CN106705857A - Automatic monitoring system of laser surface displacement - Google Patents

Abstract

The invention discloses a laser surface displacement automatic monitoring system, and relates to a laser surface displacement automatic monitoring system, which is a monitoring device that meets the needs of geotechnical engineering information construction, including the following steps: Step 1, the basic technical principle of the monitoring system, step 2, the demarcation of monitoring system, step 3, evaluation displacement measurement accuracy, step 4, the design of measuring device; The purpose of the present invention is to obtain the installation laser by obtaining the variation of the precise coordinates of the light spot projected on the target plate by the laser spot The surface displacement change solves the problem of surface displacement monitoring with a large amount and a wide area in the monitoring of foundation pit engineering and tunnel engineering.

Description

A laser surface displacement automatic monitoring system

technical field

The invention belongs to the technical field of automatic monitoring systems, and relates to a laser surface displacement automatic monitoring system, which is suitable for the monitoring of large-scale and wide-ranging surface displacements in the monitoring of foundation pit engineering and tunnel engineering.

Background technique

At present, the monitoring of geotechnical engineering is mainly based on manual measurement methods. Although there are automatic monitoring instruments such as static levels, the market price is too high, which makes it impossible to widely promote and use them in geotechnical engineering practice. If the monitoring equipment can be developed at a moderate price and can meet the needs of geotechnical engineering information construction, its market prospect is huge.

Therefore, the overall consideration is to use the embedded system to develop a sensor system that can intelligently read the displacement value online. For reading the coordinates of the laser point on the target plate, after investigation, there are several technical approaches, one is to use PSD or CCD photosensitive chip to directly read the coordinates of the photosensitive point, this method is relatively expensive. Another method is to use precision processing to form an array of optical fibers, and then use array scanning to read the coordinates of the photosensitive points. This method is more difficult to process. The last method is to directly use the camera to take pictures of the photosensitive target plate, and use the photogrammetry method to read the coordinates. The last method has lower cost and will be easy to realize. It is planned to use this method to develop the automatic monitoring system.

Contents of the invention

The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a laser surface displacement automatic monitoring system. The purpose of the present invention is to obtain The surface displacement change of the installed laser solves the problem of large-scale and wide-ranging surface displacement monitoring in foundation pit engineering and tunnel engineering monitoring.

In order to solve the above technical problems, the present invention provides a laser surface displacement automatic monitoring system, comprising the following steps:

Step (1), the basic technical principle of the monitoring system;

Using the collimation of the laser, when the laser at point A shoots to point B, there is a spot on the target board at point B. If point B (reference point) is stationary, when point A is displaced, the spot on the target board The spot also moves accordingly. Accurately measure the coordinates of the light spot on the target plate to obtain the displacement of point A. And the displacement value can be obtained in two directions of horizontal displacement and vertical displacement.

Turn off the laser, get the RGB color image photo of the target plate without light spots, turn on the laser, and get the RGB color image photo of the target plate with light spots, the camera pixel used is 2592×1944.

The two RGB color image photos of the target plate were processed in grayscale respectively.

Subtract the processed grayscale images. Since the two images are only different in the bright spots, there will be obvious grayscale differences after the subtraction operation.

The image binarization process is performed on this grayscale image photo, and the binarization threshold is 80% of the grayscale value of the brightest point to obtain a binarized image.

Under normal circumstances, the bright spots of the round spot hit by the laser are irregular. In order to accurately hit the coordinate position of the light spot, the binary image is corroded using the corrosion algorithm. In order to speed up the corrosion algorithm, according to the prior knowledge of the experiment, it can be known that , the pixel of the bright spot of the laser is not greater than 301×301, and the image is intercepted; the pixel size of the binary image after interception is not greater than 301×301.

At this time, use the corrosion algorithm to corrode the binarized image, and select the corrosion matrix as shown in the formula:

After each erosion, if the image is not all 0, the previous image will be covered; if the image is all 0, the corrosion will be stopped without covering the original image.

The pixel coordinates of the bright spots in the image are averaged to obtain the coordinates of the laser hitting the target.

The coordinates of the laser rays on the target plate are obtained many times, and the surface displacement of the structure on which the laser is installed can be measured.

Step (2), calibration of the monitoring system;

The calibration of the monitoring system includes two parts, the first is the perspective calibration of the target plate, and the second is the calibration of pixels and distances.

