CN103020629B - Method for extracting and identifying bit field data construction information - Google Patents

Abstract

The invention discloses a method for extracting and identifying bit field data construction information, which comprises the following steps: a. carrying out primary processing on original bit field raster data; b. scanning each line of data, taking an asymmetric waveform consisting of a local maximum and a local minimum adjacent to the local maximum as a period, and establishing a feature space of each waveform; c. taking each 3 rows of data as a group, taking each waveform as a unit, and calculating a value function of connection between each waveform of each group of middle rows and adjacent waveforms on the front and rear rows of data so as to judge the connection state between the two waveforms; d. c, detecting the public area of the event segments formed by waveform connection in the step c, connecting the event segments with the public area, updating the event directory and storing the connected events into an event file; e. the detection result of the event is directly displayed in the graph. The invention can quickly and accurately identify and extract the structure information in the bit field data.

Description

A kind of potential field data structural information is extracted and is known method for distinguishing

Technical field

The present invention relates to automatic identification and the extractive technique field of potential field characteristics of image, specifically a kind of potential field data structural information is extracted and is known method for distinguishing.

Background technology

At present, a lot of method for other side is applied to the automatic identification of contraposition field picture feature and extracts abroad.Radon transform is applied to the continuation up and down of gravimetric data and removes linear feature useless in gravimetric data composite diagram by Yunxuan (1992), achieve satisfied effect, and before this kind of method, be widely used in medical treatment, earthquake and remote sensing aspect; Niuetal. (2001), Vassilas, N.etal. (1999) use self-organizing clustering figure to process electromagnetic data in conjunction with Hough algorithm, automatically the border of geologic body is extracted, and Hough algorithm is also for image procossing at first, especially in remote sensing image processing; EatonandVasudevan (2004) will be used for a kind of syntax pattern distinguishment technology---Framework Arithmetic (LeandNyland, 1990 of seism processing originally; LuandCheng, 1990; Lietal., 1997) be applied to the identification of potential field digital picture neutral line shape feature and automatically extract.

At home, many scholars have also done a lot of research in automatic target seeker gyro characteristic aspect of extracting, but also do not cause enough attention, and scientific payoffs is not abroad enriched.High-strength, Cheng Fangdao (1993) generates the greatest gradient two condition pixel map of potential field digital picture; Dodd's literary composition waits people (1996) organically to be combined with rotation transformation methods by greatest gradient method for tracing, is used for automatically generating the vector image weighing magnetic potential field greatest gradient; The people such as Wang Silong (1996) utilize Hough transformation to be semi-automatically extracted the tectonic information of gravity and magnetic data; The people such as Chen Yongliang (2002) discuss the extraction method of local maximum, minimal value line in the horizontal first order derivative image in heavy magnetic potential field; The people such as Zhang Lili (2006) will be used for detection and the enhancing of heavy magnetic image neutral line feature based on the Radon conversion improved and gradient calculation, the method can highlight linear feature in domain of variation, is conducive to extraction and the enhancing of linear feature.The people such as Liu Jinlan (2007), the people such as champion Bai (2008), the people such as Xia Lingyan (2008) have used the linear feature in the digital picture of image processing techniques contraposition field to identify respectively; The people such as Yang Yushan (2003) utilize the differential characteristics of wavelet details to carry out fracture analysis to gravity field, the method overcome the shortcoming of the off-note obfuscation that there will be in upward continuation method, compared with traditional method asking horizontal first order derivative, upward continuation etc. to analyze to rupture, it makes the fracture analysis of gravity field more reliable, accurately.

For identification and the extraction of tectonic information in potential field data, classic method needs geophysicist to rely on potential field information characteristics to carry out artificial visual by experience to complete.Such work and time-consuming, require great effort again, efficiency is very low.And the requirement of work to information judging person itself is higher.When running into the situation that expert can not come to personally, the identification and extraction work of information is often absorbed in stagnation.

Although done many research both at home and abroad in the automatic identification and extraction of potential field data tectonic information, also achieve many achievements, a lot of method but could not give play to effect when designing originally in actual applications.Such result on the one hand with disposal route itself defect about: the disposal route had this in design time just could not consider the general applicability of method, cause method can only utilization in very narrow; Some methods are then do not notice the feature of target seeker gyro data, copying mechanically and applying indiscriminately other similar data processing method, causing the method in practice not reach desirable effect when designing.On the other hand then with application method data processing personnel about: the potential field data processing had explains that personnel are in the feature carrying out being unfamiliar with when processing and explain various disposal route to potential field data, various process and interpretation procedure is not noted to combine, select single when selecting data processing method, therefore can not play the advantage of each method, thus cause the error in real work.

In view of above-mentioned existing potential field data structural information extracts the problems and defect in use existed with recognition technology, the present inventor relies on working experience for many years and abundant professional knowledge are actively studied and innovate, the potential field data structural information of a kind of novelty of final invention is extracted and is known method for distinguishing, can identifying quickly and accurately and extract the tectonic information in potential field data, is paving the way fast smoothly of work.

Summary of the invention

In order to solve the above-mentioned problems in the prior art, the invention provides a kind of potential field data structural information and extract and know method for distinguishing, achieving and identify quickly and accurately and the tectonic information extracted in potential field data.

In order to solve the problems of the technologies described above, present invention employs following technical scheme:

A kind of potential field data structural information is extracted and is known method for distinguishing, comprises the steps:

A. preliminary process is carried out to original potential field raster data;

B. the potential field raster data crossed of input processing, every column data is scanned, find out each local maximum in every column data and local minimum adjacent with it, the asymmetric waveform be made up of a local maximum and an adjacent with it local minimum is one-period, sets up the feature space of each waveform;

C. it is one group with every 3 column datas, the each waveform obtained in step b is as a unit, each waveform calculating the middle ordered series of numbers of often group closes on on the column data of front and back two cost function be connected between waveform, judges the connection status between two waveforms according to cost function;

D. the public domain of the event segments that waveform is connected to form in detection steps c, will have the event segments of public domain to couple together, while event segments connects, constantly update event list, by the event that connected stored in incident file;

E. the result of detection of event is directly shown in the graphic.

Further, in described step b, each local extremum in this column data is obtained by each column data in scanning raster data, the local minimum that each local maximum is adjacent with it with one forms a waveform, local maximum do not find a local minimum corresponded then this local maximum just cast out; Equally, if a local minimum does not find a local maximum corresponded, so it also can be rejected; The dimensionless proper vector f expressed with following formula 1 defines the feature space of each waveform:

F=(λ _pApA, λ _nAnA, λ _pWpW, PO) ^tformula 1

In formula, PA is crest amplitude; NA is trough amplitude; PW is the distance between crest to trough, i.e. pulse width; λ _pA, λ _nAwith the weights that λ PW is PA, NA and PW respectively; PO is the polarity in cycle; T represents vector transpose.

