WO2020093436A1

WO2020093436A1 - Three-dimensional reconstruction method for inner wall of pipe

Info

Publication number: WO2020093436A1
Application number: PCT/CN2018/115851
Authority: WO
Inventors: 宋展
Original assignee: 深圳先进技术研究院
Priority date: 2018-11-09
Filing date: 2018-11-16
Publication date: 2020-05-14
Also published as: CN109544679B; CN109544679A

Abstract

A three-dimensional reconstruction method for the inner wall of a pipe, comprising the following steps: controlling a laser device and a photography device to move together in a pipe (S1); controlling the laser device to emit laser light to the inner wall of the pipe so as to form a laser stripe on the inner wall of the pipe, controlling the photography device to synchronously obtain image information of the laser stripe at each moment from two different angles, and recording pose parameters of the photography device at each moment (S2); obtaining, according to the image information, three-dimensional coordinates of the central point of the laser stripe at each moment in a photography device coordinate system corresponding to each moment (S3); and performing three-dimensional reconstruction of the inner wall of the pipe according to the three-dimensional coordinates and the pose parameters of the photography device at multiple moments (S4). By controlling the laser device and the photography device to move in the pipe, real-time image information of the laser stripe can be dynamically obtained, so as to obtain image information of the inner wall of the pipe in real time, thus obtaining a high-precision three-dimensional image of the inner wall of the pipe.

Description

Three-dimensional reconstruction method of pipeline inner wall

Technical field

The invention belongs to the field of computer vision technology, and in particular, relates to a three-dimensional reconstruction method of an inner wall of a pipeline.

Background technique

As an important transportation device, pipelines are widely used in power transmission, urban sewage, petrochemical and other fields. The effect of pipelines on our lives is as important as arteries to the human body. However, pipeline problems with various forms and a variety of types will continue to appear with the increase of pipeline use time, such as accidental damage of pipelines, aging of materials, chemical corrosion of pipeline walls, cracks and intrusion of debris, etc. If these cannot be found and resolved Problems will bury hidden dangers. In order to improve the service life of the pipeline and ensure safety, regular inspection and maintenance of the pipeline is imperative.

At present, the maintenance of pipelines mainly depends on manual visual inspection, and the detection accuracy, automation, and intelligence of pipeline defects are not very high. Moreover, according to statistics, more than 2/3 of the existing pipeline network in China does not have detailed three-dimensional data of the internal environment of the pipeline, and the internal situation of the pipeline is not known. In view of this, a pipeline inspection device based on a wheeled mobile robot has recently appeared. The closed-circuit television system loaded by the robot records the picture in the pipeline and is judged by ground personnel. Information is still not available, such as the distance and size of obstacles in the pipeline, the three-dimensional size of the inner wall defects of the pipeline, etc. With the development of intelligent detection technology, the three-dimensional visualization of underground pipelines has gradually become an urgent demand, but there is no Very good technical means can achieve this function. In response to the above problems, the existing literature was searched:

1: The Chinese patent application number is CN201210012237.2, which is a patent for an optical-based pipeline rapid detection method. The detection method disclosed in this patent mainly includes the following steps: pretreatment of the surface of the pipeline under test; and placement of marks on and around the pipeline surface Point; place a ruler; take a sequence of photos of global marker points; calculate the coordinates of the marker points in the photo group; collect and align the local dense point cloud on the surface of the measured pipeline; pre-process and model the dense point cloud; integrate the pipeline measurement model with the pipeline CAD design model The coordinate system is aligned, and the coordinate conversion of the pipeline measurement model is performed; the deviation of the pipeline measurement model and the pipeline CAD design model is compared. This method has fast detection speed and high precision, and will not cause damage to the pipe surface. It mainly adopts the traditional three-dimensional scanning method to carry out accurate three-dimensional scanning on the pipeline, which is not suitable for three-dimensional scanning under the condition of robot movement, and the huge amount of data and short working distance are not suitable for the 3D scanning of underground pipelines.

