Roughness detection device based on computer binocular vision

Abstract

The utility model discloses a roughness detection device based on computer binocular vision. The device comprises a mechanical support device, an image acquisition device and a calibration device, wherein the image acquisition device is arranged on the mechanical supporting device and is used for acquiring a surface image of the precast concrete member based on a plurality of image acquisition devices; the mechanical supporting device is arranged at the bottom of the image acquisition device and is used for adjusting the position and angle of the image acquisition device and fixing the image acquisition device; the calibration device is arranged on the surface of the to-be-measured target object and used for measuring the internal and external parameters of the image acquisition device based on the surface image of the precast concrete member. According to the device, the computer vision technology is applied to concrete crack detection, the problem that the actual engineering for accurately and rapidly measuring the surface roughness of the concrete in a large-area, non-contact mode does not exist in the civil engineering field at present is solved, and the device can adapt to various complex measurement working conditions. Meanwhile, the device is high in universality and wide in application range.

Description

Technical Field

[0001] The present application relates to the technical field of building structure detection, and in particular to a roughness detection device based on computer binocular vision. Background Art

[0002] Prefabricated building structures are widely promoted because of their high construction efficiency, noise reduction, dust prevention, environmental pollution reduction, energy conservation, and green construction. However, since different components must be directly installed at the construction site after they are produced in the factory, the joints and connections of these components are often prone to deformation due to stress, water seepage caused by loose joints, etc. due to their poor shear and bending resistance, and even seriously affect the safety performance of the overall structure. In addition, when repairing and reinforcing existing concrete structures, the treatment method of the bonding surface is an important factor affecting the shear bearing capacity of the concrete bonding surface. Experiments show that when other variables are controlled, the shear resistance of the bonding surface is proportional to the degree of roughness. Therefore, strictly controlling the surface roughness of concrete components is of great significance to improving the bonding performance and shear resistance of new and old concrete surfaces.

[0003] At present, the commonly used methods for detecting the surface roughness of precast concrete components include sand filling method, three-dimensional laser scanning method, depth gauge method, stylus method and stylus profilometer, depth camera method, standard plate comparison method, etc. At present, the research on concrete crack detection based on machine vision technology mainly focuses on the research of algorithms related to image processing, including image denoising, crack edge detection, threshold segmentation and other algorithms. These algorithms and image vision technologies have more or less problems such as small detection area, low accuracy, high cost, and difficult operation. Therefore, there is no universal detection method that can solve most engineering problems. In addition, the three-dimensional displacement measurement method based on machine vision technology is mainly limited to laboratory measurement, and the accuracy in actual engineering applications needs to be improved. Especially in the case of complex test environment, problems such as long recognition time and feature matching errors are prone to occur.

[0004] Therefore, in the existing roughness detection technology of building structures, there are problems such as small detection area, low accuracy, high cost and difficult operation of ordinary detection methods, and the three-dimensional displacement measurement method based on machine vision technology is mainly limited to low accuracy in laboratory measurement, especially in complex test environments, and is prone to technical problems such as long recognition time and feature matching errors. Summary of the invention

[0005] In view of the shortcomings of the prior art mentioned above, the purpose of the present application is to provide a roughness detection device based on computer binocular vision, which is used to solve the problems in the existing roughness detection technology of building structures, namely, the ordinary detection method has a small detection area, low precision, high cost, and great difficulty in operation, and the three-dimensional displacement measurement method based on machine vision technology is mainly limited to the low accuracy in laboratory measurement, especially in complex test environments, it is prone to problems such as long recognition time and feature matching errors.

[0006] To achieve the above-mentioned purpose and other related purposes, the present application provides a roughness detection device based on computer binocular vision, the device comprising: a mechanical support device, an image acquisition device, and a calibration device; the image acquisition device is arranged on the mechanical support device, and is used to obtain a surface image of a precast concrete component based on multiple image acquisition devices; the mechanical support device is arranged at the bottom of the image acquisition device, and is used to adjust the position and angle of the image acquisition device and fix the image acquisition device; the calibration device is arranged on the surface of a target object to be measured, and is used to measure the internal and external parameters of the image acquisition device based on the surface image of the precast concrete component.

[0007] In one embodiment of the present application, the image acquisition device includes: a photosensitive chip, a visual camera, and a synchronization trigger; the photosensitive chip is arranged in an M*N manner.

[0008] In one embodiment of the present application, the visual camera is connected to the photosensitive chip; the visual camera adopts an area array black and white camera; the visual camera includes: an image acquisition card, a camera sensor, and an optical lens.

