CN105069461A - Insulator string automatic positioning method based on image feature point collineation and equidistant constraint - Google Patents
Insulator string automatic positioning method based on image feature point collineation and equidistant constraint Download PDFInfo
- Publication number
- CN105069461A CN105069461A CN201510443073.2A CN201510443073A CN105069461A CN 105069461 A CN105069461 A CN 105069461A CN 201510443073 A CN201510443073 A CN 201510443073A CN 105069461 A CN105069461 A CN 105069461A
- Authority
- CN
- China
- Prior art keywords
- conllinear
- centerdot
- sigma
- image
- insulator chain
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
Abstract
The invention discloses an insulator string automatic positioning method based on image feature point collineation and equidistant constraint. The method comprises the steps of image preprocessing, curvature scale space corner extraction, collinear equidistant point extraction, hierarchical clustering and insulator string positioning. According to the invention, the method is simple and practical; insulator string curvature scale space corner collineation and equidistant constraint is used; automatic positioning of an insulator string in any main axis direction in a complex, noisy and low-resolution aerial image is realized; less time is consumed; robustness for partial occlusion and string breakage is realized; and the problems of low accuracy, false positioning and high computation complexity of the existing insulator string positioning method are solved.
Description
Technical field
The invention belongs to power transmission and transforming equipment running status maintenance field, particularly relate to a kind of insulator chain automatic positioning method based on image characteristic point conllinear and iso-distance constraint.
Background technology
Insulator is element indispensable in transmission line of electricity, possesses the function such as mechanical support and electric insulation.Once insulator is damaged, just lose its useful effect, cause irremediable massive losses.Therefore, it is necessary for carrying out detecting in time to insulator.And automatically to orient insulator chain from Aerial Images be the important prerequisite realizing its state-detection and fault diagnosis.Insulator chain automatic positioning method of taking photo by plane at present can be divided into 4 classes substantially: original image, based on the method for segmentation, is divided into multiple region, and marks interested region by (1); (2) based on the method for rim detection, find the profile of interesting target, realize location; (3) based on the method for texture, analyze the textural characteristics of interesting target, and as criterion, target location is extracted; (4) based on the method for coupling, the feature extracting template and test pattern is mated, the region in indicia matched.
Method positioning precision based on segmentation is low, and correctly can not process the close image of gray feature, is not suitable for the Aerial Images of complexity, noisy, low resolution.Based on the method at edge to noise-sensitive, correctly cannot process the pseudo-target that edge is similar, and the image that background is complicated, edge variation is various can the travelling speed of alleviative method greatly.Method computation complexity based on texture is high, cannot solve the texture difference of insulator chain and background little time insulator chain orientation problem, and some pseudo-target close with insulator chain textural characteristics is difficult to distinguish.Method based on coupling has strong dependency to template, and a large amount of templates can reduce the feature extracting and matching speed of template and test pattern greatly.
In the Aerial Images of complicated, noisy, low resolution, there is the limitation such as precision is low, location, computation complexity be high by mistake in existing insulator chain localization method, and does not consider the shape facility of insulator chain in bianry image.Insulator chain and the target such as shaft tower, circuit have obvious shape difference in bianry image.
Summary of the invention
The technical problem to be solved in the present invention is: provide a kind of insulator chain automatic positioning method of taking photo by plane based on image characteristic point conllinear and iso-distance constraint.
The technical solution used in the present invention is:
Based on an insulator chain automatic positioning method of taking photo by plane for image characteristic point conllinear and iso-distance constraint, comprise the following steps:
Step a: Image semantic classification: pre-service is carried out to described insulator chain image of taking photo by plane, obtains the bianry image that filtering noise back edge is level and smooth;
Step b: curvature scale space angle point grid: the edge image extracting described bianry image, extracts contour curve from described edge image; The curvature of each pixel on described contour curve is calculated, local curvature's maximum point alternatively angle point in yardstick σ=3 time; If the curvature value of described candidate angular is greater than preset curvature threshold value, then it is correct angle point; In former figure, the curvature scale space angle point that accurately location is corresponding with described correct angle point.
Step c: conllinear equidistant points extracts: select the curvature scale space angle point described in any two, utilize conllinear and iso-distance constraint to find thirdly, is thirdly all curvature scale space angle point if exist, then judges that as conllinear equidistant points at 3.
