CN107578446A - A kind of method for extracting remote sensing image road and device - Google Patents
A kind of method for extracting remote sensing image road and device Download PDFInfo
- Publication number
- CN107578446A CN107578446A CN201710852459.8A CN201710852459A CN107578446A CN 107578446 A CN107578446 A CN 107578446A CN 201710852459 A CN201710852459 A CN 201710852459A CN 107578446 A CN107578446 A CN 107578446A
- Authority
- CN
- China
- Prior art keywords
- mrow
- road
- msub
- remote sensing
- sensing image
- 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.)
- Pending
Links
Landscapes
- Traffic Control Systems (AREA)
Abstract
The present invention relates to a kind of method for extracting remote sensing image road and device, belongs to remote sensing information extractive technique field.The present invention calculates original remote sensing image road probability density first, and the roadway characteristic on original remote sensing image is converted to the feature of road probability density；Then according to feature construction road-center line model of the probable value on road axis higher than the probable value in other positions, and cost function is determined according to road-center line model；The maximum of cost function is finally solved using Dynamic Programming, road axis is used as using the maximum.Complicated and diversified roadway characteristic on original remote sensing image is converted to simple consistent roadway characteristic in road probability distribution graph by the present invention, enable to extract different types of high-resolution remote sensing image road, cost function need not be changed, improves the universality of method.
Description
Technical field
The present invention relates to a kind of method for extracting remote sensing image road and device, belongs to remote sensing information extractive technique field.
Background technology
Road is automatically extracted from high-resolution remote sensing image to map rejuvenation, GIS data acquisition, Image Matching and target
Detect significant, be the research emphasis of current remote sensing survey field.High-resolution remote sensing image road extraction can be divided into
Full-automatic extraction and semi-automatic extraction, method for full automatic extraction it is common mainly have based on parallel lines to, based on mathematical morphology
With knowledge, based on window model feature etc., but from the point of view of current progress, existing full-automatic algorithm robustness is poor, extraction
As a result a large amount of artificial treatments are needed, effect is undesirable, and the semi-automatic extraction using man-machine interaction mode is more actual at present
Selection.
In remote sensing image road semiautomatic extraction method, active contour model and template matching method are considered as more
Practical two methods.Road extraction algorithm based on Dynamic Programming is a kind of conventional active contour model method, its basis
Roadway characteristic structure cost function on remote sensing image, is then extracted using the maximum of Dynamic Programming solution cost function
Road.Armin Gruen are mainly this spy of the smooth curve with high gray value according to the road on low resolution remote sensing image
Sign, it is proposed that a kind of classical road extraction algorithm based on Dynamic Programming, but this method is only applicable to low resolution image,
And on middle high-resolution image, road is no longer simple line feature, becomes the strip region with one fixed width,
Factor Poz et al. have modified cost function on the basis of Gruen, add road width information, can be used in the algorithm
The road extraction of middle high-resolution image.But because high resolution image roadway characteristic is complicated and changeable, traditional algorithm is all straight
Connect and cost function defined according to the road gray feature on raw video, therefore be difficult cost function of the definition with universality,
Result in traditional algorithm can only extract the simple path of fixed gray feature, for other kinds of road, can only redefine
Corresponding cost function, there is very big limitation in actual applications.
The content of the invention
It is an object of the invention to provide a kind of method for extracting remote sensing image road, to solve current method for extracting roads only
The problem of simple path being consistent with road model can be extracted；Present invention also offers a kind of remote sensing image road extraction element.
The present invention provides a kind of method for extracting remote sensing image road, including six technical sides to solve above-mentioned technical problem
Case, method scheme one, this method comprises the following steps：
1) original remote sensing image road probability density is calculated, it is general that the roadway characteristic on original remote sensing image is converted into road
The feature of rate density；
2) according to feature construction road axis of the probable value on road axis higher than the probable value in other positions
Model, and cost function is determined according to road-center line model；
3) maximum of cost function is solved using Dynamic Programming, road axis is used as using the maximum.