[0009] In one embodiment of the present application, the synchronization trigger is connected to the visual camera locking mechanism; the synchronization trigger adopts a hard trigger method; the synchronization trigger couples the trigger source and the trigger input terminal through a built-in trigger connection node, and presses the synchronization switch to perform synchronous shooting.

[0010] In one embodiment of the present application, the image acquisition card is arranged inside the visual camera, and is used to collect the surface image of the precast concrete component; the camera sensor is arranged inside the visual camera, and is used to convert the light image on the photosensitive surface into an electrical signal proportional to the light image; the optical lens is arranged on one side of the visual camera, and is used to adjust the clarity of the precast concrete component to be measured through optical imaging.

[0011] In one embodiment of the present application, two visual cameras are used; the distance between the visual cameras is greater than or equal to 0.2 meters and less than or equal to 1 meter.

[0012] In one embodiment of the present application, the calibration device uses a double-layer calibration plate; the double-layer calibration plate includes: a first panel and a second panel; the first panel and the second panel are bonded together.

[0013] In one embodiment of the present application, the first panel is made of glass; the first panel is set as a glass substrate; the second panel is made of aluminum oxide and is used to fix the first panel.

[0014] In one embodiment of the present application, the double-layer calibration plate adopts a 12*9 array.

[0015] In one embodiment of the present application, the mechanical support device includes: a bracket, an adjustable pan-tilt head and a clamping device; the bracket is arranged at the bottom of the image acquisition device, and is used to support and fix the image acquisition device; the adjustable pan-tilt head is connected to the bracket, and is used to adjust the height, horizontal angle, and pitch angle of the image acquisition device to provide a basis for any shooting position; the clamping device is connected to the adjustable pan-tilt head, and is used to adjust other control knobs to fix the device.

[0016] As described above, the roughness detection device based on computer binocular vision described in this application has the following beneficial effects:

[0017] The roughness detection device based on computer binocular vision of the present application includes four systems: a mechanical support device, an image acquisition device, a calibration device and a matching software development system. The present application applies computer vision technology to concrete crack detection, solving the practical engineering problem that there is no large-area, non-contact, accurate and fast measurement of concrete surface roughness in the current civil engineering field, and can adapt to a variety of complex measurement conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Shown is a schematic diagram of the overall structure of a roughness detection device based on computer binocular vision as described in an embodiment of the present application.

[0019] Figure 2 Shown is a schematic diagram of the structural composition of a roughness detection device based on computer binocular vision described in an embodiment of the present application.

[0020] Figure 3 Shown is a schematic diagram of the structure of a mechanical support device for a roughness detection device based on computer binocular vision as described in an embodiment of the present application.

[0021] Figure 4 Shown is a side view of a calibration device for a roughness detection device based on computer binocular vision as described in an embodiment of the present application.

[0022] Figure 5Shown is a front view of a calibration device for a roughness detection device based on computer binocular vision as described in an embodiment of the present application.

[0023] Description of Figure Numbers:

[0024] Serial number name

[0025] 1 Roughness detection device based on computer binocular vision

[0026] 11 Mechanical support device

[0027] 12 Image acquisition device

[0028] 13 Calibration device

[0029] 111 Bracket

[0030] 112 Adjustable pan / tilt

[0031] 113 Clamping device

[0032] 121 Photosensitive chip

[0033] 122 Vision Camera

[0034] 123 Synchronous Trigger

[0035] 131 Double-layer calibration board

[0036] 1311 First Panel

[0037] 1312 Second Panel

[0038] 2 Building Component Roughness Detection System

[0039] 21 Shooting process detection and control subsystem

[0040] 22 Stereo Matching Subsystem

[0041] 23 Point Cloud Processing Subsystem

[0042] 24 Roughness detection subsystem

[0043] 231 Point cloud filtering processing module

[0044] 232 Point cloud smoothing processing module

[0045] 233 Point Cloud Simplification Module DETAILED DESCRIPTION

[0046] The present application is further described below in conjunction with the accompanying drawings, but the protection scope of the present application is not limited to the following description.

[0047] The following describes the embodiments of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0048] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0049] The roughness detection device based on computer binocular vision provided in the following embodiments of the present application solves the problems in the existing roughness detection technology of building structures, namely, the common detection method has a small detection area, low accuracy, high cost, and great difficulty in operation, and the three-dimensional displacement measurement method based on machine vision technology is mainly limited to low accuracy in laboratory measurement, especially in complex test environments, and is prone to problems such as long recognition time and feature matching errors.

