Polymer monofilament wire diameter detection device and detection method based on machine vision
Through the polymer monofilament diameter detection device based on machine vision, the problem of lack of intelligence and real-time performance of existing detection methods is solved, and the efficient and intelligent detection and data traceability of polymer monofilament production lines are realized, which improves the detection accuracy and The degree of automation of the production line.
Patent Information
- Application Number
- CN202410130774.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-31
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-01-31
AI Technical Summary
The existing polymer monofilament production quality detection methods lack intelligence and real-time performance, low detection efficiency, and cannot achieve large-scale line diameter detection and multi-group line diameter monitoring, resulting in the inability to comprehensively evaluate the quality of polymer monofilament, which reduces the detection results. accuracy and real-time early warning capabilities.
Using a polymer monofilament diameter detection device based on machine vision, the polymer is realized by installing a CCD linear array camera, a rectangular soft light backlight and a high uniformity bar light source, combined with frequency domain low-pass filtering and histogram equalization processing. Real-time shooting and high-precision detection of single-filament images, automatic adjustment of production line parameters to ensure that the line diameter is within the tolerance range, and record detection data for traceability.
Real-time online detection of polymer monofilament production lines is realized, detection efficiency and accuracy is improved, intelligent detection is realized, production process parameters can be adjusted in real time, the influence of human factors is reduced, and quality control and traceability is facilitated through networked data transmission.
Smart Images

Figure CN117928401B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a polymer monofilament wire diameter detection device and a detection method. Background Art
[0002] Machine vision measurement is to use computers, cameras, lenses and other machines to collect images of target objects for processing. It replaces the human eye to make measurements and judgments, and realizes the technology of perception, recognition and understanding of the objective world. The captured target is converted into an image signal through an image capture device and transmitted to a dedicated image processing system to obtain the geometric size information of the captured target. According to the pixel distribution, brightness, color and other information, it is converted into a digital signal to obtain image data that can be recognized and understood by the computer, so as to obtain the geometric size of the polymer monofilament we need to measure. In the process of large-scale repetitive industrial production, the use of machine vision detection methods can greatly improve the efficiency and automation of measurement and detection. Polymer monofilament is an organic non-metallic material with excellent performance, which is widely used in various fields of the national economy. Through the measurement and detection of polymer monofilament production process, the stability of product quality can be ensured, and the product quality can be differentiated and classified.
[0003] The existing polymer monofilament production quality inspection method usually involves manually randomly selecting some production products for precision inspection, appearance inspection, etc. This method lacks intelligence and real-time performance, lacks in-depth analysis of data, has low inspection efficiency, and cannot cover all products. The existing inspection methods are single, mostly traditional physical inspections, lack large-scale wire diameter inspection methods based on machine vision, can only conduct single-filament patrol inspections, and lack standards for comprehensively evaluating the quality of polymer monofilaments based on analyzing multiple sets of wire diameter monitoring data, thereby reducing the impact of human factors on the inspection results and obtaining evaluation and early warning results in a timely manner. Summary of the invention
[0004] The object of the present invention is to provide a polymer monofilament wire diameter detection device and detection method based on machine vision, which can continuously detect the real-time wire diameters of dozens to hundreds of polymer monofilaments during production.