Further, in described step b, the line number of raster data is m and columns is n, and namely raster data size is m × n, the i-th row in raster data, the data a of jth row _i,jbe one and but be much to seek the corresponding extreme point with its composition waveform by the Local Extremum found out, as i=m-1, a _i,jbe rejected as independent extreme point; As i<m-1, get a _{i+1, j}for with a _i,jcorresponding very big or minimum point, and form waveform.

Further, in described step c, before the distance between calculating two wave vector space, first the polarity of two waveforms is judged, if two polarity of waves are consistent, then calculate the distance in two wave vector spaces, then judge whether two waveforms should connect according to maximum value functional value; Otherwise, leap to next waveform, then carry out the polarity of judgement two waveform, and so forth.

Further, in step c, in periodic unit group, comprise two cycle (waveform) a and b, so connected the value Cost (a of the event formed by this two cycles, b) then by these two cycles feature space between distance D (a, b) weigh:

Cost (a, b) ≡ D (a, b)=| f _a-f _b| formula 2

F in formula 2 _aand f _bthe proper vector of indication cycle a and cycle b respectively, wherein | f| computation rule is defined as follows:

$\left|f\right|=\left|{(f_1,...,f_4)}^T\right|=\sqrt{\underset{i=1}{\overset{4}{\mathrm{\&Sigma;}}}{f}_i^2}$ Formula 3

Wherein f _ii-th element of representation vector f, if the attribute in two cycles is consistent, the value so connected between them is zero; If the attribute in two cycles is inconsistent, the value weighing apparatus so connected between them is greater than zero;

For comprise 3 cycles unit group c, d, e}, its value of combination function is represented by following formula 4:

$\mathrm{Cost}(c,d,e)={\mathrm{\&alpha;}}_0[D(c,d)+D(d,e)]+{\mathrm{\&alpha;}}_1\sqrt{A(c,d,e)}$ Formula 4

In formula 4 A (c, d, e) be used to indication cycle c, d and e tri-summit composition the area of triangle in 4 gt, parameter alpha ₀and α ₁be adjacent periods and the weight of non-conterminous cycle shared by cost function respectively, if the weight of adjacent and non-conterminous cycle shared by cost function is identical, so α be set ₀, α ₁be respectively 0.5; The formula that is expressed as follows 5 of area A:

A (c, d, e) ≡ D (c, d) D (d, e) sin θ formula 5

Wherein, formula 6

It is as follows that Inner in above formula amasss formal definition:

$(f,g)\&equiv;\underset{i=1}{\overset{4}{\mathrm{\&Sigma;}}}{f}_1\&CenterDot;g_i$ Formula 7

Therefore A (c, d, e) more goes to zero, then the cycle c, d and e linear concord better.

Further, in steps d, the event segments with common cycle is connected, until form complete event list, finally calculate the attribute of complete event, and event attribute is left in event list, event attribute to comprise in event the mean value of rhythmic attribute and event length L, event length L is obtained by following formula 8:

$L=\sqrt{{\mathrm{dx}}^2+{\mathrm{dy}}^2}$ Formula 8

Wherein, dx and dy is respectively the length that event extends eastwards and northwards.

Compared with prior art, beneficial effect of the present invention is:

The present invention avoids the problem that the different cycle of segment polarity in prior art is also joined together in cycle connection procedure.Achieve and identify quickly and accurately and the tectonic information extracted in potential field data.And achieve good effect, give strong support to the geologic interpretation of real data.The judgment rule of the extreme point adopted in the present invention, avoids the mistake produced when data boundary identification in prior art.The composition form in waveform character space adopted in addition, can independently select important feature according to the feature of data, strengthens the weight of key character, thus improves the correctness identified.

Accompanying drawing explanation

Fig. 1 is that potential field data structural information of the present invention is extracted and the process flow diagram knowing method for distinguishing;

Fig. 2 is the model for rim detection Contrast on effect;

Fig. 3 A and Fig. 3 B is respectively the gravity anomaly of above-mentioned theory model generation and the figure of magnetic anomalies;

Fig. 4 A to Fig. 4 D is respectively the gravity and magnetic anomaly data basis produced with theoretical model, carries out with HDTDR method, TDR method, THETA method, Framework Arithmetic the recognition result figure that gravity anomaly rim detection obtains respectively;

Fig. 5 A to Fig. 5 D is respectively the gravity and magnetic anomaly data basis produced with theoretical model, carries out with HDTDR method, TDR method, THETA method, Framework Arithmetic the recognition result figure that magnetic anomalies rim detection obtains respectively;

Fig. 6 is the tectonic information recognition result figure of gravity anomaly in the embodiment of the present invention;

Fig. 7 is the tectonic information recognition result figure of magnetic anomaly in the embodiment of the present invention;

It is highly the gravimetric data recognition result of 7 ~ 60km that Fig. 8 prolongs on being;

Fig. 9 is for boat magnetic prolonging height 12 ~ 50km recognition result;

Gravity recognition result, topographic and geologic and HQ5 line are explained the integrated figure under three-dimensional environment such as section by Figure 10;

Figure 11 is boat magnetic skeleton identification and HQ5 line MT sectional drawing;

Figure 12 is the skeleton recognition result to Eastern China boat magnetic;

Figure 13 is boat magnetic skeleton line and HQ10 line MT sectional drawing;

Figure 14 is Middle_Lower Yangtze metallogenic belt tectonic framework and mineral deposit corresponding relation figure;

Figure 15 is for boat magnetic prolonging height 1 ~ 7km skeleton recognition result figure.

Embodiment

Below in conjunction with the drawings and specific embodiments, the present invention is described in further detail, but not as a limitation of the invention.