(1) Perspective calibration of the target plate

The target board and the measurement camera are installed on the measurement box. Due to the process, the target board and the camera shooting surface cannot be completely parallel, so the measurement device needs to be calibrated in perspective after the production is completed.

In the process of making the target board, a layer of graph paper with a standard distance has been pasted with a spacing of 10mm. After getting the target board photo, the perspective transformation coefficient can be obtained by using the four-point perspective method. The perspective transformation process is as follows:

(a), select the four vertices of any square on the target plate photo, which are upper left, upper right, lower left, and lower right in turn, and obtain its pixel coordinates (xZS, yZS), (xYS, yYS), (xZX, yZX), (xYX, yYX);

(b), obtain new rectangle width and height from original quadrilateral according to formula;

Then the transformed coordinates of the four vertices are as follows:

(c), let B=[X(1), Y(1), X(2), Y(2), X(3), Y(3), X(4), Y(4)] ^T , Solve the matrix equation shown as:

make

The perspective transformation coefficient is obtained as shown in the formula:

Xi＝inv(A)·B \*MERGEFORMAT (6)

Among them, Xi is an 8 × 1-order matrix, so that the transformation coefficients are obtained as shown in the formula:

(d) Assuming that the coordinates of the light point obtained by the erosion algorithm are (x ₀ , y ₀ ), it is necessary to perform perspective transformation according to the formula to obtain the final coordinates.

At this point, the perspective transformation is completed.

(2) Calibration of pixel distance

After the perspective calibration of the target board is completed, it is necessary to obtain the corresponding relationship between the pixel distance and the actual distance. According to the standard line on the target board, select the two point coordinates (x _1o , y _1o ), (x _2o , y _2o ) , its distance is known as L, and its coordinates are transformed into (x ₁ , y ₁ ), (x ₂ , y ₂ ) by using the transformation coefficient, then the distance calibration coefficient is shown in the formula:

According to this coefficient, the moving distance of the two laser light pairs can be obtained.

Step (3), evaluating displacement measurement accuracy;

The shooting pixels of the measurement camera are 2592×1944, and the size of the target plate is 133mm×100mm. The ratio of the two is as follows:

Therefore, theoretically, the pixel distance between each two points is 100/1944<0.052mm. Considering the influence of spot center of gravity coordinate search and installation error, the displacement measurement accuracy of the measuring device needs to be determined in the actual stability test.

Step (4), the design of measuring device;

(1), the design of the measuring box

The material of the measurement box is 6061-T6 alloy aluminum, which is heat-treated and strengthenable alloy. It has good formability, weldability, machinability, and medium strength. It can still maintain good operability after annealing. Low density, light weight, anti-corrosion and easy to install.

The box body adopts an assembled structure, and the main body is composed of six plates.

An adjustable bracket is installed inside the box, which is designed with a card slot for installing the measurement camera. Both ends of the box body are designed with slots for inserting the shooting target board.

(2) Design of measurement target board

The material of the measurement target plate is made of common organic plastic plate on the market, the thickness of the plate is 2mm, it is white, light-transmitting and slightly transparent. The organic plastic board has excellent toughness and dimensional stability, reliable insulation performance, good heat resistance, acid and alkali resistance, chemical corrosion resistance, easy processing, non-toxic, environmentally friendly and durable. A layer of plastic with a standard distance grid can be pasted on it for calibration of the measuring system.

(3) Design of measurement hardware

The measurement hardware uses a Raspberry Pi card computer type B Rev1, which is only the size of a credit card and has functions such as video, network, and IO.

The running time of each measurement on the Raspberry Pi platform is about 5 seconds, indicating that the measurement frequency can reach 0.2Hz. In the actual measurement of the surface displacement of foundation pits, tunnels, etc., because the surface displacement changes very slowly, the measurement frequency Fully meet the measurement requirements.

The beneficial effect of the present invention is: the method designed by the present invention realizes the surface displacement automatic monitoring system suitable for large-scale and wide-area monitoring in foundation pit engineering and tunnel engineering, reduces the monitoring cost, and can automatically obtain surface displacement information without being on duty It is a major progress in the field of foundation pit engineering and tunnel engineering monitoring technology.

