Cylindrical pipe fitting surface detection device

Abstract

The utility model discloses a cylindrical pipe fitting surface detection device, and relates to the technical field of surface detection. Three depth cameras which are perpendicular to one another are arranged in a first detection area and a second detection area corresponding to a disc workbench respectively to obtain depth images corresponding to different surface areas of a cylindrical pipe to be detected; performing depth fusion on the basis of an image splicing algorithm according to the depth images obtained from each angle through an image processor to obtain a full-surface depth image corresponding to the cylindrical pipe fitting to be detected; and calculating the matching degree between the surface curvature and chromaticity of the cylindrical pipe fitting to be detected and the standard cylindrical pipe fitting by adopting a full-surface depth image feature extraction method, and carrying out grade sorting in the sorting area according to the matching degree, thereby realizing full-surface detection on the surface of the cylindrical pipe fitting. Detection errors caused by superposition of local area image detection structures can be avoided, and the detection efficiency and the detection precision of the surface defects of the cylindrical pipe fitting are improved.

Description

Technical Field

[0001] The utility model relates to the technical field of surface detection, and particularly relates to a surface detection device for cylindrical pipe fittings. Background Art

[0002] Cylindrical pipe fittings are commonly used in fields such as vehicles, buildings, and aviation. In the actual production process, due to factors such as manufacturing equipment and production processes, various defects are likely to occur on the surface of the pipe fittings. Workers need to strictly detect the surface of the pipe fittings to evaluate whether they can meet the safety, stability, and precision requirements of the corresponding fields.

[0003] The surface detection devices for cylindrical pipe fittings provided by the prior art often detect each area of the surface of the cylindrical pipe fittings through a single-sided detection light source, and then obtain detection images of multiple local areas of the cylindrical pipe fittings. Then, image processing is performed on each detection image respectively to determine the quality grade of the cylindrical pipe fittings.

[0004] In the process of implementing the present utility model, the inventor found that the related technology has at least the following problems:

[0005] The surface detection device for cylindrical pipe fittings provided by the related technology only performs individual detection on the local areas of the cylindrical pipe fittings. This detection method can only detect local defects on the surface of the cylindrical pipe fittings, such as local paint damage and local depressions, but cannot detect the defects on the entire surface of the cylindrical pipe fittings, such as differences in overall shape, structural differences, and curvature differences. Therefore, the detection accuracy of the defects on the surface of the cylindrical pipe fittings is relatively low. Summary of the Utility Model

[0006] In view of the above problems existing in the related technology, the present utility model provides a surface detection device for cylindrical pipe fittings, which realizes the precise determination of the defects on the entire surface of the cylindrical pipe fittings through the design improvement of the device structure.

[0007] According to one aspect of the embodiments of the present utility model, there is provided a surface detection device for cylindrical pipe fittings, characterized in that it includes a transparent disc workbench that rotates along a preset direction, and a loading area, a first detection area, a second detection area, and a sorting area are sequentially distributed on the surface of the disc workbench along the rotation direction;

[0008] Among them, the loading area is used to place the cylindrical pipe fittings to be detected at the position to be detected, and the sorting area is used to perform grade sorting on the cylindrical pipe fittings after detection;

[0009] The first detection area and the second detection area are perpendicular to each other at the center of the disc table. A dome light source is provided above both the first detection area and the second detection area. A first depth camera, a second depth camera, a first signal trigger, and a third depth camera are perpendicularly arranged on the left, front, rear, and above of the position to be detected corresponding to the first detection area respectively; a fourth depth camera, a fifth depth camera, a second signal trigger, and a sixth depth camera are perpendicularly arranged on the rear, left, right, and below of the position to be detected corresponding to the second detection area respectively; the first depth camera, the second depth camera, the first signal trigger, the third depth camera, the fourth depth camera, the fifth depth camera, the second signal trigger, and the sixth depth camera are electrically connected to an image processor respectively;