[0050] The embodiment of the present application provides a roughness detection device based on computer binocular vision, through a mechanical support device, an image acquisition device, a calibration device and a matching software development system, the main components of which are an adjustable pan and stop device, a binocular camera, a computer, a calibration plate and a calibration support device, an artificial light source, a component to be detected, etc. The present application applies computer vision technology to concrete crack detection, solving the practical engineering problem that there is no large-scale, non-contact, accurate and fast measurement of concrete surface roughness in the civil engineering field at present, and can adapt to a variety of complex measurement conditions, can optimize and improve the calibration method, feature extraction algorithm, and image matching algorithm, and provide a theoretical reference for application in the practical engineering field.

[0051] The following will describe in detail the implementation principle of a roughness detection device based on computer binocular vision of this embodiment in conjunction with the accompanying drawings, so that those skilled in the art can understand the roughness detection device based on computer binocular vision of this embodiment without creative work.

[0052] See also Figure 1 and Figure 2 , respectively showing the overall structure diagram of the roughness detection device based on computer binocular vision according to the embodiment of the present application and the structural composition diagram of the roughness detection device based on computer binocular vision according to the embodiment of the present application. Figure 1 and Figure 2 As shown, the roughness detection device 1 based on computer binocular vision includes: a mechanical support device 11, an image acquisition device 12, a calibration device 13 and a matching building component roughness detection system 2. The image acquisition device 12 is arranged on the mechanical support device 11, and is used to obtain the surface image of the precast concrete component based on multiple image acquisition devices; the mechanical support device 11 is arranged at the bottom of the image acquisition device 12, and is used to adjust the position and angle of the image acquisition device 12 and fix the image acquisition device 12; the calibration device 13 is arranged on the surface of the target object to be measured, and is used to measure the internal and external parameters of the image acquisition device 12 based on the surface image of the precast concrete component.

[0053] In this embodiment, the roughness detection device 1 based on computer binocular vision further includes: a computer, a calibration plate and a calibration support device, an artificial light source, a precast concrete component to be detected, and the like.

[0054] See also Figure 3 , which is a schematic diagram of the structure of the mechanical support device of the roughness detection device based on computer binocular vision according to the embodiment of the present application. Figure 3 As shown, the mechanical support device 11 includes: a bracket 111, an adjustable gimbal 112 and a clamping device 113. The bracket 111 is arranged at the bottom of the image acquisition device 12, and is used to support and fix the image acquisition device 12; the adjustable gimbal 112 is connected to the bracket 111, and is used to adjust the height, horizontal angle, and pitch angle of the image acquisition device 12, so as to provide a basis for any shooting position; the clamping device 113 is connected to the adjustable gimbal, and is used to adjust other control knobs to fix the device 1.

[0055] Please continue reading Figure 1 and Figure 2 The image acquisition device 12 includes: a photosensitive chip 121, a visual camera 122, and a synchronization trigger 123. The photosensitive chips 121 are arranged in an M*N manner.

[0056] The visual camera 122 is connected to the photosensitive chip 121; the visual camera 122 adopts a black and white area array camera; the visual camera 122 includes: an image acquisition card, a camera sensor, and an optical lens. The image acquisition card is arranged inside the visual camera 122 to collect the surface image of the precast concrete component; the camera sensor is arranged inside the visual camera 122 to convert the light image on the photosensitive surface into an electrical signal proportional to the light image; the optical lens is arranged on one side of the visual camera 122 to adjust the clarity of the precast concrete component to be measured through optical imaging. The visual camera 122 adopts two binocular visual cameras; the distance between the visual cameras 122 is greater than or equal to 0.2 meters and less than or equal to 1 meter.

[0057] The synchronization trigger 123 is connected to the locking mechanism of the visual camera 122; the synchronization trigger 123 adopts a hard trigger mode; the synchronization trigger couples the trigger source and the trigger input terminal through a built-in trigger connection node, and presses the synchronization switch to perform synchronous shooting.

[0058] In this embodiment, according to the actual shooting scene and shooting distance, the distance between the two visual cameras 122 is set to be flexibly adjusted between 0.2 meters and 1.0 meters. The adjustable gimbal 112 can also provide height adjustment of the visual camera 122, horizontal angle adjustment, and pitch angle adjustment of the visual camera 122 to ensure that the viewing angle of the binocular visual camera 122 can be rotated at almost any angle in space, which has greater flexibility for shooting components at any position in actual engineering; in addition, the stable clamping device 113 can ensure the stability of the device when adjusting other control knobs.