Description of drawings

Fig. 1 is a principle flow chart of the present invention;

Figure 2 is a schematic diagram of the basic technical principle of the monitoring system;

Figure 3(a) is a photo of the RGB color image of the target plate obtained by turning off the laser;

Fig. 3 (b) is the RGB color image photo of the target plate obtained by turning on the laser;

Fig. 4 (a) is the photo of the grayscale image of the target plate obtained by turning off the laser;

Fig. 4 (b) is the photo of the grayscale image of the target plate obtained by turning on the laser;

Fig. 5 is the photo of the grayscale image of the target plate after the subtraction algorithm;

Fig. 6 is the binary image processing result;

Fig. 7 is the flow chart of taking the steps of spot range;

Fig. 8 is the binarized image of the spot area after interception;

Figure 9 is a picture of the corrosion algorithm process;

Figure 10 (a) is a parallel photo of the ideal target plate and the camera shooting surface;

Fig. 10(b) is a photo that the actual target plate is not parallel to the photographing surface of the camera;

Figure 11 is an assembly diagram of the measurement box;

Figure 12 shows the design of the slot for the adjustable camera bracket of the measurement box;

Fig. 13 is a block diagram of the measurement hardware structure;

Figure 14 is a block diagram of the automatic operation of the measurement hardware;

Figure 15 is the monitoring system flow;

Fig. 16 (a) is the point diagram of stability test result;

Figure 16(b) is a histogram of the stability test results;

Figure 17 is a schematic diagram of the measurement results of the square path of the displacement platform.

detailed description

Example 1

This embodiment provides an automatic monitoring system for laser surface displacement, the principle of which is shown in Figure 1, including the following steps:

Step (1), the basic technical principle of the monitoring system;

As shown in Figure 2, using the collimation of the laser, when the laser at point A is directed at point B, there is a spot on the target plate at point B. If point B (reference point) is stationary, when point A is displaced , the light spot on the target plate also moves accordingly. Accurately measure the coordinates of the light spot on the target plate to obtain the displacement of point A. And the displacement value can be obtained in two directions of horizontal displacement and vertical displacement.

Turn off the laser and obtain the RGB color image photo of the target plate without light spots, as shown in Figure 3(a), turn on the laser, and obtain the RGB color image photo of the target plate with light spots, as shown in Figure 3(b), the camera used The pixels are 2592×1944.

The two RGB color image photos of the target plate were processed in grayscale respectively, as shown in Fig. 4(a) and Fig. 4(b).

Subtract the processed grayscale images. Since the two images are only different in the bright spots, after the subtraction operation, there will be obvious grayscale differences, as shown in Figure 5.

Image binarization is performed on this grayscale image photo, and the binarization threshold is 80% of the grayscale value of the brightest point. The obtained binarized image is shown in Figure 6.

Under normal circumstances, the bright spots of the round spot hit by the laser are irregular. In order to accurately hit the coordinate position of the light spot, the binary image is corroded using the corrosion algorithm. In order to speed up the corrosion algorithm, according to the prior knowledge of the experiment, it can be known that , the pixel of the bright spot of the laser is not larger than 301×301, and the image can be intercepted according to the method shown in Figure 7.

The pixel size of the intercepted binarized image is not larger than 301×301, as shown in Figure 8.

At this time, use the corrosion algorithm to corrode the binarized image, and select the corrosion matrix as shown in the formula:

After each corrosion, if the image is not all 0, then cover the previous image, if the image is all 0, then stop the corrosion, do not cover the original image, as shown in Figure 9.

The pixel coordinates of the bright spots in the image are averaged to obtain the coordinates of the laser hitting the target.

The coordinates of the laser rays on the target plate are obtained many times, and the surface displacement of the structure on which the laser is installed can be measured.

Step (2), calibration of the monitoring system;

The calibration of the monitoring system includes two parts, the first is the perspective calibration of the target plate, and the second is the calibration of pixels and distances.

(1) Perspective calibration of the target plate

The target board and the measurement camera are installed on the measurement box. Due to the process, the target board and the camera shooting surface cannot be completely parallel, as shown in Figure 10. Therefore, the measurement device needs to be calibrated in perspective after the production is completed.

