A strain gauge defect recognition device and method based on machine vision

Abstract

The invention discloses a machine vision-based strain gauge defect recognition device and method, comprising a vibration isolation table, an XY axis motion unit, a camera mounting bracket, an AB axis camera adjustment mechanism and a shooting camera; the XY axis motion unit and the camera mounting bracket are arranged on the vibration isolation table, the AB axis camera adjustment mechanism is arranged on the top of the camera mounting bracket, the AB axis camera adjustment mechanism is located directly above the XY axis motion unit, the shooting camera is arranged on the AB axis camera adjustment mechanism, and the shooting camera lens is arranged toward the top of the XY axis motion unit. The stability of the strain gauge and the shooting camera during the movement process is improved, the image acquisition effect is improved, and thus the accuracy of defect recognition is improved.

Description

Technical Field

[0001] The invention belongs to the field of resistive strain gauges and relates to a strain gauge defect recognition device and method based on machine vision. Background Art

[0002] The resistance strain gauge (hereinafter referred to as the strain gauge) is a sensitive element that uses the "resistance-strain" principle and is the core element of the force sensor. During the production process, the strain gauge needs to go through the appearance inspection process. The traditional appearance inspection method is to identify it with the human eye under a microscope or fluorescent light. This method can still be used for some strain gauges with relatively large specifications and simple textures. Once the demand for products increases or the size of the strain gauge decreases, manual inspection cannot meet the production requirements. In addition, the cost of manual inspection continues to rise. The application of intelligent recognition systems based on machine vision in strain gauge production is an inevitable trend and has gradually been applied in the industrial field. The defect recognition system based on machine vision uses optical means to obtain the image of the object being measured and digitize it, and then compares, analyzes and judges it through some algorithms. Therefore, according to the function, the system can be divided into two parts: the optical part and the image processing part. At present, the 08FG strain gauge has the characteristics of large production volume and strict requirements on the error detection rate. The traditional production process is mainly completed by manual defect detection. There are problems such as high labor cost, high labor intensity, low degree of automation, and difficulty in controlling the defective rate of products. Therefore, it is particularly important to realize the full automation of the appearance defect detection of strain gauges.

[0003] At present, there are many types of automated equipment based on machine vision at home and abroad, and the technical levels vary greatly, but most of them are used to identify larger objects. Specialized detection equipment suitable for strain gauge production is rare, and there are even fewer automated equipment that can truly recognize the appearance defects of strain gauges. The appearance size of a single strain gauge is small, with a length and width of only a few millimeters. The smallest detection object inside is the sensitive grid, which is only a few tens of microns wide. Various defects are randomly distributed throughout the sensitive grid, and the shape, size, and position of the defects are uncertain. The minimum defect size is only 5 microns, which places strict requirements on the system performance of machine vision. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a strain gauge defect recognition device and method based on machine vision, which improves the stability of the strain gauge during movement and the shooting camera, improves the image acquisition effect, and thus improves the accuracy of defect recognition.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A machine vision-based strain gauge defect recognition device, comprising a vibration isolation table, an XY axis motion unit, a camera mounting bracket, an AB axis camera adjustment mechanism and a shooting camera;

[0007] The XY-axis motion unit and the camera mounting bracket are arranged on a vibration isolation table, the AB-axis camera adjustment mechanism is arranged on the top of the camera mounting bracket, the AB-axis camera adjustment mechanism is located directly above the XY-axis motion unit, the shooting camera is arranged on the AB-axis camera adjustment mechanism, and the shooting camera lens is arranged toward the top of the XY-axis motion unit.

[0008] Preferably, a plurality of air springs are arranged inside the vibration isolation platform in the horizontal, longitudinal and vertical directions respectively.

[0009] Preferably, the XY axis motion unit comprises a driving mechanism, a slide rail and a slider, the slider is slidably connected to the slide rail, and the driving mechanism adopts a magnetic driving mechanism.

[0010] Furthermore, a linear grating ruler is arranged on the slide rail.

[0011] Preferably, the AB axis camera adjustment mechanism includes a pitch adjustment mechanism, a yaw adjustment mechanism, a vertical translation adjustment mechanism and a base which are connected in sequence, and the shooting camera is arranged on the pitch adjustment mechanism.

