CN104568003A - Remote monitoring system and method for ice coating process of power transmission lines - Google Patents

Abstract

The invention discloses a remote monitoring system and method for ice coating process of power transmission lines and belongs to the field of power industry monitoring system products. The remote monitoring system includes a simulation wire, a tension sensor connected with the simulation wire, a binocular camera and a control box; a data acquisition module connected with the binocular camera, a control module connected with the tension sensor, a communication module connected with a central control terminal and the like are arranged in the control box; the remote monitoring method is a method for obtaining three-dimensional geometric information of an object from multiple images based on the principle of parallax, wherein two digital images of the measured object can be obtained simultaneously by the binocular camera from different angles, the three-dimensional geometric information of the object is restored based on the principle of parallax, the three-dimensional profile and position of the object are reconstructed, and the three-dimensional image data is transmitted to the central control terminal for calculation, so as to master the dynamic changes of ice coating in real time. According to the invention, the ice coating data monitoring accuracy and convenience can be satisfied, the expansibility is excellent, and the data monitoring sensitivity is improved.

Description

A kind of remote supervision system of powerline ice-covering process and method

Technical field

The present invention relates to a kind of remote supervision system and method for powerline ice-covering process, belong to power industry monitoring system product scope.

Background technology

Within 2013, according to State Grid Corporation of China's requirement, artificial sight ice device is installed by office of prefectures and cities in icing key monitoring point, be used for measuring icing.But during icing, bad weather, and sight territory, ice formation is generally positioned at the area that personnel are difficult to arrive, artificial sight ice mode in the past needs to consume a large amount of manpowers, and when weather extreme, personnel cannot arrive observation station at all and measure.In addition due to the scrambling of icing, thickness and the shape of icing cannot be accurately measured by the mode of manual measurement.The position of artificial sight ice device installation is lower and actual icing situation is inconsistent.

At present, by monitoring the device at wire hitch point inclination angle and suspension insulator deviation angle measuring circuit icing; By the device of monitoring stress of conductor measure of the change line ice coating; Adopt the device of distributed optical fiber sensing system Monitoring Line icing; Adopt the icing monitoring installing the mode widespread use transmission lines of electricity such as the device of analog conducting wire Monitoring Line icing additional.But after line ice coating, the comprehensive load of wire changes, and common several monitoring modes well can not react the situation of line ice coating, limited to actual production job guide meaning.

In view of this, the present inventor studies this, and develop a kind of remote supervision system and method for powerline ice-covering process specially, this case produces thus.

Summary of the invention

The object of this invention is to provide a kind of remote supervision system and method for powerline ice-covering process, accuracy and the convenience of icing data monitoring can be met, there is excellent expansibility, improve data monitoring sensitivity.

To achieve these goals, solution of the present invention is:

A kind of remote supervision system of powerline ice-covering process, comprise analog conducting wire, the pulling force sensor be connected with analog conducting wire, binocular camera and control box, wherein, the data acquisition module be connected with binocular camera is provided with in described control box, the control module be connected with pulling force sensor, the communication module of holding and being connected is controlled with middle, and be the power module that in control box, modules and binocular camera are powered, described binocular camera camera lens and analog conducting wire are positioned in same level measuring surface, above-mentioned analog conducting wire and actual track wire are same material and same model.

As preferably, described pulling force sensor is arranged in cylinder type measurement component, and analog conducting wire is located in the web member of cylinder type measurement component, and the measuring head of pulling force sensor is positioned at web member, for the gravity of measure analog wire.

As preferably, the control module in described control box adopts Microprocessor S3C44B0X, after relevant image data is sent to control module by data acquisition module and pulling force sensor, is sent to middle control end by microprocessor after initial analysis process by communication module.

As preferably, described microprocessor Intel atom processor.

As preferably, described binocular camera inside is provided with DSP processing unit, described DSP processing unit by optical system measuring to analog conducting wire on icing image transform can the composite vision signal format of standard, be transported to the data acquisition module of control box by signal cable.