In the process of making the target board, a layer of graph paper with a standard distance has been pasted with a spacing of 10mm. After getting the target board photo, the perspective transformation coefficient can be obtained by using the four-point perspective method. The perspective transformation process is as follows:

(a), select the four vertices of any square on the target plate photo, which are upper left, upper right, lower left, and lower right in turn, and obtain its pixel coordinates (xZS, yZS), (xYS, yYS), (xZX, yZX), (xYX, yYX);

(b), obtain new rectangle width and height from original quadrilateral according to formula;

Then the transformed coordinates of the four vertices are as follows:

(c), let B=[X(1), Y(1), X(2), Y(2), X(3), Y(3), X(4), Y(4)] ^T , Solve the matrix equation shown as:

make

The perspective transformation coefficient is obtained as shown in the formula:

Xi＝inv(A)·B \*MERGEFORMAT (6)

Among them, Xi is an 8 × 1-order matrix, so that the transformation coefficients are obtained as shown in the formula:

(d) Assuming that the coordinates of the light point obtained by the erosion algorithm are (x ₀ , y ₀ ), it is necessary to perform perspective transformation according to the formula to obtain the final coordinates.

At this point, the perspective transformation is completed.

(2) Calibration of pixel distance

After the perspective calibration of the target board is completed, it is necessary to obtain the corresponding relationship between the pixel distance and the actual distance. According to the standard line on the target board, select the two point coordinates (x _1o , y _1o ), (x _2o , y _2o ) , its distance is known as L, and its coordinates are transformed into (x ₁ , y ₁ ), (x ₂ , y ₂ ) by using the transformation coefficient, then the distance calibration coefficient is shown in the formula:

According to this coefficient, the moving distance of the two laser light pairs can be obtained.

Step (3), evaluating displacement measurement accuracy;

The shooting pixels of the measurement camera are 2592×1944, and the size of the target plate is 133mm×100mm. The ratio of the two is as follows:

Therefore, theoretically, the pixel distance between each two points is 100/1944<0.052mm. Considering the influence of spot center of gravity coordinate search and installation error, the displacement measurement accuracy of the measuring device needs to be determined in the actual stability test.

Step (4), the design of measuring device;

(1), the design of the measuring box

The material of the measurement box is 6061-T6 alloy aluminum, which is heat-treated and strengthenable alloy. It has good formability, weldability, machinability, and medium strength. It can still maintain good operability after annealing. Small density, light weight, anti-corrosion and easy to install.

The box body adopts an assembled structure, and the main body is composed of six plates, as shown in Figure 11.

An adjustable bracket is installed inside the box, which is designed with a card slot for installing the measurement camera. Both ends of the box are designed with slots for inserting the shooting target board; as shown in Figure 12.

(2) Design of measurement target board

The material of the measurement target plate is made of common organic plastic plate on the market, the thickness of the plate is 2mm, white, slightly transparent. The organic plastic board has excellent toughness and dimensional stability, reliable insulation performance, good heat resistance, acid and alkali resistance, chemical corrosion resistance, easy processing, non-toxic, environmentally friendly and durable. A layer of plastic with a standard distance grid can be pasted on it for calibration of the measuring system.

(3) Design of measurement hardware

The measurement hardware uses Raspberry Pi card type computer B-type Rev1, which is only the size of a credit card and has functions such as video, network, and IO. The hardware block diagram is shown in Figure 13.

As shown in Figure 14, the running time for each measurement on the Raspberry Pi platform is about 5 seconds, indicating that the measurement frequency can reach 0.2 Hz. Very slowly, this measurement frequency fully meets the measurement requirements.

The technical solution of the present invention will be described in detail below in conjunction with accompanying drawings 15-17:

The process is shown in Figure 15. Firstly, the measurement box and the target board should be produced, and then the written measurement program should be burned on the Raspberry Pi, the distance between the camera and the target board should be determined, the measurement platform should be built, and then the initial photos should be taken. The perspective transformation is used to solve the calibration factor, determine the relationship between the pixel distance and the experimental distance, and finally conduct a stability test to obtain the actual measurement accuracy.

The implementation process of coordinate extraction of laser pixel points in the technical solution of the present invention will be described below through specific calculation examples.

After the measurement box is installed, the perspective transformation calibration can be carried out, the transformation factor can be calculated, and the relationship between the pixel distance and the actual distance can be determined. Randomly take four vertices of a square, and finally calculate the perspective transformation factor as:

a=0.96142384

b=0.00401253

c=168.45083955

d=-0.01557080

e=0.98860270

f=-149.21030551

g=-0.00001872

h=0.00000839

At the same time, the actual distance between every two adjacent pixels is 0.04755435mm

Importing these parameters and measurement programs into the measurement hardware completes the preparation of the measurement hardware and software.