Fig. 1 is that potential field data structural information of the present invention is extracted and the process flow diagram knowing method for distinguishing.As shown in Figure 1, a kind of potential field data structural information is extracted and is known method for distinguishing, comprises the steps:

A. preliminary process is carried out to original potential field raster data; Due to the data that original target seeker gyro is a superimposed field, containing having plenty of underground deep until the integrated information of superficial part.Therefore, as required the process such as some separation are carried out to target seeker gyro, reach the object of a certain feature in outstanding data.As airborne magnetic survey data, in order to make the horizontal level of magnetic anomaly and the underground magnetic objective body recorded can have certain spatial correspondence, usually all can process its doing pole before use magnetic data.Gravimetric data needs to do various correction, obtains bouguer gravity anomaly, according to the detected degree of depth, carries out continuation to bouguer gravity anomaly, obtains the gravitation information of different depth, to reach the object of detection different depth tectonic information.Proven technique in prior art owing to carrying out preliminary process to potential field raster data, so repeat no more.

B. the potential field raster data crossed of input processing, every column data is scanned, find out each local maximum in every column data and local minimum adjacent with it, the asymmetric waveform be made up of a local maximum and an adjacent with it local minimum is one-period, sets up the feature space of each waveform;

Find in the process of extreme point at the every column data of scanning, do every column data first and last raster data when extreme value judges can not often row interlude data be the same together, consider separately.Suppose the raster data size inputted is m × n, m, n distinguishes raster data line number and columns.A _i,jfor the i-th row in raster data, the data of jth row, i, j are respectively line number and the row number of a.If a _i,jbe one by the Local Extremum found out, but be much to seek the extreme point of the correspondence forming with it waveform.As i=m-1, a _i,jbe rejected as independent extreme point; As i<m-1, get a _{i+1, j}for with a _i,jcorresponding very big or minimum point, and form waveform.Now the pulse width of ripple is 1 raster width.

C. it is one group with every 3 column datas, the each waveform obtained in step b is as a unit, each waveform calculating the middle ordered series of numbers of often group closes on on the column data of front and back two cost function be connected between waveform, judges the connection status between two waveforms according to cost function;

For the Rule of judgment of two waveform connection status, except cost function value, give another Rule of judgment, the connection status be used between judgement two waveform: before the distance between calculating two wave vector space, first judge the polarity of two waveforms.If two polarity of waves are consistent, then calculate the distance of their vector spaces, then judge whether two waveforms should connect according to maximum value functional value; Otherwise, the distance in two wave vector spaces will not be calculated, leap to next waveform, then judge, and so forth.Such way, to not only avoid between opposed polarity waveform because maximum value function chooses excessive and incorrect link that is that cause; Meanwhile, judge that two polarity of waves are wanted easily more than cost function between calculating two waveform, thus also improve operation efficiency.

D. the public domain of the event segments that waveform is connected to form in detection steps c, will have the event segments of public domain to couple together, while event segments connects, constantly update event list, by the event that connected stored in incident file;

In event segments connection procedure, in the data structure representing waveform, add a mark, be used for the ending of flag event.Namely the ending of waveform can be established to be designated trail, and to be initially vacation.Often finish an event segments to connect, then whether last waveform detecting new event segments is connected with arbitrary waveform of its next column.If connected, then the mark trail of this waveform is still true, continues the connection of this event segments; Otherwise the trail of this waveform is false, by event segments stored in incident file, then is deleted from fragment list by this event segments, discharges its space.The object done like this can save memory headroom when processing big data quantity as far as possible, raises the efficiency.

E. the result of detection of event is directly shown in the graphic.

Preferred as one, after steps d, by grid 90-degree rotation, repeat step c, d and then carry out step e.To improve the accuracy of result.

In the present invention, the asymmetric waveform (for simplicity, claim local maximum to be crest in the present invention, local minimum is trough) that one-period is made up of a local maximum and an adjacent with it local minimum.Each local extremum in this column data is obtained by each column data in scanning raster data.The trough that each crest is adjacent with it with one forms a waveform, and the amplitude of this waveform is exactly the difference between crest and trough.In target seeker gyro, if the trough of one-period than the crest corresponded closer to the arctic, so this cycle is just decided to be positive pole, i.e. PO=1; Otherwise this cycle is identified as negative pole, i.e. PO=-1.If a crest does not find a trough corresponded to form a waveform, so this crest is just cast out; Equally, if a trough does not find a crest corresponded to form a waveform, so it also can be rejected.Because now can not determine their polarity.

Once the crest in certain cycle and trough are identified, 3 adeditive attributes so will be had again to be used for describing this cycle: crest amplitude PA; Trough amplitude NA; Distance between crest to trough, or be pulse width PW.This group attribute combines just composition and defines the feature space of each waveform with the dimensionless proper vector f that following formula 1 is expressed:

F=(λ _pApA, λ _nAnA, λ _pWpW, PO) ^tformula 1

In formula, PA is crest amplitude; NA is trough amplitude; PW is the distance between crest to trough, i.e. pulse width; λ _pA, λ _nAand λ _pWthe weights of PA, NA and PW respectively; PO is the polarity in cycle; T represents vector transpose.

Be be based upon to do on the basis scanned to every 3 row raster datas to the connection in cycle, the connection work in cycle launches around centre one column data of every 3 row raster datas.For each cycle on this column data middle, close in two column datas at it and respectively select 2 ~ 3 cycles to connect with it.Select the cycle that a kind of optimization method based on cost function minimum principle is selected for attended operation.Cost function itself is a kind of criterion of closing on cycle similarity, and the interpretation of similarity is then based on the dimensionless proper vector f described in formula 1.The form of concrete cost function is then decide according to the number of cycles in periodic unit group.If comprise cycle a and two cycles of cycle b in unit group, the value Cost (a, b) so connecting the event formed by this two cycles then by these two cycles feature space between distance D (a, b) weigh:

Cost (a, b) ≡ D (a, b)=| f _a-f _b| formula 2

F in formula 2 _aand f _bthe proper vector of indication cycle a and cycle b respectively, wherein | f| computation rule is defined as follows:

$\left|f\right|=\left|{(f_1,...,f_4)}^T\right|=\sqrt{\underset{i=1}{\overset{4}{\mathrm{\&Sigma;}}}{f}_i^2}$ Formula 3

Wherein f _ii-th element of representation vector f, if the attribute in two cycles is consistent, the value so connected between them is zero; If the attribute in two cycles is inconsistent, the value weighing apparatus so connected between them is greater than zero;

For comprise 3 cycles unit group c, d, e}, its value of combination function is represented by following formula 4:

$\mathrm{Cost}(c,d,e)={\mathrm{\&alpha;}}_0[D(c,d)+D(d,e)]+{\mathrm{\&alpha;}}_1\sqrt{A(c,d,e)}$ Formula 4

In formula 4 A (c, d, e) be used to indication cycle c, d and e tri-summit composition the area of triangle in 4 gt, parameter alpha ₀and α ₁be adjacent periods and the weight of non-conterminous cycle shared by cost function respectively, if the weight of adjacent and non-conterminous cycle shared by cost function is identical, so α be set ₀, α ₁be respectively 0.5; The formula that is expressed as follows 5 of area A:

A (c, d, e) ≡ D (c, d) D (d, e) sin θ formula 5

Wherein, formula 6

It is as follows that Inner in above formula amasss formal definition:

$(f,g)\&equiv;\underset{i=1}{\overset{4}{\mathrm{\&Sigma;}}}{f}_1\&CenterDot;g_i$ Formula 7

Therefore A (c, d, e) more goes to zero, then the cycle c, d and e linear concord better.