The present invention is converted to complicated and diversified roadway characteristic on original remote sensing image simple one in road probability distribution graph
The roadway characteristic of cause, enable to extract different types of high-resolution remote sensing image road, it is not necessary to change cost function, carry
The high universality of method.
Method scheme two：On the basis of method scheme one, the remote sensing image road probability density of the step 1) is to adopt
Determined with SVMs and Density Estimator, detailed process is as follows：
A. remote sensing image is classified using SVMs, obtains series of road sample point；
B. the probability density of each road sample point is calculated using Density Estimator.
The present invention is classified using SVMs, and required training sample from roadway characteristic database by manually choosing
Choosing, it is bright by a small amount of artificial road information acquisition tasks for participating in completing that there is one fixed width on high score resolution remote sense image
It is aobvious to shorten drafting period, substantially increase remote sensing image automatic business processing degree.
Method scheme three：On the basis of method scheme two, the probability density obtained in the step B is：
Wherein xiFor i-th obtained of road sample point,It is the probability density valuation at point x, h is Density Estimator
Bandwidth, K (x) is kernel function.
Method scheme four：On the basis of method scheme three, described kernel function uses gaussian kernel function.
Method scheme five：On the basis of method scheme one or two, the road-center line model of structure in the step 2)
For：
Ep=∫ { G [f (s)] }2Ds=max
Eg=∫ [f " (s)]2Ds=min
Cg=| f " (s) |≤T1
Wherein G (x) represents road probability-distribution function, and f (s) represents road axis, T1For given threshold.
Method scheme six：On the basis of method scheme five, the cost function according to road axis model construction is：
Wherein p={ p1,...,pn},pi=(xi,yi) be broken line on road axis n summit, S is line segment
pipi+1Pixel set,For line segment p in probability distribution graphipi+1Gamma function, aiFor line segment pipi+1Direction, | Δ si
| it is line segment pipi+1Length.
Present invention also offers a kind of remote sensing image road extraction element, including following six scheme, device scheme one：Should
Road extraction device includes memory and processor and is stored in the meter run on the memory and on the processor
Calculation machine program, the processor are coupled with the memory, and realization is following during computer program described in the computing device
Instruction：
1) original remote sensing image road probability density is calculated, it is general that the roadway characteristic on original remote sensing image is converted into road
The feature of rate density；
2) according to feature construction road axis of the probable value on road axis higher than the probable value in other positions
Model, and cost function is determined according to road-center line model；
3) maximum of cost function is solved using Dynamic Programming, road axis is used as using the maximum.
Device scheme two：On the basis of device scheme one, the remote sensing image road probability density of the step 1) is to adopt
Determined with SVMs and Density Estimator, detailed process is as follows：
A. remote sensing image is classified using SVMs, obtains series of road sample point；
B. the probability density of each road sample point is calculated using Density Estimator.
Device scheme three：On the basis of device scheme two, the probability density obtained in the step B is：
Wherein xiFor i-th obtained of road sample point,It is the probability density valuation at point x, h is Density Estimator
Bandwidth, K (x) is kernel function.
Device scheme four：On the basis of device scheme three, described kernel function uses gaussian kernel function.
Device scheme five：On the basis of device scheme one or two, the road-center line model of structure in the step 2)
For：
Ep=∫ { G [f (s)] }2Ds=max
Eg=∫ [f " (s)]2Ds=min
Cg=| f " (s) |≤T1
Wherein G (x) represents road probability-distribution function, and f (s) represents road axis, T1For given threshold.
Device scheme six：On the basis of device scheme five, the cost function according to road axis model construction is：
Wherein p={ p1,...,pn},pi=(xi,yi) be broken line on road axis n summit, S is line segment
pipi+1Pixel set,For line segment p in probability distribution graphipi+1Gamma function, aiFor line segment pipi+1Direction, | Δ si
| it is line segment pipi+1Length.