During the installation process of the laser surface displacement automatic monitoring system, the installation conditions are limited as follows:

The distance between the laser and the measuring box should not exceed 10 meters. If the distance is too far, it will be difficult to adjust the laser light spot to the target plate. The laser light spot will fly out of the target plate if there is a slight twist.

The measurement box needs to remain stationary, the target plate surface is parallel to the displacement surface, and the laser light is perpendicular to the displacement surface and the target plate surface at the same time.

The laser is installed on the precision displacement platform and can move in three directions with the displacement platform. For this automatic monitoring system, it only needs to move on the surface perpendicular to the horizontal plane. The displacement precision of the displacement platform is 1 μm. When the laser is turned off, the monitoring system will take a picture, then control the laser to turn on, and then the measurement system will process the two pictures in grayscale, binarize, subtract and then intercept the area near the brightest point for corrosion processing , until it cannot be corroded. The average coordinates of the remaining bright spots are considered to be the pixel coordinates on the laser target board.

In the stability test of the measurement system, the position of the laser was kept unchanged, and 1000 consecutive measurements were made. The results obtained are shown in Figure 16. In Figure 16(a), it can be seen that the hormone bright spots are concentrated in 6 pixels. within the square. Taking the average value of this pixel point as the center, the pixel distance distribution probability of each measurement result from this center point is shown in Figure 16(b). It can be seen that the probability that the distance between each pixel point of a pixel and the center is less than 3 pixels is 99.8%, the probability of less than 4 pixels is 100%, that is, it can be considered that the displacement measurement accuracy of this monitoring system is 4 pixels, combined with the actual distance between two adjacent pixels in the calibration result, the automatic monitoring system can be obtained The measurement accuracy is 4×0.04755435<0.2mm.

In the actual measurement experiment, the displacement platform is controlled to move in a square path, each displacement is 1mm, and the final measurement result is shown in Figure 17.

The above embodiments are only to illustrate the technical ideas of the present invention, and cannot limit the protection scope of the present invention with this. All technical ideas proposed according to the present invention, any changes made on the basis of technical solutions, all fall within the protection scope of the present invention. Inside.

Claims (9)

1. a kind of laser surface displacement automatic monitoring system, it is characterised in that including step in detail below：

Step (1), the basic technique principle of monitoring system；

Using the collimation of laser, when A dot laser directive B points, a hot spot is there is on the target plate of B points, if B points are base Transfixion on schedule, when A points are subjected to displacement, the hot spot on target plate is also accordingly moved；Luminous point on accurate measurement target plate Coordinate, you can obtain the size of A point displacements；And shift value can be obtained in horizontal displacement and vertical displacement both direction；

Laser is closed, the target plate RGB color image photo without luminous point is obtained, laser is opened, the target plate with luminous point is obtained RGB color image photo；

Two target plate RGB color image photos are carried out into gray processing treatment respectively；

Gray scale picture after treatment is carried out into phase reducing, the image only having due to two images at bright spot is different, is subtracting each other behaviour After work, it may appear that obvious gray difference；

Image binaryzation treatment is carried out to this gray level image photo, the binary image for obtaining；

Now, binary image is corroded using erosion algorithm, selection corrosion matrix is as shown in formula：

$\begin{array}{cc}b=\left[\begin{array}{ccc}0& 1& 0\\ 1& 1& 1\\ 0& 1& 0\end{array}\right]& \backslash *MERGEFORMAT\end{array}---\left(1\right)$

Bright spot pixel coordinate in image is averaged, that is, obtains the coordinate that laser is beaten on target plate；

Repeatedly obtain coordinate of the laser radiation on target plate, you can measurement obtains installing the body structure surface displacement of laser；

The demarcation of step (2), monitoring system；

The demarcation of monitoring system includes：The perspective of target plate is demarcated, the demarcation of pixel and distance；

(1) perspective of target plate is demarcated

Target plate and measurement camera are arranged on measuring box, and perspective demarcation is carried out after measurement apparatus complete；

Make target plate during, paste one layer of squared paper of gauged distance, after target plate photo is obtained, can be used 4 points it is saturating Perspective transform coefficient is obtained depending on method, the process of perspective transform is as follows：

