 | 埃洛克航空科技（北京）有限公司 | Building image processing method, building image processing device, building image processing equipment and storage medium |
| CN114896564A (en) * | 2022-05-23 | 2022-08-12 | 武汉市市政建设集团有限公司 | Transient electromagnetic two-dimensional Bayesian inversion method adopting self-adaptive Thiessen polygon parameterization |
| CN117095134A (en) * | 2023-10-18 | 2023-11-21 | 中科星图深海科技有限公司 | Three-dimensional marine environment data interpolation processing method |
| CN117197398A (en) * | 2023-11-07 | 2023-12-08 | 中国电建集团昆明勘测设计研究院有限公司 | Statistical method, device, equipment and storage medium of mine resource model |
| CN118469740A (en) * | 2024-07-09 | 2024-08-09 | 成都理工大学 | Ore body data three-dimensional ore formation prediction method and system based on GOCAD software |
| CN118469740B (en) * | 2024-07-09 | 2024-10-25 | 成都理工大学 | Ore body data three-dimensional ore formation prediction method and system based on GOCAD software |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101055631A (en) * | 2006-04-10 | 2007-10-17 | 中国地质大学（武汉） | Space data fuzzy evidence weight analysis method |
| CN102622526A (en) * | 2012-04-06 | 2012-08-01 | 中国矿业大学(北京) | Digital mine tunneling search prediction method |
| CN102651144A (en) * | 2012-03-30 | 2012-08-29 | 北京师范大学 | Method for geologic body to quickly and dynamically generate linear octree |
-
2014
- 2014-03-04 CN CN201410077136.2A patent/CN103903061B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101055631A (en) * | 2006-04-10 | 2007-10-17 | 中国地质大学（武汉） | Space data fuzzy evidence weight analysis method |
| CN102651144A (en) * | 2012-03-30 | 2012-08-29 | 北京师范大学 | Method for geologic body to quickly and dynamically generate linear octree |
| CN102622526A (en) * | 2012-04-06 | 2012-08-01 | 中国矿业大学(北京) | Digital mine tunneling search prediction method |
Non-Patent Citations (2)
| Title |
|---|
| 唐艳华: "三维成矿信息分析若干关键技术研究", 《中国优秀硕士学位论文全文数据库》 * |
| 朱大明: "基于地理信息系统（GIS）的个旧高松矿田成矿预测", 《中国博士学位论文全文数据库》 * |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104167021A (en) * | 2014-07-31 | 2014-11-26 | 北京大学 | Space target indexing method based on terrestrial space three-dimensional partition |
| CN106295848A (en) * | 2015-06-26 | 2017-01-04 | 中国石油化工股份有限公司 | A kind of method and system determining that geology risk domain is distributed |
| CN105678399A (en) * | 2015-12-29 | 2016-06-15 | 中国地质科学院矿产资源研究所 | Regional mineral resource quantity estimation analysis method and system |
| US10839598B2 (en) | 2016-07-26 | 2020-11-17 | Hewlett-Packard Development Company, L.P. | Indexing voxels for 3D printing |
| CN108241593A (en) * | 2018-01-22 | 2018-07-03 | 西南林业大学 | A kind of Spatial sampling object localization method and system |
| CN108241593B (en) * | 2018-01-22 | 2020-05-15 | 西南林业大学 | Space sampling target positioning method and system |
| CN111739164A (en) * | 2019-03-19 | 2020-10-02 | 北京京东尚科信息技术有限公司 | Box body modeling method and device, robot picking system, electronic device and medium |
| CN110826122A (en) * | 2019-10-12 | 2020-02-21 | 中广核工程有限公司 | Voxel method and system for nuclear power three-dimensional layout design model |
| CN110826122B (en) * | 2019-10-12 | 2023-02-28 | 中广核工程有限公司 | Voxel method and system for nuclear power three-dimensional layout design model |
| CN111583423A (en) * | 2020-05-19 | 2020-08-25 | 北京数字绿土科技有限公司 | Method and device for extracting cross section line of point cloud data |
| CN113703068B (en) * | 2021-10-26 | 2022-02-15 | 核工业北京地质研究院 | Method for determining uranium ore distribution |
| US11460601B1 (en) | 2021-10-26 | 2022-10-04 | Beijing Research Institute Of Uranium Geology | Method for determining distribution of uranium deposit |
| CN113703068A (en) * | 2021-10-26 | 2021-11-26 | 核工业北京地质研究院 | Method for determining uranium ore distribution |
| CN114332369A (en) * | 2021-12-28 | 2022-04-12 | 埃洛克航空科技（北京）有限公司 | Building image processing method, building image processing device, building image processing equipment and storage medium |
| CN114896564A (en) * | 2022-05-23 | 2022-08-12 | 武汉市市政建设集团有限公司 | Transient electromagnetic two-dimensional Bayesian inversion method adopting self-adaptive Thiessen polygon parameterization |
