CN106339985B - A method of selection is inlayed line and is inlayed to aviation image from vector house data - Google Patents
A method of selection is inlayed line and is inlayed to aviation image from vector house data Download PDFInfo
- Publication number
- CN106339985B CN106339985B CN201610752518.XA CN201610752518A CN106339985B CN 106339985 B CN106339985 B CN 106339985B CN 201610752518 A CN201610752518 A CN 201610752518A CN 106339985 B CN106339985 B CN 106339985B
- Authority
- CN
- China
- Prior art keywords
- line
- image
- inlays
- air strips
- inlayed
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4038—Scaling the whole image or part thereof for image mosaicing, i.e. plane images composed of plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of from vector house data chooses and inlays the method that line inlays aviation image, which comprises the following steps: 1) according in the air strips, in the boundary line computation air strips of every image effective coverage adjacent image overlap area;2) according to adjacent air strips overlay region between the boundary line computation air strips of the effective coverage of every obtained air strips;3) existing vector house data are based on, is found in adjacent air strips overlay region between adjacent image overlap area and air strips in air strips and inlays line;4) in all air strips and line will be inlayed between air strips mutually cut, individually effectively inlay polygon for the building of each image;5) data of every image effectively inlayed in polygon are merged, formation includes the final of all single images and inlays image.The present invention initially inlays polygon using vector data building to the method that aviation image is inlayed, then with Raster Images itself come local optimum, can be shortened and inlay the polygon building time, avoids passing through the prominent object such as house, significantly reduces later period human-edited amount.
Description
Technical field
The present invention relates to the processing technology field of aviation image, in particular to a kind of choose from vector house data is inlayed
The method that line inlays aviation image.
Background technique
" National Program for Medium-to Long-term Scientific and Technological Development (2006~the year two thousand twenty) " will clearly develop high-resolution pair
Ground observation system (referred to as " high score is special ") is classified as one of 16 key special subjects.Front and back, the system will be built up comprehensively to 2020
And come into operation, the earth observation ability in round-the-clock China, round-the-clock, Global coverage will be obviously improved.However, with remote sensing shadow
The resolution ratio of picture is higher and higher, remotely-sensed data in all trades and professions using more and more extensive, cover given area (such as Wuhan
City) remote sensing image data amount present explosive growth, promptly increase TB number till now from the GB order of magnitude of early 20th century
Magnitude (Chen Jiechen, 2012;2008;Xu Difeng, 2009;Primary outstanding person, 2013), it is intended that compared with than the past, obtain now
The high resolution image of identical ground region needs to inlay more remote sensing images.This inlays quick, the intelligence of remote sensing image
More stringent requirements are proposed.Its difficult point is how to inlay line and avoids passing through the digital elevation models such as house from correcting
Object (Wang et al., 2012).
Fall behind relatively however, the theory and method that remote sensing image is inlayed but develops, causes many surveying and mapping units and department must
More people must be employed, or reduces and inlays quality, are inlayed to meet increasingly increased remote sensing image.Such as, Wuhan City's production is primary
The main city zone image of 1:2000 scale bar needs to shoot and inlay about 10000 images, runs in single PC machine and know in the world
The tessellation software OrthoVista (Inpho, 2014) of name inlays these images, needs day and night ceaselessly operation 10 days or so,
Single technician's modification, which inlays line and is set into figure, also needs 6 months or so (Wang et al., 2012).Such efficiency makes
It obtains relevant departments to have to put into a large amount of high-performance computer and manpower, can be completed in one month to inlay work.With
This simultaneously, Surveying and Mapping Industry has accumulated the very strong vector data of a large amount of actuality within the past few decades, such as house and road, these
Vector data can be used as priori knowledge to serve the production of many mapping products, such as image mosaic.Wang et al.