Steps d: hierarchical clustering: carry out hierarchical clustering to the direction of all conllinear equidistant points, makes the conllinear equidistant points direction in every class change and is less than preset direction threshold value, and the class selecting quantity maximum is as the equidistant point set of conllinear of insulator chain.
Step e: insulator chain is located: the equidistant point set of conllinear marking described insulator chain with minimum enclosed rectangle, thus realize the automatic location of insulator chain in described complicated Aerial Images.
Concrete steps in described step a are as follows:
Step a-1: insulator chain image carries out binary conversion treatment to taking photo by plane, and obtains insulator chain bianry image;
Step a-2: carry out morphological erosion and expansion to described insulator chain bianry image, obtains filtered insulator chain bianry image;
Step a-3: in filtered insulator chain bianry image described in filtering, area is less than the zonule of preset area threshold value, obtains pre-processed results image.
In described step b, concrete steps are as follows:
Step b-1: extract the canny edge in described pre-processed results image, generates edge image;
Step b-2: extract contour curve from described edge image, to be expressed as under yardstick σ with the functional form Γ (μ, σ) that arc length μ is parameter by described contour curve:
Γ(μ,σ)=(x(μ,σ),y(μ,σ))(1)
001"/>
002"/>
The wherein Gaussian function of g (μ, σ) to be yardstick be σ, the coordinate representation that x (μ), y (μ) are is parameter with arc length μ,
for convolution operation;
Step b-3: the curvature calculating each pixel on described contour curve in yardstick σ=3 time, finds out local curvature's maximum point, alternatively angle point;
003"/>
Wherein,
004"/>
represent single order and the second derivative of g (μ, σ) respectively,
for convolution operation.
Step b-4: if the curvature value of described candidate angular is greater than preset curvature threshold value, described candidate angular is correct angle point;
Step b-5: in former figure, the curvature scale space angle point that accurately location is corresponding with described correct angle point.
Concrete steps in described step c are as follows:
Step c-1: set up two-dimensional array A (N, 2), N is the number of angle point, its array element is the coordinate of original image mean curvature metric space angle point;
Step c-2: give point (x in order by each array element in described array A
p, y
p), (x
p, y
p) ∈ A, and to every bit (x
p, y
p), repeat step c-3 ~ c-4;
Step c-3: (x is different to each
p, y
p) point (x
q, y
q) ∈ A, calculate (x
p, y
p) and (x
q, y
q) between distance d
pqwith direction o
pq;
Step c-4: get successively and be different from (x
p, y
p) and (x
q, y
q) point (x, y) ∈ A, calculate (x, y) and (x
p, y
p) between distance d
pwith direction o
pif meet d
pand d
pqrelative mistake is less than certain minimum value ε
1, o
pand o
pqabsolute difference is less than certain minimum value ε
2:
005"/>
Then judge 3 X={ (x
p, y
p), (x
q, y
q), (x, y) } be conllinear equidistant points, turn to step c-3, otherwise repeated execution of steps c-4.
Concrete steps in described steps d are as follows:
Steps d-1: to often organizing conllinear equidistant points X
j={ (x
j1, y
j1), (x
j2, y
j2), (x
j3, y
j3), 1≤j≤M, calculates its direction o
j, M is conllinear equidistant points group number;
006"/>
Steps d-2: by each group conllinear equidistant points X
jbe set to cluster;
Steps d-3: calculate the root mean square between any two bunches, obtains the distance matrix O={o in direction
ij, 1≤i, j≤M;
o
ij=o
i-o
j(8)
Steps d-4: by o
ijcorresponding two bunches of minimum value merge into one new bunch;
Steps d-5: repeat steps d-3 ~ d-4, when bunch in the difference of maxima and minima in direction of conllinear equidistant points be greater than preset direction threshold value time, ending cluster merges.
The beneficial effect adopting technique scheme to produce is:
1, the present invention utilizes conllinear and the iso-distance constraint of insulator chain curvature scale space angle point, the automatic precision realizing the insulator chain of any major axes orientation in complicated Aerial Images determines position, solves the problem that existing insulator chain positioning method accuracy is low, mistake is located and computation complexity is high;
2, the present invention is consuming time few, and to partial occlusion and fall string and have robustness, can improve positioning precision, improve robotization performance.