[0059] In this embodiment, the photosensitive chip 121 preferably uses the IMX183 photosensitive chip provided by Sony. The array camera chips are arranged in an M×N manner. Therefore, the photographed photo is 5472×3648 pixels, and the size of each pixel is only 2.4μm×2.4μm, and the image has an extremely high resolution.

[0060] Similarly, there are many types of visual cameras 122, and an array black and white camera can be used. It can flexibly realize autonomous operation, soft triggering and hard triggering. The exposure control time can be set by a program or hard triggered by an adjustable image acquisition trigger device. Due to the strict accuracy requirements of binocular shooting, unexpected errors may be caused by environmental factors (ground disturbances, light changes, etc.) in a very short time. Therefore, a synchronization trigger 123 is needed to ensure the simultaneity of shooting. The synchronization trigger 123 can be hard-triggered by coupling the trigger source and the trigger input terminal through the built-in trigger connection node, and pressing the synchronization switch can achieve synchronous shooting.

[0061] See also Figure 4 and Figure 5 , respectively showing a side view of the calibration device of the roughness detection device based on computer binocular vision according to the embodiment of the present application and a front view of the calibration device of the roughness detection device based on computer binocular vision according to the embodiment of the present application. Figure 4 and Figure 5 As shown, the calibration device 1 uses a double-layer calibration plate 131; the double-layer calibration plate 131 includes: a first panel 1311 and a second panel 1312; the first panel 1311 and the second panel 1312 are bonded. The first panel 1311 is made of glass; the first panel 1311 is set as a glass substrate; the second panel 1312 is made of alumina material, and is used to fix the first panel 1311.

[0062] In this embodiment, the double-layer calibration plate 131 preferably uses a calibration plate with a size of 12x9 grids and a side length of 15 mm. Figure 3 It can be seen that the double-layer calibration plate 131 is divided into two layers, the first layer is a glass substrate for fixing the panel, and the second layer is an opaque, non-reflective aluminum oxide panel. Therefore, the double-layer calibration plate 131 is composed of a glass substrate and an aluminum oxide panel.

[0063] Please continue reading Figure 2 The roughness detection device 1 based on computer binocular vision also includes: a building component roughness detection system 2 matched with the above-mentioned image acquisition hardware device.

[0064] In this embodiment, the building component roughness detection system includes 2: a shooting process detection and control subsystem 21, a stereo matching subsystem 22, a point cloud processing subsystem 23, and a roughness detection subsystem 24. Among them, the point cloud processing subsystem 23 includes: a point cloud filtering processing module 231, a point cloud smoothing processing module 232, and a point cloud simplification module 233.

[0065] Specifically, the shooting process detection and control subsystem can use the pylon Viewer64-Bit software, which provides a programming interface independent of the camera interface being used, so that applications created using pylon can be ported to another camera interface technology. The software can also be used to adjust the frequency, time interval and shooting system of the camera, adjust the camera aperture and exposure time according to the actual shooting scene, select the shutter trigger method (soft trigger and hard trigger), and adjust the shutter mode and many other supporting functions.

[0066] The following is an example of the detection of the surface roughness of a precast concrete component by a roughness detection device based on computer binocular vision.

[0067] In this embodiment, first determine a more accurate shooting distance according to the model of the binocular vision camera 122, set up the precast concrete slab to be tested at this distance, mark and fix the position of the precast concrete component to be tested. Through the mechanical support device 11, the position and angle of the binocular vision camera 122 can be adjusted, and the precast concrete component to be tested can be adjusted to the center of the visual camera 122 shooting. After adjusting to the appropriate position, tighten the three directions of each pair of visual cameras 122. After adjusting the aperture of the visual camera 122 to the maximum, adjust the aperture size of the two visual cameras 122 to moderate light and dark according to the size of the natural light source in the environment (the aperture size of the two visual cameras 122 is adjusted to the same). If the ambient light is too dark or there is reflection, ghosting, etc., it is necessary to block the natural light source and use a symmetrical artificial light source. Then adjust the focal length of the binocular vision camera 122 until the object to be tested can be clearly displayed in the picture.

[0068] Furthermore, the binocular vision camera 122 can be calibrated. First, the double-layer calibration plate 131 is placed on the surface of the precast concrete component to be measured. At this time, the double-layer calibration plate 131 is within the field of view of the binocular vision camera 122, so the double-layer calibration plate 131 in the binocular vision camera 122 is clear. Then the double-layer calibration plate 131 is slowly moved away from the surface of the object to be measured until the image displayed in the shooting process detection and control subsystem 21 begins to become blurred, and the position where the image just begins to become blurred is recorded. The range from this position to the position of the object to be measured is the field of view of the binocular vision camera 122. The double-layer calibration plate 131 is placed within the field of view of the visual camera 122, and the position of the double-layer calibration plate 131 is continuously adjusted to shoot 20 to 40 sets of binocular vision image pairs, and the internal and external parameters of the binocular vision camera 122 are measured through the image pairs.

[0069] Finally, through the supporting building component roughness detection system 2, the entire shooting process can be monitored in real time, the storage path can be controlled, and the grid can be manually set to more accurately determine the relative position of the precast concrete components to be tested in the image; at the same time, taking into account the shooting and video recording functions, it is also well applicable to binocular shooting of dynamic scenes.

[0070] In summary, the roughness detection device based on computer binocular vision of the present application includes four systems: a mechanical support device, an image acquisition device, a calibration device and a matching software development system. The main components include an adjustable pan and stop device, a binocular camera, a computer, a calibration plate and a calibration support device, an artificial light source, and components to be detected. The present application applies computer vision technology to concrete crack detection, which solves the practical engineering problem that there is no large-area, non-contact, accurate, and fast measurement of concrete surface roughness in the civil engineering field, and can adapt to a variety of complex measurement conditions. At the same time, the structure of this device is highly versatile, has a wide range of applications, and has high practical value.

[0071] The above embodiments are only for explaining the principle and effect of the present application, and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.

Claims

1. A roughness detection device based on computer binocular vision, characterized in that: The device comprises: a mechanical support device, an image acquisition device, and a calibration device; The image acquisition device is arranged on the mechanical support device and is used to acquire the surface image of the precast concrete component based on a plurality of image acquisition devices; The mechanical support device is arranged at the bottom of the image acquisition device, and is used to adjust the position and angle of the image acquisition device and fix the image acquisition device; The calibration device is arranged on the surface of the target object to be measured, and is used to measure the internal and external parameters of the image acquisition device based on the surface image of the precast concrete component.

2. The roughness detection device based on computer binocular vision according to claim 1 is characterized in that: The image acquisition device includes: a photosensitive chip, a visual camera, and a synchronous trigger; The photosensitive chips are arranged in an M*N manner.

3. The roughness detection device based on computer binocular vision according to claim 2 is characterized in that: The visual camera is connected to the photosensitive chip; the visual camera adopts an area array black and white camera; The visual camera comprises: an image acquisition card, a camera sensor, and an optical lens.

4. The roughness detection device based on computer binocular vision according to claim 2 is characterized in that: The synchronization trigger is connected to the visual camera locking mechanism; The synchronization trigger adopts a hard trigger mode; The synchronization trigger couples the trigger source and the trigger input terminal through the built-in trigger connection node, and the synchronization switch is pressed to perform synchronous shooting.

5. The roughness detection device based on computer binocular vision according to claim 3 is characterized in that: The image acquisition card is arranged inside the visual camera and is used to acquire the surface image of the precast concrete component; The camera sensor is arranged inside the visual camera and is used to convert the light image on the photosensitive surface into an electrical signal proportional to the light image; The optical lens is arranged at one side of the visual camera and is used to adjust the clarity of the precast concrete component to be measured through optical imaging.

6. The roughness detection device based on computer binocular vision according to claim 3 is characterized in that: The visual cameras are 2; The distance between the visual cameras is greater than or equal to 0.2 meters and less than or equal to 1 meter.

7. The roughness detection device based on computer binocular vision according to claim 1 is characterized in that: The calibration device adopts a double-layer calibration plate; The double-layer calibration plate comprises: a first panel and a second panel; The first panel and the second panel are attached to each other.

8. The roughness detection device based on computer binocular vision according to claim 7 is characterized in that: The first panel is made of glass; the first panel is set as a glass substrate; The second panel is made of anodized aluminum and is used to fix the first panel.

9. The roughness detection device based on computer binocular vision according to claim 7, characterized in that: The double-layer calibration board adopts a 12*9 array.

10. The roughness detection device based on computer binocular vision according to claim 1, characterized in that: The mechanical support device comprises: a bracket, an adjustable pan / tilt head and a clamping device; The bracket is arranged at the bottom of the image acquisition device, and is used to support and fix the image acquisition device; The adjustable gimbal is connected to the bracket and is used to adjust the height, horizontal angle, and pitch angle of the image acquisition device to provide a basis for any shooting position; The clamping device is connected to the adjustable pan / tilt head and is used to adjust other control knobs to fix the device.

Images