(2013) propose that a kind of global optimum's aviation image based on vector road data inlays line selection selection method, to make to inlay line
It is walked as far as possible along road axis, to avoid house.However, also rarely having scholar's research how by vector house data application so far
It is inlayed in aviation image, and research is applied to the efficiency and effect of image mosaic.
Summary of the invention
It is inlayed in view of the above-mentioned problems, goal of the invention of the invention is to provide a kind of choose from vector house and road data
The method that polygon inlays aviation image is constructed using vector data and initially inlays polygon, then with Raster Images itself
Carrying out local optimum, these by what vector data generated inlay polygon, inlay the polygon building time to shorten, and ensure any
One section final, which to inlay line, can avoid passing through the prominent object such as house, to significantly reduce human-edited's amount in later period.
The specific technical solution of the present invention is that a kind of choose from vector house data is inlayed line and inlayed to aviation image
Method, which comprises the following steps:
1) it is based on left-turn algorithm, to the side of every Extraction of Image effective coverage in every air strips of the aviation image of shooting
Boundary line, in every air strips, according to adjacent image overlap in the boundary line computation air strips of every image effective coverage in the air strips
Area;
2) merge the effective coverage in every air strips, the boundary line of the effective coverage of every air strips is obtained, according to what is obtained
Adjacent air strips overlay region between the boundary line computation air strips of the effective coverage of every air strips;
3) existing vector house data, the adjacent air strips overlay region between adjacent image overlap area and air strips in air strips are based on
Line is inlayed in interior searching;
4) in all air strips and line will be inlayed between air strips mutually cut, left-turn algorithm is then based on, by inlaying after cutting
Line joins end to end, and individually effectively inlays polygon for every image building in every air strips, finally inlays image and be used only
The data of every image effectively inlayed in polygon, rest part are given up;
5) data of every image effectively inlayed in polygon are merged, is formed and includes all single images most
Image is inlayed eventually.
Further, it is one by one that the method for line is inlayed in searching in adjacent image overlap area in air strips in the step 3)
It extracts and inlays line between all adjacent images, found in adjacent air strips overlay region between air strips when inlaying line, adjacent air strips are worked as
Make adjacent image, inlay line between adjacent air strips using finding to inlay the identical method of line and extract in adjacent image overlap area,
If two adjacent images are m degree superimposed image, i.e., except this two images itself in addition to, also opened with m-2 other images have it is overlapping, m >=
2, extract two adjacent images between inlay line method the following steps are included:
(1) between vector house vacant lot in the two adjacent image overlap areas extracted based on constraint Delaunay triangulation network algorithm
Medium line and overlay region skeleton line, then by the skeleton line of the medium line in vacant lot and overlay region between vector house in overlay region
Merge, composition candidate inlays line library;
(2) effective width that line is inlayed according to candidate is inlayed line to candidate and is assigned through cost, is sought based on dijkstra's algorithm
The minimum candidate of the cost looked between origin-to-destination inlays line;
(3) the minimum candidate of cost step (2) extracted using m degree superimposed image data inlay line in overlay region into
Row optimization processing forms and final inlays line, it is ensured that any one section final, which to inlay line, can avoid passing through in image data
Protrusion object that is existing, and not including in vector house data.