3, the inventive method is simple, practical, and achieves higher positioning precision, and required time is shorter, without the need to artificial participation.
Accompanying drawing explanation
Fig. 1 is process flow diagram of the present invention;
Fig. 2 is conllinear and the iso-distance constraint schematic diagram of the curvature scale space angle point of insulator chain image of the present invention;
Fig. 3 is that the embodiment of the present invention 1 is taken photo by plane insulator chain image;
Fig. 4 is the embodiment of the present invention 1 curvature scale space angle point grid result;
Fig. 5 is that the embodiment of the present invention 1 conllinear equidistant points extracts result;
Fig. 6 is the embodiment of the present invention 1 hierarchical clustering result;
Fig. 7 is the embodiment of the present invention 1 minimum enclosed rectangle mark frame;
Fig. 8 is the embodiment of the present invention 1 positioning result.
Embodiment
Below in conjunction with drawings and Examples, the invention will be further described.
Embodiment 1:
As shown in Figure 1, a kind of insulator chain automatic positioning method of taking photo by plane based on image characteristic point conllinear and iso-distance constraint, comprises the following steps:
Step a: Image semantic classification: pre-service is carried out to described insulator chain image of taking photo by plane, obtains the bianry image that filtering noise back edge is level and smooth;
Step b: curvature scale space angle point grid: the edge image extracting described bianry image, extracts contour curve from described edge image; The curvature of each pixel on described contour curve is calculated, local curvature's maximum point alternatively angle point in yardstick σ=3 time; If the curvature value of described candidate angular is greater than preset curvature threshold value, then it is correct angle point; In former figure, the curvature scale space angle point that accurately location is corresponding with described correct angle point.
Step c: conllinear equidistant points extracts: select the curvature scale space angle point described in any two, utilize conllinear and iso-distance constraint to find thirdly, is thirdly all curvature scale space angle point if exist, then judges that as conllinear equidistant points at 3.For 3 umbrella disks, as shown in Figure 2, l is insulator chain major axes orientation, and A, B, C are insulator chain curvature scale space angle point, d
aB, d
bCbe respectively AB, the distance between BC.A, B, C are approximate is positioned at straight line l
1on, and l
1parallel with l; d
aB, d
bCapproximately equal.Therefore judge A, B, C are conllinear equidistant points.
Steps d: hierarchical clustering: carry out hierarchical clustering to the direction of all conllinear equidistant points, makes the conllinear equidistant points direction in every class change and is less than preset direction threshold value, and the class selecting quantity maximum is as the equidistant point set of conllinear of insulator chain.
Step e: insulator chain is located: the equidistant point set of conllinear marking described insulator chain with minimum enclosed rectangle, thus realize the automatic location of insulator chain in described complicated Aerial Images.
Concrete steps in described step a are as follows:
Step a-1: insulator chain image carries out binary conversion treatment to taking photo by plane, and obtains insulator chain bianry image;
Step a-2: carry out morphological erosion and expansion to described insulator chain bianry image, obtains filtered insulator chain bianry image;
Step a-3: in filtered insulator chain bianry image described in filtering, area is less than the zonule of preset area threshold value, obtains pre-processed results image.
In described step b, concrete steps are as follows:
Step b-1: extract the canny edge in described pre-processed results image, generates edge image;
Step b-2: extract contour curve from described edge image, to be expressed as under yardstick σ with the functional form Γ (μ, σ) that arc length μ is parameter by described contour curve:
Γ(μ,σ)=(x(μ,σ),y(μ,σ))(1)
007"/>
008"/>
The wherein Gaussian function of g (μ, σ) to be yardstick be σ, the coordinate representation that x (μ), y (μ) are is parameter with arc length μ,
for convolution operation;
Step b-3: the curvature calculating each pixel on described contour curve in yardstick σ=3 time, finds out local curvature's maximum point, alternatively angle point;
009"/>
Wherein,
010"/>
represent single order and the second derivative of g (μ, σ) respectively,
for convolution operation.
Step b-4: if the curvature value of described candidate angular is greater than preset curvature threshold value, described candidate angular is correct angle point;
Step b-5: in former figure, the curvature scale space angle point that accurately location is corresponding with described correct angle point.