Brief description of the drawings
Fig. 1 is the flow chart of method for extracting remote sensing image road of the present invention；
Fig. 2 is the event handling flow chart of remote sensing image road extraction element of the present invention.
Embodiment
The embodiment of the present invention is described further below in conjunction with the accompanying drawings.
The present invention is divided remote sensing image using support vector machine method according to high-resolution remote sensing image road POP feature
For road class and non-rice habitats class, series of road sample point is obtained, road probability distribution graph is calculated using Density Estimator；According to
The probable value on center line in road probability distribution graph will be apparently higher than the feature construction road of the probable value in other positions
Center line model and cost function, utilize the road axis on Dynamic Programming extraction remote sensing image.The flow of this method such as Fig. 1
Shown, specific implementation process is as follows：
1. obtaining remote sensing image, and remote sensing image will be divided into road class and non-rice habitats class, and obtain series of road sample
Point.
Because high-resolution remote sensing image has the problems such as foreign matter is with spectrum different with jljl is composed, from soft margin support vector machine
Classified.SVMs is a kind of two classification mould proposed by Corinna Cortes and Vapnik et al. in nineteen ninety-five
Type, it shows many distinctive advantages in text classification, Handwritten Digit Recognition, target identification and Face datection, recognized
To be one of current best learning algorithm.High-resolution remote sensing image is divided into road by the present invention using SVMs
Class and non-rice habitats class, to obtain series of road sample point, wherein when being classified using SVMs, required training
Sample from roadway characteristic database by manually selecting.
2. calculate the road probability distribution graph of acquired road sample point.
Each pixel is the probability of road on road probability distribution graph expression remote sensing image, and probable value is according to pixel
What gray scale and other combined factors calculated, the observation data of unknown distribution are obeyed in processing, it usually needs are estimated from given data
Its probability density function is counted, this is referred to as Multilayer networks.Currently used Multilayer networks method has parameter Estimation, Nogata
Figure estimation and Density Estimator.Although the Multilayer networks based on histogram can describe in data the regularity of distribution,
Still suffer from three major defects：1. figure is rough；2. the shape of histogram is easily by the position of starting point and the shadow of interval width
Ring；3. when data are three-dimensional or more higher-dimension, there is very big limitation in histogram estimation.Density Estimator is the one of histogram estimation
Kind is promoted, but unlike histogram estimation, Density Estimator is given accordingly according to observation data close to estimation point x degree
Weight, overcome histogram estimation it is rough, rely on starting point the shortcomings of.The present invention calculates each road using Density Estimator
The probability distribution graph of sample point.The mathematical modeling that Density Estimator uses for：
Wherein, xiFor i-th of road sample point,It is the Multilayer networks value at road sample point x, h is that core is close
The bandwidth of estimation is spent, K (x) is kernel function.In Density Estimator, conventional kernel function have gaussian kernel function,
Epanechnikov kernel functions, triangle kernel function and rectangle kernel function.The present invention use gaussian kernel function, its mathematic(al) representation and
Standardized normal distribution is similar：
3. according to the probable value on road axis apparently higher than in the feature construction road of the probable value in other positions
Center line shape and cost function.
Roadway characteristic in Selection utilization road probability distribution graph of the present invention builds road-center line model and cost letter
Number, it is according to the road-center line model that road Probability Characteristics are established：
1) in road probability distribution graph, the probabilistic estimated value of the road sample point on road axis is than other road samples
The probabilistic estimated value of this point is big, and therefore, the quadratic sum of the probabilistic estimated value of all sample points is up to one on road axis
Maximum, i.e.,：
Ep=∫ { G [f (s)] }2Ds=max
Wherein G (x) represents road probability-distribution function, and f (s) represents road axis.
2) according to the geometrical property of road, road axis should be a smooth curve, i.e.,：
Eg=∫ [f " (s)]2Ds=min
3) required according to traffic safety code, the local curve rate of road has a upper bound, i.e.,：
Cg=| f " (s) |≤T1
Wherein T1For given threshold value.