Further, line tax is inlayed to candidate according to the effective width that candidate inlays line in the step (2) to pass through into
This, the specific method for inlaying line based on the minimum candidate of the cost between dijkstra's algorithm searching origin-to-destination is, if liFor
Candidate in i-th two adjacent image overlap areas inlays line, which inlays line liEffective width vadW (li) as the following formula
(I) it calculates:
Wherein, d (li) represent and inlay line l with candidateiMinimum range between adjacent room, projD (hi1,hi2) it is two phases
Formula (II) calculating is pressed by the sum of the height displacement in next door room:
projD(hi1,hi2)=tan θj1*hi1(li)+tanθi2*hi2(li) (II)
Wherein, θi1And θi2Line l is being inlayed with candidate respectively for two adjacent imagesiThe downwards angle of visibility in two adjacent houses;
hi1(li)、hi2(li) be the two houses height,
Candidate inlays line liPass through cost Tcost (li) press formula (III) calculating:
Tcost(li)=κ (li)×len(li) (III)
Wherein, len (li) indicate that candidate inlays line liLength;κ(li) indicate that candidate inlays line liWidth to cost
Contribution, κ (li) press formula (IV) calculating:
Wherein, c is set as constant 0.1, and minW is the minimum widith of the medium line in vacant lot between house in overlay region,
If SL indicates to inlay line set, CS by all candidates that the medium line and overlay region skeleton line in vacant lot between house form
It indicates to connect candidate in several SL and inlays line and the path by inlaying line beginning and end, inlaying line beginning and end is phase
Two intersection points of adjacent image effective coverage, press formula (V) calculate CS in either path cs by cost f (cs):
Based on dijkstra's algorithm, the candidate that cost is minimum in CS can be found and inlay line ms, be min [f by cost
(cs)],cs∈CS∈SL。
Further, the minimum candidate of cost inlays line and optimizes processing in overlay region in the step (3)
Method is, it is first determined the minimum candidate of cost inlays the optimization processing range of line, and in distance costs, minimum candidate inlays line
In range in optimization distance BufDis (ms), presses formula (VI) and it is optimized:
If the line of inlaying in optimization is rs, formula (VII) calculating is pressed by cost Tcost (rs):
In formula, D refers to the search range defined by upper formula (VI), and Tcost (u, v) refers to that the line rs that inlays in optimization passes through pixel
The cost of (u, v), Tcost (u, v) are the overlay region of m image in the maximum value of pixel (u, v) and the difference of minimum value, are pressed
Formula (VIII) calculates:
Tcost (u, v)=max [Lj(u,v)]-min[Lj(u, v)], j=1 ..., m (VIII)
In formula, Lj(u, v) refers to j-th of image in the brightness value of pixel (u, v), and brightness value is by red, green, blue three of image
Wave band presses formula (IX) and calculates gained:
Wherein,WithRefer to pixel (u, v) at red, green, blue three of j-th of image
The pixel value of wave band,
Based on grid dijkstra's algorithm, can find in rs by cost it is minimum inlay line, as finally inlay line.
The invention adopts the above technical scheme, which has the following advantages: 1) method of the invention can be by history
Mapping big data, which significantly improves, inlays efficiency.Experiment shows to generally acknowledge the best edge based on dijkstra's algorithm of effect with current
Embedding selection method is compared, and 80-90% operation time can be saved.This is because the node of vector data is usually far fewer than grid number
According to pixel, although the later period is also required to inlay line based on raster data optimization, optimization range is insufficient original 1/10, thus efficiency
It is significantly higher.2) it can be significantly improved by mapping big data and inlay quality, pass through less house.It is built based on three groups of differences
The experiment of inlaying of density shows: with based on vector road inlay line selection selection method compared with, the present invention can pass through 10-40% less
House;And with it is current generally acknowledge effect it is best based on dijkstra's algorithm inlay line selection selection method compared with, the present invention can less wear
The more house of 1-15%.The method inlayed to aviation image of the invention is particularly suitable for the completed region of the city domain of house data rich
Aviation image inlay.
Detailed description of the invention
Fig. 1 is multiple boats of a plurality of air strips based on vector house data in the method for the invention inlayed to aviation image
Empty image mosaic flow diagram;
Fig. 2 is to inlay the method flow of line image is opened in the extraction two of the invention between in the method that aviation image is inlayed to show
It is intended to;
Fig. 3 is that the candidate by vacant lot two houses of the invention between in the method that aviation image is inlayed inlays line and has
Imitate width diagram;
Fig. 4 is the optimization processing range schematic diagram that line is inlayed to minimum candidate in method of the invention, gray area
For the optimization processing range for inlaying line.
Specific embodiment
Technical solution of the present invention is further described with reference to the accompanying drawings of the specification.
