A high-precision measurement method for ceramic splitter aperture based on machine vision

Abstract

The invention discloses a high-precision measurement method for the aperture of a ceramic splitting tool based on machine vision, which relates to the technical field of ceramic splitting tool aperture measurement and comprises the following steps: fixing the ceramic splitting tool to a splitting tool fixture and lighting the ceramic splitting tool; using a side camera to collect height profile information of the splitting tool; controlling an upper camera to move to a focusing range; the upper camera performs height position compensation according to a focus clarity to achieve focus clarity; selecting ROI and performing sub-pixel extraction on the aperture profile of the splitting tool; using a Huber method to fit a profile edge curve and perform measurement; a host computer saves the measurement result, controls a ejector pin to eject unqualified NG materials, and completes the measurement; the invention adopts the cooperation of a side camera and an upper camera, records visual height difference information of different splitting tools through the side camera, and then uses a motion control card to control the upper camera to move to a corresponding compensation position according to the height difference, so as to achieve an effect of automatic clear focusing, thereby improving the measurement accuracy and efficiency of the ceramic splitting tool.

Description

Technical Field

[0001] The invention relates to the technical field of ceramic splitting cutter aperture measurement, and in particular to a high-precision ceramic splitting cutter aperture measurement method based on machine vision. Background Art

[0002] A ceramic splitter is an axisymmetric ceramic tool with a vertical hole. Its end face diameter is about 0.11-0.12mm, and the hole center diameter is about 0.02-0.03mm. It is a precision microstructure ceramic component. During the processing of the ceramic splitter, it is necessary to measure the size of the aperture of each ceramic splitter end face. The traditional measurement method is to clamp the splitter on a proprietary fixture, then put it under a microscope with a smart camera and manually click the mouse to take points to measure the aperture. The ceramic splitter workpiece is extremely small, and a slightly floating end face will be slightly out of focus, making detection difficult. In addition, manual labor is prone to fatigue and distraction during long-term measurement work, which will lead to problems such as blurred focus and inaccurate point selection, resulting in inaccurate measurement results and low measurement efficiency. Therefore, the present invention proposes a high-precision ceramic splitter aperture measurement device based on machine vision to realize automated measurement operations, thereby saving labor costs and further improving production efficiency. Summary of the invention

[0003] The purpose of the present invention is to provide a high-precision measurement method for the aperture of a ceramic splitting cutter based on machine vision, so as to solve the problems existing in the prior art mentioned in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A high-precision measurement method for the aperture of a ceramic splitter based on machine vision comprises the following steps:

[0006] S1: Fix the ceramic splitting knife to the splitting knife fixture and polish the ceramic splitting knife;

[0007] S2: Autofocus:

[0008] (1) The side camera collects the height profile information of the riving cutter;

[0009] (2) Control the upper camera to move within the focus range;

[0010] (3) The upper camera performs height compensation according to the focus clarity to achieve focus clarity;

[0011] S3: Aperture measurement:

[0012] (1) Select ROI and perform sub-pixel extraction on the aperture contour of the wedge;

[0013] (2) Use Huber method to fit the contour edge curve and make measurements;

[0014] (3) The host computer saves the measurement results, controls the ejector to eject unqualified NG materials, and completes the measurement.

[0015] Preferably, the specific steps of fitting the contour edge curve in S3 are:

[0016] The least squares method is used to fit the curve. The idea of ​​the least squares method is to require the observed value y i The weighted sum of squares of the deviations is minimized, that is, the fitting function The value of is the smallest, which is to minimize the mean square error as the cost function to determine the optimal function matching of the data;

[0017] The Huber loss function is used to improve the robustness of the least squares method to noise outliers. The function formula is as follows:

[0018]

[0019] In the above formula, α is the residual, which is the distance between the generated straight line and the discrete point in straight line fitting; δ is the residual threshold, which is used to determine whether it is an error point.

[0020] Preferably, the lighting in S2 is performed by installing a highlight light source under the ceramic splitting knife to highlight the holes on the ceramic splitting knife.

[0021] Preferably, the side camera in S2 uses an industrial camera with a telecentric coaxial metallographic lens to ensure clear imaging of the height profile of the riving cutter.