[0012] Furthermore, the pitch adjustment mechanism is connected to the yaw adjustment mechanism by a rotation mechanism; the yaw adjustment mechanism is connected to the vertical translation adjustment mechanism by a flipping mechanism; the vertical translation adjustment mechanism is connected to the base by a linear motion mechanism, and the rotation axis of the rotation mechanism, the flip axis of the flip mechanism and the movement direction of the linear motion mechanism are perpendicular to each other.

[0013] Preferably, a strain gauge positioning fixture is provided on the top of the XY axis motion unit, the strain gauge positioning fixture comprises a strain gauge pressure plate and a strain gauge positioning plate, the strain gauge positioning fixture is a square ring, and the strain gauge positioning fixture is connected to the strain gauge positioning plate by a pin.

[0014] Furthermore, a plurality of negative pressure adsorption holes are arranged on the strain gauge positioning plate, the negative pressure adsorption holes are connected to the negative pressure device, and the negative pressure adsorption holes are located inside the area surrounded by the strain gauge pressure plate.

[0015] A strain gauge defect recognition method of a machine vision-based strain gauge defect recognition device according to any one of the above items comprises the following process:

[0016] The strain gauge is placed on the top of the XY-axis motion unit. Driven by the XY-axis motion unit, it performs uniaxial motion according to the set position and speed. When it reaches the edge of the strain gauge, the motion axis perpendicular to the uniaxial direction starts to move, realizing the strain gauge wrapping operation.

[0017] After completing the line feed, the strain gauge continues to move along the set route, and so on, until the collection of the entire page of strain gauges is completed;

[0018] During the movement of the strain gauge, the camera collects image information of the strain gauge and pre-processes the image information through the image acquisition card. After the pre-processing is completed, the data is identified by defect algorithm;

[0019] When the camera completes the image acquisition of the entire strain gauge, the strain gauge returns to the coordinate origin. When the defect calculation is completed, the strain gauge is removed to complete the single strain gauge defect identification process.

[0020] Preferably, the process of placing the strain gauge on top of the XY-axis motion unit is as follows: the strain gauge positioning fixture is driven by the XY-axis motion unit to move to the loading station, the strain gauge pressure plate above the strain gauge positioning fixture is removed, and the direction of the strain gauge is identified at the same time, and the strain gauge to be measured is fixed on the strain gauge positioning fixture, and then the strain gauge pressure plate is placed on the strain gauge positioning fixture again, and the strain gauge pressure plate flattens the edge of the strain gauge and starts the negative pressure device.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] The present invention improves the vibration isolation performance by being arranged on a vibration isolation table, effectively avoiding the influence of vibration of surrounding equipment on the imaging function of the shooting camera; the AB axis camera adjustment mechanism facilitates the three-axis precise adjustment of the camera, and at the same time greatly improves the adjustment efficiency of the camera, ensures the imaging quality and puzzle quality, and can adjust the camera and the lens so that the camera axis is perpendicular to the strain gauge placement platform, and ensures that the lens and the object maintain a precise distance.

[0023] Furthermore, the vibration isolation table is provided with a plurality of air springs distributed in the three directions of X, Y and Z, respectively, which can effectively block the vibration transmitted from the frame.

[0024] Furthermore, the XY axis motion unit utilizes a magnetic drive mechanism to move, making the movement more stable and meeting the requirements of accurate photography.

[0025] Furthermore, the XY axis motion unit uses a linear grating ruler to achieve precise positioning, with higher motion accuracy, meeting the requirements of accurate photography.

[0026] Furthermore, by adjusting the pitch adjustment mechanism, the pitch position of the shooting camera can be achieved; by adjusting the yaw adjustment mechanism, the yaw position of the shooting camera can be achieved; by adjusting the vertical translation adjustment mechanism, the vertical height of the shooting camera can be achieved; by adjusting the AB axis camera adjustment mechanism, two rotations and one translation in space, a total of three degrees of freedom of movement of the shooting camera can be achieved, so as to adapt to the relative position of the shooting camera and the strain gauge positioning tooling, thereby ensuring that the shooting camera can take clear and accurate photos.

[0027] Furthermore, the strain gauge positioning tooling uses a strain gauge pressure plate to press the strain gauge, which effectively avoids the warping of the strain gauge in its natural state, avoids the distortion of collected data and unclear photos caused by the warping of the strain gauge, and thus avoids image blur caused by reflection and refraction of the light source, thereby improving the detection accuracy and the accuracy of defect identification.

[0028] Furthermore, the strain gauge positioning tooling uses negative pressure adsorption to fix the entire plate of strain gauges, thereby achieving stable positioning of the strain gauges and effectively ensuring the accuracy of photography.