If attached drawing 1 shows, a method of selection is inlayed line and is inlayed to aviation image from vector house data, and feature exists
In, comprising the following steps:
1) it is based on left-turn algorithm, to the side of every Extraction of Image effective coverage in every air strips of the aviation image of shooting
Boundary line, effective coverage refer to the region of non-null value in image.In every air strips, according to every image effective coverage in the air strips
Adjacent image overlap area in the line computation air strips of boundary;
2) merge the effective coverage in every air strips, the boundary line of the effective coverage of every air strips is obtained, according to what is obtained
Adjacent air strips overlay region between the boundary line computation air strips of the effective coverage of every air strips;
3) existing vector house data, the adjacent air strips overlay region between adjacent image overlap area and air strips in air strips are based on
Line is inlayed in interior searching;
As shown in Fig. 2, it is to extract all phases one by one that the method for line is inlayed in searching in adjacent image overlap area in air strips
Line is inlayed between adjacent image, is found when inlaying line in adjacent air strips overlay region between air strips, by adjacent air strips as adjacent image,
Line is inlayed between adjacent air strips using finding to inlay the identical method of line and extract in adjacent image overlap area, if two adjacent shadows
As being m degree superimposed image, i.e., in addition to this two images itself, also other images have overlapping, m >=2 with m-2, extract two phases
The method of line is inlayed between adjacent image the following steps are included:
(1) between vector house vacant lot in the two adjacent image overlap areas extracted based on constraint Delaunay triangulation network algorithm
Medium line and overlay region skeleton line, then by the skeleton line of the medium line in vacant lot and overlay region between vector house in overlay region
Merge, composition candidate inlays line library;
(2) effective width that line is inlayed according to candidate is inlayed line to candidate and is assigned through cost, is sought based on dijkstra's algorithm
The minimum candidate of the cost looked between origin-to-destination inlays line;
(3) the minimum candidate of cost step (2) extracted using m degree superimposed image data inlay line in overlay region into
Row optimization processing forms and final inlays line, it is ensured that any one section final, which to inlay line, can avoid passing through in image data
Protrusion object that is existing, and not including in vector house data, prominent object include house, temporary construction stacking object etc..
4) in all air strips and line will be inlayed between air strips mutually cut, left-turn algorithm is then based on, by inlaying after cutting
Line joins end to end, and individually effectively inlays polygon for every image building in every air strips, finally inlays image and be used only
The data of every image effectively inlayed in polygon, rest part are given up;
5) data of every image effectively inlayed in polygon are merged, forms the final edge for including all single images
Embedding image.
The effective width for inlaying line in the step (2) according to candidate is inlayed line to candidate and is assigned through cost, is based on
Dijkstra's algorithm finds the minimum candidate of cost between origin-to-destination and inlays the specific method of line, if liIt is i-th two
Candidate in Zhang Xianglin image overlap area inlays line, which inlays line liEffective width vadW (li) press formula (I) calculating:
Wherein, d (li) represent and inlay line l with candidateiMinimum range between adjacent room, as shown in Fig. 3, projD
(hi1,hi2) be two adjacent rooms the sum of height displacement, press formula (II) calculating:
projD(hi1,hi2)=tan θi1*hi1(li)+tanθi2*hi2(li) (II)
Wherein, θi1And θi2Line l is being inlayed with candidate respectively for two adjacent imagesiThe downwards angle of visibility in two adjacent houses;
hi1(li)、hi2(li) be the two houses height,
Candidate inlays line liPass through cost Tcost (li) press formula (III) calculating:
Tcost(li)=κ (li)×len(li) (III)
Wherein, len (li) indicate that candidate inlays line liLength;κ(li) indicate that candidate inlays line liWidth to cost
Contribution, κ (li) press formula (IV) calculating:
Wherein, it is much higher than in overlay region the medium line in vacant lot between house to guarantee that the skeleton line of overlay region passes through cost
C is set as normal to make the medium line of least cost inlayed line and preferentially track vacant lot between house by cost by minimum widith
Number 0.1, minW is the minimum widith of the medium line in vacant lot between house in overlay region, if SL indicates the medium line by vacant lot between house
Line set is inlayed with all candidates of overlay region skeleton line composition, CS indicates to connect candidate in several SL and inlays line and pass through edge
Two intersection points that line beginning and end is adjacent image effective coverage are inlayed, in the prior art in the path for embedding the wire beginning and end
If number of hits is more than two in, two farthest intersection points of selected distance.Press passing through for the either path cs in formula (V) calculating CS
Cost f (cs):
Based on dijkstra's algorithm, the candidate that cost is minimum in CS can be found and inlay line ms, be min [f by cost
(cs)],cs∈CS∈SL。
The minimum candidate of cost, which inlays the method that line optimizes processing in overlay region, in the step (3) is, first
Determine that the minimum candidate of cost inlays the optimization processing range of line, as shown in Fig. 4, minimum candidate inlays line in distance costs
In range in optimization distance BufDis (ms), presses formula (VI) and it is optimized:
I.e. when the minimum candidate of cost inlays the effective width of line between 10M (rice) -60M (rice), it is believed that two rooms
It is likely to road between room, the range for optimizing processing using image data is defined as 5M (rice), otherwise, when candidate inlays
The effective width of line segment be less than 10M (rice) or be greater than 60M (rice) when, it is believed that there will more than likely be between two houses other buildings or
Prominent object, is defined as 50M (rice) for the range for optimizing processing using image data.
If the line of inlaying in optimization is rs, formula (VII) calculating is pressed by cost Tcost (rs):
In formula, D refers to the search range defined by upper formula (VI), and Tcost (u, v) refers to that the line rs that inlays in optimization passes through pixel
The cost of (u, v), Tcost (u, v) are the overlay region of m image in the maximum value of pixel (u, v) and the difference of minimum value.Optimization
In inlay the initial value that line is rs and can be set as the candidate that cost is minimum in CS and inlay line ms by the sum of the costs of all pixels,
Other any and paths by inlay line beginning and end interior positioned at search range D can also be set, the beginning and end of line is inlayed
Two intersection points of respectively adjacent image effective coverage.Formula (VIII) is pressed by cost to calculate:
Tcost (u, v)=max [Lj(u,v)]-min[Lj(u, v)], j=1 ..., m (VIII)
In formula, Lj(u, v) refers to j-th of image in the brightness value of pixel (u, v), and brightness value is by red, green, blue three of image
Wave band presses formula (IX) and calculates gained:
Wherein,WithRefer to pixel (u, v) at red, green, blue three of j-th of image
The pixel value of wave band is 0.3,0.59 and 0.11 by prior art setting three wave band weights of red, green, blue.
Based on grid dijkstra's algorithm, can find in rs by cost it is minimum inlay line, as finally inlay line.
Below with three groups of aviation orthography data to the present invention --- it is chosen from vector house data and inlays line pair
The method that aviation image is inlayed --- efficiency and effect be tested, and and based on vector road and based on Dijkstra calculate
The method of method is compared.The relevant parameter of three group image data is as shown in table 1.The present invention with based on vector road and being based on
The method accuracy comparison of dijkstra's algorithm is as shown in Table 2-4.
The relevant parameter of 1 three groups of aviation images of table
As shown in Table 1, three groups of aviation images are respectively from three kinds of different regions of site coverage, including city, suburb
And rural area, the data precision in vector house are from 0.5-5M.
Table 2 is compared based on three kinds of methods of 36 city images
*Indicate: dijkstra's algorithm does not depend on vector data.Candidate inlays between the intersection point that line is two image overlay regions
Straight line, therefore calculate the time and leveled off to 0s.