Once the connection in cycle completes, the event segments namely comprising 2 or 3 tectonic elements is determined.Just connect these event segments again, make it to form complete event.If cycle c, d, e and d, e, g are connected to two discrete event segments in back.So, because two fragments have common cycle de, so they also can be joined together, thus the event segments cdeg that formation one is longer.This operation by common cycle connection event fragment is all performed to all cycles, until form complete event list.Finally calculate the attribute of complete event, and event attribute is left in event list.Event attribute to comprise in event the mean value of rhythmic attribute: and event length L, event length L is obtained by following formula 8:

$L=\sqrt{{\mathrm{dx}}^2+{\mathrm{dy}}^2}$ Formula 8

Wherein, dx and dy is respectively the length that event extends eastwards and northwards.

For test effect, set up the model for rim detection Contrast on effect as shown in Figure 2.Observation area size is Middle_Lower Yangtze metallogenic belt scope, and survey line is north-south, line-spacing 5km, and point is apart from 200m.The field source of gravity-magnetic anomaly is made up of 7 buried depths, size, physical property (magnetic susceptibility and density) all not identical model, and 1,2, No. 3 model simulation zone of fracture, the geologic body of 4,5,6, No. 7 model simulation different depths, design parameter is in table 1.

Table 1

Getting background density is 2.7g/cm-3, and background magnetic susceptibility is 0 × 10-6cgsm, and magnetic dip is 90 °, and magnetic declination is 0 °, and Fig. 3 A and Fig. 3 B is respectively the gravity anomaly of above-mentioned theory model generation and the figure of magnetic anomalies, as shown in Figure 3 A and Figure 3 B.The gravity anomaly boundary line that high density difference wire model (2, No. 3) produces is clear, Spring layer and model top profile substantially identical, but the negative G that produces of low-density difference No. 1 model is abnormal the trend of dispersing.4,5,6, No. 7 model generations is superimposed anomaly, 4, the gravity anomaly border of No. 5 model generations is more clear, 6, the exception boundary of No. 7 two superficial part field sources generations is then not clear, and one and 6, anomaly peak that the gravity anomaly scale that produces of 7 models is suitable is there is 6, No. 7 model horizontal direction centers, infer that this is caused by 4,5,6, No. 7 models superpositions.1, the positive magnetic anomaly boundary line of 2,3 wire models generations is clear, and the magnetic anomaly peak value of 4,5,6, No. 7 model generations overlaps substantially with model boundary.

The gravity and magnetic anomaly data basis produced with theoretical model, carries out with HDTDR method, TDR method, THETA method, Framework Arithmetic the recognition result that gravity anomaly rim detection obtains as shown in Fig. 4 A to Fig. 4 D respectively.As can be seen from the figure HDTDR method is to having the Boundary Detection of positive density difference model and model boundary coincide better, but has false border to superimposed anomaly region and occur, as No. 4 with No. 5 model intermediate detection to the border do not conformed to realistic model.HDTDR method also can distinguish (No. 1 model) the model with negative density difference, but two false borders in " eight " font detected in these model both sides.TDR method detection boundaries and realistic model difference larger, to positive density difference model, its border detected is in the middle of model, poor to the discriminating power at abnormal edge, to the model of negative density difference, the border detected, on the endless belt of outer 2-3 times of model, does not all detect effective boundary information in model edge and inside.The Boundary Detection accuracy that the model of Theta method to positive density difference is corresponding is higher, only outwards departs from about 10% to No. 5 model borders of Additive Model.Theta method is also comparatively accurate to the model Boundary Detection of negative density difference, but the false border of the distribution in " eight " font can be detected in its both sides.Framework Arithmetic aligns, the model Boundary Detection of negative density difference is all comparatively accurate, and do not detect false border, the density of skeleton line also reflects the tendency of model, tendency as 2, No. 3 models is northwest (NW), inclination angle is 75 °, skeleton line is mild in model northwest (NW) side, then becomes intensive in east southeast side.For gravity anomaly, rim detection effect trap queuing is: skeleton method, HDTDR method, Theta method, TDR method, skeleton method result is accurate, false border is few, HDTDR method and Theta method to the border interpretation of positive anomaly better, but there will be false border to the detection of negative anomaly, and the editor that TDR method aligns negative anomaly all can not accurately detect, therefore, rim detection is carried out to gravity abnormal data and should preferentially adopt skeleton method.

Fig. 5 A to Fig. 5 D is respectively the gravity and magnetic anomaly data basis produced with theoretical model, carries out with HDTDR method, TDR method, THETA method, Framework Arithmetic the recognition result figure that magnetic anomalies rim detection obtains respectively; As shown in Fig. 5 A to Fig. 5 D, HDTDR method is coincide better to the Boundary Detection of single model and model boundary, but have false border to superimposed anomaly region to occur, as detected two borders do not conformed to realistic model between No. 4 and No. 5 models, between No. 5 with 6, No. 7 models more.The border that TDR method detects and realistic model difference are comparatively large, and single model border appears at model center, and the border of Additive Model is in the inboard of actual boundary.Theta method is higher to the Boundary Detection accuracy that single model is corresponding, but outside model, there will be false border, for Additive Model, detect with model border while also detect false border, as detected two borders do not conformed to realistic model between No. 4 and No. 5 models, between No. 5 with 6, No. 7 models more.Skeleton method model Boundary Detection is all comparatively accurate, and the position of skeleton line dense distribution and model border coincide better.When continuation is highly excessive, there will be boundary diffusion phenomenon, as the skeleton line occurred in the outside of 1,2, No. 3 model.The density of skeleton line also reflects the tendency of model, and the tendency as 2, No. 3 models is northwest (NW), and inclination angle is 75 °, and skeleton line is mild in model northwest (NW) side, then becomes intensive in east southeast side.For magnetic anomalies, rim detection effect trap queuing is: skeleton method, Theta method, HDTDR method, TDR method, skeleton method result is accurate, when continuation highly arranges reasonable, false border is few, HDTDR method and the border interpretation of Theta method to magnetic anomaly exception better, but false border be there will be to the detection of built-up pattern, TDR method can not accurately detect the border of magnetic anomaly, and therefore, carrying out rim detection to magnetic anomalies data should preferentially adopt skeleton method.