4. solving road-center line model using dynamic programming algorithm, the extraction of road is realized.
In specific solution procedure, road axis is represented with a broken line containing n summit, and on broken line
Summit is around its initial position (xi,yi) mobile, if the summit of broken line is p={ p1,...,pn},pi=(xi,yi), road-center
The property 1 of line model) discrete form is：
Wherein S is line segment pipi+1Pixel set,For line segment p in probability distribution graphipi+1Gamma function, property
2) and property 3) discrete form be：
Cg=| ai-ai+1| ＜ T1
Wherein aiFor line segment pipi+1Direction, | Δ si| it is line segment pipi+1Length, i.e.,
According to road-center line model, following cost function is built：
Wherein, cost function E is a series of function item EiSum, each EiOnly depend on three adjacent vertexs of broken line
{pi-1,pi,pi+1},pi=(xi,yi), while cost function E must is fulfilled for restrictive condition formula | ai-ai+1| ＜ T1。
Dynamic Programming is mainly used to solve optimization problem, use the solution procedure of Dynamic Programming for：
For a cost function：
G=g (x1,x2,...,xn),0≤xi≤mi, i=1,2 ..., n (1)
As its independent variable (x1,x2,...,xn) it is centrifugal pump and when cost function g is following form：
g(x1,x2,...,xn)=g1(x1,x2,x3)+g2(x2,x3,x4)+...+gn-2(xn-2,xn-1,xn) (2)
Function maxima M can be solved with dynamic programming algorithm, and its process is：
1. for aleatory variable x2,x3, solved function f1(x2,x3)：
2. the first step is copied to continue to eliminate variable x2, i.e., for aleatory variable x3,x4, solved function f2(x3,x4)：
3. repeating the above steps, can finally obtain：
Then the maximum M of cost function is：
A kind of embodiment of remote sensing image road extraction element of the present invention
Road extraction device in the present embodiment includes memory and processor and storage on a memory and resonable device
The computer program of upper operation, processor are coupled with memory, realize to give an order during computing device computer program：1)
Original remote sensing image road probability density is calculated, the roadway characteristic on original remote sensing image is converted to the spy of road probability density
Sign；2) it is higher than the feature construction road-center line model of the probable value in other positions according to the probable value on road axis,
And cost function is determined according to road-center line model；3) maximum of cost function is solved using Dynamic Programming, it is very big with this
Value is used as road axis.Processor can use single-chip microcomputer, DSP, PLC or MCU etc., memory can use RAM memory,
Flash memory, ROM memory, eprom memory, eeprom memory, register, hard disk, mobile disk, CD-ROM or this area
The storage medium of any other known form, the specific implementation means respectively instructed are said in the embodiment of method
It is bright, repeat no more here.The event handling flow of the device is as shown in Figure 2.Event flow in device implementation process is mainly divided
For 3 parts, main frame window, remote sensing image road extraction element master control entrance, database engine access component.User's root first
According to need by main frame window carry out corresponding operating, set road gather design parameter；Then device calls external data
Data or model among the engine of storehouse, obtain corresponding roadway characteristic；Finally, entered by remote sensing image road extraction element master control
Mouth carries out real road extraction process and last result is evaluated.
Among high-resolution remote sensing image road information Collecting operation, change existing geography information and protect
Barrier pattern, the road information collection for completing to have on high score resolution remote sense image one fixed width by a small amount of artificial participation are appointed
Business, hence it is evident that shorten drafting period, greatly improve remote sensing image automatic business processing degree.Have to economic construction of China and social development
Play an important role.In addition, present invention may also apply to an information to earthquake monitoring, GIS renewals, map making etc., tool
There are extensive social application prospect and important application value.