As preferably, described power module is connected with solar electric power supply system, adopts photovoltaic power supply.

A remote monitoring method for powerline ice-covering process, comprises the steps:

Step 1, gravity auxiliary monitoring step:

Step 1-1: the transducing signal of detection sends to middle control to hold by control box by pulling force sensor, middle control end first calculates the load w1 (N/m) under analog conducting wire normality according to formula (1),

w ₁＝qg≈qg _n＝9.80665q (1)

In formula, gn is normal acceleration of gravity, gn=9.80665 (m/s2);

Step 1-2: according to formula (2), formula (3) calculates analog conducting wire icing time load

Suppose that the icing of all kinds and different section profile is all converted as density is the circular glaze section of 0.9g/cm3, when oneself knows wire diameter D (mm) and ice covering thickness b (mm), its unit length ice load w2 (N/m) is

$w_2=\frac{0.9\mathrm{\π}{g}_n}{4}[{(D+2b)}^2-D^2]\×{10}^{-3}=\mathrm{\κb}(b+D)---\left(2\right)$

κ=0.027728 in formula

During wire icing, vertical total load (TL) w3 is wire gravity load w1 and ice load w2 sum, namely

w ₃＝w ₁+w ₂(3)

Step 1-3: the ice covering thickness b (mm) calculating analog conducting wire according to formula (4):

$b=\frac{-\mathrm{\κD}\±\sqrt{{\left(\mathrm{\κD}\right)}^2+4\×\mathrm{\κ}\×(w_3-w_1)}}{2\×\mathrm{\κ}}---\left(4\right)$

Step 2, binocular measuring process:

Step 2-1: the binocular camera demarcated gathers the icing image (left image and right image) on same environment Imitating wire respectively, and sends to middle control to hold by control box 2 icing images;

Step 2-2: middle control end extracts this node visual signature within the vision according to vision node disjoint each in icing image;

Step 2-3: middle control end according in vision node calculate step 2-2 each in icing image extract the three-dimensional coordinate of visual signature, and caching process;

Step 2-4: middle control end according to the locations of contours impact point of analog conducting wire, the scope of ice covering thickness on calculating simulation wire, on described calculating simulation wire, the concrete grammar of ice covering thickness scope is:

(1) three-dimensional coordinate of each point on analog conducting wire profile is extracted: according to the same place on the icing image of left and right, carry out intersection and obtain spatial point three-dimensional coordinate;

(2) according to the imaging equation (collinearity equation) of central projection, the principle of three point on a straight line is adopted to build measuring and calculation formula:

$x_l=x-x_0=-f\frac{a_1(X_A-X_S)+b_1(Y_A-Y_S)+c_1(Z_A-Z_S)}{a_3(X_A-X_S)+b_3(Y_A-Y_S)+c_3(Z_A-Z_S)}$

$y_l=y-y_0=-f\frac{a_2(X_A-X_S)+b_2(Y_A-Y_S)+c_2(Z_A-Z_S)}{a_3(X_A-X_S)+b_3(Y_A-Y_S)+c_3(Z_A-Z_S)}$

$r=\frac{\sqrt[2]{x_l^2+y_l^2}-d_1}{2}$

Wherein r is ice covering thickness, d ₁for known analog wire radius;

(3) according to the value calculating the ice covering thickness selected, take out maximum effective value and minimum effective value, and calculate the mean value of ice covering thickness.

Step 3, middle control end calculate the ice covering thickness mean value of analog conducting wire respectively by step 1 and step 2, when both ice covering thickness differences are within setting range, and in the valid value range of ice covering thickness, then think that monitoring result is error free, directly export the ice covering thickness of the analog conducting wire that step 2 obtains; When both ice covering thickness differences are greater than setting value, then resampling, repeats step 1 and step 2, until detect thickness difference is within the scope of permissible error.

As preferably, user can according to the actual requirements, and icing image click any 2 ice covering thickness values calculating current measurement point.