[0029] The method of the present invention collects strain gauge images through a camera, analyzes the collected images, and performs defect algorithm recognition, thereby realizing an automatic defect recognition function, which has great advantages in realizing batch and automated operations. At the same time, it effectively avoids the occurrence of misdetection and missed detection, greatly improves the efficiency of strain gauge defect recognition and detection, and reduces the intensity of manual labor. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a structural schematic diagram of the defect identification device of the present invention;

[0031] Figure 2 It is a structural schematic diagram of the AB axis camera adjustment mechanism of the present invention;

[0032] Figure 3 It is a structural schematic diagram of the strain gauge positioning tooling of the present invention.

[0033] Including: 1. Frame, 2. Vibration isolation table, 3. Y-axis motion unit, 4. X-axis motion unit, 5. Strain gauge positioning fixture, 6. Shooting camera, 7. AB axis camera adjustment mechanism, 8. Camera mounting bracket, 9. Pitch adjustment mechanism, 10. Yaw adjustment mechanism, 11. Vertical translation adjustment mechanism, 12. Strain gauge positioning plate, 13. Negative pressure adsorption hole, 14. Strain gauge pressure plate. DETAILED DESCRIPTION

[0034] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0035] like Figure 1As shown, the strain gauge defect recognition device based on machine vision described in the present invention includes a frame 1, a vibration isolation table 2, an XY axis motion unit, a strain gauge positioning tool 5, a camera mounting bracket 8, an AB axis camera adjustment mechanism 7 and a shooting camera 6.

[0036] The bottom of the vibration isolation table 2 is connected to the frame 1 by screws, and the vibration isolation table 2 can effectively suppress the environmental vibration from being transmitted to the camera components. The upper platform of the vibration isolation table 2 is respectively provided with air springs in the three directions of X, Y, and Z, which realizes the high rigidity of the workbench and has a high level of vibration isolation performance, which can effectively block the vibration transmitted from the frame and effectively avoid the influence of the vibration of the surrounding equipment on the imaging function of the shooting camera 6.

[0037] The XY-axis motion unit includes a Y-axis motion unit 3 and an X-axis motion unit 4. The Y-axis motion unit 3 is connected to the vibration isolation table 2 by screws to realize the Y-direction motion of the strain gauge positioning fixture 5. The X-axis motion unit 4 is connected to the Y-axis motion unit 3 by screws to realize the X-direction motion of the strain gauge positioning fixture 5. The strain gauge positioning fixture 5 is connected to the X-axis motion unit by screws. Through the movement of the X- and Y-axis motion units, the X- and Y-direction motions of the strain gauge positioning fixture 5 are realized, thereby realizing the movement of the entire plate of strain gauges.

[0038] The XY axis motion unit includes a driving mechanism, a slide rail and a slider. The slider is slidably connected to the slide rail. The driving mechanism adopts a magnetic driving mechanism. A linear grating ruler is arranged on the slide rail to achieve precise positioning, so that the movement is smoother and the movement accuracy is higher, meeting the requirements of accurate photography. At the same time, it also ensures the accuracy of image acquisition and improves the imaging effect of the shooting camera 6. On this basis, the shooting efficiency is improved, thereby improving the defect recognition efficiency of the whole machine.

[0039] The camera mounting bracket 8 is connected to the vibration isolation table 2 by screws. The AB axis camera adjustment mechanism 7 is connected to the camera mounting bracket 8 by a connector, and the shooting camera 6 is connected to the AB axis camera adjustment mechanism 7 by a connector, and the shooting camera 6 lens is set toward the top of the XY axis motion unit.

[0040] like Figure 2As shown, the AB axis camera adjustment mechanism 7 is composed of a pitch adjustment mechanism 9, a yaw adjustment mechanism 10, a vertical translation adjustment mechanism 11 and a base. The pitch adjustment mechanism 9 is connected to the yaw adjustment mechanism 10 by a rotation mechanism; the yaw adjustment mechanism 10 is connected to the vertical translation adjustment mechanism 11 by a flip mechanism; the vertical translation adjustment mechanism 11 is connected to the base by a linear motion mechanism, and the rotation axis of the rotation mechanism, the flip axis of the flip mechanism and the movement direction of the linear motion mechanism are perpendicular to each other. By adjusting the pitch adjustment mechanism 9, the pitch posture of the shooting camera 6 can be achieved; by adjusting the yaw adjustment mechanism 10, the yaw posture of the shooting camera 6 can be achieved; by adjusting the vertical translation adjustment mechanism 11, the vertical height of the shooting camera 6 can be achieved; by adjusting the AB axis camera adjustment mechanism 7, two rotations and one translation in space can be achieved for the shooting camera 6, a total of three degrees of freedom, so as to adapt to the relative position of the shooting camera 6 and the strain gauge positioning fixture 5, thereby ensuring that the shooting camera 6 can take clear and accurate photos.