[0022] Preferably, the upper camera in S2 uses a 4K industrial smart camera with a microscope lens to build a top visual environment.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. The present invention adopts the cooperation of the side camera and the upper camera. The side camera records the visual height difference information of different splitting knives, and then uses the motion control card to control the upper camera to move to the corresponding compensation position according to the height difference, so as to achieve the effect of automatic clear focus, thereby improving the measurement accuracy and efficiency of the ceramic splitting knives.

[0025] 2. The present invention realizes curve fitting of the workpiece by introducing Huber loss function and least square method, which can reduce the influence of abnormal points on the fitting circle equation without losing the information of important contour points. The fitting result of the quasi-circular area is ideal, thereby improving the accuracy of the ceramic splitting cutter aperture measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 Flow chart of the measuring method of the present invention.

[0027] Figure 2 It is an end view of the ceramic splitting knife of the present invention.

[0028] Figure 3 It is a schematic diagram of measuring the height of the ceramic splitting cutter of the present invention.

[0029] Figure 4 This is a schematic diagram of extracting ROI according to the present invention.

[0030] Figure 5 This is a schematic diagram of the front and back of the edge of the ceramic splitting cutter contour fitted by the Huber method in the present invention.

[0031] Figure 6 It is a schematic diagram of the structure of the measuring device of the present invention. DETAILED DESCRIPTION

[0032] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0033] See also Figure 1-6 , the present invention provides the following technical solutions:

[0034] A high-precision measurement method for the aperture of a ceramic splitter based on machine vision comprises the following steps:

[0035] S1: Fix the ceramic splitting knife to the splitting knife fixture and polish the ceramic splitting knife.

[0036] Since there will be grinding marks left over from previous operations around the end surface diameter of the ceramic splitter, direct processing will have a certain impact. Therefore, in order to remove the interference of factors around the aperture, this project uses a high-point light source installed under the splitter to highlight the hole, making the aperture measurement result more accurate.

[0037] S2: Autofocus:

[0038] (1) The side camera collects the height profile information of the riving cutter;

[0039] (2) Control the upper camera to move within the focus range;

[0040] (3) The upper camera performs height compensation based on the focus clarity to achieve focus clarity.

[0041] In the construction of the imaging solution, it is divided into two parts: the measurement of the hole diameter of the end face of the chopper and the measurement of the height difference of the chopper profile. For the imaging of the height profile of the side of the chopper, because the ceramic chopper is extremely small, it is difficult for a general FA lens to recognize this accuracy. Therefore, the present invention adopts an industrial camera with a telecentric coaxial metallographic lens to ensure clear imaging of the height profile of the chopper; for the imaging of the hole diameter of the end face of the chopper, in order to ensure a clear display of the end face of the chopper, the present invention adopts a 4K industrial smart camera with a microscope lens and a high-light source to build the top visual environment.

[0042] During the focusing stage, first manually focus the first splitter, record its height information according to the splitter profile captured by the side camera, and use this as a reference to move the material tray to the second splitter through the control card, and then measure the height difference between it and the previous one. For example, if the height information of the first splitter is 249.81um, if the height of the second splitter is 247.56um, that is, the offset is equal to -2.25um, then according to the high offset difference, control the upper axis to drive the upper camera to compensate 2.25um downward to the corresponding focus position, thereby realizing automatic focusing of the equipment and obtaining a clear splitter end face image for the next step of processing.

[0043] S3: Aperture measurement:

[0044] (1) Select ROI and perform sub-pixel extraction on the aperture contour of the wedge;

[0045] (2) Use Huber method to fit the contour edge curve and make measurements;

[0046] After focusing, the aperture measurement phase begins. The upper camera is used to collect images and the ROI region of interest is selected to lock the area where the aperture contour is located, thereby effectively reducing the amount of image calculations and improving image processing efficiency. Image ROI extraction is as follows: Figure 4 shown.

[0047] When using a machine vision system to measure the size of the chopper hole diameter, direct measurement will cause a large error between the final measured data and the actual data of the workpiece due to the grinding marks around the end hole diameter. Taking this into account, corresponding mathematical processing must be implemented to reduce the impact of the error.