As preferably, the Visual Feature Retrieval Process concrete grammar described in step 2-2 is:

1) each vision node binocular solid camera of adopting one to demarcate, contains left and right two cameras, can the left and right stereo-picture of Real-time Obtaining; Then adopt PCA-SIFT method, extract the SIFT feature descriptor of left and right image respectively, each feature descriptor is 36 n dimensional vector ns, and each n dimensional vector n represents with 8;

2) SIFT feature adopting KD-Tree method to realize between left and right image is mated, and the Matching power flow of record often pair of SIFT feature coupling;

3) epipolar line restriction filtering Mismatching point is adopted; For the binocular solid camera demarcated, epipolar line restriction requires that the vertical coordinate difference of correct matching double points in left and right image is less than 1;

4) with the benchmark image that left image is Visual Feature Retrieval Process, left image is divided into 20 × 20 equal-sized regions, only choose the visual signature of SIFT feature as this region that has maximum Matching power flow in each area, namely from left image, 400 visual signatures can be extracted at most, and using the visual signature of these visual signatures as node present frame; Therefore the descriptor maximum amount of data that each vision node extracts visual signature from each frame stereo image pair is 400 × 36 × 8/1024=112.5kb;

As preferably, the visual signature three-dimensional coordinate described in step 2-3 calculates concrete grammar and is:

In step 2-2 extract the vertical coordinate difference of visual signature in left and right image and be less than 1, and horizontal coordinate difference is called parallax d, according to binocular stereo vision three-dimensional reconstruction formula, can obtain the three-dimensional coordinate (X of visual signature i under left camera coordinate system ⁱ, Y ⁱ, Z ⁱ):

$\left\{\begin{array}{c}{X}^i=b(u_i-u_0)/d_i\\ Y^i=b(v_0-v_i)/d_i\\ Z^i=\mathrm{bf}/d_i\end{array}\right.---\left(5\right)$

Wherein b, f, (u0, v0) is respectively the baseline, focal length and the left optical picture centre coordinate that obtain in binocular solid camera calibration.(ui, vi) for the coordinate of visual signature i in left image, di be the parallax of visual signature i.

The remote supervision system of powerline ice-covering process of the present invention and method, remote supervision system, by building Binocular Stereo Vision System, adopts based on principle of parallax and is obtained the method for object dimensional geological information by multiple image.Binocular Stereo Vision System obtains two width digital pictures of measured object from different perspectives by binocular camera simultaneously, by single camera in the two width digital pictures not obtaining measured object in the same time from different perspectives, and the three-dimensional geometric information of object is recovered based on principle of parallax, rebuild object three-dimensional contour outline and position, and 3 d image data is sent to middle control end by 3 dimension measurement models calculating, grasp icing dynamic change situation in real time; Simultaneously, remote supervision system adopts gravity sensor Real-Time Monitoring, measurement is arranged on the stressed situation of change of the analog conducting wire of the unit length (1 meter) under overhead transmission line identical icing environment, and dynamic contrast analyzes icing dynamic change situation, makes monitoring result more accurate.

Below in conjunction with drawings and the specific embodiments, the present invention is described in further detail.

Accompanying drawing explanation

Fig. 1 is the remote supervision system control block diagram of the powerline ice-covering process of the present embodiment;

Fig. 2 is the scheme of installation of the analog conducting wire of the present embodiment, pulling force sensor and binocular camera;

Fig. 3 is the basic schematic diagram of vision measurement of the present embodiment.