[0010] The image processor is configured to, when the first signal trigger detects that there is a cylindrical pipe fitting to be detected at the position to be detected corresponding to the first detection area, start the first depth camera, the second depth camera, and the third depth camera to obtain depth images of corresponding areas of the cylindrical pipe fitting to be detected. When the second signal trigger detects that there is a cylindrical pipe fitting to be detected at the position to be detected corresponding to the second detection area, start the fourth depth camera, the fifth depth camera, and the sixth depth camera to obtain depth images of corresponding areas of the cylindrical pipe fitting to be detected. Then, based on the depth images obtained from various angles, perform depth fusion based on an image stitching algorithm to obtain a full-surface depth image corresponding to the cylindrical pipe fitting to be detected. Finally, calculate the surface curvature and chromaticity of the cylindrical pipe fitting to be detected and the matching degree with the standard cylindrical pipe fitting according to the full-surface depth image, and perform grade sorting in the sorting area according to the matching degree.

[0011] In a preferred embodiment, the first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are respectively used to obtain depth images of at least 1 / 4 cylindrical surface areas of the cylindrical pipe fitting to be detected.

[0012] In a preferred embodiment, the rotation speed of the disc table is less than a rotation speed threshold, and the calculation formula of the rotation speed threshold is:

[0013] ;

[0014] where, is the rotation speed threshold, is the preset depth image accuracy value, is the rotation radius corresponding to the point where the position to be detected is located, is the exposure time of the depth camera.

[0015] In a preferred embodiment, cylindrical grooves for fixing the cylindrical pipe fittings to be detected are provided at each position to be detected on the disk table, and the bottom thickness of the cylindrical grooves is less than 1 mm.

[0016] In a preferred embodiment, the third depth camera acquires depth images of the upper surface of the cylindrical pipe fitting to be detected through the central through hole of the dome light source.

[0017] In a preferred embodiment, the sixth depth camera acquires depth images of the lower surface of the cylindrical pipe fitting to be detected through the bottom of the cylindrical groove.

[0018] In a preferred embodiment, the sixth depth camera uses a telecentric lens.

[0019] In a preferred embodiment, the first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are located on the same horizontal plane.

[0020] Compared with the prior art, a cylindrical pipe fitting surface detection device provided by the present utility model has the following advantages:

[0021] A cylindrical pipe fitting surface detection device provided by the present utility model acquires depth images corresponding to different surface regions of the cylindrical pipe fitting to be detected by respectively arranging three mutually perpendicular depth cameras in the corresponding first detection area and second detection area of the disk table, and then the image processor deeply fuses the depth images acquired from each angle based on the image stitching algorithm to obtain the full-surface depth image corresponding to the cylindrical pipe fitting to be detected. Finally, according to the full-surface depth image, the surface curvature and chromaticity of the cylindrical pipe fitting to be detected are calculated by using the full-surface depth image feature extraction method, and the matching degree with the standard cylindrical pipe fitting is obtained, and the grade sorting is carried out in the sorting area according to the matching degree, so as to realize the full-surface detection of the surface of the cylindrical pipe fitting, avoid the detection error caused by the superposition of local area image detection structures, and improve the detection efficiency and detection accuracy of the surface defects of the cylindrical pipe fitting. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present utility model, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present utility model, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0023] Figure 1 It is a schematic diagram of a cylindrical pipe fitting surface detection device provided by an embodiment of the present utility model.

[0024] Figure 2It is a schematic diagram of another cylindrical pipe fitting surface detection device provided by an embodiment of the present invention.

[0025] Figure 3 It is a working schematic diagram of a cylindrical pipe fitting surface detection device provided by an embodiment of the present invention.

[0026] Figure 4 It is a working schematic diagram of another cylindrical pipe fitting surface detection device provided by an embodiment of the present invention. Specific implementation manners

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for description in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0028] The existing surface detection of cylindrical pipe fittings often focuses on whether the surface paint is damaged and the flatness degree, etc. It conducts partition detection on local areas of the cylindrical pipe fitting through unilateral detection light. This detection method can detect most obvious pipe fitting surface defects, but it cannot accurately detect the overall tiny differences on the pipe fitting surface. For example, the existing surface detection device detects that the adjacent detection areas of the cylindrical pipe fitting are all flat areas, but the surface actually formed by the two detection areas may have a certain curvature, making this common area not completely flat, that is, there is an error between the detection result of the local area of the cylindrical pipe fitting and the detection result of the actual full surface area of the cylindrical pipe fitting, and this error increases with the increase in the number of local areas.

[0029] To solve the above problems existing in the prior art, the present invention provides Figure 1 、 Figure 2 The schematic diagram of a cylindrical pipe fitting surface detection device shown in, such as Figure 1 、 2As shown in the figure, the surface detection device for cylindrical pipe fittings includes a transparent disk workbench 10 that rotates along a preset direction. On the surface of the disk workbench 10, a feeding area 100, a first detection area 200, a second detection area 300, and a sorting area 400 are sequentially distributed along the rotation direction. Among them, the feeding area 100 is used to place the cylindrical pipe fittings to be detected at the position to be detected, and the sorting area 400 is used to perform grade sorting on the cylindrical pipe fittings after detection. The first detection area 200 and the second detection area 300 are perpendicular to each other at the center of the disk workbench 10. A dome light source 500 is provided above both the first detection area 200 and the second detection area 300. A first depth camera 210, a second depth camera 220, a first signal trigger 230, and a third depth camera 240 are respectively and perpendicularly provided on the left, front, rear, and above of the position to be detected corresponding to the first detection area 200. A fourth depth camera 310, a fifth depth camera 320, a second signal trigger 330, and a sixth depth camera 340 are respectively and perpendicularly provided on the rear, left, right, and below of the position to be detected corresponding to the second detection area 300. The first depth camera 210, the second depth camera 220, the first signal trigger 230, the third depth camera 240, the fourth depth camera 310, the fifth depth camera 320, the second signal trigger 330, and the sixth depth camera 340 are respectively electrically connected to an image processor 600. The image processor 600 is configured to, when the first signal trigger 230 detects that there is a cylindrical pipe fitting to be detected at the position to be detected corresponding to the first detection area 200, start the first depth camera 210, the second depth camera 220, and the third depth camera 230 to obtain depth images of corresponding areas of the cylindrical pipe fitting to be detected. When the second signal trigger 330 detects that there is a cylindrical pipe fitting to be detected at the position to be detected corresponding to the second detection area 300, start the fourth depth camera 310, the fifth depth camera 320, and the sixth depth camera 330 to obtain depth images of corresponding areas of the cylindrical pipe fitting to be detected. Then, based on the depth images obtained from each angle, perform depth fusion based on an image stitching algorithm to obtain the full-surface depth image corresponding to the cylindrical pipe fitting to be detected. Finally, calculate the matching degree between the surface curvature and chromaticity of the cylindrical pipe fitting to be detected and those of a standard cylindrical pipe fitting according to the full-surface depth image, and perform grade sorting in the sorting area 400 according to the matching degree.

[0030] In a feasible implementation manner, the matching degree between the surface curvature and chromaticity of the cylindrical pipe fitting to be detected and those of a standard cylindrical pipe fitting can be calculated by using a full-surface depth image feature extraction method.

[0031] Further, the image processor is further configured to display a full-surface depth image and adjust the exposure time and gain parameters of each depth camera.

[0032] The utility model uses an area array depth camera to take multi-angle pictures of the cylindrical pipe fittings to be detected to obtain local depth images corresponding to each area of the cylindrical pipe fittings to be detected, and then fuses each local depth image to obtain a full-surface depth image corresponding to the cylindrical pipe fittings to be detected for defect detection on the surface of the cylindrical pipe fittings.

[0033] Specifically, after the cylindrical pipe fittings to be detected are placed at the detection position in the loading area of the disk workbench, the disk workbench rotates itself to the first detection area. When the first signal trigger detects that there are cylindrical pipe fittings to be detected at the detection position corresponding to the first detection area, the first depth camera, the second depth camera, and the third depth camera are started to obtain depth images of the corresponding areas of the cylindrical pipe fittings to be detected: the first depth camera obtains a depth image of the left area of the cylindrical pipe fittings to be detected. At the same time, the second depth camera obtains a depth image of the front area of the cylindrical pipe fittings to be detected, and the third depth camera obtains a depth image of the upper surface of the cylindrical pipe fittings to be detected; then, the cylindrical pipe fittings to be detected are sent to the second detection area by the rotation of the disk workbench itself. When the second signal trigger detects that there are cylindrical pipe fittings to be detected at the detection position corresponding to the second detection area, the fourth depth camera, the fifth depth camera, and the sixth depth camera are started to obtain depth images of the corresponding areas of the cylindrical pipe fittings to be detected: the fourth depth camera obtains a depth image of the rear area of the cylindrical pipe fittings to be detected. At the same time, the fifth depth camera obtains a depth image of the left area of the cylindrical pipe fittings to be detected, and the third depth camera obtains a depth image of the lower surface of the cylindrical pipe fittings to be detected; after the image processor detects the acquisition of 6 depth images, each depth image is fused by an algorithm into a full-surface depth image corresponding to the cylindrical pipe fittings to be detected. Finally, the surface curvature and chromaticity of the cylindrical pipe fittings to be detected are calculated according to the full-surface depth image, and the matching degree with the standard cylindrical pipe fittings is calculated, and the grade sorting is carried out in the sorting area according to the matching degree.

[0034] It should be noted that in the present utility model, each depth camera in each detection area is vertically distributed respectively, so that different depth cameras can at least obtain one of the depth images of the upper bottom surface, the lower bottom surface, and at least 1 / 4 cylindrical surface of the cylindrical pipe fittings to be detected.

[0035] In a preferred embodiment, the first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are respectively used to obtain depth images of at least 1 / 4 cylindrical surface areas of the cylindrical pipe fittings to be detected.

[0036] In a feasible implementation, each depth camera acquires a single depth image only when the cylindrical pipe fitting to be detected rotates to a preset position in the detection area. At this time, the first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are respectively used to acquire depth images of 1 / 4 cylindrical surface areas of the cylindrical pipe fitting to be detected.

[0037] Taking Figure 3 the shown working schematic diagram as an example, when the cylindrical pipe fitting to be detected is in the first detection area, the first depth camera acquires a depth image a of the A-side cylindrical surface area of the cylindrical pipe fitting to be detected, and the second depth camera acquires a depth image b of the B-side cylindrical surface area. When the cylindrical pipe fitting to be detected is in the second detection area, the fourth depth camera acquires a depth image c of the C-side cylindrical surface area of the cylindrical pipe fitting to be detected, and the fifth depth camera acquires a depth image d of the D-side cylindrical surface area. Subsequently, the image processor can obtain the full-surface cylindrical depth image corresponding to the cylindrical pipe fitting to be detected based on the depth images a, b, c, and d.

[0038] In another feasible implementation, each depth camera starts acquiring depth images for a preset duration when the cylindrical pipe fitting to be detected rotates to the detection area. At this time, the first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are respectively used to acquire depth images of more than 1 / 4 cylindrical surface areas of the cylindrical pipe fitting to be detected. In this depth image acquisition method, there are overlapping areas in the local cylindrical depth images acquired by each depth camera. The image processor can perform fusion of the depth images based on the depth information corresponding to each overlapping area in the subsequent process to obtain the full-surface depth image corresponding to the cylindrical pipe fitting to be detected, thereby reducing the boundary error caused by the fusion of different local depth images.

[0039] Taking Figure 4 the shown working schematic diagram as an example, when the cylindrical pipe fitting to be detected is in the first detection area, the first depth camera acquires a depth image a of the A-side cylindrical surface area of the cylindrical pipe fitting to be detected, and the second depth camera acquires a depth image b of the B-side cylindrical surface area. When the cylindrical pipe fitting to be detected is in the second detection area, the fourth depth camera acquires a depth image c of the C-side cylindrical surface area of the cylindrical pipe fitting to be detected, and the fifth depth camera acquires a depth image d of the D-side cylindrical surface area. After the image processor detects the overlapping area e existing in the depth images a, b, c, and d, high-precision fusion of each depth image can be achieved, avoiding information loss caused by the fusion between local depth images.

[0040] In a preferred embodiment, the rotation speed of the disk workbench is less than the rotation speed threshold, and the calculation formula of the rotation speed threshold is:

[0041] ;

[0042] Wherein, is the rotational speed threshold value, is the preset depth image precision value, is the rotation radius corresponding to the point where the position to be detected is located, is the exposure time of the depth camera.

[0043] To ensure the precision of the depth image obtained by the depth camera, the rotational speed of the disk table is less than the rotational speed threshold value.

[0044] In a preferred embodiment, cylindrical grooves for fixing the cylindrical pipe fittings to be detected are provided at each position to be detected on the disk table, and the bottom thickness of the cylindrical grooves is less than 1 mm.

[0045] The smaller the bottom thickness of the cylindrical groove is, the higher the precision of the depth image obtained by the sixth depth camera for the lower surface of the cylindrical pipe fitting to be detected is.

[0046] In a preferred embodiment, the third depth camera obtains a depth image of the upper surface of the cylindrical pipe fitting to be detected through the central through hole of the dome light source.

[0047] In a preferred embodiment, the sixth depth camera obtains a depth image of the lower surface of the cylindrical pipe fitting to be detected through the bottom of the cylindrical groove.

[0048] In a preferred embodiment, the sixth depth camera adopts a telecentric lens.

[0049] Although the bottom of the cylindrical groove is thin, the transparent material will still cause precision interference to the shooting of the depth camera, and the image will be distorted due to problems such as light refraction. In the present utility model, the sixth depth camera adopts a telecentric lens, which can reduce the distortion of the captured image and eliminate parallax through the imaging principle of the telecentric lens.

[0050] In a preferred embodiment, the first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are located on the same horizontal plane.

[0051] In summary, a surface detection device for cylindrical pipe fittings provided by the present utility model obtains depth images corresponding to different surface areas of the cylindrical pipe fittings to be detected by respectively arranging three depth cameras perpendicular to each other in the first detection area and the second detection area corresponding to the disc workbench. Then, the image processor depth-fuses the depth images obtained from each angle based on the image stitching algorithm to obtain the full-surface depth image corresponding to the cylindrical pipe fittings to be detected. Finally, according to the full-surface depth image, the surface curvature and chromaticity of the cylindrical pipe fittings to be detected are calculated by using the full-surface depth image feature extraction method, and the matching degree with the standard cylindrical pipe fittings is determined. According to the matching degree, the grade sorting is carried out in the sorting area, so as to realize the full-surface detection of the surface of the cylindrical pipe fittings, avoid the detection error caused by the superposition of the local area image detection structures, and improve the detection efficiency and detection accuracy of the surface defects of the cylindrical pipe fittings.

[0052] Although the present utility model has been described in detail above with general descriptions, specific embodiments and experiments, it can be modified or improved on the basis of the present utility model, which is obvious to those skilled in the art. Therefore, these modifications or improvements made without departing from the spirit of the present utility model fall within the scope of protection required by the present utility model.