Three kind methods comparison of the table 3 based on 6 suburb images
Three kind methods comparison of the table 4 based on 110 rural images
By table 2-4 it is found that provided by the invention choose from vector house and road data inlays polygon to aviation shadow
As the method inlayed has been compared obviously compared with the method based on vector road with the line selection selection method of inlaying based on dijkstra's algorithm
Advantage: 1, with it is current generally acknowledge effect it is best based on dijkstra's algorithm inlay line selection selection method compared with, 80-90% can be saved
Operation time.2, the present invention can be significantly improved by mapping big data inlays quality, passes through less house.Be based on vector
The line selection selection method of inlaying of road is compared, and the present invention can pass through the house of 10-40% less;And the best base of effect is generally acknowledged with current
It is compared in the line selection selection method of inlaying of dijkstra's algorithm, the present invention can pass through the house of 1-15% less.It can be seen that this method
Greater advantages are inlaid with to the urban district aviation image of house data rich, is avoided that and passes through almost all of house.3, based on arrow
The maximum memory that the method for measuring house needs is close with the method based on vector road, but is significantly less than based on dijkstra's algorithm
Method (the only 26-30% or so of the latter).
The various embodiments described above are merely to illustrate the present invention, and the connection of each component and structure may be changed,
On the basis of technical solution of the present invention, improvement that all principles according to the present invention carry out the connection of individual part and structure and equivalent
Transformation, should not exclude except protection scope of the present invention.
Claims (2)
1. a kind of choose from vector house data inlays the method that line inlays aviation image, which is characterized in that including following
Step:
1) it is based on left-turn algorithm, to the boundary of every Extraction of Image effective coverage in every air strips of the aviation image of shooting
Line, in every air strips, according in the air strips, in the boundary line computation air strips of every image effective coverage adjacent image overlap area;
2) merge the effective coverage in every air strips, the boundary line of the effective coverage of every air strips is obtained, according to every obtained
Adjacent air strips overlay region between the boundary line computation air strips of the effective coverage of air strips;
3) existing vector house data are based on, are sought in adjacent air strips overlay region between adjacent image overlap area and air strips in air strips
It looks for and inlays line;
4) in all air strips and line will be inlayed between air strips mutually cut, be then based on left-turn algorithm, line head will be inlayed after cutting
Tail is connected, and individually effectively inlays polygon for every image building in every air strips, finally inlays image using only every
The data of image effectively inlayed in polygon, rest part are given up;
5) data of every image effectively inlayed in polygon are merged, formation includes the final of all single images and inlays shadow
Picture;
It is to extract all adjacent shadows one by one that the method for line is inlayed in searching in adjacent image overlap area in air strips in the step 3)
Line is inlayed as between, when line is inlayed in searching in adjacent air strips overlay region between air strips, by adjacent air strips as adjacent image, is used
Line is inlayed between adjacent air strips with finding to inlay the identical method of line and extract in adjacent image overlap area, if two adjacent images are
M degree superimposed image, i.e., in addition to this two images itself, also other images have overlapping, m >=2 with m-2, extract two adjacent shadows
As between inlay line method the following steps are included:
(1) in two adjacent image overlap areas based on constraint Delaunay triangulation network algorithm extraction between vector house in vacant lot
Between line and overlay region skeleton line, then the medium line in vacant lot between vector house in overlay region and the skeleton line of overlay region are closed
And it forms candidate and inlays line library;