[0041] By adjusting the AB axis camera adjustment mechanism 7, the camera axis is perpendicular to the strain gauge placement platform, and the lens is kept at an accurate distance from the object, thereby ensuring the image quality and puzzle quality. Compared with the traditional adjustment mechanism, the AB axis camera adjustment mechanism 7 has the characteristics of high precision and easy adjustment. Through coarse adjustment and fine adjustment, the three-axis precise adjustment of the shooting camera 6 is achieved, and the adjustment efficiency and adjustment accuracy of the camera are greatly improved.

[0042] A strain gauge positioning fixture 5 is provided on the top of the XY axis motion unit. Since the strain gauge itself is very thin, it will be warped when placed on a horizontal plane. This will cause distortion in the collected data, making the photos unclear and not conducive to image collection. Figure 3 As shown, the strain gauge positioning fixture 5 includes a strain gauge pressing plate 14 and a strain gauge positioning plate 12. The strain gauge positioning fixture 5 is a square ring, and the strain gauge positioning fixture 5 is connected to the strain gauge positioning plate 12 by a pin. The strain gauge pressing plate 14 is used to press the strain gauge, which effectively avoids the warping of the strain gauge in the natural state, avoids the distortion of the collected data caused by the warping of the strain gauge, and avoids unclear photography, thereby avoiding image blur caused by reflection and refraction of the light source, improving the detection accuracy, and thus improving the accuracy of defect recognition.

[0043] The strain gauge positioning plate 12 is provided with a plurality of negative pressure adsorption holes 13, which are connected to a negative pressure device and are located inside the area surrounded by the strain gauge pressing plate 14. The negative pressure adsorption holes 13 are used to position and fix the entire strain gauge with the help of the strain gauge pressing plate 14 and the positioning pins, so that the strain gauge is accurately and firmly positioned, effectively ensuring the accuracy of the photographing, and at the same time, facilitating the loading and unloading of the operator.

[0044] The strain gauge defect recognition method of the strain gauge defect recognition device based on machine vision includes the following processes:

[0045] When the system is powered on, first power on the host computer, mechanical system and corresponding motors. After opening the host computer software, the strain gauge positioning fixture 5 is driven by the X and Y axis motion units to move to the manual loading station. Manually remove the strain gauge pressure plate 14 above the strain gauge positioning fixture 5, identify the direction of the strain gauge, and fix the strain gauge to be tested on the strain gauge positioning fixture 5. Then re-place the strain gauge pressure plate 14 on the strain gauge positioning fixture 5, and then press the start button to complete the manual loading.

[0046] After the loading is completed and the start button is pressed, the strain gauge positioning fixture 5 with the strain gauge is driven by the X and Y axis motion units to perform uniaxial motion according to the position and speed set by the host computer. When it reaches the edge of the strain gauge, the vertical motion axis starts to move to realize the strain gauge line break operation. After completing the line break, the strain gauge positioning fixture 5 continues to move uniaxially, and so on, moving along the route set by the host computer until the collection of the entire page of strain gauges is completed. At the same time, during the movement of the strain gauge, the camera 6 collects image information of the strain gauge, and pre-processes the image information through the image acquisition card. After the pre-processing is completed, the data is sent to the host computer for defect algorithm recognition of the image analysis software. When the camera 6 completes the image collection of the entire page of strain gauges, the strain gauge positioning fixture 5 returns to the coordinate origin, that is, the manual loading station, waits for the image processing and manual confirmation results, analyzes the collected image, compares the current strain gauge defect image with the image in the defect database, determines whether it is qualified, and marks the strain gauge according to the judgment result. When the host computer completes the defect calculation and manually confirms the calculation results, the strain gauge is manually removed and waits for the next version of the strain gauge to be manually loaded, and the process is repeated alternately. Due to the use of the host computer communication and control method, the present invention can realize data interaction and interconnection with other devices, providing a good foundation for realizing full-line automation of strain gauge production and smart factories.