[0005] The technical solution of the present invention is:
[0006] A polymer monofilament wire diameter detection device based on machine vision is characterized in that: a controller is installed at the front end of the left side of the mounting bracket, a camera hood is installed on the right side, two CCD linear array cameras and a telecentric lens used in combination with the two CCD linear array cameras are installed in the camera hood, a rectangular soft light backlight is arranged directly below the two CCD linear array cameras, the soft light backlight projects the outline of the polymer monofilament into the linear array camera, two high-uniformity strip light sources are installed below both sides of the camera hood to improve the ambient light in the camera imaging space, a grating ruler is installed on the rack platform to the right of the rectangular soft light backlight for calibrating the two linear array cameras, and the left side of the rectangular soft light backlight is installed through two thread splitting rods The base is provided with a wire splitter rod; two CCD linear array cameras respectively photograph the left and right parts of the polymer monofilaments uniformly arranged along the wire splitter rod along the width direction of the polymer monofilament arrangement; the linear array camera photographs the image of the polymer monofilament in uniform motion in real time, and transmits the image data to the controller, and the image processing software inside the controller can perform online detection of the polymer monofilament wire diameter based on the image; the controller extracts the wire diameter size of the polymer monofilament using the image of the polymer monofilament photographed by the linear array camera, and controls the speed difference of the two drawing machines in the polymer monofilament production line and the temperature of the hot air oven based on the wire diameter size, so as to ensure that the wire diameter of the finished polymer monofilament is within the tolerance range, thereby realizing the adjustment of the polymer monofilament production working conditions.
[0007] A polymer monofilament diameter detection method using a polymer monofilament diameter detection device based on machine vision is characterized by comprising the following steps:
[0008] (1) First, calibrate the camera. Use a high-precision calibration target grating ruler to calibrate the target. Define the initial values of the camera's internal and external parameters. Initialize the matrix used to store the image coordinates of the calibration mark and the initial pose of the calibration target. During the calibration process, correct the initial values to determine the position of the calibration mark with sub-pixel accuracy and the pose of the calibration target relative to the camera.
[0009] The use of high-uniformity strip light source can adjust the color of the irradiated light according to the color of the single fiber being measured, forming a complementary color with the color of the single fiber being measured, thereby improving the contrast of the image during molding;
[0010] (2) After the camera calibration is completed, the camera starts taking pictures and performs image preprocessing. The image will be affected by different types of noise during the acquisition, transmission, and display processes. The images captured by the camera are seriously affected by discrete pulses, salt and pepper noise, and zero-mean Gaussian noise, which brings many difficulties to subsequent image segmentation, image feature extraction, and image recognition. Therefore, it is necessary to smooth and filter the images collected in real time to eliminate the noise components in the image; try not to damage important information such as image contours and edges, so that the image is clear and convenient for subsequent image processing. This technical solution adopts a frequency domain low-pass filtering method, packages the transformation function into a low-pass filter, and transmits it to the controller after the camera takes a picture. The high-frequency components within a certain range are attenuated in the frequency domain to achieve the purpose of removing noise and smoothing the image. Since the contrast of the collected polymer monofilament image is not high, the image is processed by histogram equalization, the grayscale of the original image is listed, the grayscale pixels of the original histogram are counted, the original histogram and the cumulative distribution function are calculated, the mapping correspondence is determined, and then the grayscale pixels of the new histogram are counted, and the new histogram is calculated to obtain a globally equalized histogram, so that the image grayscale spans a larger range, and the image is automatically enhanced to obtain a The size information of polymer monofilaments in the image is obtained, and the image segmentation is performed to extract the area corresponding to the polymer monofilaments in the image. An image is used as the input segmentation operation, and the bilinear interpolation method is used for adaptive threshold segmentation. The canny edge detection operator and the connection operator are used to extract the sub-pixel contour as the output to improve the accuracy and extraction efficiency of the edge of the polymer monofilament contour in the image. The edge contours on both sides of all polymer monofilaments are selected from the segmentation results, and the Ramer algorithm is used to recursively subdivide the edge contours until the maximum distance from all the line segments to the corresponding contour segments is less than the specified threshold, and the axial contour collected by the camera is fitted into a continuous contour.