(2) effective width that line is inlayed according to candidate is inlayed line to candidate and is assigned through cost, is found based on dijkstra's algorithm
Point inlays line to the minimum candidate of the cost between terminal;
(3) the minimum candidate of cost step (2) extracted using m degree superimposed image data inlay line carried out in overlay region it is excellent
Change processing forms and final inlays line, it is ensured that any one section final, which to inlay line, can avoid passing through in image data existing
, and the protrusion object for not including in vector house data;
The effective width for inlaying line in the step (2) according to candidate is inlayed line to candidate and is assigned through cost, and Dijkstra is based on
Algorithm finds the minimum candidate of cost between origin-to-destination and inlays the specific method of line, if liFor i-th two adjacent shadow
As the candidate in overlay region inlays line, which inlays line liEffective width vadW (li) press formula (I) calculating:
Wherein, d (li) represent and inlay line l with candidateiMinimum range between adjacent room, projD (hi1, hi2) it is two phase next doors
Formula (II) calculating is pressed by the sum of the height displacement in room:
projD(hi1, hi2)=tan θi1*hi1(li)+tanθi2*hi2(li) (II)
Wherein, θi1And θi2Line l is being inlayed with candidate respectively for two adjacent imagesiThe downwards angle of visibility in two adjacent houses;hi1
(li)、hi2(li) be the two houses height,
Candidate inlays line liPass through cost Tcost (li) press formula (III) calculating:
Tcost(li)=κ (li)×len(li) (III)
Wherein, len (li) indicate that candidate inlays line liLength;κ(li) indicate that candidate inlays line liContribution of the width to cost,
κ(li) press formula (IV) calculating:
Wherein, c is set as constant 0.1, and minW is the minimum widith of the medium line in vacant lot between house in overlay region, if SL is indicated by room
All candidates of the medium line in vacant lot and overlay region skeleton line composition inlay line set between room, and CS indicates to connect candidate in several SL
Line and the path by inlaying line beginning and end are inlayed, two that line beginning and end is adjacent image effective coverage are inlayed
Intersection point, press formula (V) calculate CS in either path cs by cost f (cs):
F (cs)=Σ Tcost (li), li∈CS∈SL (V)
Based on dijkstra's algorithm, the candidate that cost is minimum in CS can be found and inlay line ms, be min [f by cost
(cs)], cs ∈ CS ∈ SL.
2. a kind of choose from vector house data as described in claim 1 inlays the method that line inlays aviation image,
Be characterized in that: the minimum candidate of cost, which inlays the method that line optimizes processing in overlay region, in the step (3) is, first
It determines that the minimum candidate of cost inlays the optimization processing range of line, inlays line optimization distance Buf in the minimum candidate of distance costs
In range in Dis (ms), presses formula (VI) and it is optimized:
If the line of inlaying in optimization is rs, formula (VII) calculating is pressed by cost Tcost (rs):
Tcost (rs)=Σ Tcost (u, v), (u, v) ∈ rs ∈ D (VII)
In formula, D refers to the search range defined by upper formula (VI), Tcost (u, v) refer in optimization inlay line rs by pixel (u,
V) cost, Tcost (u, v) are the overlay region of m image in the maximum value of pixel (u, v) and the difference of minimum value, as the following formula
(VIII) it calculates:
Tcost (u, v)=max [Lj(u, v)]-min [Lj(u, v)], j=1 ..., m (VIII)
In formula, Lj(u, v) refer to j-th of image in the brightness value of pixel (u, v), brightness value by image three wave bands of red, green, blue
It presses formula (IX) and calculates gained:
Wherein,WithRefer to pixel (u, v) in three wave bands of red, green, blue of j-th of image
Pixel value,
Based on grid dijkstra's algorithm, can find in rs by cost it is minimum inlay line, as finally inlay line.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610752518.XA CN106339985B (en) | 2016-08-29 | 2016-08-29 | A method of selection is inlayed line and is inlayed to aviation image from vector house data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610752518.XA CN106339985B (en) | 2016-08-29 | 2016-08-29 | A method of selection is inlayed line and is inlayed to aviation image from vector house data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106339985A CN106339985A (en) | 2017-01-18 |
CN106339985B true CN106339985B (en) | 2019-03-15 |
Family
ID=57822555