Concrete steps in described step c are as follows:
Step c-1: set up two-dimensional array A (N, 2), N is the number of angle point, its array element is the coordinate of original image mean curvature metric space angle point;
Step c-2: give point (x in order by each array element in described array A
p, y
p), (x
p, y
p) ∈ A, and to every bit (x
p, y
p), repeat step c-3 ~ c-4;
Step c-3: (x is different to each
p, y
p) point (x
q, y
q) ∈ A, calculate (x
p, y
p) and (x
q, y
q) between distance d
pqwith direction o
pq;
Step c-4: get successively and be different from (x
p, y
p) and (x
q, y
q) point (x, y) ∈ A, calculate (x, y) and (x
p, y
p) between distance d
pwith direction o
pif meet d
pand d
pqrelative mistake is less than certain minimum value ε
1, o
pand o
pqabsolute difference is less than certain minimum value ε
2:
011"/>
Then judge 3 X={ (x
p, y
p), (x
q, y
q), (x, y) } be conllinear equidistant points, turn to step c-3, otherwise repeated execution of steps c-4.
Concrete steps in described steps d are as follows:
Steps d-1: to often organizing conllinear equidistant points X
j={ (x
j1, y
j1), (x
j2, y
j2), (x
j3, y
j3), 1≤j≤M, calculates its direction o
j, M is conllinear equidistant points group number;
012"/>
Steps d-2: by each group conllinear equidistant points X
jbe set to cluster;
Steps d-3: calculate the root mean square between any two bunches, obtains the distance matrix O={o in direction
ij, 1≤i, j≤M;
o
ij=o
i-o
j(8)
Steps d-4: by o
ijcorresponding two bunches of minimum value merge into one new bunch;
Steps d-5: repeat steps d-3 ~ d-4, when bunch in the difference of maxima and minima in direction of conllinear equidistant points be greater than preset direction threshold value time, ending cluster merges.
In the present embodiment, single insulator string string takes photo by plane original image as shown in Fig. 3 (a), and two insulator string string takes photo by plane original image as shown in Fig. 3 (b).After respectively pre-service being carried out to two width images, extract its curvature scale space angle point as shown in Fig. 4 (a) He 4 (b), can find out that this angle point contains abundant local feature and shape information, tentatively can describe shape facility.Utilize the distinctive conllinear of insulator chain and iso-distance constraint condition, remove the angle point not meeting this constraint, extract the conllinear equidistant points meeting this constraint, as shown in Fig. 5 (a) He 5 (b).Utilize hierarchical clustering to obtain maximum kind, and the direction of the conllinear equidistant points of maximum kind is approximate consistent with major axes orientation, as shown in Fig. 6 (a) He 6 (b).Finally, with the conllinear equidistant points of minimum enclosed rectangle mark insulator chain, as shown in Fig. 7 (a) He 7 (b); Rectangle posting is presented in original image, realizes the automatic positioning result of insulator chain as shown in Fig. 8 (a) He 8 (b).
The above; be only the present invention's preferably embodiment, but protection scope of the present invention is not limited thereto, is anyly familiar with those skilled in the art in the technical scope that the present invention discloses; the change that can expect easily or replacement, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.
Claims (5)
1., based on an insulator chain automatic positioning method of taking photo by plane for image characteristic point conllinear and iso-distance constraint, comprise the following steps:
Step a: Image semantic classification: pre-service is carried out to described insulator chain image of taking photo by plane, obtains the bianry image that filtering noise back edge is level and smooth;
Step b: curvature scale space angle point grid: the edge image extracting described bianry image, extracts contour curve from described edge image; The curvature of each pixel on described contour curve is calculated, local curvature's maximum point alternatively angle point in yardstick σ=3 time; If the curvature value of described candidate angular is greater than preset curvature threshold value, then it is correct angle point; In former figure, the curvature scale space angle point that accurately location is corresponding with described correct angle point.
Step c: conllinear equidistant points extracts: select the curvature scale space angle point described in any two, utilize conllinear and iso-distance constraint to find thirdly, is thirdly all curvature scale space angle point if exist, then judges that as conllinear equidistant points at 3.
Steps d: hierarchical clustering: carry out hierarchical clustering to the direction of all conllinear equidistant points, makes the conllinear equidistant points direction in every class change and is less than preset direction threshold value, and the class selecting quantity maximum is as the equidistant point set of conllinear of insulator chain.