[0011] (3) Feature extraction is then performed to extract the contours of the polymer monofilaments on both sides of the image. The one-dimensional boundary is obtained by calculating the local extreme value of the first-order derivative of the gray value contour. The system is only interested in the edges on both sides of the axial direction of the polymer monofilament. The metric pair tool is used to automatically pair all the monofilament contours on the image. To ensure that the return value is the correct pair of edges, the first pair of edges is required to have a positive transformation. In order to eliminate the influence of distortion and obtain results in metric units, the edge points on the image are transformed into the CCD linear camera coordinate system, and the line of sight through the edge points on the image is constructed to complete the task. For two edge points, due to the use of a telecentric lens, the line of sight is returned in earth units between the two points on the line, and all lines of sight are Are parallel, determine that the coordinates of the two points of the line are the same, the polymer monofilament diameter is the distance between the parallel lines passing through this edge point, and this distance value is measured in pixels and converted into metric units; and determine whether this size is within the set tolerance range. If the polymer monofilament size diameter error exceeds the set value, and the number of monofilaments that exceed the tolerance exceeds 60% of the total number of roots, the production line control system based on machine vision adjusts the production line process strategy to ensure the diameter of the polymer monofilament product; in addition, the system will record the size data of all monofilaments, which will be used for the subsequent process to classify the polymer monofilament products by quality, and to facilitate the traceability of later products when they are used, and to find the data of related products.
[0012] The specific method for calibrating the camera is as follows: a plane grid-shaped calibration template is established, and the grid intersection points on the grid-shaped target are used as calibration control points. The coordinate value of this control point on the grid calibration target is known, and this coordinate value is used as the coordinate value of the world coordinate system. After establishing this calibration target, the target is photographed from different angles to obtain multiple calibration template images, and the image coordinates of the control points on the calibration target are detected. Then, the analytical solution of the camera parameters is estimated based on the target coordinates and image coordinates of the control points. Then, the deflection coefficient is obtained by minimum variance, and the final iterative result is obtained according to nonlinear programming.
[0013] The internal parameters of the camera include lens focal length, lens distortion coefficient, sensor pixel width, height, movement speed, etc.; the external parameters include the rotation angle and translation amount from the camera coordinates to the world coordinates XYZ three axes.
[0014] The present invention: (1) The production of single filaments can realize online real-time detection in large quantities, improve detection efficiency, and reduce the use of personnel. For online real-time detection, embedding the visual measurement system into the production line to make the measurement speed consistent with the production rhythm of the production line is the key to improving the detection efficiency. The execution time of visual measurement depends to a large extent on the speed of the underlying image processing (image smoothing, filtering, segmentation, etc.). Therefore, using dedicated hardware to implement an environment-independent processing algorithm can greatly improve the image processing speed.
[0015] (2) Realize intelligent detection. Machine vision measurement technology enables the transition of single-filament wire diameter detection from traditional detection to online, dynamic, active, and large-scale real-time detection, and adjusts process (operating condition) parameters in real time online based on detection data.
[0016] (3) Achieve high-precision detection with a measurement accuracy of 0.02mm, which can meet the measurement requirements of various wire diameters of single-filament products.
[0017] (4) Measurement data is networked, and detection data can be transmitted and traced. Measurement data is transmitted to the network cloud in real time for multiple end users to query data, which is convenient for production process personnel to adjust production process parameters. It is also convenient for product quality control. The specifications and dimensions of different batches of products can be queried in the cloud, which is convenient for tracing product quality problems.
[0018] (5) Realize flexible measurement technology. Products of different specifications and sizes can be measured using the same visual measurement system.
[0019] (6) Highly sensitive wire breakage detection and alarm. If one polymer monofilament breaks in a large batch of polymer monofilaments, the system will lose the wire diameter detection data of this wire and will determine that the wire is broken and issue an alarm to remind production personnel to deal with it in time to reduce losses.
[0020] The advantages of the present invention also lie in:
[0021] (1) Realize online intelligent and high-precision detection of large quantities of polymer monofilaments.
[0022] (2) Measurement data is networked and detection data can be transmitted and traced.
[0023] (3) Realize the polymer monofilament flexibility measurement technology, and adjust the process (operating condition) parameters online in real time according to the real-time production conditions.