Embodiment

As shown in Figure 1-2, a kind of remote supervision system of powerline ice-covering process, comprise analog conducting wire 1, the pulling force sensor 2 be connected with analog conducting wire 1, binocular camera 3 and control box, wherein, the data acquisition module 4 be connected with binocular camera 3 is provided with in described control box, the control module 5 be connected with pulling force sensor 2, the communication module 6 be connected with middle control end 7, and be the power module that in control box, modules and binocular camera 3 are powered, described binocular camera camera lens 3 is positioned in same level measuring surface with analog conducting wire 1, both distances are 600MM-1200MM, above-mentioned analog conducting wire 1 is same material and same model with actual track wire.In the present embodiment, described pulling force sensor 2 is arranged in cylinder type measurement component, and analog conducting wire 1 is located in the web member 8 of cylinder type measurement component, and the measuring head of pulling force sensor 2 is positioned at web member 8, for the gravity of measure analog wire 1.

Control module 5 in described control box adopts Microprocessor S3C44B0X, be specifically as follows the embedded system that Intel atom processor controls, after relevant image data is sent to control module 5 by data acquisition module 4 and pulling force sensor 2, after initial analysis process, be sent to middle control end 7 by communication module 6 by microprocessor.

Described binocular camera 3 inside is provided with DSP processing unit, for image procossing, described DSP processing unit by optical system measuring to analog conducting wire 1 on icing image transform can the composite vision signal format of standard, be transported to the data acquisition module 4 of control box by signal cable.

Described power module is connected with solar electric power supply system, adopts photovoltaic power supply, ensures the continued power of control box and binocular camera 3.

Adopt the monitoring method of the remote supervision system of above-mentioned powerline ice-covering process, comprise the steps:

Step 1, gravity auxiliary monitoring step:

Step 1-1: the transducing signal of detection sends to middle control to hold by control box by pulling force sensor, middle control end first calculates the load w1 (N/m) under analog conducting wire normality according to formula (1),

Under normal conditions analog conducting wire from gravitational unit load w1 (N/m) for unit linear mass q and gravity acceleration g long-pending.

w ₁＝qg≈qg _n＝9.80665q (1)

In formula, gn is normal acceleration of gravity, gn=9.80665 (m/s2).

Step 1-2: according to formula (2), formula (3) calculates analog conducting wire icing time load:

Suppose that the icing of all kinds and different section profile is all converted as density is the circular glaze section of 0.9g/cm3.When oneself knows wire diameter D (mm) and ice covering thickness b (mm), its unit length ice load w2 (N/m) is

$w_2=\frac{0.9\mathrm{\π}{g}_n}{4}[{(D+2b)}^2-D^2]\×{10}^{-3}=\mathrm{\κb}(b+D)---\left(2\right)$

κ=0.027728 in formula

During wire icing, vertical total load (TL) w3 is wire gravity load w1 and ice load w2 sum, namely

w ₃＝w ₁+w ₂(3)

Step 1-3: the ice covering thickness calculating analog conducting wire according to formula (4):

$b=\frac{-\mathrm{\κD}\±\sqrt{{\left(\mathrm{\κD}\right)}^2+4\×\mathrm{\κ}\×(w_3-w_1)}}{2\×\mathrm{\κ}}---\left(4\right)$

Step 2, binocular measuring process:

Step 2-1: the binocular stereo camera demarcated gathers the icing image (left image and right image) on same environment Imitating wire respectively, and sends to middle control to hold by control box 2 icing images;

Step 2-2: middle control end extracts this node visual signature within the vision according to vision node disjoint each in icing image;

Described Visual Feature Retrieval Process concrete grammar is:

1) each vision node adopts a binocular stereo camera demarcated, and contains left and right two video cameras, can the left and right stereo-picture of Real-time Obtaining; Then adopt PCA-SIFT method, extract the SIFT feature descriptor of left and right image respectively, each feature descriptor is 36 n dimensional vector ns, and each n dimensional vector n represents with 8;

2) SIFT feature adopting KD-Tree method to realize between left and right image is mated, and the Matching power flow of record often pair of SIFT feature coupling;

3) epipolar line restriction filtering Mismatching point is adopted; For the binocular stereo camera demarcated, epipolar line restriction requires that the vertical coordinate difference of correct matching double points in left and right image is less than 1;

4) with the benchmark image that left image is Visual Feature Retrieval Process, left image is divided into 20 × 20 equal-sized regions, only choose the visual signature of SIFT feature as this region that has maximum Matching power flow in each area, namely from left image, 400 visual signatures can be extracted at most, and using the visual signature of these visual signatures as node present frame; Therefore the descriptor maximum amount of data that each vision node extracts visual signature from each frame stereo image pair is 400 × 36 × 8/1024=112.5kb;

Step 2-3: middle control end according in vision node calculate step 2-2 each in icing image extract the three-dimensional coordinate of visual signature, and caching process.