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610752518.XA Active CN106339985B (en) | 2016-08-29 | 2016-08-29 | A method of selection is inlayed line and is inlayed to aviation image from vector house data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106339985B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106846251B (en) * | 2017-01-24 | 2018-02-06 | 重庆市地理信息中心 | Orthography based on building roof vector inlays gauze network automatic selecting method |
CN111080520A (en) * | 2019-12-10 | 2020-04-28 | 西安中科星图空间数据技术有限公司 | Satellite image fast embedding method and device based on main frame search |
CN111311622B (en) * | 2020-01-30 | 2023-06-23 | 桂林理工大学 | Optimization selection method for embedding wire nets of multiple images based on ant colony algorithm |
CN112669459B (en) * | 2020-12-25 | 2023-05-05 | 北京市遥感信息研究所 | Satellite image optimal mosaic line generation method based on feature library intelligent decision |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8588547B2 (en) * | 2008-08-05 | 2013-11-19 | Pictometry International Corp. | Cut-line steering methods for forming a mosaic image of a geographical area |
CN103247055B (en) * | 2013-05-27 | 2015-08-19 | 武汉大学 | Based on the seam line optimization method of large span extracted region |
CN103871072B (en) * | 2014-04-04 | 2017-06-27 | 武汉大学 | Orthography based on project digital elevation model inlays line extraction method |
CN105761291B (en) * | 2016-02-14 | 2019-09-10 | 华浩博达(北京)科技股份有限公司 | Dijkstra's algorithm based on Weight inlays line automatic generation method |
-
2016
- 2016-08-29 CN CN201610752518.XA patent/CN106339985B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN106339985A (en) | 2017-01-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106339985B (en) | A method of selection is inlayed line and is inlayed to aviation image from vector house data | |
CN106127204B (en) | A kind of multi-direction meter reading Region detection algorithms of full convolutional neural networks | |
Malinverni et al. | Deep learning for semantic segmentation of 3D point cloud | |
CN105069843A (en) | Rapid extraction method for dense point cloud oriented toward city three-dimensional modeling | |
CN107092877A (en) | Remote sensing image roof contour extracting method based on basement bottom of the building vector | |
CN104952107A (en) | Three-dimensional bridge reconstruction method based on vehicle-mounted LiDAR point cloud data | |
CN102938066A (en) | Method for reconstructing outer outline polygon of building based on multivariate data | |
Hofmann et al. | Knowledge-based building detection based on laser scanner data and topographic map information | |
JP2011501301A (en) | Geospatial modeling system and related methods using multiple sources of geographic information | |
CN102842044B (en) | Method for detecting variation of remote-sensing image of high-resolution visible light | |
CN109658431A (en) | Rock mass point cloud plane extracting method based on region growing | |
CN109583302A (en) | A kind of mobile robot area of feasible solutions training dataset extending method | |
CN113361499B (en) | Local object extraction method and device based on two-dimensional texture and three-dimensional attitude fusion | |
CN103871072A (en) | Method for automatic extraction of orthoimage embedding line based on projection digital elevation models | |
Gennaretti et al. | A methodology proposal for land cover change analysis using historical aerial photos | |
CN112700358A (en) | Intelligent monitoring system for ocean space resources monitored by ocean ecological piles | |
CN109871812A (en) | A kind of multi-temporal remote sensing image urban vegetation extracting method neural network based | |
CN106910218B (en) | Earth's surface surface area calculation method based on big region grid digital elevation model | |
CN109031235A (en) | The method of quick obtaining radar baseis reflectivity three-dimensional isopleth data | |
Wang et al. | Seamline determination for high resolution orthoimage mosaicking using watershed segmentation | |
Price | Urban street grid description and verification | |
Sammartano et al. | High scale 3D modelling and orthophoto of curved masonries for a multipurpose representation, analysis and assessment | |
Gong et al. | Roof-cut guided localization for building change detection from imagery and footprint map | |
Condorelli et al. | Photogrammetry and medieval architecture: Using black and white analogic photographs for reconstructing the foundations of the lost rood screen at Santa Croce, Florence | |
Ali et al. | Automatic extractions of road intersections from satellite imagery in urban areas |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |