A pipeline measurement method and device thereof

Abstract

The present invention discloses a pipeline measurement method and its device, including calibrating a multi-camera system and a measurement platform, calculating the effective area of the actually captured images of the multi-camera system, and simultaneously calculating the conversion relationship between the coordinate matrix of the marking points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system. Move the calibrated multi-camera system along the extension direction of the pipeline to be measured on the calibrated measurement platform, continuously capture multiple images of the pipeline to be measured in the effective area in segments and then perform image preprocessing, calculate the position area of each section of the pipeline in its respective image, and use the detection cylinder model to retrieve and match with the cylinder center point of the cylinder model to obtain the central axis position of each section of the pipeline, and reconstruct the pipeline models of each section based on the central axis position. After coordinate conversion, splice the images corresponding to the pipeline models of each section and calculate the pipeline data of each section, and output the total pipeline model and pipeline data.

Description

Technical Field

[0001] The present invention relates to the field of pipeline measurement, and specifically to a pipeline measurement method and device thereof. Background Art

[0002] As a structure for transporting materials and connecting, pipelines are widely used in fields such as automobiles, ships, and aerospace. Due to their wide range of usage scenarios, the detection of pipelines has become a popular research direction in recent years. Because of the various sizes of pipelines, it poses a significant challenge to the detection means. Currently, most measurement means are for pipe fittings with a length within 2m, and there is still no good solution for the measurement of pipe fittings beyond 2m.

[0003] The measurement of pipelines can be classified into contact type and non-contact type according to the implementation method. Currently, the most widely used and highest-precision contact measurement is coordinate measurement. The probe gently touches the pipe wall through a mechanical structure, and the coordinate of this position point will be automatically recorded and locked at the back end. By this contact-recording method, the position coordinates of some characteristic points on the pipeline are determined, thus completing the pipeline measurement. Since this measurement method takes a long time and requires high technical difficulty, it can only be applied to the occasions where the instrument can reach the measurement scene and cannot meet the measurement requirements of long pipelines.

[0004] Structured light scanning, as a typical representative of the active vision method of non-contact measurement, can project modulated laser or speckle onto the object surface, and an industrial camera captures the pipeline with a characteristic image on the surface, divides and extracts the characteristic regions of the image, and splices the point cloud to achieve pipeline reconstruction. Although it significantly shortens the measurement time compared with coordinate measurement, the measurement accuracy is easily affected by the detection accuracy of the device. Summary of the Invention

[0005] The purpose of the present invention is to provide a pipeline measurement method and device thereof with high recognition accuracy, strong adaptability, and large field of view for long pipelines.

[0006] The present invention provides a pipeline measurement method, including the following steps:

[0007] S1. Calibrate a multi-camera system composed of multiple cameras;

[0008] S2. According to the calibrated multi-camera system, calculate the area of the image captured by each camera, and thus calculate the effective area of the image actually captured by the multi-camera system;

[0009] S3. Calibrate a measurement platform provided with multiple marking points through the calibrated multi-camera system, and calculate the conversion relationship between the coordinate matrix of the marking points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system;

[0010] S4. Move the calibrated multi-camera system along the extension direction of the pipeline to be measured on the calibrated measurement platform, and continuously capture multiple images of the pipeline to be measured in the effective area section by section. Among them, each image of the pipeline to be measured in the effective area contains the image of each section of the pipeline and the image of the measurement platform marking points corresponding to each section of the pipeline;

[0011] S5. Perform image preprocessing on each image of the pipeline to be measured in the effective area, calculate the position area of each section of the pipeline in its respective image, and then use the detection cylinder model to retrieve and match with the cylinder center point of the cylinder model to obtain the central axis position of each section of the pipeline, and reconstruct the pipeline models of each section based on this central axis position;

[0012] S6. According to the conversion relationship between the coordinate matrix of the marking points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system, calculate the pipeline data and pipeline models of each section of the reconstructed pipeline models in the measurement platform coordinate system;

[0013] S7. According to the pipeline data and pipeline models of each section of the pipeline models in the measurement platform coordinate system, splice the images corresponding to each section of the pipeline models in sequence and calculate the pipeline data of each section, and output the total pipeline model and pipeline data.

[0014] Preferably, the step S1 includes the following steps:

[0015] S1-1. Place a calibration board printed with identification points directly below the multi-camera system composed of multiple cameras;

[0016] S1-2. Obtain the image of the calibration board printed with identification points through the multi-camera system;

[0017] S1-3. For the obtained image of the calibration board printed with identification points, identify the spatial positions of the centers of each identification point through the multi-vision calibration principle to calculate the internal and external parameters of each camera;

[0018] S1-4. Take the position of the center point of the calibration board corresponding to the calibration board in the image of the calibration board printed with identification points as the origin of the multi-camera system coordinate system, and unify the internal and external parameters of each camera to this origin of the multi-camera system coordinate system to complete the calibration of the multi-camera system.

[0019] Preferably, the calibration of the measurement platform provided with multiple marking points in the step S3 includes the following steps:

[0020] S3-1: Based on the principle of close-range photogrammetry, obtain multiple angular images of the marking points on the measurement platform by the calibrated multi-camera system;

[0021] S3-2: Perform image edge detection processing on the obtained multiple angular marking point images;

[0022] S3-3: Perform sub-pixel edge extraction on the multiple angular marker point images after edge detection processing.

[0023] S3-4: Determine the spatial positions of the centers of the circles of each marker point on the measurement platform by the method of ellipse center fitting for the multiple angular marker point images after sub-pixel edge extraction, and establish a measurement platform coordinate system based on this.

[0024] Preferably, the obtaining of multiple angular images of the marker points on the measurement platform by the calibrated multi-camera system includes:

[0025] The calibrated multi-camera system obtains images of the marker points on the measurement platform from three angles of 90°, 60°, and 45° with respect to the plane of the measurement platform.

[0026] Preferably, the image preprocessing of each pipeline image to be measured in the effective region in step S5 includes the following steps:

[0027] S5-1: Set a limited gray threshold for each pipeline image to be measured in the effective region.

[0028] S5-2: Perform feature extraction on each pipeline image to be measured in the effective region after setting the limited gray threshold.

[0029] Preferably, the reconstruction of each pipeline model in step S5 further includes:

[0030] S5-3: Determine the intersection points of the ends of each segmented pipeline and the boundary of the effective region by establishing a detection cylinder model, and divide a circular region with the intersection point as the center and the cylinder radius of the detection cylinder model.

[0031] S5-4: If there are three or more small cylinder points in the pipeline image within the circular region outside the effective region of the image captured by the multi-camera system, then this point is a break point; conversely, if there is no intersection point between the ends of each segmented pipeline and the boundary, or the intersection point is within the divided circular region and the number of points outside the effective region of the image captured by the multi-camera system is less than three, then this point is the end point of each segmented pipeline.

[0032] Preferably, step S7 further includes: If there are cases of the same bending point or multiple spatial coordinates during the splicing process of the images corresponding to each pipeline model, then filter out the regions of the duplicate pipeline models.

[0033] Preferably, filtering out the regions of the duplicate pipeline models specifically includes:

[0034] Perform an averaging process on the coordinates of each bending point in the image within the bending duplicate region.

[0035] Calculate the mean coordinates of the bending points after homogenization processing, and use the mean coordinates as the coordinates of the bending points in the image within the bending repetition area, so as to correct the coordinates of the bending points in the overall pipeline model.

[0036] The present invention also provides a pipeline measurement device, including a measurement platform, a multi-camera system installed on the measurement platform. The multi-camera system at least includes a memory and a processor. A plurality of marking points are arranged on the measurement platform. The memory includes at least one executable program stored therein;

[0037] When the executable program is executed by the processor, the method described above is implemented.

[0038] Preferably, a backlight is also arranged on the measurement platform.

[0039] The present invention provides a pipeline measurement method and its device. By moving the calibrated multi-camera system along the extension direction of the pipeline to be measured on the calibrated measurement platform, continuously taking multiple images of the pipeline to be measured in the effective area in segments, and using the marking point recognition technology to complete the alignment and splicing of the pipeline, the blind area existing in the measurement of long pipelines can be reduced. And this measurement method has a larger visual field range and higher recognition accuracy, can solve the problem of false matching in traditional visual measurement, and thus effectively improves the accuracy of pipeline splicing and measurement. At the same time, this method preprocesses the image of the pipeline to be measured in each effective area, calculates the position area of each section of the pipeline in its respective image, then uses the detection cylinder model, retrieves and matches with the center point of the cylinder of the cylinder model to obtain the central axis position of each section of the pipeline, and reconstructs the pipeline models of each section with this central axis position, which can quickly establish the pipeline model in a short time, is easy to operate, has no special restrictions on the pipeline size and shape, and thus can achieve the purpose of adapting to pipeline measurement in various scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic flow chart of a pipeline measurement method provided by an embodiment of the present invention;

[0041] Figure 2 It is a schematic structural diagram of a multi-camera system obtaining an image of a calibration board printed with identification points provided by an embodiment of the present invention;

[0042] Figure 3 It is a schematic diagram of pipeline implementation measurement provided by an embodiment of the present invention;

[0043] Figure 4 It is a schematic diagram of determining the break point position of a segmented pipeline provided by an embodiment of the present invention;

[0044] Figure 5 It is a schematic diagram of pipeline corresponding splicing provided by an embodiment of the present invention. Specific Embodiments

[0045] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only for explaining the present invention and not for limiting the present invention. Additionally, it should be noted that for the sake of description, only the parts related to the present invention rather than all the structures are shown in the accompanying drawings.

[0046] Figure 1 FIG. is a schematic flow chart of a pipeline measurement method provided according to an embodiment of the present invention. The pipeline measurement method includes the following steps:

[0047] S1. Calibrate a multi-camera system composed of multiple cameras;

[0048] S2. According to the calibrated multi-camera system, calculate the area of the image captured by each camera, so as to calculate the effective area of the image actually captured by the multi-camera system;

[0049] S3. Calibrate a measurement platform provided with multiple marking points through the calibrated multi-camera system, and calculate the conversion relationship between the coordinate matrix of the marking points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system;

[0050] S4. Move the calibrated multi-camera system along the extension direction of the pipeline to be measured on the calibrated measurement platform, and continuously capture multiple images of the pipeline to be measured in the effective area in segments. Among them, each image of the pipeline to be measured in each effective area includes the image of each section of the pipeline and the image of the measurement platform marking points corresponding to each section of the pipeline;

[0051] S5. Perform image preprocessing on each image of the pipeline to be measured in each effective area, calculate the position area of each section of the pipeline in its respective image, and then use a detection cylinder model to retrieve and match with the cylinder center point of the cylinder model to obtain the central axis position of each section of the pipeline, and reconstruct the pipeline models of each section based on the central axis position;

[0052] S6. According to the conversion relationship between the coordinate matrix of the marking points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system, calculate the pipeline data and pipeline models of each section of the reconstructed pipeline models in the measurement platform coordinate system;

[0053] S7. According to the pipeline data and pipeline models of each section of the pipeline models in the measurement platform coordinate system, and sequentially splice the images corresponding to the pipeline models of each section and calculate the pipeline data of each section, and output the total pipeline model and pipeline data.

[0054] The invention is a solution for measuring pipelines within the length range exceeding 1.2m and not exceeding 6m. It consists of a measurement platform and a multi-camera system, which can solve the measurement work of pipelines with a length exceeding the conventional 1.2m and within 6m, reduce labor input, and improve production efficiency.

[0055] The pipeline measurement device includes a measurement platform and a multi-camera system installed on the measurement platform. The multi-camera system includes at least one memory and one processor. Multiple marking points are set on the measurement platform. The memory includes at least one executable program stored therein. Among them, the multi-camera system is responsible for collecting and transmitting pipeline images. It is a frame structure composed of at least four cameras. Each camera in the system is distributed at different positions on the top of the frame to obtain images of the pipeline to be measured on the measurement table. The measurement platform is equipped with a backlight, which can better extract the edge features of the catheter during image processing. Marking points are printed around the surface of the measurement platform. Through the marking point recognition technology, the pipeline alignment and splicing process are completed. Its detection accuracy is directly related to the final pipeline measurement accuracy.

[0056] Calibration of the multi-camera system and the measurement platform:

[0057] Calibration of the multi-camera system:

[0058] Place a calibration board with identification points printed on it directly below the multi-camera system composed of multiple cameras. Obtain the image of the calibration board with identification points through the multi-camera system. Then, through the principle of multi-view vision calibration, identify the spatial positions of the centers of each identification point to calculate the internal and external parameters of each camera. Use the position of the center point of the corresponding calibration board in the image of the calibration board with identification points as the origin of the coordinate system of the multi-camera system, and unify the internal and external parameters of each camera to the origin of the coordinate system of the multi-camera system to complete the calibration of the multi-camera system. After the calibration of the multi-camera system is successful, use the position of the center point corresponding to the calibration board in the image as the origin of the camera coordinate system, establish a camera coordinate system at this center point position, and calculate the area of the image captured by each camera, so as to calculate the effective area of the image actually captured by the multi-camera system. As Figure 2 shown, the area within the calibration board format is the effective area of the image captured by the multi-camera system.

[0059] Calibration of the measurement platform:

[0060] Based on the principle of close-range photogrammetry, multiple-angle images of the marked points on the measurement platform are obtained by a calibrated multi-camera system. Then, image edge detection processing is carried out. For the multiple-angle marked point images after edge detection processing, sub-pixel edge extraction is performed. Then, by the method of ellipse center fitting, the spatial positions of the centers of the marked points on the measurement platform are determined, and a coordinate system of the measurement platform is established accordingly. Obtaining multiple-angle images of the marked points on the measurement platform includes, for example, obtaining images of the marked points on the measurement platform by the calibrated multi-camera system at three angles of 90°, 60°, and 45° with respect to the plane of the measurement platform.

[0061] Pipeline measurement:

[0062] Turn on the backlight of the measurement platform, place the pipeline to be measured on the measurement platform, and move the multi-camera system along the extension direction of the pipeline to be measured. As Figure 3 shown, continuously capture multiple images of the pipeline to be measured in the effective area in segments. The measurement platform is 6 m long, 1 m wide, and 0.8 m high; the multi-camera system is 3 m high, which basically meets the measurement requirements for long pipelines within 6 m in length. Among them, preprocessing the images of the pipeline to be measured in each effective area includes setting a gray-scale threshold for the images of the pipeline to be measured in each effective area, and then performing feature extraction.

[0063] Determination of the break point positions of each segmented pipeline:

[0064] By establishing a detection cylinder model, the intersection points of the ends of each segmented pipeline and the boundary of the effective area are determined. Taking this intersection point as the center and the cylinder radius of the detection cylinder model, a circular area is divided. If there are three or more small cylinder points in the pipeline image within the circular area outside the effective area of the image captured by the multi-camera system, then this point is a break point. On the contrary, if there is no intersection point between the ends of each segmented pipeline and the boundary, or the intersection point is within the divided circular area and the number of points outside the effective area of the image captured by the multi-camera system is less than three, then this point is the end point of each segmented pipeline. As Figure 4 shown.

[0065] Pipeline model reconstruction:

[0066] For the image data of each segmented pipeline containing the marked points on the measurement platform, set a gray-scale threshold, perform feature extraction, match the pipeline position areas in each image, and use the detection cylinder model to retrieve and match the center position of the cylinder to obtain the position of the pipeline axis, so as to reconstruct the pipeline models at each segmented position.

[0067] Pipeline model alignment:

[0068] After the measurement platform is calibrated, the coordinate matrix of the marked points on the measurement platform in its own coordinate system is Q. During the pipeline measurement stage of the multi-camera system, the coordinate matrix of the marked points on the platform obtained under the multi-camera system is S. Based on SVD (Singular Value Decomposition), the transformation relationship between matrices Q and S is solved: the rotation matrix R and the translation matrix T. The specific calculation process is as follows:

[0069] Solve for the matrix centroid:

[0070] Calculate the centroids of matrices Q and S respectively as Then: (both matrices are n*n matrices)

[0071]

[0072] Translation matrix:

[0073] Translate matrices Q and S respectively relative to their own centroid positions, and the new matrices are Q' and S'. Then:

[0074]

[0075] SVD decomposition:

[0076] Use matrices Q' and S' to construct matrix M and perform SVD decomposition on it:

[0077]

[0078] Solve for RS and TS

[0079]

[0080] Therefore, according to the transformation relationship between the coordinate matrix of the marked points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system, the pipeline data and pipeline models of each reconstructed pipeline segment in the measurement platform coordinate system are calculated.

[0081] Pipe type splicing

[0082] According to the pipeline data and pipeline models of each pipeline segment in the measurement platform coordinate system, and sequentially splice the images corresponding to each pipeline segment and calculate the pipeline data of each segment, output the total pipeline model and pipeline data. If there are cases of the same bending point or multiple spatial coordinates during the splicing process of the images corresponding to each pipeline segment, then filter out the areas of duplicate pipeline models. The areas of filtering out duplicate pipeline models specifically include: performing averaging processing on the coordinates of each bending point in the image within the bending repetition area, calculating the average coordinate of the averaged bending point coordinates, and using this average coordinate as the coordinate of the bending point in the image within the bending repetition area, so as to correct the bending point coordinates in the total pipeline model, such as Figure 5As shown, there are overlapping measurement areas in pipe section A and pipe section B. The corresponding bending points are Q2, Q2'; Q3, Q3'. The break points K1 of pipe section A and break points K2, K3 of pipe section B do not participate in pipeline splicing. For the overlapping area, the coordinates of each bending point are averaged, and the obtained average coordinates are recorded as the coordinates of that bending point. Taking Figure 5 Q2 and Q2' in [0000184] as an example, let Q2(X, Y, Z) and Q2'(X', Y', Z'), then:

[0083]

[0084] Finally, the position coordinates of this point are Q20(X0, Y0, Z0). In the pipeline model, the coordinates of this type of bending point are corrected, and the images corresponding to each section of the pipeline model are spliced in sequence to output the overall pipeline model and pipeline data.

[0085] The above content is a further detailed description of the present invention in combination with specific preferred embodiments. It cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention pertains, without departing from the concept of the present invention, several simple deductions or substitutions can be made, and all should be regarded as belonging to the protection scope of the present invention.

Claims

1. A pipeline measurement method, characterized in that, It includes the following steps: S1. Calibrate the multi-camera system composed of multiple cameras; S2. According to the calibrated multi-camera system, calculate the image area captured by each camera, so as to calculate the effective area of the image actually captured by the multi-camera system; S3. Calibrate the measurement platform with multiple marking points through the calibrated multi-camera system, and calculate the conversion relationship between the coordinate matrix of the marking points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system; S4. Move the calibrated multi-camera system along the extension direction of the pipeline to be measured on the calibrated measurement platform, and continuously capture multiple images of the pipeline to be measured in the effective area in segments. Among them, each image of the pipeline to be measured in the effective area contains the image of each section of the pipeline and the image of the measurement platform marking points corresponding to each section of the pipeline; S5. Perform image preprocessing on each image of the pipeline to be measured in the effective area, calculate the position area of each section of the pipeline in its respective image, and then use the detection cylinder model to retrieve and match with the center point of the cylinder of the cylinder model to obtain the central axis position of each section of the pipeline, and reconstruct the pipeline models of each section based on this central axis position. Specifically, it includes: by establishing a detection cylinder model, determining the intersection points of the ends of each segmented pipeline and the boundary of the effective area, using this intersection point as the center of the circle, and using the cylinder radius of the detection cylinder model to divide the circular area; if there are three or more small cylinder points in the pipeline image within the circular area outside the effective area of the image captured by the multi-camera system, then this point is a break point. On the contrary, if there is no intersection point between the ends of each segmented pipeline and the boundary, or the intersection point is within the circular area, and the number of points outside the effective area of the image captured by the multi-camera system is less than three, then this point is the end point of each segmented pipeline; S6. According to the conversion relationship between the coordinate matrix of the marking points in the measurement platform coordinate system and the coordinate matrix in the camera coordinate system, calculate the pipeline data and pipeline models of each section of the reconstructed pipeline models in the measurement platform coordinate system; S7. According to the pipeline data and pipeline models of each section of the pipeline models in the measurement platform coordinate system, and sequentially splice the images corresponding to the pipeline models of each section and calculate the pipeline data of each section, and output the total pipeline model and pipeline data; among them, if there are the same bending points or multiple spatial coordinates during the splicing process of the images corresponding to the pipeline models of each section, then filter out the areas of the duplicate pipeline models; filtering out the areas of the duplicate pipeline models specifically includes: Performing averaging processing on the bending point coordinates of the images in the bending duplicate area; Calculating the average coordinate of the bending point coordinates after averaging processing, and using this average coordinate as the bending point coordinate of the image in the bending duplicate area, so as to correct the bending point coordinates in the total pipeline model.

2. The pipeline measurement method according to claim 1, wherein The step S1 includes the following steps: S1-1. Place a calibration board printed with identification points directly below the multi-camera system composed of multiple cameras; S1-2. Obtain the image of the calibration board printed with identification points through the multi-camera system; S1-3. Use the obtained image of the calibration board printed with identification points, and through the principle of multi-view vision calibration, identify the spatial positions of the centers of each identification point to calculate the internal and external parameters of each camera. S1-4. Use the position corresponding to the center point of the calibration board in the calibration board image printed with identification points as the origin of the multi-camera system coordinate system, and unify the internal and external parameters of each camera to the origin of the multi-camera system coordinate system to complete the calibration of the multi-camera system.

3. The pipeline measurement method according to claim 1, characterized in that In step S3, the calibration of the measurement platform provided with multiple marking points includes the following steps: S3-1: Based on the principle of close-range photogrammetry, obtain multiple angular images of the marking points on the measurement platform by the calibrated multi-camera system. S3-2: Perform image edge detection processing on the obtained multiple angular marking point images. S3-3: Perform sub-pixel edge extraction on the multiple angular marking point images after edge detection processing. S3-4: Use the method of ellipse center fitting for the multiple angular marking point images after sub-pixel edge extraction to determine the spatial positions of the centers of each marking point on the measurement platform, and establish a measurement platform coordinate system based on this.

4. The pipeline measurement method according to claim 3, wherein, The obtaining of multiple angular images of the marking points on the measurement platform by the calibrated multi-camera system includes: The calibrated multi-camera system obtains images of the marking points on the measurement platform from three angles of 90°, 60°, and 45° with respect to the plane of the measurement platform.

5. The pipeline measurement method according to claim 1, characterized in that In step S5, the image preprocessing of each image of the pipeline to be measured within the effective area includes the following steps: S5-1. Set a limited gray threshold for each image of the pipeline to be measured within the effective area. S5-2. Perform feature extraction on each image of the pipeline to be measured within the effective area after setting the limited gray threshold.

6. A pipeline measuring device, characterized in that, It includes a measurement platform, a multi-camera system installed on the measurement platform. At least one memory and one processor are included in the multi-camera system. Multiple marking points are provided on the measurement platform. The memory includes at least one executable program stored therein. When the executable program is executed by the processor, it implements the method described in any one of claims 1 to 5.

7. The pipeline measuring device according to claim 6, characterized in that, A backlight is also provided on the measurement platform.

Images