[0024] (4) Highly sensitive line break detection and alarm. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0026] Figure 1 It is a front view of the detection device of the present invention.
[0027] Figure 2 It is a left side view of the detection device of the present invention.
[0028] Figure 3 It is a top view of the detection device of the present invention.
[0029] Figure 4 It is an axial view of the detection device of the present invention.
[0030] Figure 5 It is a control principle block diagram of the detection device of the present invention.
[0031] In the figure: 1. Rack; 2. Camera and controller mounting bracket; 3. Alarm light; 4. Wire separator; 5. Rectangular soft backlight; 6. Wire guide mounting bracket; 7. Controller tray; 8. Controller; 9. Camera hood; 10. Telecentric lens; 11. Highly uniform strip light source; 12. Detected single wire; 13. Grating ruler; 14. CCD linear array camera on the left; 15. CCD linear array camera on the right. DETAILED DESCRIPTION
[0032] A polymer monofilament wire diameter detection device based on machine vision, wherein a controller is installed at the front end of the left side of the mounting bracket, and a camera hood is installed on the right side. Two CCD linear array cameras and a telecentric lens used in combination with the two CCD linear array cameras are installed in the camera hood. A rectangular soft light backlight is arranged directly below the two CCD linear array cameras. The soft light backlight projects the outline of the polymer monofilament into the linear array camera. Two high-uniformity strip light sources are installed below both sides of the camera hood to improve the ambient light in the camera imaging space. A grating ruler is installed on the rack platform on the right side of the rectangular soft light backlight for calibrating the two linear array cameras. The left side of the rectangular soft light backlight is installed through two split wire rod mounting seats. Install the wire rod; two CCD linear array cameras respectively photograph the left and right parts of the polymer monofilaments evenly arranged along the wire rod along the width direction of the polymer monofilament arrangement; the linear array camera takes the image of the polymer monofilament in uniform motion in real time, and transmits the image data to the controller, and the image processing software inside the controller can perform online detection of the polymer monofilament diameter based on the image; the controller uses the polymer monofilament image taken by the linear array camera to extract its wire diameter size, and based on this wire diameter size, controls the speed difference of the two drawing machines in the polymer monofilament production line and the temperature of the hot air oven, so as to ensure that the wire diameter of the polymer monofilament finished product is within the tolerance range, thereby realizing the adjustment of the polymer monofilament production conditions.
[0033] The visual inspection device of this technical solution is installed between the third and fourth drawing machines in the monofilament production line. According to the production requirements, there will be dozens to hundreds of monofilaments uniformly pulled in parallel. As shown in the figure, the monofilament to be tested passes through the wire rod 4. Since the monofilament 12 to be tested has tension on it in this state, there are evenly distributed grooves on the wire rod. The monofilament 12 to be tested passes through the wire rod and is evenly arranged to enter the shooting area of the CCD linear array camera. The rectangular soft light backlight 5 is located at the bottom of the monofilament 12 to be tested, and the CCD linear array cameras 13 and 14; the telecentric lens 10 and the high uniform strip light source are located above the monofilament 12 to be tested. The high uniform strip light source is used to improve the ambient light irradiation conditions. The backlight illuminates the monofilament 12 to be tested and forms an image in the CCD linear array camera. Since the visual width of the arrangement of the monofilament 12 to be tested is relatively large, the technical solution uses two high-resolution and high-frame-rate CCD linear array cameras to detect the monofilament 12 to be tested in the left and right areas respectively, so that the detection efficiency is higher, the image processing speed is higher, and the detection accuracy is more accurate. The device provides a frame 1, on which a camera and controller mounting bracket 2 are mounted. The front side of the bracket is cantilevered, and an alarm light is mounted near the support. Figure 2 As shown: a controller 8 is installed at the front end of the left side of the mounting bracket, and a camera hood 9 is installed on the right side, in which two CCD line array cameras and a telecentric lens used with the two CCD line array cameras are installed, a rectangular soft light backlight is arranged directly below the CCD line array cameras 13 and 14, and two high-uniformity strip light sources 11 are installed below the two sides of the hood, a grating ruler 13 is installed on the platform of the frame 1 on the right side of the rectangular soft light backlight 5, and a splitter rod is installed on the left side of the rectangular soft light backlight 5 through two splitter rod mounting seats.