| CN117095134A (en) * | 2023-10-18 | 2023-11-21 | 中科星图深海科技有限公司 | Three-dimensional marine environment data interpolation processing method |
| CN117095134B (en) * | 2023-10-18 | 2023-12-22 | 中科星图深海科技有限公司 | Three-dimensional marine environment data interpolation processing method |
| CN117197398A (en) * | 2023-11-07 | 2023-12-08 | 中国电建集团昆明勘测设计研究院有限公司 | Statistical method, device, equipment and storage medium of mine resource model |
| CN117197398B (en) * | 2023-11-07 | 2024-02-09 | 中国电建集团昆明勘测设计研究院有限公司 | Statistical method, device, equipment and storage medium of mine resource model |
| CN118469740A (en) * | 2024-07-09 | 2024-08-09 | 成都理工大学 | Ore body data three-dimensional ore formation prediction method and system based on GOCAD software |
| CN118469740B (en) * | 2024-07-09 | 2024-10-25 | 成都理工大学 | Ore body data three-dimensional ore formation prediction method and system based on GOCAD software |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103903061B (en) | 2016-08-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103903061B (en) | Information comprehensive processing device and method in three-dimensional mineral resource prediction evaluation | |
| Natali et al. | Modeling Terrains and Subsurface Geology. | |
| Xu et al. | Block modeling and segmentally iterative ray tracing in complex 3D media | |
| EP4030198B1 (en) | Building accurate training images for automatic seismic interpretation | |
| WO2014099204A1 (en) | Method and system for geophysical modeling of subsurface volumes based on computed vectors | |
| Jia et al. | A triangulated irregular network constrained ordinary Kriging method for three-dimensional modeling of faulted geological surfaces | |
| Galley et al. | Geophysical inversion for 3D contact surface geometry | |
| Pouliot et al. | Reasoning about geological space: Coupling 3D GeoModels and topological queries as an aid to spatial data selection | |
| CA2939957C (en) | Multi-z polyline to single-z horizons conversion | |
| CN115494560A (en) | Novel modeling method based on combination of CT (computed tomography) and drilling data | |
| Ji et al. | An automated method to build 3D multi-scale geological models for engineering sedimentary layers with stratum lenses | |
| CN112634447B (en) | Outcrop stratum layering method, device, equipment and storage medium | |
| de Oliveira Miranda et al. | Finite element mesh generation for subsurface simulation models | |
| Dong et al. | RETRACTED ARTICLE: Research on North Gulf distributed big data submarine 3D terrain computing system based on remote sensing and multi-beam | |
| Michalak et al. | Using Delaunay triangulation and cluster analysis to determine the orientation of a sub-horizontal and noise including contact in Kraków-Silesian Homocline, Poland | |
| Shen et al. | Three-dimensional modeling of loose layers based on stratum development law | |
| Zuo | The construction of stratigraphic structure model in mining area under virtual reality–geographic information system | |
| Zhou et al. | Fault surface extraction from a global perspective | |
| Taranchuk et al. | The integrated program complex of the composer of geological models. Examples of results | |
| Zoraster | A surface modeling algorithm designed for speed and ease of use with all petroleum industry data | |
| Patel et al. | Modeling terrains and subsurface geology | |
| CN112053437B (en) | Three-dimensional modeling method for geophysical exploration based on contour line | |
| Guo et al. | Pattern-Based Multiple-point Geostatistics for 3D Automatic Geological Modeling of Borehole Data | |
| He et al. | Research on 3D Product Service System Based on Spherical Model | |
| Westerteiger et al. | Interactive Retro-Deformation of Terrain for Reconstructing 3D Fault Displacements |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20180823 Address after: 100037 26 million village street, Xicheng District, Beijing Patentee after: Mining Resources Inst., Chinese Academy of Geology Sciences Address before: 100037 26 million village street, Xicheng District, Beijing Co-patentee before: Li Nan Patentee before: Mining Resources Inst., Chinese Academy of Geology Sciences |