2: The Chinese patent application number is CN201210365683.1, which is a patent for a three-dimensional reconstruction method of pipeline structure. The reconstruction method of this patent includes: acquiring the point cloud data of the pipeline and the unit normal vector of the ground; transforming the point cloud data to The normal vector of the ground is the coordinate system of the Z axis; the normal vector of the point cloud data after the coordinate transformation is calculated; the point cloud data whose normal vector is parallel to the ground is separated from it, and projected onto the plane parallel to the ground; the normal vector is not parallel The point cloud data on the ground is projected onto the Gaussian sphere, projected onto the equator along the latitude direction, and the peak point is detected on the equator; the point cloud data within the area of the peak point is projected onto the plane formed by the direction of the peak and the Z axis Get the point cloud circle formed by the point cloud data; get the three-dimensional map of the pipeline through the point cloud circle. The method of the present invention avoids the noise interference to the greatest extent. It can still clearly classify the point cloud data under the condition that the point cloud data is not completely obtained, and the calculation amount is small, which is suitable for the three-dimensional reconstruction of various pipelines. It is mainly for the processing of CAD pipeline models, and does not involve the 3D imaging part inside the actual pipeline.

3: The Chinese patent application number is CN201110001166.1, the name is the patent of the automatic non-destructive testing method for the inner wall of the binocular vision pipeline. The detection method of this patent includes the following steps: Step 1: Calibration of the binocular CCD camera; Step 2: Setting the measurement parameters ; Step three, real-time measurement; Step four, data management. The measurement method does not cause damage to the inner wall of the pipeline, has a high degree of automation, is simple and flexible to use, and has high measurement accuracy; the measurement device is light in weight, small in size, and convenient to use. This binocular vision measurement method can only obtain very sparse three-dimensional point cloud data, and cannot obtain an accurate 3D model of the inner wall of the pipeline.

None of the existing technologies disclosed above can dynamically scan the inner wall of the pipeline, so that a high-precision three-dimensional model of the inner wall of the pipeline cannot be obtained.

Summary of the invention

(1) Technical problems to be solved

The technical problem to be solved by the present invention is: how to perform dynamic three-dimensional scanning on the inner wall of the pipeline to obtain a high-precision three-dimensional model of the inner wall of the pipeline.

(2) Technical solution

In order to achieve the above objectives, the present invention adopts the following technical solutions:

A three-dimensional reconstruction method of a pipeline inner wall includes the following steps:

Control the laser device and camera device to move synchronously in the pipeline;

The laser device is controlled to emit laser light to the inner wall of the pipe to form a laser stripe on the inner wall of the pipe, and the camera device is simultaneously controlled to acquire the image information of the laser stripe at each moment from two different angles and record each The pose parameters of the camera device at the moment;

Obtaining the three-dimensional coordinates of the center point of the laser stripe at each moment in the coordinate system of the camera device corresponding to each moment according to the image information;

Three-dimensionally reconstruct the inner wall of the pipeline according to the three-dimensional coordinates at multiple times and the pose parameters of the camera device.

Optionally, the method of controlling the laser device to emit laser light to the inner wall of the pipe to form a laser stripe on the inner wall of the pipe specifically includes:

The laser device is controlled to emit a ring laser to the inner wall of the pipe to form a ring-shaped laser stripe on the inner wall of the pipe.

Optionally, the method for controlling the camera device to synchronously acquire the image information of the laser stripes formed by the laser on the inner wall of the pipeline from two different angles specifically includes:

Acquiring the first image of the laser stripe from a first angle using the first camera of the camera device;

Using the second camera of the camera device to synchronously acquire the second image of the laser stripes from a second angle;

The first angle and the second angle are different.

Optionally, before controlling the laser device and the camera device to move synchronously in the pipeline, the method further includes:

Perform dual target setting on the camera device and obtain dual target setting parameters.

Optionally, the method for obtaining the three-dimensional coordinates of the center point of the laser stripe at each moment in the coordinate system of the camera device corresponding to each moment according to the image information specifically includes:

Obtain the coordinates of the first skeleton point of the laser stripe in the first image and the coordinates of the second skeleton point in the second image, respectively;

The three-dimensional coordinates of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system are obtained according to the first skeleton point coordinates, the second skeleton point coordinates, and the dual-target fixed parameters.

Optionally, the center point of the laser stripe is acquired in the first camera coordinate system or the second camera coordinate system according to the first skeleton point coordinates, the second skeleton point coordinates and the dual target fixed parameters Before the three-dimensional coordinates in, the method further includes:

Obtaining the first sub-pixel coordinates of the center point of the laser stripe in the first image according to the coordinates of the first skeleton point;

Obtain the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the coordinates of the second skeleton point.

Optionally, the acquiring the center point of the laser stripe in the first camera coordinate system or the second camera coordinate according to the first skeleton point coordinate, the second skeleton point coordinate and the dual target fixed parameter The method of three-dimensional coordinates in the system is as follows:

Acquiring the three-dimensional of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system according to the first sub-pixel coordinates, the second sub-pixel coordinates and the dual-target fixed parameters coordinate.

Optionally, the method for obtaining the first sub-pixel coordinates of the center point of the laser stripe in the first image according to the coordinates of the first skeleton point is:

Acquiring a first image gradient corresponding to the coordinates of the first skeleton point;

The first sub-pixel coordinates of the center point of the laser stripe in the first image are obtained according to the first image gradient and the first grayscale image corresponding to the first image.

Optionally, the method for obtaining the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the coordinates of the second skeleton point is:

Acquiring a second image gradient corresponding to the coordinates of the second skeleton point;

Obtain the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the second image gradient and the second grayscale image corresponding to the second image.

Optionally, the method for three-dimensionally reconstructing the inner wall of the pipeline according to the three-dimensional coordinates and the pose parameters of the camera device specifically includes:

Obtain the coordinates of the center point of the laser stripe in the starting global coordinate system at each moment according to the three-dimensional coordinates at each moment and the pose parameters of the camera device, where the starting global coordinate system is The coordinate system where the camera device in the initial position is located;

Three-dimensionally reconstruct the inner wall of the pipeline according to the coordinates of the center point of the laser stripes in the starting global coordinate system at multiple moments.

(3) Beneficial effects

The three-dimensional reconstruction method of the inner wall of the pipeline disclosed by the present invention can dynamically obtain real-time image information of the laser stripes by controlling the laser device and the camera device to move in the pipeline, thereby acquiring the image information of the inner wall of the pipeline in real time, and realizing high-precision pipeline Three-dimensional image of the inner wall. In addition, by using a ring laser to achieve 360-degree scanning, the image information of the inner wall of the pipeline can be obtained more quickly and comprehensively, and the cost is also reduced.

BRIEF DESCRIPTION

FIG. 1 is a structural block diagram of a three-dimensional reconstruction system according to an embodiment of the present invention;

2 is a flowchart of a three-dimensional reconstruction method according to an embodiment of the present invention.

detailed description

In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not intended to limit the present invention.

As shown in FIG. 1, the three-dimensional reconstruction method of the inner wall of the pipeline according to the embodiment of the present invention is performed on the three-dimensional reconstruction system of the inner wall of the pipeline. The system includes: a laser device 10, a camera device 20, a mobile device 30, and a processor 40, in which processing The device 40 includes an image processing module 41 and a three-dimensional reconstruction module 42, and the processor 40 is communicatively connected to the laser device 10, the camera device 20, and the mobile device 30.

FIG. 2 shows a flowchart of a three-dimensional reconstruction method of an inner wall of a pipe according to an embodiment of the present invention. The three-dimensional reconstruction method includes steps S1 to S4:

Step S1: Control the laser device 10 and the camera device 20 to move synchronously in the pipeline.

Specifically, as a preferred embodiment, the laser device 10 and the camera device 20 are mounted on the mobile device 30, and the mobile device 30 can move within the pipeline. Among them, the mobile device 30 may be a pipeline inspection robot. The movement of the moving device 30 in the pipeline drives the laser device 10 and the camera device 20 to move synchronously in the pipeline.

Of course, in other embodiments, the laser device 10 and the camera device 20 can be separately installed, as long as the laser device 10 and the camera device 20 are moved synchronously.

Step S2: The laser device 10 is controlled to emit laser light to the inner wall of the pipe to form laser stripes on the inner wall of the pipe, and the camera device 20 is simultaneously controlled to acquire image information of the laser stripes from two different angles synchronously, and the posture parameters of the camera device 20 are recorded.

As a preferred embodiment, the laser device 10 is controlled to emit a ring laser to the inner wall of the pipe to form a ring-shaped laser stripe on the inner wall of the pipe. The position of the center point of the laser stripe is the same as the position of the inner wall surface of the pipe corresponding to the center point of the laser stripe. Specifically, the laser device 10 includes a point laser and a grating provided at the front of the point laser. After the point laser emits laser light, a ring or circular laser can be generated after the conversion of the grating, so that a ring-shaped laser stripe can be formed on the inner wall of the pipe. For example, for the laser device 10, a circle laser of brand Anford and model T850AY16100 can be used.

Further, as a preferred embodiment, as shown in FIG. 1, the imaging device 20 includes a first camera 21 and a second camera 22. The first camera 21 and the second camera 22 are installed on opposite sides of the laser device 10, respectively. The one camera 21 and the second camera 22 are arranged in parallel, and ensure that the shooting angles of the first camera 21 and the second camera 22 can cover the laser stripes. The first camera 21 is used to capture a first image of a laser stripe from a first angle, and the second camera 22 is used to capture a second image of a laser stripe from a second angle, where the first angle and the second angle are different, and the first camera 21 It is synchronized with the second camera 22 when shooting.

As a preferred embodiment, the cameras of the first camera 21 and the second camera 22 use cameras with a resolution of 1280 * 720. The shooting frame rate of the first camera 21 and the second camera 22 is 30 FPS. The simultaneous shooting of the two cameras 22 can collect 30 binocular images per second. The laser emitted by the laser device may be a laser in the visible band or a laser in the non-visible band, such as an infrared band laser. In this case, the first camera 21 and the second camera 22 need to be equipped with red light filters to improve Imaging effect.

Since this embodiment uses the principle of binocular stereo vision imaging for three-dimensional reconstruction, before controlling the laser device 10 and the camera device 20 to move synchronously in the pipeline, the first camera 21 and the second camera 22 need to be bi-targeted, and The image processing module 41 is used to acquire the dual target fixed parameters. The bi-targeting method is a common method in binocular stereo vision imaging, which will not be repeated in the embodiments of the present invention.

Step S3: Obtain the three-dimensional coordinates of the center point of the laser stripe at each moment in the coordinate system of the camera 20 at each moment according to the image information.

Specifically, this step includes steps S31 to S32:

Step S31: Obtain the coordinates of the first skeleton point of the laser stripe in the first image and the coordinates of the second skeleton point in the second image, respectively;

The first image and the second image are the original images obtained by the first camera 21 and the second camera 22 directly shooting the laser stripes.

Specifically, this step S21 includes the following steps:

Step S311: Gaussian filtering is performed on the first image and the second image to smooth the image and remove image noise, respectively, to obtain a first gray image I ₁ (x ₁ , y ₁ ) and a second gray image I ₁ (x ₂ , Y ₂ ).

Step S312: Perform binarization processing on the first gray image I ₁ (x ₁ , y ₁ ) and the second gray image I ₁ (x ₂ , y ₂ ), respectively, to obtain the binarized image I ₂ (x ₁ , y ₁ ) and I ₂ (x ₂ , y ₂ ). The threshold of the binarization process can be set according to the brightness of the laser stripes in the original image, for example, the threshold is set to 150.

Step S313: Refine the images I ₂ (x ₁ , y ₁ ) and I ₂ (x ₂ , y ₂ ) obtained after the binarization process to obtain the first skeleton image L (x ₁ , y ₁ ) And the second skeleton image L (x ₂ , y ₂ ), wherein the skeleton image can be obtained by using a conventional binary image skeleton extraction algorithm, which will not be repeated here. In this way, the skeleton point in the first skeleton image L (x ₁ , y ₁ ) is the first skeleton point coordinate of the center point of the laser stripe in the first image, and the second skeleton image L (x ₂ , y ₂ ) The skeleton point of is the second skeleton point coordinate of the center point of the laser stripe in the second image. It should be noted that the positioning accuracy of the skeleton image is pixel-level accuracy.

Step S32: Acquire the three-dimensional coordinates of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system according to the first skeleton point coordinates, the second skeleton point coordinates and the dual-target fixed parameters.

Specifically, taking the coordinates of the first skeleton point as an example, the coordinate system where the first skeleton point coordinates are located is the first image coordinate system O ₁ X ₁ Y ₁ , and the first image coordinate system O ₁ X ₁ Y ₁ is located in the first camera In the focal plane of 21, the origin O ₁ of the first image coordinate system O ₁ X ₁ Y ₁ is the focal point of the first camera 21, and the first skeleton point coordinate is the connection between the optical center of the first camera and the center point of the laser stripe The coordinates of the intersection point of the line and the focal plane, that is, the coordinates of the first skeleton point are two-dimensional coordinates. The second skeleton point coordinates can also be analyzed similarly, and will not be repeated here.

Further, the dual target fixed parameters include the first focal length f _{1 of the} first camera 21, the second focal length f _{2 of the} second camera 22, the optical center distance B of the first camera 21 and the second camera 22 and the second The pixel coordinates (u ₁ , v ₁ ) of the origin O ₁ of an image coordinate system O ₁ X ₁ Y ₁ . According to the principle of binocular visual stereo imaging, the first skeleton point coordinates, the second skeleton point coordinates and the dual target fixed parameters can be obtained to obtain the three-dimensional coordinates of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system.

It should be noted that the three-dimensional coordinates obtained above are the three-dimensional coordinates of the center point of the laser stripe at each moment corresponding to the first camera coordinate system and the second camera coordinate system at each moment. According to the same method described above, the three-dimensional coordinates of the center point of the laser stripe at multiple times in the first camera coordinate system or the second camera coordinate system can be obtained, and the three-dimensional coordinate point cloud can be obtained after the laser device scans the inner wall of the pipeline data. The calculation process in step S3 is completed in the image processing module 41.

In other embodiments, since the coordinates of the first skeleton point and the coordinates of the second skeleton point are both coordinates with pixel accuracy, the positioning accuracy is not accurate enough. In order to further improve the positioning accuracy of the center point of the laser stripes, a sub-pixel calculation strategy is used to accurately position the coordinates of the skeleton points. Specifically, the sub-pixel calculation method specifically includes steps one to two:

Step 1: Obtain the first image gradient q _{1 of the} first skeleton point coordinate and the second image gradient q _{2 of the} second skeleton point coordinate, respectively.

Specifically, taking the coordinates of the first skeleton point as an example, the first skeleton point p _{1 is taken} as the center, and 2m skeleton points adjacent to the first skeleton point p ₁ are found along the first skeleton image, where m is a defined local neighborhood radius, the value of m may be set according to the actual accuracy required by skeletal point 2m and the first skeleton point p ₁ can be calculated first skeleton point p at _a gradient image a first q _1. Similarly, the second image gradient q ₂ at the second skeleton point p ₂ can be calculated using the above method.

Step 2: Obtain the first sub-pixel coordinates of the center point of the laser stripe according to the first image gradient q ₁ and the first gray image I ₁ (x ₁ , y ₁ ), and according to the second image gradient q ₂ and the first gray image I ₁ (x ₂ , y ₂ ) acquires the second sub-pixel coordinates of the center point of the laser stripe.

Specifically, in the first grayscale image I ₁ (x ₁ , y ₁ ), with the first skeleton point p ₁ as the center, 2n adjacent image points are searched in a direction perpendicular to the first image gradient q ₁ , Taking the luminance value of 2n adjacent image points and the luminance value of the first skeleton point p ₁ as weighting coefficients, the first sub-pixel coordinates of the center point of the laser stripe are calculated. Similarly, in the first grayscale image I ₁ (x ₂ , y ₂ ), with the second skeleton point p ₂ as the center, 2n adjacent image points are searched in the direction perpendicular to the second image gradient q ₂ , Taking the brightness value of 2n adjacent image points and the brightness value of the second skeleton point p ₂ as weighting coefficients, the second sub-pixel coordinates of the center point of the laser stripe are calculated, which can improve the positioning accuracy of the center point of the laser stripe.

Further, the three-dimensional coordinates of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system are calculated according to the first sub-pixel coordinates, the second sub-pixel coordinates and the dual-target fixed parameters, which can improve the center of the laser stripes The three-dimensional coordinate accuracy of the point. The calculation method is similar to the calculation method in step S32, and will not be repeated here.

Step S4: Three-dimensionally reconstruct the inner wall of the pipeline according to the three-dimensional coordinates at multiple times and the pose parameters of the camera device.

Specifically, as a preferred embodiment, the camera device in this embodiment is installed on a mobile device, and the mobile device and the camera device have the same coordinate system, so the posture parameters of the mobile device are acquired through the sensor, and then the camera device and the mobile The relative relationship of the devices can indirectly obtain the pose parameters of the mobile device. Among them, when the mobile device selects the pipeline inspection robot, it has a built-in encoder and attitude sensor, which can obtain the initial position of the mobile device, the position at each time, and the direction of the mobile device at each time, so that the initial position of the camera device can be obtained , The location at each moment and the direction of the mobile device at each moment.

Further, coordinate transformation is performed according to the three-dimensional coordinates of the laser stripe center at each moment in the camera coordinate system and the pose parameter of the camera to obtain the three-dimensional coordinates of the laser stripe center at each moment in the coordinate system where the camera is located. Further, the three-dimensional coordinates of the laser stripe center at each moment in the coordinate system where the camera device is located are converted into a starting global coordinate system, where the starting global coordinate system is the coordinate system where the camera device is at the initial position. After the mobile device and the camera device scan the inner wall of the pipeline, a three-dimensional contour line is drawn according to the coordinates of the laser stripe center in the starting global coordinate system at multiple times to realize three-dimensional reconstruction of the inner wall of the pipeline. The calculation process of step S3 is completed in the three-dimensional reconstruction module 42.

The embodiment of the present invention discloses a three-dimensional reconstruction method for the inner wall of a pipeline. By controlling the movement of the laser device and the camera device in the pipeline, the real-time image information of the laser stripes can be dynamically obtained, so that the image information of the inner wall of the pipeline can be obtained in real time. Accurate 3D image of the inner wall of the pipe. In addition, by using a ring laser to achieve 360-degree scanning, the image information of the inner wall of the pipeline can be obtained more quickly and comprehensively, and the cost is also reduced.

The specific embodiments of the present invention have been described in detail above, and although some embodiments have been shown and described, those skilled in the art should understand that the principle and spirit of the present invention defined by the claims and their equivalents are not deviated from Under the circumstances, these embodiments can be modified and improved, and these modifications and improvements should also fall within the protection scope of the present invention.

Claims

A three-dimensional reconstruction method of the inner wall of a pipeline, which includes the following steps:

Control the laser device and camera device to move synchronously in the pipeline;

The laser device is controlled to emit laser light to the inner wall of the pipe to form a laser stripe on the inner wall of the pipe, and the camera device is simultaneously controlled to acquire the image information of the laser stripe at each moment from two different angles and record each The pose parameters of the camera device at the moment;

Obtaining the three-dimensional coordinates of the center point of the laser stripe at each moment in the coordinate system of the camera device corresponding to each moment according to the image information;

Three-dimensionally reconstruct the inner wall of the pipeline according to the three-dimensional coordinates at multiple times and the pose parameters of the camera device.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 1, wherein the method for controlling the camera device to acquire the image information of the laser stripes formed by the laser on the inner wall of the pipeline from two different angles specifically includes:

Acquiring the first image of the laser stripe from a first angle using the first camera of the camera device;

Using the second camera of the camera device to synchronously acquire the second image of the laser stripes from a second angle;

The first angle and the second angle are different.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 1, wherein before the laser device and the camera device are controlled to move synchronously in the pipeline, the method further comprises:

Perform dual target setting on the camera device and obtain dual target setting parameters.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 3, wherein the center point of the laser stripe at each moment corresponding to the three-dimensional coordinates of the camera coordinate system at each moment is obtained from the image information The methods include:

Obtain the coordinates of the first skeleton point of the laser stripe in the first image and the coordinates of the second skeleton point in the second image, respectively;

The three-dimensional coordinates of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system are obtained according to the first skeleton point coordinates, the second skeleton point coordinates, and the dual-target fixed parameters.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 4, wherein the center point of the laser stripe obtained based on the coordinates of the first skeleton point, the coordinates of the second skeleton point, and the dual target fixed parameter is Before the three-dimensional coordinates in a camera coordinate system or the second camera coordinate system, the method further includes:

Obtaining the first sub-pixel coordinates of the center point of the laser stripe in the first image according to the coordinates of the first skeleton point;

Obtain the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the coordinates of the second skeleton point.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 5, wherein the center point of the laser stripe obtained from the first skeleton point coordinate, the second skeleton point coordinate and the dual target fixed parameter is The method of the three-dimensional coordinates in the first camera coordinate system or the second camera coordinate system is specifically:

Acquiring the three-dimensional of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system according to the first sub-pixel coordinates, the second sub-pixel coordinates and the dual-target fixed parameters coordinate.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 6, wherein the method of acquiring the first sub-pixel coordinates of the center point of the laser stripe in the first image according to the coordinates of the first skeleton point is :

Acquiring a first image gradient corresponding to the coordinates of the first skeleton point;

The first sub-pixel coordinates of the center point of the laser stripe in the first image are obtained according to the first image gradient and the first grayscale image corresponding to the first image.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 6, wherein the method of acquiring the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the coordinates of the second skeleton point is :

Acquiring a second image gradient corresponding to the coordinates of the second skeleton point;

Obtain the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the second image gradient and the second grayscale image corresponding to the second image.
The three-dimensional reconstruction method for the inner wall of the pipeline according to claim 1, wherein the method for three-dimensionally reconstructing the inner wall of the pipeline according to the three-dimensional coordinates and the posture parameters of the camera device specifically includes:

Obtain the coordinates of the center point of the laser stripe in the starting global coordinate system at each moment according to the three-dimensional coordinates at each moment and the pose parameters of the camera device, where the starting global coordinate system is The coordinate system where the camera device in the initial position is located;

Three-dimensionally reconstruct the inner wall of the pipeline according to the coordinates of the center point of the laser stripes in the starting global coordinate system at multiple moments.
The three-dimensional reconstruction method of the inner wall of the pipe according to claim 1, wherein the method of controlling the laser device to emit laser light to the inner wall of the pipe to form laser stripes on the inner wall of the pipe specifically includes:

The laser device is controlled to emit a ring laser to the inner wall of the pipe to form a ring-shaped laser stripe on the inner wall of the pipe.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 10, wherein the method of controlling the camera device to simultaneously acquire the image information of the laser stripes formed on the inner wall of the pipeline from two different angles specifically includes:

Acquiring the first image of the laser stripe from a first angle using the first camera of the camera device;

Using the second camera of the camera device to synchronously acquire the second image of the laser stripes from a second angle;

The first angle and the second angle are different.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 10, wherein before the laser device and the camera device are controlled to move synchronously in the pipeline, the method further comprises:

Perform dual target setting on the camera device and obtain dual target setting parameters.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 12, wherein the center point of the laser stripe at each moment corresponding to the three-dimensional coordinates in the coordinate system of the camera device at each moment is obtained from the image information The methods include:

Obtain the coordinates of the first skeleton point of the laser stripe in the first image and the coordinates of the second skeleton point in the second image, respectively;

The three-dimensional coordinates of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system are obtained according to the first skeleton point coordinates, the second skeleton point coordinates, and the dual-target fixed parameters.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 13, wherein the center point of the laser fringe obtained based on the coordinates of the first skeleton point, the coordinates of the second skeleton point and the dual-target fixed parameter is in the first Before the three-dimensional coordinates in a camera coordinate system or the second camera coordinate system, the method further includes:

Obtaining the first sub-pixel coordinates of the center point of the laser stripe in the first image according to the coordinates of the first skeleton point;

Obtain the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the coordinates of the second skeleton point.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 14, wherein the center point of the laser stripe obtained according to the coordinates of the first skeleton point, the coordinates of the second skeleton point and the dual-target fixed parameter is located in the The method of the three-dimensional coordinates in the first camera coordinate system or the second camera coordinate system is specifically:

Acquiring the three-dimensional of the center point of the laser stripe in the first camera coordinate system or the second camera coordinate system according to the first sub-pixel coordinates, the second sub-pixel coordinates and the dual-target fixed parameters coordinate.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 15, wherein the method of acquiring the first sub-pixel coordinates of the center point of the laser stripe in the first image according to the coordinates of the first skeleton point is :

Acquiring a first image gradient corresponding to the coordinates of the first skeleton point;

The first sub-pixel coordinates of the center point of the laser stripe in the first image are obtained according to the first image gradient and the first grayscale image corresponding to the first image.
The three-dimensional reconstruction method of the inner wall of the pipeline according to claim 15, wherein the method for obtaining the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the coordinates of the second skeleton point is :

Acquiring a second image gradient corresponding to the coordinates of the second skeleton point;

Obtain the second sub-pixel coordinates of the center point of the laser stripe in the second image according to the second image gradient and the second grayscale image corresponding to the second image.
The three-dimensional reconstruction method for the inner wall of the pipeline according to claim 10, wherein the method for three-dimensionally reconstructing the inner wall of the pipeline according to the three-dimensional coordinates and the posture parameters of the camera device specifically includes:

Obtain the coordinates of the center point of the laser stripe in the starting global coordinate system at each moment according to the three-dimensional coordinates at each moment and the pose parameters of the camera device, where the starting global coordinate system is The coordinate system where the camera device in the initial position is located;

Three-dimensionally reconstruct the inner wall of the pipeline according to the coordinates of the center point of the laser stripes in the starting global coordinate system at multiple moments.