[0034] When any one or several of the tested monofilaments 12 are broken, the alarm light on the device will sound an alarm to remind workers to deal with production abnormalities in a timely manner. When the device detects that the diameter of the tested monofilament exceeds the allowable error range with the set value, the production device of the production line of the tested monofilament 12 adjusts the process parameters of the production device online in real time based on the visual inspection data to improve the stability of the diameter of the tested monofilament 12. The principle is as follows: Figure 5 : Image-based visual production line control method.
[0035] Before the device is tested, we first need to calibrate the camera. The application scenario of this technical solution has high requirements for real-time performance, so a high-precision calibration target grating ruler 13 is used to calibrate the target, define the initial values of the camera's internal parameters (lens focal length, lens distortion coefficient, sensor pixel width, height, movement speed, etc.) and external parameters (rotation angle and translation of the three axes from the camera coordinates to the world coordinates XYZ), and initialize the matrix used to store the image coordinates of the calibration mark and the initial posture of the calibration target. During the calibration process, the initial value is corrected to determine the position of the calibration mark with sub-pixel accuracy and the position of the calibration target relative to the camera. The specific method is: establish a plane grid-shaped calibration template, and the grid intersection on the grid-shaped target is used as the calibration control point. The coordinate value of this control point on the grid calibration target is known, and this coordinate value is used as the coordinate value of the world coordinate system; after establishing this calibration target, the target is photographed from different angles and positions to obtain multiple calibration template images, and the image coordinates of the control points on the calibration target are detected, and then the analytical solution of the camera parameters is estimated based on the target coordinates and image coordinates of the control points. Then, the deflection coefficient is obtained through the minimum square error, and the final iterative result is obtained according to nonlinear programming.
[0036] Since the polymer monofilaments to be measured will have different colors during the production process, the technical solution uses a high-uniformity strip light source to adjust the color of the irradiated light according to the color of the monofilaments to be measured, forming a complementary color with the color of the monofilaments to be measured, which can improve the contrast of the image during molding.
[0037] After the camera calibration is completed, the camera starts taking pictures and preprocessing the images. The images will be affected by different types of noise during the acquisition, transmission, and display process. The images captured by the camera are seriously affected by discrete pulses, salt and pepper noise, and zero-mean Gaussian noise, which brings many difficulties to the subsequent image segmentation, image feature extraction, and image recognition. Therefore, it is necessary to smooth and filter the images collected in real time to eliminate the noise components in the image; try not to damage important information such as image contours and edges, so that the image is clear and convenient for subsequent image processing. This technical solution adopts a frequency domain low-pass filtering method to package the transformation function into a low-pass filter, which is transmitted to the controller when the camera takes a picture, and the high-frequency components within a certain range are attenuated in the frequency domain to achieve the purpose of removing noise and smoothing the image. Since the contrast of the collected polymer monofilament image is not high, the image is processed by histogram equalization. The grayscale of the original image is listed, the grayscale pixels of the original histogram are counted, the original histogram and the cumulative distribution function are calculated, the mapping correspondence is determined, and the grayscale pixels of the new histogram are counted, and the new histogram is calculated to obtain a globally equalized histogram, so that the image grayscale spans a larger range and automatically enhances the image. In order to obtain the size information of the polymer monofilament in the image, we must perform image segmentation to extract the area corresponding to the polymer monofilament in the image. An image is used as the input segmentation operation, and the bilinear interpolation method is used for adaptive threshold segmentation. The canny edge detection operator and the connection operator are used to extract the sub-pixel contour as the output. Improve the accuracy and extraction efficiency of the edge of the polymer monofilament contour in the image. Select the edge contours on both sides of all polymer monofilaments from the segmentation results, and use the Ramer algorithm to recursively subdivide the edge contours until the maximum distance from all