By the inventive method, the regional gravity in Middle_Lower Yangtze metallogenic belt study area, Aeromagnetic data are carried out to the tectonic information identification and extraction of target seeker gyro.Regional gravity data are 1 ﹕ 200,000 Bouguer gravity data; The engineer's scale of Aeromagnetic data is 1 ﹕ 200,000.

To original potential field raster data through on prolong, after the rough handling such as ask level to lead, the gravity anomaly data investigation of different continuation height identified together.Fig. 6 is the result figure to gravity anomalous identification.As shown in Figure 6, along with the change of continuation height, some grow small scale, cut not dark Surface geological structures border to fade away, and the signal of the long wavelength remained reflects substantially, and growth in region is huge, cutting is comparatively dark or the tectonic structure border (as Tan-Lu fault, plate boundary) of deep concealed.The border person's movements and expression of these linear spreads, demonstrate the change of the trend of each structure, continuity and each aligned structure different depth level.The block split between the individual border that deep retains just represents the discrepant geologic body distribution of underground deep physical property.

To original potential field raster data through change pole, on prolong, after a series of process rough handling such as ask level to lead, the abnormal data of the different continuation height of the Middle_Lower Yangtze metallogenic belt aeromagnetic anomaly of acquisition is superimposed and identifies.If Fig. 7 is magnetic anomaly Tectonic Framework identification stereographic map.As shown in Figure 7, the recognition result implication of differing heights continuation and the implication representated by aforesaid Bouguer anomaly skeleton recognition result similar, bouguer gravity anomaly skeleton result reflects the border of different depth density variation geologic body, and deep fracture such as Tan-Lu fault, plate separatrix, Volcanic Basin and rock mass osculatory are all well portrayed.

The principle that multiple scale detecting result is explained is as follows:

In same map sheet, different continuation height detection result is superimposed together, the line style of employing different colours, size represents the signal boundary person's movements and expression of different depth yardstick, the similarity that different depth sideband signal person's movements and expression embody on distributing position can know Growth Depth and the tendency feature of this structural feature: wire harness more intensive expression boundary formation depth of cut is larger, and wire harness is sparse, represents that the degree of depth that its cutting is grown is more shallow; This structure of the wider expression of wire harness combination tendency is more slow, otherwise then represents that this boundary formation growth occurrence is comparatively steep, inclination angle is less.Often organize wire harness combination and all can reflect the features such as the trend of boundary formation and swing, tendency, depth of cut.

According to above-mentioned canons of construction, in conjunction with other geophysicses, geologic information, the inventive method reflects following information to heavy in Middle_Lower Yangtze metallogenic belt region (being called for short study area once), magnetic data conformation identification result:

Plate, terrain border

It is highly the gravimetric data recognition result of 7 ~ 60km that Fig. 8 prolongs on being; Fig. 9 is for boat magnetic prolonging height 12 ~ 50km recognition result.The larger macroscopical geology tectonic framework of the study area degree of depth is reflected by Fig. 8 and Fig. 9.

Label in Fig. 8 is 1 Quaternary system; 2 water systems; 3 Triassic system limestones; 4 Silurian siltstones; 5 ordovician limestones, pierite; 6 the Cambrian period shale, limestone; First ancient epimetamorphic rock in 7; First ancient metamorphosed rock in 8; 9 early first ancient metamorphosed rocks; 10 Archean era gneiss; 11 late sieve dwarf, volcanics; In 12 Jurassic period diorite, grouan; 13 late sieve dwarf-morning chalk grouan, diorite.Shown in Fig. 8, there is a shape northwestward as the density boundary of dove shape, and intensive the surrounded regional body part of its wire harness is large other upwelling area, and northwards east orientation shrinks gradually, is the reflection on Dabie-Sulu UHPM belt border.The southeast of facing each other across the river with the northwestward, detects an annular boundary, and its border occurrence that is western and south is more slow, and northern occurrence is comparatively steep, and this border reflects the border of south of the River terrain substantially.Intensive wire harness along the distribution of Hefei-Feixi County-six Annan one line reflects the border of North China Plate and Yangtze craton, and its north is North China Plate, and its south is lower Yangtze craton.Gravity recognition result, topographic and geologic and HQ5 line are explained the integrated Figure 10 under three-dimensional environment such as section, by interactive analysis finds the skeleton wire harness that identifies and greatly not Long Qi, under raise sub-down warping region and the south of the River swell between boundary line one_to_one corresponding.The dense degree of skeleton wire harness additionally provides terrain Contact Boundary occurrence information: large not and under the contact boundary line of raising between son comparatively steep, the density wire harness of its correspondence is intensive, under raise son and the south of the River swell between contact boundary line more slow, the wire harness of its correspondence is sparse.Figure 11 navigates the identification of magnetic skeleton and HQ5 line MT sectional drawing, Figure 11 well supports above-mentioned viewpoint, south of the River protuberance general performance is high resistant, under to raise sub-down warping region be low-resistance, south of the River protuberance with under to raise sub-down warping region contact interface change in resistance angle more slow, and large not Long Qi with under to raise sub-down warping region contact interface change in resistance steeper.

Label in Fig. 9 is: 1 Quaternary system; 2 water systems; 3 Triassic system limestones; 4 Silurian siltstones; 5 ordovician limestones, pierite; 6 the Cambrian period shale, limestone; First ancient epimetamorphic rock in 7; First ancient metamorphosed rock in 8; 9 early first ancient metamorphosed rocks; 10 Archean era gneiss; 11 late sieve's dwarf volcanics; In 12 Jurassic period diorite, grouan; 13 late sieve dwarf-morning chalk grouan, diorite.Boat magnetic skeleton recognition result and the gravity recognition result of Fig. 9 display are existing similarly also has any different: in district, the northwestward corresponding dove shape density boundary also exists the magnetic boundary of shape as dove shape, but this border and density boundary slightly offset, its scope is slightly little compared with density boundary, in conjunction with the physical property characteristic that the magnetic of Dabie-Sulu UHPM belt is higher, can think that this magnetic boundary is the reflection on large other upwelling area and Dabie-Sulu UHPM belt border, meanwhile, also illustrate that within the scope of this density and magnetic boundary, geological mass has the character of heavy magnetic homology.Boat magnetic recognition result does not significantly show (in Figure 10) to the border that the south of the River swell, illustrates the south of the River swell and under to raise sub-magnetic contrast less.