Specific embodiment is presented above, but the present invention is not limited to described embodiment, it is general to this area
For logical technical staff, according to the teachings of the present invention, design the models of various modifications, formula, parameter and wound need not be spent
The property made is worked, change, modification, replacement and the change carried out without departing from the principles and spirit of the present invention to embodiment
Type is still within the scope of the present invention.
Claims (10)
1. a kind of method for extracting remote sensing image road, it is characterised in that the method for extracting roads comprises the following steps：
1) original remote sensing image road probability density is calculated, it is close that the roadway characteristic on original remote sensing image is converted into road probability
The feature of degree；
2) it is higher than the feature construction road-center line model of the probable value in other positions according to the probable value on road axis,
And cost function is determined according to road-center line model；
3) maximum of cost function is solved using Dynamic Programming, road axis is used as using the maximum.
2. method for extracting remote sensing image road according to claim 1, it is characterised in that the remote sensing image of the step 1)
Road probability density determines that detailed process is as follows using SVMs and Density Estimator：
A. remote sensing image is classified using SVMs, obtains series of road sample point；
B. the probability density of each road sample point is calculated using Density Estimator.
3. method for extracting remote sensing image road according to claim 2, it is characterised in that obtained in the step B general
Rate density is：
<mrow>
<mover>
<mi>f</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<msup>
<mi>nh</mi>
<mi>d</mi>
</msup>
</mrow>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<mi>K</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mi>x</mi>
<mo>-</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
</mrow>
<mi>h</mi>
</mfrac>
<mo>)</mo>
</mrow>
</mrow>
Wherein xiFor i-th obtained of road sample point,It is the probability density valuation at point x, h is the band of Density Estimator
Width, K (x) are kernel function.
4. method for extracting remote sensing image road according to claim 1 or 2, it is characterised in that built in the step 2)
Road-center line model be：
Ep=∫ { G [f (s)] }2Ds=max
Eg=∫ [f " (s)]2Ds=min
Cg=| f " (s) |≤T1
Wherein G (x) represents road probability-distribution function, and f (s) represents road axis, T1For given threshold.
5. method for extracting remote sensing image road according to claim 4, it is characterised in that according to road-center line model structure
The cost function built is：
<mrow>
<mi>E</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mi>i</mi>
</munder>
<msub>
<mi>E</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>p</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mi>i</mi>
</munder>
<mo>&lsqb;</mo>
<mn>1</mn>
<mo>+</mo>
<mi>cos</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>a</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>a</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>s</mi>
<mo>&Element;</mo>
<mi>S</mi>
</mrow>
</munder>
<msub>
<msup>
<mi>g</mi>
<mn>2</mn>
</msup>
<mrow>
<msub>
<mi>P</mi>
<mi>i</mi>
</msub>
<msub>
<mi>P</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>s</mi>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mo>|</mo>
<msub>
<mi>&Delta;s</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
</mrow>
Wherein p={ p1,...,pn},pi=(xi,yi) be broken line on road axis n summit, S is line segment pipi+1's
Pixel set,For line segment p in probability distribution graphipi+1Gamma function, aiFor line segment pipi+1Direction, | Δ si| it is line segment
pipi+1Length.
A kind of 6. remote sensing image road extraction element, it is characterised in that the road extraction device include memory and processor with
And the computer program run on the memory and on the processor is stored in, the processor and the memory phase
Coupling, realized described in the computing device during computer program to give an order：
1) original remote sensing image road probability density is calculated, it is close that the roadway characteristic on original remote sensing image is converted into road probability
The feature of degree；
2) it is higher than the feature construction road-center line model of the probable value in other positions according to the probable value on road axis,
And cost function is determined according to road-center line model；
3) maximum of cost function is solved using Dynamic Programming, road axis is used as using the maximum.