the line segments to the corresponding contour segments is less than the specified threshold, and fit the axial contour collected by the camera into a continuous contour. Then feature extraction is performed to extract the contours of the polymer monofilaments on both sides of the image. The one-dimensional boundary is obtained by calculating the local extreme value of the first-order derivative of the gray value contour. The system is only interested in the edges on both sides of the axial direction of the polymer monofilament. The metric pair tool is used to automatically pair all the monofilament contours on the image. To ensure that the return value is the correct pair of edges, the first pair of edges is required to have a positive transformation. In order to eliminate the influence of distortion and obtain results in metric units, the edge points on the image are transformed into the CCD linear camera coordinate system, and the line of sight through the edge points on the image is constructed to complete it.For two edge points, due to the use of telecentric lenses, the line of sight returns in earth units of the two points on the line, all lines of sight are parallel, and the coordinates of the two points of the line are determined to be the same. The diameter of the polymer monofilament is the distance between the parallel lines passing through the edge points. This distance value is measured in pixels and converted into metric units; and it is determined whether this size is within the set tolerance range. If the error of the polymer monofilament size line diameter exceeds the set value, and the number of monofilaments that exceed the tolerance exceeds 60% of the total number of roots, the production line control system based on machine vision adjusts the production line process strategy to ensure the line diameter of the polymer monofilament product. In addition, the system will record the size data of all monofilaments for the subsequent process to classify the polymer monofilament products by quality. And it is convenient for the later products to be traced when used. Find the data of related products.
Claims
1. A polymer monofilament diameter detection method using a polymer monofilament diameter detection device based on machine vision, characterized in that: The polymer monofilament wire diameter detection device based on machine vision comprises: a controller is installed at the front end of the left side of the mounting bracket, a camera hood is installed on the right side, two CCD linear array cameras and a telecentric lens used with the two CCD linear array cameras are installed in the camera hood, a rectangular soft light backlight is arranged directly below the two CCD linear array cameras, the soft light backlight projects the outline of the polymer monofilament into the linear array camera, two high-uniformity strip light sources are installed below both sides of the camera hood to improve the ambient light in the camera imaging space, a grating ruler is installed on the rack platform to the right of the rectangular soft light backlight for calibrating the two linear array cameras, and two thread splitter rod mounting seats are provided on the left side of the rectangular soft light backlight. Install the wire splitter rod; two CCD linear array cameras respectively photograph the left and right parts of the polymer monofilaments evenly arranged along the wire splitter rod along the width direction of the polymer monofilament arrangement; the linear array camera takes the image of the polymer monofilaments moving at a constant speed in real time, and transmits the image data to the controller, and the image processing software inside the controller can perform online detection of the diameter of the polymer monofilament based on the image; the controller uses the polymer monofilament images photographed by the linear array camera to extract the wire diameter size, and based on this wire diameter size, controls the speed difference of the two drawing machines in the polymer monofilament production line and the temperature of the hot air oven, so as to ensure that the wire diameter of the polymer monofilament finished product is within the tolerance range, thereby realizing the adjustment of the polymer monofilament production working conditions; The detection method includes the following steps: (1) First, calibrate the camera. Use a high-precision calibration target grating ruler to calibrate the target. Define the initial values of the camera's internal and external parameters. Initialize the matrix used to store the image coordinates of the calibration mark and the initial pose of the calibration target. During the calibration process, correct the initial values to determine the position of the calibration mark with sub-pixel accuracy and the pose of the calibration target relative to the camera. The use of high-uniformity strip light source can adjust the color of the irradiated light according to the color of the single fiber being measured, forming a complementary color with the color of the single fiber being measured, thereby improving the contrast of the image during molding; (2) After the camera calibration is completed, the camera starts taking pictures and performs image preprocessing. The frequency domain low-pass filtering method