Described visual signature three-dimensional coordinate calculates concrete grammar:

In step 2-2 extract the vertical coordinate difference of visual signature in left and right image and be less than 1, and horizontal coordinate difference is called parallax d, according to binocular stereo vision three-dimensional reconstruction formula, can obtain the three-dimensional coordinate (X of visual signature i under left camera coordinate system ⁱ, Y ⁱ, Z ⁱ):

$\left\{\begin{array}{c}{X}^i=b(u_i-u_0)/d_i\\ Y^i=b(v_0-v_i)/d_i\\ Z^i=\mathrm{bf}/d_i\end{array}\right.---\left(5\right)$

Wherein b, f, (u0, v0) is respectively the baseline, focal length and the left optical picture centre coordinate that obtain in binocular solid camera calibration.(ui, vi) for the coordinate of visual signature i in left image, di be the parallax of visual signature i.

Step 2-4: middle control end according to the locations of contours impact point of analog conducting wire, the scope of ice covering thickness on calculating simulation wire.

On described calculating simulation wire, the concrete grammar of ice covering thickness scope is:

(1) three-dimensional coordinate of each point on analog conducting wire profile is extracted: according to the same place on the icing image of left and right, carry out intersection and obtain spatial point three-dimensional coordinate; As shown in Figure 3,

(2) according to the imaging equation (collinearity equation) of central projection, the principle of three point on a straight line is adopted to build measuring and calculation formula:

$x_l=x-x_0=-f\frac{a_1(X_A-X_S)+b_1(Y_A-Y_S)+c_1(Z_A-Z_S)}{a_3(X_A-X_S)+b_3(Y_A-Y_S)+c_3(Z_A-Z_S)}$

$y_l=y-y_0=-f\frac{a_2(X_A-X_S)+b_2(Y_A-Y_S)+c_2(Z_A-Z_S)}{a_3(X_A-X_S)+b_3(Y_A-Y_S)+c_3(Z_A-Z_S)}$

$r=\frac{\sqrt[2]{x_l^2+y_l^2}-d_1}{2}$

Wherein r is ice covering thickness, d ₁for known analog wire radius;

(3) according to the value calculating the ice covering thickness selected, take out maximum effective value and minimum effective value, and calculate the mean value selecting ice covering thickness.

Step 3, middle control end calculate the ice covering thickness mean value of analog conducting wire respectively by step 1 and step 2, when both ice covering thickness differences are within 2 millimeters, and in the valid value range of ice covering thickness, then think that monitoring result is error free, the ice covering thickness of the analog conducting wire that direct output step 2 obtains, when both ice covering thickness differences are greater than 2 millimeters, then resampling, repeat step 1 and step 2, until detect thickness difference is within the scope of permissible error.

Step 4: simultaneously, user can be according to the actual requirements, icing image clicks the ice covering thickness value of any 2 calculating current measurement point, 2 impact points selected in icing image according to user are held in middle control, navigate to three-dimensional coordinate corresponding in three dimensions, realize three-dimensional measurement with cognitive.

Above-described embodiment and graphic and non-limiting product form of the present invention and style, any person of an ordinary skill in the technical field, to its suitable change done or modification, all should be considered as not departing from patent category of the present invention.