7. remote sensing image road extraction element according to claim 6, it is characterised in that the remote sensing image of the step 1)
Road probability density determines that detailed process is as follows using SVMs and Density Estimator：
A. remote sensing image is classified using SVMs, obtains series of road sample point；
B. the probability density of each road sample point is calculated using Density Estimator.
8. remote sensing image road extraction element according to claim 7, it is characterised in that obtained in the step B general
Rate density is：
<mrow>
<mover>
<mi>f</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<msup>
<mi>nh</mi>
<mi>d</mi>
</msup>
</mrow>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<mi>K</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mi>x</mi>
<mo>-</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
</mrow>
<mi>h</mi>
</mfrac>
<mo>)</mo>
</mrow>
</mrow>
Wherein xiFor i-th obtained of road sample point,It is the probability density valuation at point x, h is the band of Density Estimator
Width, K (x) are kernel function.
9. the remote sensing image road extraction element according to claim 6 or 7, it is characterised in that built in the step 2)
Road-center line model be：
Ep=∫ { G [f (s)] }2Ds=max
Eg=∫ [f " (s)]2Ds=min
Cg=| f " (s) |≤T1
Wherein G (x) represents road probability-distribution function, and f (s) represents road axis, T1For given threshold.
10. remote sensing image road extraction element according to claim 9, it is characterised in that according to road-center line model
The cost function of structure is：
<mrow>
<mi>E</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mi>i</mi>
</munder>
<msub>
<mi>E</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>p</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mi>i</mi>
</munder>
<mo>&lsqb;</mo>
<mn>1</mn>
<mo>+</mo>
<mi>c</mi>
<mi>o</mi>
<mi>s</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>a</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>a</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>s</mi>
<mo>&Element;</mo>
<mi>S</mi>
</mrow>
</munder>
<msub>
<msup>
<mi>G</mi>
<mn>2</mn>
</msup>
<mrow>
<msub>
<mi>P</mi>
<mi>i</mi>
</msub>
<msub>
<mi>P</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>s</mi>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mo>|</mo>
<msub>
<mi>&Delta;s</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
</mrow>
Wherein p={ p1,...,pn},pi=(xi,yi) be broken line on road axis n summit, S is line segment pipi+1's
Pixel set,For line segment p in probability distribution graphipi+1Gamma function, aiFor line segment pipi+1Direction, | Δ si| it is line segment
pipi+1Length.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710852459.8A CN107578446A (en) | 2017-09-19 | 2017-09-19 | A kind of method for extracting remote sensing image road and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710852459.8A CN107578446A (en) | 2017-09-19 | 2017-09-19 | A kind of method for extracting remote sensing image road and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107578446A true CN107578446A (en) | 2018-01-12 |
Family
ID=61033597
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710852459.8A Pending CN107578446A (en) | 2017-09-19 | 2017-09-19 | A kind of method for extracting remote sensing image road and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107578446A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110175574A (en) * | 2019-05-28 | 2019-08-27 | 中国人民解放军战略支援部队信息工程大学 | A kind of Road network extraction method and device |
| CN112749453A (en) * | 2020-12-16 | 2021-05-04 | 安徽三禾一信息科技有限公司 | Complex equipment residual service life prediction based on improved SVR |
| CN113408457A (en) * | 2021-06-29 | 2021-09-17 | 西南交通大学 | Road information intelligent extraction method combining high-resolution image and video image |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101714252A (en) * | 2009-11-26 | 2010-05-26 | 上海电机学院 | Method for extracting road in SAR image |
| CN103258203A (en) * | 2013-05-20 | 2013-08-21 | 武汉大学 | Method for automatically extracting road centerline of remote-sensing image |
| CN104915636A (en) * | 2015-04-15 | 2015-09-16 | 北京工业大学 | Remote sensing image road identification method based on multistage frame significant characteristics |