is used to package the transformation function into a low-pass filter. When the camera takes a picture and transmits it to the controller, the high-frequency components within a certain range are attenuated in the frequency domain to achieve the purpose of removing noise and smoothing the image. Since the contrast of the collected polymer monofilament image is not high, the image is processed by histogram equalization. The grayscale of the original image is listed, the grayscale pixels of the original histogram are counted, the original histogram and the cumulative distribution function are calculated, the mapping correspondence is determined, and then the grayscale pixels of the new histogram are counted, and the new histogram is calculated to obtain a globally equalized histogram, so that the image grayscale spans a larger range. To obtain the size information of the polymer monofilaments in the image, the image segmentation is performed to extract the area corresponding to the polymer monofilaments in the image. An image is used as the input segmentation operation, and the bilinear interpolation method is used for adaptive threshold segmentation. The canny edge detection operator and the connection operator are used to extract the sub-pixel contour as the output to improve the accuracy and extraction efficiency of the edge of the polymer monofilament contour in the image. The edge contours on both sides of all polymer monofilaments are selected from the segmentation results, and the Ramer algorithm is used to recursively subdivide the edge contours until the maximum distance from all the obtained line segments to the corresponding contour segments is less than the specified threshold, and the axial contours collected by the camera are fitted into a continuous contour. (3) Feature extraction is then performed to extract the contours of the polymer monofilaments on both sides of the image. The one-dimensional boundary is obtained by calculating the local extreme value of the first-order derivative of the gray value contour. The system is only interested in the edges on both sides of the axial direction of the polymer monofilament. The metric pair tool is used to automatically pair all the monofilament contours on the image. To ensure that the return value is the correct pair of edges, the first pair of edges is required to have a positive transformation. In order to eliminate the influence of distortion and obtain results in metric units, the edge points on the image are transformed into the CCD linear camera coordinate system, and the line of sight through the edge points on the image is constructed to complete the task. For two edge points, due to the use of a telecentric lens, the line of sight is returned in earth units between the two points on the line, and all lines of sight are Are parallel, determine that the coordinates of the two points of the line are the same, the polymer monofilament diameter is the distance between the parallel lines passing through this edge point, and this distance value is measured in pixels and converted into metric units; and determine whether this size is within the set tolerance range. If the polymer monofilament size diameter error exceeds the set value, and the number of monofilaments that exceed the tolerance exceeds 60% of the total number of roots, the production line control system based on machine vision adjusts the production line process strategy to ensure the diameter of the polymer monofilament product; in addition, the system will record the size data of all monofilaments, which will be used for the subsequent process to classify the polymer monofilament products by quality, and to facilitate the traceability of later products when they are used, and to find the data of related products.
2. The polymer monofilament diameter detection method using a polymer monofilament diameter detection device based on machine vision according to claim 1, characterized in that: The specific method for calibrating the camera is as follows: a plane grid-shaped calibration template is established, and the grid intersection points on the grid-shaped target are used as calibration control points. The coordinate value of this control point on the grid calibration target is known, and this coordinate value is used as the coordinate value of the world coordinate system. After establishing this calibration target, the target is photographed from different angles to obtain multiple calibration template images, and the image coordinates of the control points on the calibration target are detected. Then, the analytical solution of the camera parameters is estimated based on the target coordinates and image coordinates of the control points. Then, the deflection coefficient is obtained by minimum variance, and the final iterative result is obtained according to nonlinear programming.
3. The polymer monofilament diameter detection method using a polymer monofilament diameter detection device based on machine vision according to claim 1, characterized in that: The internal parameters of the camera include lens focal length, lens distortion coefficient, sensor pixel width, height, and movement speed; the external parameters include the rotation angle and translation amount from the camera coordinates to the world coordinates XYZ three axes.