[0048] The least squares method is used to fit the curve. The idea of ​​the least squares method is to require the observed value y i The weighted sum of squares of the deviations is minimized, that is, the fitting function The value of is the smallest, which is to minimize the mean square error as the cost function to determine the optimal function matching of the data; it can easily and efficiently calculate the unknown data, while minimizing the sum of square errors between the results and the actual data. However, when fitting the curve in practice, there will be more or less abnormal values ​​x that deviate from the normal value. i, these outliers x i This will result in a large error when using the least squares method for straight line fitting. To this end, the Huber loss function is used to improve the robustness of the least squares method to noise outliers. The function formula is as follows:

[0049]

[0050] In the above formula, α is the residual, which is the distance between the generated straight line and the discrete point in straight line fitting; δ is the residual threshold, which is used to determine whether it is an error point.

[0051] After the Huber loss function is introduced, the abnormal points in the boundary curve will not be as sensitive as the least square fitting function. No matter how large the fitting loss is, the loss function of the least square method always grows as a quadratic function, while the loss function of the Huber algorithm grows linearly only when the residual value is greater than the residual threshold δ. Taking the residual threshold δ as 1 as an example, if the fitting residual α is 3 during the fitting process, the loss that the least square method can tolerate is much higher than the Huber algorithm, that is, it will treat more abnormal points as reference points for fitting in the fitting, and the fitting result obtained is much lower than the fitting accuracy of the Huber algorithm. Therefore, the present invention uses the Huber algorithm to fit the aperture profile. The application of the Huber weighted algorithm can reduce the influence of abnormal points on the fitting circle equation, while not losing the information of important contour points, and the fitting result for the circular area is ideal. Through testing, it is found that the result of the wedge end hole diameter measurement technology of the present invention is within 1 μ of the result of the two-dimensional measuring instrument, which meets the detection requirements.

[0052] (3) The host computer saves the measurement results, controls the ejector to eject unqualified NG materials, and completes the measurement.

[0053] After obtaining the aperture measurement value, the host computer displays the result in real time on the software interface and controls the rotation of the splitter fixture to the next ceramic splitter through the motion control card. After one plate of inspection is completed, the bottom ejector is controlled to move a small section of the defective product, thereby realizing high-precision automated focus compensation measurement and sorting technology for the splitter aperture.

[0054] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A high-precision measurement method for the aperture of a ceramic splitter based on machine vision, characterized in that: The following steps are involved: S1: Fix the ceramic splitting knife to the splitting knife fixture and polish the ceramic splitting knife; S2: Autofocus: (1) The side camera collects the height profile information of the riving cutter; (2) Control the upper camera to move within the focus range; (3) The upper camera performs height compensation according to the focus clarity to achieve focus clarity; S3: Aperture measurement: (1) Select ROI and perform sub-pixel extraction on the aperture contour of the wedge; (2) Use Huber method to fit the contour edge curve and make measurements; (3) The host computer saves the measurement results, controls the ejector to eject unqualified NG materials, and completes the measurement.

2. According to claim 1, a high-precision measurement method for ceramic splitting cutter aperture based on machine vision is characterized in that: The specific steps of fitting the contour edge curve in S3 are: The least squares method is used to fit the curve. The idea of ​​the least squares method is to require the observed value y i The weighted sum of squares of the deviations is minimized, that is, the fitting function The value of is the smallest, which is to minimize the mean square error as the cost function to determine the optimal function matching of the data; The Huber loss function is used to improve the robustness of the least squares method to noise outliers. The function formula is as follows: In the above formula, α is the residual, which is the distance between the generated straight line and the discrete point in straight line fitting; δ is the residual threshold, which is used to determine whether it is an error point.

3. The method for high-precision measurement of ceramic splitting cutter aperture based on machine vision according to claim 1 is characterized in that: The lighting in S1 is performed by installing a highlight light source under the ceramic splitting knife to highlight the holes on the ceramic splitting knife.

4. The method for high-precision measurement of ceramic splitting cutter aperture based on machine vision according to claim 1 is characterized in that: The side camera in the S2 uses an industrial camera with a telecentric coaxial metallographic lens to ensure clear imaging of the height profile of the splitting cutter.

5. The method for high-precision measurement of ceramic splitting cutter aperture based on machine vision according to claim 1 is characterized in that: The upper camera in the S2 uses a 4K industrial smart camera with a microscope lens to build a top visual environment.

Images