Rift structure

The linear feature of gravity, boat magnetic recognition result provides the abundant information of rift structure, all in all, district's internal linear feature can be divided into three types: the large other upwelling area wire harness trend of northwest based on NW to, on the south Yangxin-Jiujiang, the west and south, wire harness NW is to for main, and all the other regions wire harness NE is to for main.These phenomenons reflect fracture that different earth history formed therebetween, construct reverse change, to the analysis of wire harness person's movements and expression, tentatively determine following rift structure distribution characteristics:

Tan-Lu fault:

Tancheng--Lujiang fault zone be a series of east northeasts on eastern Asiatic continent to the main fault band of in huge fault system, cut and wear the different geotectonic element in Eastern China, magnificent in scale, complex structure, forefathers' research and argue more (Wu Genyao, 2007; Hou Mingjin, 2007).Plate tectonic theory viewpoint roughly has three kinds of views to it: one, think that Zheng's hut zone of fracture is the ancient suture line in continent; Its two, think that it is a transform fault; Another view, thinks that it is a continental rift belt.It is extended to southwards to the larger dispute of also existence of where, forefathers (Wang Jingbin, 1991) think that it prolongs Guanji, Hubei, and through xiushui, Lianyunshan In Hunan Province, Heng Shan, Jiang Yong direct reach a deep fracture zone of great Rong mountain one, Guangxi Liuwan Dashan.

On gravity, boat magnetic skeleton diagram, along interrupted appearance one wire harness of yellow plum-Susong-Taihu Lake-Lujiang one line, in Fig. 10, with by methods such as MT, heavy magnetic, this wire harness explains that the Tancheng--Lujiang fault zone center in the HQ5 section of gained coincides, therefore, can think that this wire harness is the Tancheng--Lujiang fault zone reflection domestic in Anhui, Tancheng--Lujiang fault zone is also the border of lower Yangtze craton and Dab ie orogenic belt and North China Plate simultaneously.This wire harness south is intensive, sparse between Feidong-Chaohu, and illustrate that southern occurrence is comparatively steep, northern occurrence is relatively slow.On more macroscopical yardstick, Figure 12 is the skeleton recognition result to Eastern China boat magnetic, found by Figure 12, along Tancheng--Lujiang fault zone intermittently occurs east northeast to wire harness, substantially reflect trend and the distributional pattern of Tan-Lu fault, it intermittently occurs, the intensive or sparse alternate feature of wire harness shows the complicacy of Tancheng--Lujiang fault zone structure.

No matter gravity or boat magnetic, be showed no the density be connected with Tancheng--Lujiang fault zone or magnetic wire border in southern side, the Changjiang river, and it is less that the conclusion obtained by the result of the inventive method is that Tancheng--Lujiang fault zone passes the possibility in the Changjiang river.

The Changjiang river profound fault

Team of Ministry of Geology aerial survey group 902 analyzed according to the middle and lower reach of Yangtze River aeromagnetic anomaly of Wuhan to Zhenjiang section in 1958, thought along direction, riverbed, the Changjiang river, there is a regional fault (or zone of fracture).Geology bureau of Ministry of Geology Jiangsu Province, based on the data such as above-mentioned, also to think by Anqing through Wuhu, Nanjing to Zhenjiang, there is fracture along the Changjiang river, and is referred to as the Changjiang river shatter belt (or under raise sub-shatter belt).There is dispute about the definition of Yangtze Fault Zone, character always.Forefathers are referred to as the Changjiang river profound fault, the Changjiang river shatter belt, the Changjiang river fault depression and rift valley etc., Qin great Zheng (1983) and Liu Xiangpei etc. 1988) think Yangtze Fault Zone, by Jiujiang, Anqing after Wuhu, Nan Jingdong are extended down to Zhenjiang, do not stop, but continue to extend eastward.Its total length nearly thousand kilometers (comprising marine site part), wide about 10-15 kilometers, form a huge graben formula tensile continental rift belt, this continental rift belt until the modern times still in continuation activity.Li Qitong (1984) etc. then thinks that namely Yangtze Fault Zone stops to Zhenjiang, no longer extends toward east.Tang Yongcheng etc. (1998) etc. utilize the subtrace ruptured in Huaining-Quanjiang fracture, the fracture of east stream-Ma'an Mountain, Tongcheng-Lujiang to determine the Position Approximate ruptured in the Changjiang river, and thinking that first two break to form comparatively early, is product orogenic phase.

The gravity skeleton recognition result figure of Fig. 8 can find out, in Yangxin-Ruichang-Jiujiang-Wangjiang-Anqing-Guichi-inaction-Ma'an Mountain one be with occur wire harness, wire harness is intensive, form is continuous, and the Changjiang river profound fault that its main body section and Tang Yongcheng etc. (1998) etc. infer is substantially identical.Telluric electromagnetic sounding also reflects the existence of the Changjiang river profound fault, HQ5(Figure 10), HQ10(Figure 11) on line MT inverting resistivity resistivity section figure, all there is obvious low-resistivity body below heavy magnetic skeleton wire harness near the Changjiang river, reflect the existence of low-resistance zone of fracture.The Changjiang river profound fault form that gravity skeleton recognition result reflects is clear, the complete spread featuring the Changjiang river profound fault.From more macroscopical dimensional analysis, east boat magnetic knot fruit (Figure 12), complete reflects the Changjiang river profound fault, and it is better along Anqing-Tong Lingbei-Wuhu-Ma'an Mountain-Nanjing one line continuity, always toward east continuity after turning to Zhenjiang from Nanjing, prolong to Nantong Hou Dong and enter the East Sea.