| CN106709465A (en) * | 2016-12-29 | 2017-05-24 | 武汉大学 | Polarization SAR image road extraction method based on conditional random field |
| CN106778605A (en) * | 2016-12-14 | 2017-05-31 | 武汉大学 | Remote sensing image road net extraction method under navigation data auxiliary |
-
2017
- 2017-09-19 CN CN201710852459.8A patent/CN107578446A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101714252A (en) * | 2009-11-26 | 2010-05-26 | 上海电机学院 | Method for extracting road in SAR image |
| CN103258203A (en) * | 2013-05-20 | 2013-08-21 | 武汉大学 | Method for automatically extracting road centerline of remote-sensing image |
| CN104915636A (en) * | 2015-04-15 | 2015-09-16 | 北京工业大学 | Remote sensing image road identification method based on multistage frame significant characteristics |
| CN106778605A (en) * | 2016-12-14 | 2017-05-31 | 武汉大学 | Remote sensing image road net extraction method under navigation data auxiliary |
| CN106709465A (en) * | 2016-12-29 | 2017-05-24 | 武汉大学 | Polarization SAR image road extraction method based on conditional random field |
Non-Patent Citations (1)
| Title |
|---|
| 曹帆之，徐杨斌，朱宝山，李润生: "利用动态规划半自动提取高分辨率遥感影像道路中心线", 《测绘科学技术学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110175574A (en) * | 2019-05-28 | 2019-08-27 | 中国人民解放军战略支援部队信息工程大学 | A kind of Road network extraction method and device |
| CN112749453A (en) * | 2020-12-16 | 2021-05-04 | 安徽三禾一信息科技有限公司 | Complex equipment residual service life prediction based on improved SVR |
| CN112749453B (en) * | 2020-12-16 | 2023-10-13 | 安徽三禾一信息科技有限公司 | Complex equipment residual service life prediction method based on improved SVR |
| CN113408457A (en) * | 2021-06-29 | 2021-09-17 | 西南交通大学 | Road information intelligent extraction method combining high-resolution image and video image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108428229B (en) | Lung texture recognition method based on appearance and geometric features extracted by deep neural network | |
| Yuan et al. | Gated CNN: Integrating multi-scale feature layers for object detection | |
| Yu et al. | Mixed pooling for convolutional neural networks | |
| CN110059698B (en) | Semantic segmentation method and system based on edge dense reconstruction for street view understanding | |
| CN110569033B (en) | Method for generating basic codes of digital transaction type intelligent contracts | |
| CN102592268B (en) | Method for segmenting foreground image | |
| CN103116766B (en) | A kind of image classification method of encoding based on Increment Artificial Neural Network and subgraph | |
| CN109002841A (en) | A kind of building element extracting method based on Faster-RCNN model | |
| CN110263673A (en) | Human facial expression recognition method, apparatus, computer equipment and storage medium | |
| CN109753949B (en) | Multi-window traffic sign detection method based on deep learning | |
| CN107480575A (en) | The training method of model, across age face identification method and corresponding device | |
| CN104361096B (en) | The image search method of a kind of feature based rich region set | |
| CN107578446A (en) | A kind of method for extracting remote sensing image road and device | |
| CN105975913A (en) | Road network extraction method based on adaptive cluster learning | |
| CN107633226A (en) | A kind of human action Tracking Recognition method and system | |
| CN107992807A (en) | A kind of face identification method and device based on CNN models | |
| CN111259950B (en) | Method for training YOLO neural network based on 3D model | |
| CN103489000A (en) | Achieving method of human movement recognition training system | |
| CN113920148B (en) | Building boundary extraction method and equipment based on polygon and storage medium | |
| CN106844785A (en) | Saliency segmentation-based content-based image retrieval method | |
| CN105844667A (en) | Structural target tracking method of compact color coding | |
| CN114842507A (en) | Reinforced pedestrian attribute identification method based on group optimization reward | |
| Sun et al. | An improved vehicle detection algorithm based on YOLOV3 | |
| CN114241231A (en) | Identification method and device based on hierarchical label attention | |
| CN107832732A (en) | Method for detecting lane lines based on ternary tree traversal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180112 |
|
| RJ01 | Rejection of invention patent application after publication |