[0047] The image analysis software has a learning and memory function. By identifying different strain gauge defects, the software will learn and memorize each different type of strain gauge defect, and continuously expand the types and specific characteristics of defects in its database, thereby achieving the purpose of efficient and accurate detection.

[0048] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A strain gauge defect recognition device based on machine vision, characterized in that: It includes a vibration isolation table (2), an XY axis motion unit, a camera mounting bracket (8), an AB axis camera adjustment mechanism (7) and a shooting camera (6); The XY-axis motion unit and the camera mounting bracket (8) are arranged on the vibration isolation table (2), the AB-axis camera adjustment mechanism (7) is arranged on the top of the camera mounting bracket (8), the AB-axis camera adjustment mechanism (7) is located directly above the XY-axis motion unit, the shooting camera (6) is arranged on the AB-axis camera adjustment mechanism (7), and the shooting camera (6) is arranged with its lens facing the top of the XY-axis motion unit; The AB axis camera adjustment mechanism (7) comprises a pitch adjustment mechanism (9), a yaw adjustment mechanism (10), a vertical translation adjustment mechanism (11) and a base which are connected in sequence, and the shooting camera (6) is arranged on the pitch adjustment mechanism (9); A strain gauge positioning fixture (5) is provided on the top of the XY-axis motion unit. The strain gauge positioning fixture (5) comprises a strain gauge pressure plate (14) and a strain gauge positioning plate (12). The strain gauge positioning fixture (5) is a square ring. The strain gauge positioning fixture (5) is connected to the strain gauge positioning plate (12) by a pin. A plurality of negative pressure adsorption holes (13) are provided on the strain gauge positioning plate (12), the negative pressure adsorption holes (13) are connected to a negative pressure device, and the negative pressure adsorption holes (13) are located inside the area surrounded by the strain gauge pressure plate (14).

2. The machine vision-based strain gauge defect recognition device according to claim 1, characterized in that: A plurality of air springs are arranged inside the vibration isolation platform (2) in the horizontal, longitudinal and vertical directions respectively.

3. The machine vision-based strain gauge defect recognition device according to claim 1, characterized in that: The XY axis motion unit comprises a driving mechanism, a slide rail and a slider. The slider is slidably connected to the slide rail, and the driving mechanism adopts a magnetic driving mechanism.

4. The machine vision-based strain gauge defect recognition device according to claim 3, characterized in that: A linear grating ruler is arranged on the slide rail.

5. The machine vision-based strain gauge defect recognition device according to claim 1, characterized in that: The pitch adjustment mechanism (9) is connected to the yaw adjustment mechanism (10) by a rotation mechanism; the yaw adjustment mechanism (10) is connected to the vertical translation adjustment mechanism (11) by a flipping mechanism; the vertical translation adjustment mechanism (11) is connected to the base by a linear motion mechanism, and the rotation axis of the rotation mechanism, the flip axis of the flip mechanism and the motion direction of the linear motion mechanism are perpendicular to each other.

6. A strain gauge defect recognition method according to any one of claims 1 to 5, characterized in that: The process includes: The strain gauge is placed on the top of the XY-axis motion unit. Driven by the XY-axis motion unit, it performs uniaxial motion according to the set position and speed. When it reaches the edge of the strain gauge, the motion axis perpendicular to the uniaxial direction starts to move, realizing the strain gauge wrapping operation. After completing the line feed, the strain gauge continues to move along the set route, and so on, until the collection of the entire page of strain gauges is completed; During the movement of the strain gauge, the camera (6) collects image information of the strain gauge, and pre-processes the image information through an image acquisition card. After the pre-processing is completed, the data is identified by a defect algorithm; When the camera (6) completes the image acquisition of the entire strain gauge, the strain gauge returns to the coordinate origin. When the defect calculation is completed, the strain gauge is removed to complete the single strain gauge defect identification process.

7. The method for identifying strain gauge defects based on machine vision according to claim 6, characterized in that: The process of placing the strain gauge on the top of the XY-axis motion unit is as follows: the strain gauge positioning fixture (5) is driven by the XY-axis motion unit to move to the loading station, the strain gauge pressure plate (14) above the strain gauge positioning fixture (5) is removed, and the direction of the strain gauge is identified at the same time, and the strain gauge to be measured is fixed on the strain gauge positioning fixture (5), and then the strain gauge pressure plate (14) is placed on the strain gauge positioning fixture (5) again, and the strain gauge pressure plate (14) flattens the edge of the strain gauge, and the negative pressure device is started.

Images