Step e: insulator chain is located: the equidistant point set of conllinear marking described insulator chain with minimum enclosed rectangle, thus realize the automatic location of insulator chain in described complicated Aerial Images.
2. the insulator chain automatic positioning method of taking photo by plane based on image characteristic point conllinear and iso-distance constraint according to claim 1, is characterized in that: the concrete steps in described step a are as follows:
Step a-1: insulator chain image carries out binary conversion treatment to taking photo by plane, and obtains insulator chain bianry image;
Step a-2: carry out morphological erosion and expansion to described insulator chain bianry image, obtains filtered insulator chain bianry image;
Step a-3: in filtered insulator chain bianry image described in filtering, area is less than the zonule of preset area threshold value, obtains pre-processed results image.
3., according to the insulator chain automatic positioning method based on image characteristic point conllinear and iso-distance constraint according to claim 1, it is characterized in that: in described step b, concrete steps are as follows:
Step b-1: extract the canny edge in described pre-processed results image, generates edge image;
Step b-2: extract contour curve from described edge image, to be expressed as under yardstick σ with the functional form Γ (μ, σ) that arc length μ is parameter by described contour curve:
Γ(μ,σ)=(x(μ,σ),y(μ,σ))(1)
The wherein Gaussian function of g (μ, σ) to be yardstick be σ, the coordinate representation that x (μ), y (μ) are is parameter with arc length μ,
for convolution operation;
Step b-3: the curvature calculating each pixel on described contour curve in yardstick σ=3 time, finds out local curvature's maximum point, alternatively angle point;
Wherein,
represent single order and the second derivative of g (μ, σ) respectively,
for convolution operation.
Step b-4: if the curvature value of described candidate angular is greater than preset curvature threshold value, described candidate angular is correct angle point;
Step b-5: in former figure, the curvature scale space angle point that accurately location is corresponding with described correct angle point.
4., according to the insulator chain automatic positioning method based on image characteristic point conllinear and iso-distance constraint according to claim 1, it is characterized in that: the concrete steps in described step c are as follows:
Step c-1: set up two-dimensional array A (N, 2), N is the number of angle point, its array element is the coordinate of original image mean curvature metric space angle point;
Step c-2: give point (x in order by each array element in described array A
p, y
p), (x
p, y
p) ∈ A, and to every bit (x
p, y
p), repeat step c-3 ~ c-4;
Step c-3: (x is different to each
p, y
p) point (x
q, y
q) ∈ A, calculate (x
p, y
p) and (x
q, y
q) between distance d
pqwith direction o
pq;
Step c-4: get successively and be different from (x
p, y
p) and (x
q, y
q) point (x, y) ∈ A, calculate (x, y) and (x
p, y
p) between distance d
pwith direction o
pif meet d
pand d
pqrelative mistake is less than certain minimum value ε
1, o
pand o
pqabsolute difference is less than certain minimum value ε
2:
Then judge 3 X={ (x
p, y
p), (x
q, y
q), (x, y) } be conllinear equidistant points, turn to step c-3, otherwise repeated execution of steps c-4.