Claims (10)

1. the remote supervision system of a powerline ice-covering process, it is characterized in that: comprise analog conducting wire, the pulling force sensor be connected with analog conducting wire, binocular camera and control box, wherein, the data acquisition module be connected with binocular camera is provided with in described control box, the control module be connected with pulling force sensor, the communication module of holding and being connected is controlled with middle, and be the power module that in control box, modules and binocular camera are powered, described binocular camera camera lens and analog conducting wire are positioned in same level measuring surface, above-mentioned analog conducting wire and actual track wire are same material and same model.

2. the remote supervision system of a kind of powerline ice-covering process as claimed in claim 1, it is characterized in that: described pulling force sensor is arranged in cylinder type measurement component, analog conducting wire is located in the web member of cylinder type measurement component, the measuring head of pulling force sensor is positioned at web member, for the gravity of measure analog wire.

3. the remote supervision system of a kind of powerline ice-covering process as claimed in claim 1, it is characterized in that: the control module in described control box adopts Microprocessor S3C44B0X, after relevant image data is sent to control module by data acquisition module and pulling force sensor, after initial analysis process, be sent to middle control by communication module by microprocessor and hold.

4. the remote supervision system of a kind of powerline ice-covering process as claimed in claim 3, is characterized in that: described microprocessor Intel atom processor.

5. the remote supervision system of a kind of powerline ice-covering process as claimed in claim 1, it is characterized in that: described binocular camera inside is provided with DSP processing unit, described DSP processing unit by optical system measuring to analog conducting wire on icing image transform can the composite vision signal format of standard, be transported to the data acquisition module of control box by signal cable.

6. the remote supervision system of a kind of powerline ice-covering process as claimed in claim 1, is characterized in that: described power module is connected with solar electric power supply system, adopts photovoltaic power supply.

7. a remote monitoring method for powerline ice-covering process, is characterized in that comprising the steps:

Step 1, gravity auxiliary monitoring step:

Step 1-1: the transducing signal of detection sends to middle control to hold by control box by pulling force sensor, middle control end first calculates the load w1 (N/m) under analog conducting wire normality according to formula (1),

w ¹＝qg≈qg _n＝9.80665q (1)

In formula, gn is normal acceleration of gravity, gn=9.80665 (m/s2);

Step 1-2: according to formula (2), formula (3) calculates analog conducting wire icing time load

Suppose that the icing of all kinds and different section profile is all converted as density is the circular glaze section of 0.9g/cm3, when oneself knows wire diameter D (mm) and ice covering thickness b (mm), its unit length ice load w2 (N/m) is

$w_2=\frac{0.9\mathrm{\π}{g}_n}{4}[{(D+2b)}^2-D^2]\×{10}^{-3}=\mathrm{\κb}(b+D)---\left(2\right)$

κ=0.027728 in formula

During wire icing, vertical total load (TL) w3 is wire gravity load w1 and ice load w2 sum, namely

w ³＝w ¹+w ²(3)

Step 1-3: the ice covering thickness b (mm) calculating analog conducting wire according to formula (4):

$b=\frac{-\mathrm{\κD}\±\sqrt{{\left(\mathrm{\κD}\right)}^2+4\×\mathrm{\κ}\×(w_3-w_1)}}{2\×\mathrm{\κ}}---\left(4\right)$

Step 2, binocular measuring process:

Step 2-1: the binocular camera demarcated gathers the icing image (left image and right image) on same environment Imitating wire respectively, and sends to middle control to hold by control box 2 icing images;

Step 2-2: middle control end extracts this node visual signature within the vision according to vision node disjoint each in icing image;

Step 2-3: middle control end according in vision node calculate step 2-2 each in icing image extract the three-dimensional coordinate of visual signature, and caching process;

Step 2-4: middle control end according to the locations of contours impact point of analog conducting wire, the scope of ice covering thickness on calculating simulation wire, on described calculating simulation wire, the concrete grammar of ice covering thickness scope is:

Extract the three-dimensional coordinate of each point on analog conducting wire profile: according to the same place on the icing image of left and right, carry out intersection and obtain spatial point three-dimensional coordinate;