(a), any one foursquare four summit on target plate photo is chosen, be successively upper left, upper right, lower-left, bottom right, acquisition Its pixel coordinate (xZS, yZS), (xYS, yYS), (xZX, yZX), (xYX, yYX)；

(b), obtain that new rectangle is wide and height from former quadrangle according to formula:

$\begin{array}{cc}\begin{array}{c}w=round\left(\sqrt{{\left(xZS-xYS\right)}^2+{\left(yZS-yYS\right)}^2}\right)\\ h=round\left(\sqrt{{\left(xZS-xZX\right)}^2+{\left(yZS-yZX\right)}^2}\right)\end{array}& \backslash *MERGEFORMAT\end{array}---\left(2\right)$

Four apex coordinates after then converting are as shown in formula：

$\begin{array}{cc}\begin{array}{c}X=\left(xZS,xZS+w,xZS,xZS+w\right)\\ Y=\left(yZS,yZS,yZS+h,yZS+h\right)\end{array}& \backslash *MERGEFORMAT\end{array}---\left(3\right)$

(c), make B=[X (1), Y (1), X (2), Y (2), X (3), Y (3), X (4), Y (4)]^T, matrix equation of the solution as shown in formula：

$\begin{array}{c}\left[\begin{array}{cccccccc}xZS& yZS& 1& 0& 0& 0& -X\left(1\right)*xZS& -X\left(1\right)*yZS\\ 0& 0& 0& xZS& yZS& 1& -Y\left(1\right)*xZS& -Y\left(1\right)*yZS\\ xZS& yZS& 1& 0& 0& 0& -X\left(2\right)*xYS& -X\left(2\right)*yYS\\ 0& 0& 0& xYS& yYS& 1& -Y\left(2\right)*xYS& -Y\left(2\right)*yYS\\ xZX& yZX& 1& 0& 0& 0& -X\left(3\right)*xZX& -X\left(3\right)*yZX\\ 0& 0& 0& xZX& yZX& 1& -Y\left(3\right)*xZX& -Y\left(3\right)*yZX\\ xYX& yYZ& 1& 0& 0& 0& -X\left(4\right)*xYX& -X\left(4\right)*yYX\\ 0& 0& 0& xYX& yYX& 1& -Y\left(4\right)*xYX& -Y\left(4\right)*yYX\end{array}\right]\&CenterDot;Xi=B\\ \backslash *MERGEFORMAT\end{array}---\left(4\right)$

Order

$\begin{array}{c}A=\left[\begin{array}{cccccccc}xZS& yZS& 1& 0& 0& 0& -X\left(1\right)*xZS& -X\left(1\right)*yZS\\ 0& 0& 0& xZS& yZS& 1& -Y\left(1\right)*xZS& -Y\left(1\right)*yZS\\ xZS& yZS& 1& 0& 0& 0& -X\left(2\right)*xYS& -X\left(2\right)*yYS\\ 0& 0& 0& xYS& yYS& 1& -Y\left(2\right)*xYS& -Y\left(2\right)*yYS\\ xZX& yZX& 1& 0& 0& 0& -X\left(3\right)*xZX& -X\left(3\right)*yZX\\ 0& 0& 0& xZX& yZX& 1& -Y\left(3\right)*xZX& -Y\left(3\right)*yZX\\ xYX& yYZ& 1& 0& 0& 0& -X\left(4\right)*xYX& -X\left(4\right)*yYX\\ 0& 0& 0& xYX& yYX& 1& -Y\left(4\right)*xYX& -Y\left(4\right)*yYX\end{array}\right]\\ \backslash *MERGEFORMAT\end{array}---\left(5\right)$

Perspective transform coefficient is obtained as shown in formula：

Xi=inv (A) B * MERGEFORMAT (6)

Wherein, Xi is 8 × 1 rank matrixes, so as to obtain conversion coefficient as shown in formula：

$\begin{array}{cc}\begin{array}{ccc}a=Xi\left(1\right)& b=Xi\left(2\right)& c=Xi\left(3\right)\\ d=Xi\left(4\right)& e=Xi\left(5\right)& f=Xi\left(6\right)\\ g=Xi\left(7\right)& h=Xi\left(8\right)& \end{array}& \backslash *MERGEFORMAT\end{array}---\left(7\right)$

(d), assume that the light point coordinates that erosion algorithm draws is (x₀, y₀), then need to carry out perspective transform according to formula, obtain final Coordinate：