5., according to the insulator chain automatic positioning method based on image characteristic point conllinear and iso-distance constraint according to claim 1, it is characterized in that: the concrete steps in described steps d are as follows:
Steps d-1: to often organizing conllinear equidistant points X
j={ (x
j1, y
j1), (x
j2, y
j2), (x
j3, y
j3), 1≤j≤M, calculates its direction o
j, M is conllinear equidistant points group number;
Steps d-2: by each group conllinear equidistant points X
jbe set to cluster;
Steps d-3: calculate the root mean square between any two bunches, obtains the distance matrix O={o in direction
ij, 1≤i, j≤M;
o
ij=o
i-o
j(8)
Steps d-4: by o
ijcorresponding two bunches of minimum value merge into one new bunch;
Steps d-5: repeat steps d-3 ~ d-4, when bunch in the difference of maxima and minima in direction of conllinear equidistant points be greater than preset direction threshold value time, ending cluster merges.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510443073.2A CN105069461B (en) | 2015-07-24 | 2015-07-24 | Based on image characteristic point collinearly with the insulator chain automatic positioning method of iso-distance constraint |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510443073.2A CN105069461B (en) | 2015-07-24 | 2015-07-24 | Based on image characteristic point collinearly with the insulator chain automatic positioning method of iso-distance constraint |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105069461A true CN105069461A (en) | 2015-11-18 |
CN105069461B CN105069461B (en) | 2018-11-06 |
Family
ID=54498822
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510443073.2A Active CN105069461B (en) | 2015-07-24 | 2015-07-24 | Based on image characteristic point collinearly with the insulator chain automatic positioning method of iso-distance constraint |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105069461B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107369162A (en) * | 2017-07-21 | 2017-11-21 | 华北电力大学(保定) | A kind of generation method and system of insulator candidate target region |
CN115512252A (en) * | 2022-11-18 | 2022-12-23 | 东北电力大学 | Unmanned aerial vehicle-based power grid inspection automation method and system |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103714342B (en) * | 2013-12-20 | 2016-10-05 | 华北电力大学(保定) | Insulator chain automatic positioning method of taking photo by plane based on bianry image shape facility |
CN104021394B (en) * | 2014-06-05 | 2017-12-01 | 华北电力大学(保定) | Insulator image-recognizing method based on AdaBoost algorithms |
-
2015
- 2015-07-24 CN CN201510443073.2A patent/CN105069461B/en active Active
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107369162A (en) * | 2017-07-21 | 2017-11-21 | 华北电力大学(保定) | A kind of generation method and system of insulator candidate target region |
CN107369162B (en) * | 2017-07-21 | 2020-07-10 | 华北电力大学(保定) | Method and system for generating insulator candidate target area |
CN115512252A (en) * | 2022-11-18 | 2022-12-23 | 东北电力大学 | Unmanned aerial vehicle-based power grid inspection automation method and system |
CN115512252B (en) * | 2022-11-18 | 2023-02-21 | 东北电力大学 | Unmanned aerial vehicle-based power grid inspection automation method and system |
Also Published As
Publication number | Publication date |
---|---|
CN105069461B (en) | 2018-11-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109598290A (en) | A kind of image small target detecting method combined based on hierarchical detection | |
CN103295232B (en) | Based on the SAR image registration method in straight line and region | |
CN106803253B (en) | A kind of three-dimensional rock image crack identification method | |
CN107346420A (en) | Text detection localization method under a kind of natural scene based on deep learning | |
CN109543676B (en) | Character wheel type water meter digital character recognition method based on image processing | |
CN110334578B (en) | Weak supervision method for automatically extracting high-resolution remote sensing image buildings through image level annotation | |
CN102938066A (en) | Method for reconstructing outer outline polygon of building based on multivariate data | |
CN111754465B (en) | Insulator positioning and string dropping detection method | |
CN103258037A (en) | Trademark identification searching method for multiple combined contents | |
CN103310195A (en) | LLC-feature-based weak-supervision recognition method for vehicle high-resolution remote sensing images | |
CN107392141A (en) | A kind of airport extracting method based on conspicuousness detection and LSD straight-line detections | |
CN103345760B (en) | A kind of automatic generation method of medical image object shapes template mark point | |
CN104574401A (en) | Image registration method based on parallel line matching | |
CN103106658A (en) | Island or reef coastline rapid obtaining method | |
CN109344820A (en) | Digital electric meter Recognition of Reading method based on computer vision and deep learning | |
Mishchenko et al. | Chart image understanding and numerical data extraction | |
CN110298227A (en) | A kind of vehicle checking method in unmanned plane image based on deep learning | |
CN105069395A (en) | Road marking automatic identification method based on terrestrial three-dimensional laser scanning technology | |
Zhu et al. | Integrating saliency and ResNet for airport detection in large-size remote sensing images | |
CN102081740B (en) | 3D image classification method based on scale invariant features | |
CN105069461A (en) | Insulator string automatic positioning method based on image feature point collineation and equidistant constraint | |
CN101488182B (en) | Image characteristics extraction method used for handwritten Chinese character recognition | |
CN104036295A (en) | Road center line auto-detection method employing multispectral remote sensing images | |
CN117115390A (en) | Three-dimensional model layout method of power transformation equipment in transformer substation | |
CN116452604A (en) | Complex substation scene segmentation method, device and storage medium |
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 |