Comprehensive above information, the result that the inventive method obtains supports the viewpoints such as (1988) such as Qin great Zheng (1983) and Liu Xiangpei, and namely Yangtze Fault Zone is by Jiujiang, Anqing after Wuhu, Nan Jingdong are extended down to Zhenjiang, does not stop, but continues to extend eastward.The Changjiang river deep fracture zone swells with middle and lower reach of Yangtze River Mohorovicic discontinuity and echoes mutually, infer that its formation is relevant with mantle bulge, be the tensional fault of a geochronology of rift-type, this band controls riverine Volcanic Basin, and it is to middle and lower reach of Yangtze River polymetallic area important as seen.

Other rift structure:

According to heavy magnetic skeleton recognition result, also can determine a series of rift structure, by trend mainly can be divided into NW to NE to two serial rift structures.Main Fracture Structures as shown in figure 12, the fracture of NE trend mainly contain F1, F2, F3, F4, F5, F6 and F7:F1, F2 respectively corresponding aforesaid Tancheng--Lujiang fault zone and the Changjiang river profound fault, F3 and forefathers (Tang Yongcheng, 1998) the smooth fracture of height inferred of research is basically identical, its to the east of-one section, Qingyang for south of the River protuberance and under raise the separatrix of sub-down warping region.To be that south of the River protuberance is inner grow more shallow fracture for F4, F5, and these two break on MT section and also have corresponding display (Figure 11), and F4 and F5 is corresponding two low-resistance belts respectively, topple to northwest (NW), and inclination angle is comparatively steep, and the degree of depth, at 7,8 km, should be in shell and ruptures.F6 and F7 is fractured into down two component in fault structure zone of raising in sub-down warping region.Figure 13 is boat magnetic skeleton line and HQ10 line MT sectional drawing.In figure, wire harness is skeleton recognition result, and geologic map is superimposed on DEM, MT section: dark 70km, resistivity isoline logarithmic interval, unit: Ω m.On Figure 13, correspond to low-resistance belt, detect wire harness signature analysis from gravity, F6 occurrence is comparatively steep, and F7 inclines to NW, and produce and turn more slow, these two fracture depths are more shallow, under raise local fracture system in sub-down warping region.NW has F7, F8, F9, F10, F11 and F12 to rift structure, F12 is a discordogenic fault across plate, linear person's movements and expression and Chen Husheng etc. (1999) propose large not-Zhoushan is ruptured basically identical, only there is certain displacement at individual lot, its-Shui Hou ridge, Yingshan Mountains fracture that one section of corresponding forefathers (Tonga is rich, 2003) is inferred in Tan-Lu fault west side.F9 is a part for Xiangfan-guangji Fault, F10 and F11 is the secondary fragmentation that Dab ie orogenic belt inside is grown.

Volcanic Basin border

Since In Middle And Lower Reaches of Changjiang River secondary era, this area's magmation is strong, and various volcanics-subvolcano rock area coverage is at 5000km ²above.Forefathers' research is thought, in the Metals Deposit Ore of this metallogenic belt and Volcanic Basin, Magmatic Activity of Yanshan Period has close genetic relationship (Chang Yinfo etc., 1991; Yu Yao, 2009).Mesozoic volcanic rock is distributed in multiple continental volcanic rock rifted-basin, mainly comprises Lujiang―Zongyang basin, flourishing basin, peaceful overgrown with weeds basin, Lishui Basin, Huaining-basin, Anqing etc.Basin boundary is geological work emphasis always, once once there is arguement to some of them boundary of basin, as Lujiang―Zongyang basin with or without second half etc.The heavy magnetic multi-scale morphology of the present invention is the border of main method recourse Volcanic Basin.

Figure 15, for boat magnetic prolonging height 1 ~ 7km skeleton recognition result figure, is the reflection on the relatively shallow magnetic border of the degree of depth.Label in figure is: 1 Quaternary system; 2 water systems; 3 Triassic system limestones; 4 Silurian siltstones; 5 ordovician limestones, pierite; 6 the Cambrian period shale, limestone; First ancient epimetamorphic rock in 7; First ancient metamorphosed rock in 8; 9 early first ancient metamorphosed rocks; 10 Archean era gneiss; 11 late sieve's dwarf volcanics; In 12 Jurassic period diorite, grouan; 13 late sieve dwarf-morning chalk grouan, diorite.Along the Changjiang river profound fault one line, from north to south, there is multiple ring-type wire harness in boat magnetic skeleton line, in conjunction with the feature that volcanics magnetic is higher, thinks that these wire harness represent the border of middle and lower reach of Yangtze River Volcanic Basin successively.The annular wire harness that Jiangning county-Ma'an Mountain-Wuhu one is with is peaceful overgrown with weeds basin boundary, annular wire harness residing for Fanchang County-Nanling County is flourishing boundary of basin, the annular wire harness that Lu Jiangnan-Congyang one is with is the border of Lujiang―Zongyang basin, and the wire harness in its Anqing-Huaining, southern side is boundary of basin, Huaining.Wherein peaceful sharpness of border that is overgrown with weeds, flourishing, Lujiang―Zongyang basin is complete, and basin, Huaining is by the impact of its scale, and the border of reflection is very not complete.Figure 14 Middle_Lower Yangtze metallogenic belt tectonic framework and mineral deposit corresponding relation figure, by gravity, boat magnetic skeleton interpretation result and mineral deposit comprehensively on a figure.In figure, F-ruptures; B-Volcanic Basin.According to weight, the multiple dimensioned Volcanic Basin border determined of magnetic as shown in figure 14, B1, B2, B3, B4 are respectively peaceful overgrown with weeds basin, flourishing basin, Lujiang―Zongyang basin and basin, Huaining.

Tectonic framework and ore deposit collection cell relation

See Figure 14, visible middle and lower reach of Yangtze River metalliferous deposit (point) distribution has significantly the feature of minute band in groups.The four Ge Kuangji districts such as the peaceful grassland that four Volcanic Basins are corresponding, hut fir, flourishing, Huaining have concentrated the mineral deposit (point) of Middle_Lower Yangtze metallogenic belt nearly 50%, and deposit type is based on Fe ore deposit, and Cu takes second place in ore deposit.All the other are distributed in grand recessed transitional zone Tongling, nine auspicious, Qingyang-Guichi and Southeastern Hubei Kuang Ji district, and deposit type is based on Cu, polymetallic ore.East northeast is distinguished to aligned structure (fracture) and doughnut structure (volcanic rocks basin) and ore deposit collection and is furnished with direct relation, especially the Changjiang river profound fault controls the distribution of Volcanic Basin (peaceful overgrown with weeds, Huaining is at Qi Bei, hut fir, flourishing in its southern side), can think that the background that this and the Changjiang river profound fault are formed has substantial connection: this tensional fault is that the geochronology of rift-type that asthenosphere upwelling causes ruptures, become and lead ore deposit passage, deep mineral resources material from lower crust or earth mantle is upwards migrated by profound fault, forms seen metalogeny at present in place.Large not-Zhoushan breaks to form the epoch and may be later than the Changjiang river profound fault, the latter is with bad break by it in Tongling one, form crossing of discordogenic fault, also have certain contribution to the formation in Tongling, hut fir, Ning Wukuangji district.At local area, the coal controlling of Tancheng--Lujiang fault zone is also not obvious, but near Lujiang magmation do not get rid of by Tan-Lu fault affect may, as the mineral deposit points such as husky small stream copper mine likely have certain associating with Tan-Lu fault.