$\begin{array}{cc}\begin{array}{c}x=\frac{d\&CenterDot;x_0+e\&CenterDot;y_0+f}{g\&CenterDot;x_0+h\&CenterDot;y_0+1}\\ y=\frac{a\&CenterDot;x_0+b\&CenterDot;y_0+c}{g\&CenterDot;x_0+h\&CenterDot;y_0+1}\end{array}& \backslash *MERGEFORMAT\end{array}---\left(8\right)$

So far, perspective transform is completed；

(2) demarcation of pixel distance

, it is necessary to obtain the corresponding relation of pixel distance and actual range after the perspective demarcation of target plate terminates, according on target plate Normal line, chooses two point coordinates (x of distance_1o, y_1o), (x_2o, y_2o), it is known that its distance be L, using conversion coefficient by its Coordinate is transformed to (x₁, y₁), (x₂, y₂), then distance calibration coefficient is as shown in formula：

$\begin{array}{cc}{b}_d=\frac{L}{\sqrt{{(x_1-x_2)}^2+{(y_1-y_2)}^2}}& \backslash *MERGEFORMAT\end{array}---\left(9\right)$

According to this coefficient, the displacement of laser optical pair twice can be obtained；

Step (3), assessment displacement measurement accuracy；

Step (4), the design of measurement apparatus；

(1) design of box body, is measured

Measurement box body uses splicing structure, and main body is made up of six blocks of sheet materials, and tray interior installs adjustable support, is set using draw-in groove Meter, for installing measurement camera, box body two ends use socket designs, and target plate is shot for inserting；

(2) design of target plate, is measured

Measurement target plate is organic plastic plate, and sheet metal thickness 2mm, white, printing opacity is sub-translucent, and its upper surface pastes one layer of gauged distance The plastics of grid；

(3) design of hardware, is measured

Measurement hardware uses Raspberry Pi card form computer Type Bs Rev1；

The run time is measured each time in Raspberry Pi platforms to be about 5 seconds, shown that measurement frequency can reach 0.2Hz, in the measurement of the surface displacements such as actual foundation ditch, tunnel, due to surface displacement change slowly, the measurement is frequently Rate fully meets measurement request.

2. laser surface displacement automatic monitoring system according to claim 1, it is characterised in that in the step (1), The camera pixel that acquisition target plate RGB color image photo is used is 2592 × 1944.

3. laser surface displacement automatic monitoring system according to claim 1, it is characterised in that in the step (1), Binary-state threshold is the 80% of most bright spot gray value.

4. laser surface displacement automatic monitoring system according to claim 1, it is characterised in that in the step (1), Generally, the round spot bright spot that laser is beaten is irregular, in order to accurately get to the coordinate position of luminous point, uses corrosion Algorithm is corroded binary image, and in order to accelerate erosion algorithm, the facular point pixel of laser is not more than 301 × 301, And carry out image interception；Binary image pixel size after interception is not more than 301 × 301.

5. laser surface displacement automatic monitoring system according to claim 1, it is characterised in that in the step (1), Every time after corrosion, if image is not all 0, a preceding image is covered, if image is full 0, stop corrosion, do not cover artwork Picture.

6. laser surface displacement automatic monitoring system according to claim 1, it is characterised in that in the step (2), The compartment of the squared paper of described gauged distance is away from being 10mm.

7. laser surface displacement automatic monitoring system according to claim 2, it is characterised in that in the step (3), Assess displacement measurement accuracy specific method be：Measurement camera shoot pixel be 2592 × 1944, target plate size be 133mm × 100mm, both ratios such as formula depicted：

$\begin{array}{cc}\frac{2592}{1944}=\frac{4}{3}\&ap;\frac{133}{100}& \backslash *MERGEFORMAT\end{array}---\left(10\right)$

Then, every 2 points of pixel distance is 100/1944<0.052mm, it is considered to the barycentric coodinates searching of hot spot and alignment error etc. Influence, the measurement apparatus displacement measurement accuracy need to then determine in actual stability test.

8. laser surface displacement automatic monitoring system according to claim 1, it is characterised in that in the step (4), Measurement box body material uses 6061-T6 alloy aluminums.

9. laser surface displacement automatic monitoring system according to claim 1, it is characterised in that in the step (4), The measurement target plate thickness is 2mm, and white, printing opacity is sub-translucent.