[0053] After considering the specification and practicing the present utility model herein, those skilled in the art will readily think of other embodiments of the present utility model. The present utility model is intended to cover any variations, uses or adaptations of the present utility model, which follow the general principles of the present utility model and include the common general knowledge or conventional technical means in the technical field not disclosed by the present utility model. It should be understood that the present utility model is not limited to the exact structures and methods described above and shown in the drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A surface inspection device for cylindrical pipe fittings, characterized in that, It includes a transparent disk worktable that rotates along a preset direction. Along the rotation direction on the surface of the disk worktable, there are successively a loading area, a first detection area, a second detection area, and a sorting area; Among them, the loading area is used to place the cylindrical pipe fittings to be detected at the position to be detected, and the sorting area is used to perform grade sorting on the cylindrical pipe fittings after detection; The first detection area and the second detection area are perpendicular to each other at the center of the disk worktable. There are dome lights above both the first detection area and the second detection area. A first depth camera, a second depth camera, a first signal trigger, and a third depth camera are respectively and perpendicularly arranged on the left side, the front side, the rear side, and the upper side of the position to be detected corresponding to the first detection area; A fourth depth camera, a fifth depth camera, a second signal trigger, and a sixth depth camera are respectively and perpendicularly arranged on the rear side, the left side, the right side, and the lower side of the position to be detected corresponding to the second detection area; The first depth camera, the second depth camera, the first signal trigger, the third depth camera, the fourth depth camera, the fifth depth camera, the second signal trigger, and the sixth depth camera are respectively electrically connected to an image processor; The image processor is configured to start the first depth camera, the second depth camera, and the third depth camera to obtain depth images of corresponding areas of the cylindrical pipe fitting to be detected when the first signal trigger detects that there is a cylindrical pipe fitting to be detected at the position to be detected corresponding to the first detection area. When the second signal trigger detects that there is a cylindrical pipe fitting to be detected at the position to be detected corresponding to the second detection area, start the fourth depth camera, the fifth depth camera, and the sixth depth camera to obtain depth images of corresponding areas of the cylindrical pipe fitting to be detected. Then, based on the depth images obtained from each angle, perform depth fusion based on the image stitching algorithm to obtain the full-surface depth image corresponding to the cylindrical pipe fitting to be detected. Finally, calculate the matching degree of the surface curvature and chromaticity of the cylindrical pipe fitting to be detected with those of the standard cylindrical pipe fitting according to the full-surface depth image, and perform grade sorting in the sorting area according to the matching degree.

2. The cylindrical pipe fitting surface detection device according to claim 1, characterized in that, The first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are respectively used to obtain depth images of at least 1 / 4 cylindrical surface areas of the cylindrical pipe fitting to be detected.

3. The cylindrical pipe fitting surface detection device according to claim 1, characterized in that, The rotation speed of the disk worktable is less than a rotation speed threshold, and the calculation formula of the rotation speed threshold is: ； wherein, is the rotational speed threshold value, is the preset depth image accuracy value, is the rotation radius corresponding to the point where the position to be detected is located, is the exposure time of the depth camera.

4. The cylindrical pipe fitting surface detection device according to claim 1, characterized in that There are cylindrical grooves for fixing the cylindrical pipe fittings to be detected at each position to be detected on the disk worktable, and the bottom thickness of the cylindrical grooves is less than 1 mm.

5. The cylindrical pipe fitting surface detection device according to claim 1, characterized in that, The third depth camera obtains a depth image of the upper surface of the cylindrical pipe fitting to be detected through the central through hole of the dome light.

6. The cylindrical pipe fitting surface detection device according to claim 4, characterized in that, The sixth depth camera obtains a depth image of the lower surface of the cylindrical pipe fitting to be detected through the bottom of the cylindrical groove.

7. The cylindrical pipe fitting surface detection device according to claim 6, characterized in that, The sixth depth camera uses a telecentric lens.

8. The cylindrical pipe fitting surface detection device according to claim 1, characterized in that, The first depth camera, the second depth camera, the fourth depth camera, and the fifth depth camera are located on the same horizontal plane.

Images