Above embodiment is only exemplary embodiment of the present invention, and be not used in restriction the present invention, protection scope of the present invention is defined by the claims.Those skilled in the art can in essence of the present invention and protection domain, and make various amendment or equivalent replacement to the present invention, this amendment or equivalent replacement also should be considered as dropping in protection scope of the present invention.

Claims (6)

1. potential field data structural information is extracted and is known a method for distinguishing, it is characterized in that, comprises the steps:

A. preliminary process is carried out to original potential field raster data;

B. the potential field raster data crossed of input processing, every column data is scanned, find out each local maximum in every column data and local minimum adjacent with it, the asymmetric waveform be made up of a local maximum and an adjacent with it local minimum is one-period, sets up the feature space of each waveform;

C. it is one group with every 3 column datas, the each waveform obtained in step b is as a unit, each waveform calculating the middle ordered series of numbers of often group closes on on the column data of front and back two cost function be connected between waveform, judges the connection status between two waveforms according to cost function;

D. the public domain of the event segments that waveform is connected to form in detection steps c, will have the event segments of public domain to couple together, while event segments connects, constantly update event list, by the event that connected stored in incident file;

E. the result of detection of event is directly shown in the graphic.

2. potential field data structural information according to claim 1 is extracted and is known method for distinguishing, it is characterized in that, in described step b, each local extremum in this column data is obtained by each column data in scanning raster data, the local minimum that each local maximum is adjacent with it with one forms a waveform, local maximum do not find a local minimum corresponded then this local maximum just cast out; Equally, if a local minimum does not find a local maximum corresponded, so it also can be rejected; The dimensionless proper vector f expressed with following formula 1 defines the feature space of each waveform:

F=(λ _pApA, λ _nAnA, λ _pWpW, PO) ^tformula 1

In formula, PA is crest amplitude; NA is trough amplitude; PW is the distance between crest to trough, i.e. pulse width; λ _pA, λ _nAand λ _pWthe weights of PA, NA and PW respectively; PO is the polarity in cycle; T represents vector transpose.

3. potential field data structural information according to claim 1 is extracted and is known method for distinguishing, and it is characterized in that, in described step b, the line number of raster data is m and columns is n, and namely raster data size is m × n, the i-th row in raster data, the data a of jth row _i,jbe one and but be much to seek the corresponding extreme point with its composition waveform by the Local Extremum found out, as i=m-1, a _i,jbe rejected as independent extreme point; As i<m-1, get a _{i+1, j}for with a _i,jcorresponding very big or minimum point, and form waveform.

4. potential field data structural information according to claim 1 is extracted and is known method for distinguishing, it is characterized in that, in described step c, before distance between calculating two wave vector space, first the polarity of two waveforms is judged, if two polarity of waves are consistent, then calculate the distance in two wave vector spaces, then judge whether two waveforms should connect according to maximum value functional value; Otherwise, leap to next waveform, then carry out the polarity of judgement two waveform, and so forth.

5. potential field data structural information according to claim 1 is extracted and is known method for distinguishing, it is characterized in that, in step c, two cycle a and b are comprised in periodic unit group, value Cost (a of the event formed so is connected by this two cycles, b) then by these two cycles feature space between distance D (a, b) weigh:

Cost (a, b) ≡ D (a, b)=| f _a-f _b| formula 2

F in formula 2 _aand f _bthe proper vector of indication cycle a and cycle b respectively, wherein | f| computation rule is defined as follows:

$\left|f\right|=\left|{(f,...,f_4)}^T\right|=\sqrt{\underset{i=1}{\overset{4}{\&Sigma;}}{f_i}^2}$ Formula 3

Wherein f _ii-th element of representation vector f, if the attribute in two cycles is consistent, the value so connected between them is zero; If the attribute in two cycles is inconsistent, the value weighing apparatus so connected between them is greater than zero;

For comprise 3 cycles unit group c, d, e}, its value of combination function is represented by following formula 4:

$Cost(c,d,e)={\mathrm{\&alpha;}}_0\&lsqb;D(c,d)+D(d,e)\&rsqb;+{\mathrm{\&alpha;}}_1\sqrt{A(c,d,e)}$ Formula 4

In formula 4 A (c, d, e) be used to indication cycle c, d and e tri-summit composition the area of triangle in 4 gt, parameter alpha ₀and α ₁be adjacent periods and the weight of non-conterminous cycle shared by cost function respectively, if the weight of adjacent and non-conterminous cycle shared by cost function is identical, so α be set ₀, α ₁be respectively 0.5; The formula that is expressed as follows 5 of area A:

A (c, d, e) ≡ D (c, d) D (d, e) sin θ formula 5

Wherein, $\mathrm{\&theta;}={\cos }^{-1}\&lsqb;\frac{(f_c-f_d,f_d-f_a)}{D(c,d)D(d,e)}\&rsqb;$ Formula 6

It is as follows that Inner in above formula amasss formal definition:

$(f,g)\&equiv;\underset{i=1}{\overset{4}{\&Sigma;}}{f}_i\&CenterDot;g_i$ Formula 7

Therefore A (c, d, e) more goes to zero, then the cycle c, d and e linear concord better.

6. potential field data structural information according to claim 1 is extracted and is known method for distinguishing, it is characterized in that, in steps d, the event segments with common cycle is connected, until form complete event list, finally calculate the attribute of complete event, and event attribute is left in event list, event attribute to comprise in event the mean value of rhythmic attribute and event length L, event length L is obtained by following formula 8:

$L=\sqrt{{\mathrm{dx}}^2+{\mathrm{dy}}^2}$ Formula 8

Wherein, dx and dy is respectively the length that event extends eastwards and northwards.