According to the imaging equation (collinearity equation) of central projection, the principle of three point on a straight line is adopted to build measuring and calculation formula:

$x_l=x-x_0=-f\frac{a_1(X_A-X_S)+b_1(Y_A-Y_S)+c_1(Z_A-Z_S)}{a_3(X_A-X_S)+b_3(Y_A-Y_S)+c_3(Z_A-Z_S)}$

$y_l=y-y_0=-f\frac{a_2(X_A-X_S)+b_2(Y_A-Y_S)+c_2(Z_A-Z_S)}{a_3(X_A-X_S)+b_3(Y_A-Y_S)+c_3(Z_A-Z_S)}$

$r=\frac{\sqrt[2]{x_l^2+y_l^2}-d_1}{2}$

Wherein r is ice covering thickness, d ₁for known analog wire radius;

According to the value calculating the ice covering thickness selected, take out maximum effective value and minimum effective value, and calculate the mean value of ice covering thickness.

Step 3, middle control end calculate the ice covering thickness mean value of analog conducting wire respectively by step 1 and step 2, when both ice covering thickness differences are in setting range, and in the valid value range of ice covering thickness, then think that monitoring result is error free, directly export the ice covering thickness of the analog conducting wire that step 2 obtains; When both ice covering thickness differences are greater than setting value, then resampling, repeats step 1 and step 2, until detect thickness difference is within the scope of permissible error.

8. the remote monitoring method of a kind of powerline ice-covering process as claimed in claim 7, is characterized in that: user according to the actual requirements, icing image can click the ice covering thickness value of any 2 calculating current measurement point.

9. the remote monitoring method of a kind of powerline ice-covering process as claimed in claim 7, is characterized in that: the Visual Feature Retrieval Process concrete grammar described in step 2-2 is:

1) each vision node binocular solid camera of adopting one to demarcate, contains left and right two cameras, can the left and right stereo-picture of Real-time Obtaining; Then adopt PCA-SIFT method, extract the SIFT feature descriptor of left and right image respectively, each feature descriptor is 36 n dimensional vector ns, and each n dimensional vector n represents with 8;

2) SIFT feature adopting KD-Tree method to realize between left and right image is mated, and the Matching power flow of record often pair of SIFT feature coupling;

3) epipolar line restriction filtering Mismatching point is adopted; For the binocular solid camera demarcated, epipolar line restriction requires that the vertical coordinate difference of correct matching double points in left and right image is less than 1;

4) with the benchmark image that left image is Visual Feature Retrieval Process, left image is divided into 20 × 20 equal-sized regions, only choose the visual signature of SIFT feature as this region that has maximum Matching power flow in each area, namely from left image, 400 visual signatures can be extracted at most, and using the visual signature of these visual signatures as node present frame; Therefore the descriptor maximum amount of data that each vision node extracts visual signature from each frame stereo image pair is 400 × 36 × 8/1024=112.5kb.

10. the remote monitoring method of a kind of powerline ice-covering process as claimed in claim 7, is characterized in that: the visual signature three-dimensional coordinate described in step 2-3 calculates concrete grammar and is:

In step 2-2 extract the vertical coordinate difference of visual signature in left and right image and be less than 1, and horizontal coordinate difference is called parallax d, according to binocular stereo vision three-dimensional reconstruction formula, can obtain the three-dimensional coordinate (X of visual signature i under left camera coordinate system ⁱ, Y ⁱ, Z ⁱ):

$\left\{\begin{array}{c}{X}^i=b(u_i-u_0)/d_i\\ Y^i=b(v_0-v_i)/d_i\\ Z^i=\mathrm{bf}/d_i\end{array}\right.---\left(5\right)$

Wherein b, f, (u0, v0) is respectively the baseline, focal length and the left optical picture centre coordinate that obtain in binocular solid camera calibration.(ui, vi) for the coordinate of visual signature i in left image, di be the parallax of visual signature i.