A method for detecting machining errors of eyepiece lenses of fundus camera lenses

Abstract

A method for detecting processing errors of eyepiece lenses of fundus camera lenses, the method comprising the following steps: step 1, building a fundus camera lens eyepiece lens error detection device; step 2, placing the fundus camera lens eyepiece lens to be tested on a tooling; step 3, a dimension measuring instrument emits a parallel light beam through a light source to illuminate the sample to be tested; step 4, preprocessing the image data collected by the dimension measuring instrument; step 5, detecting the corner points of the segmented contour; step 6, fitting the contour corner points, determining the boundary of the image to be tested, and calculating the error of the eyepiece lens of the fundus camera lens. The present invention can detect the qualification of the eyepiece lens of the fundus camera lens, and the detection system has strong versatility. The data is preprocessed by the proposed smoothing filter and Laplace sharpening to improve the accuracy of the detection result, the workpiece corner points are obtained by using the Harris corner point extraction method, and finally the contour of the workpiece to be tested is fitted, and the workpiece is detected from the processing error according to the fitting curve.

Description

Technical Field

[0001] The invention relates to the field of lens detection, in particular to a method for detecting processing errors of an eyepiece lens of a fundus camera. Background Art

[0002] The lens has the advantages of low deformation and can provide high visual resolution under the same conditions, and has a wide range of application prospects. However, since the normal vector and curvature of each point on the lens surface are different, the processing of the lens has high requirements on the processing technology, and certain deviations will occur during the processing. The causes of the deviations include: processing principle error, process system geometric error, and measurement error during the processing. As the requirements for product quality are getting higher and higher, higher requirements are also put forward for the error detection of the lens, and the manual detection method of the lens can no longer meet the online detection requirements of modern production. In view of this problem, the present invention proposes a processing error detection method for the eyepiece lens of a fundus camera lens. Summary of the invention

[0003] In order to solve the above problems, the present invention proposes a method for detecting machining errors of eyepiece lenses of fundus camera lenses, which can detect the eligibility of eyepiece lenses of fundus camera lenses, aiming to improve the efficiency and detection accuracy of sample detection. To achieve this purpose, the present invention provides a method for detecting machining errors of eyepiece lenses of fundus camera lenses, the specific steps are as follows, and the characteristics are as follows:

[0004] Step 1, build a fundus camera lens eyepiece lens error detection device, which consists of a lens rotation control module, an optical imaging module and an image processing module;

[0005] Step 2, placing the fundus camera lens eyepiece lens to be tested on the tooling, placing the tooling on a turntable, and placing the turntable on the measuring position of the dimension measuring instrument;

[0006] Step 3, the dimension measuring instrument emits a parallel light beam through the light source to illuminate the sample to be measured, and the linear array CCD collects the shadow image of the sample, and then uploads the image data to the industrial computer;

[0007] Step 4, preprocessing the image data collected by the dimension measuring instrument, including smoothing filtering and Laplace sharpening;

[0008] Step 5, segment the image to be tested according to the negative film in the database, and detect the corner points of the segmentation contour;

[0009] Step 6, fitting the contour corner points, determining the boundary of the image to be measured, and calculating the error of the eyepiece lens of the fundus camera lens.

[0010] Further, the process of the fundus camera lens eyepiece lens error detection device in step 1 can be expressed as:

[0011] The lens rotation control module includes a turntable and tooling; the optical imaging module includes a dimension measuring instrument; the image processing module includes an industrial computer and dimension data detection model software;

[0012] During the inspection, the lens is placed on the tooling of the turntable and the turntable is rotated at a constant speed. The dimension measuring instrument collects images of the lens rotated clockwise by 0°, 45°, 90° and 135° respectively. The images are transmitted to the industrial computer via USB and PCIE bus. The dimension data detection model software is used in the industrial computer to detect the lens processing error. Once an unqualified lens is detected, the industrial computer will issue an alarm to remind the user.

[0013] Furthermore, the process of the detection method in step 3 can be expressed as follows:

[0014] The dimension measuring instrument illuminates the workpiece to be measured through parallel light emitted by the light source, and then forms an enlarged shadow image on the linear array CCD by the imaging lens and the aperture. The actual size of the workpiece to be measured can be obtained by dividing the measured value of the shadow image by the magnification factor; the CCD sensor uses the size of the discrete voltage signal corresponding to the intensity of the light received by the photosensitive element to represent the image signal corresponding to the Y-axis data in the image signal, and the timing of the signal output corresponds to the X-axis data in the image signal. The collected high-speed data acquisition card performs A / D conversion and converts it into a digital signal.

[0015] Furthermore, in step 4, the process of Laplace sharpening can be expressed as follows:

[0016] The image is sharpened by Laplace, and the formula is as follows:

[0017]

[0018] In the formula, f(x,y) is the original data of the image, c is the scale factor, is the Laplace operator of the original image data, and g(x,y) is the sharpened image data.

[0019] Furthermore, in step 5, the process of detecting the corner points of the segmentation contour can be expressed as follows:

[0020] The window is slid on the sharpened image, and the change in the grayscale of the image pixels during sliding is compared. The grayscale change E of the image pixels during sliding of the window is described as follows:

[0021]

[0022] Where u and v are the offsets of the sliding window, x and y are the coordinate positions of the pixel points of the image data, ω(x, y) is the window function, I(x, y) is the pixel value of a point in the window before translation, and I(x+u, y+v) is the pixel value of the corresponding point after the window is translated;

[0023] Taylor expansion of the grayscale part in the grayscale change description formula:

[0024]

[0025] In the formula, I x is the gradient coordinate of the image data on the x-axis, I y is the gradient coordinate of the image data on the y-axis, where M can be expressed as:

[0026]

[0027] In the formula, I x (x,y) is the gradient coordinate of the image data point (x,y) on the x-axis, I y (x,y) is the gradient coordinate of the image data point (x,y) on the y-axis;

[0028] Calculate the corner point response R based on M:

[0029] R=det|M|-k·T r (5)

[0030] k is between 0.04 and 0.06, det means the determinant, when R is greater than the threshold, the pixel is considered to be a corner point, T r Represents the trace of the matrix.

[0031] Furthermore, in step 6, the process of fitting the contour corner points can be expressed as follows:

[0032] Step 6.1, sort the detected corner points according to the clockwise contour of the image, and match the sample corner points according to the position of the first corner point;

[0033] Step 6.2, fit the matched corner points and perform ellipse fitting according to the fitting formula. The ellipse fitting formula is as follows:

[0034] x 2 +Axy+By 2 +Cx+Dy+E=0 (6)

[0035] In the formula, x represents the x-axis data of the ellipse, y represents the y-axis data of the ellipse, A, B, C, D, E represent the parameters of the ellipse, and the target formula of the fitted ellipse is ε 1 :

[0036]

[0037] In the formula, (x i ,y i ) is the i measurement point on the ellipse contour, x i is the horizontal coordinate value of the detected corner point, y i is the ordinate value of the detected corner point;

[0038] Step 6.3, determine whether the corner point is an arc point set. If the aspect ratio of the fitted ellipse is greater than the threshold, it is determined to be a straight line point set, otherwise it is determined to be an arc point set. The fitting formula of the straight line point set is as follows:

[0039] ax+by+g=0 (8)

[0040] In the formula, a, b, g represent the parameters of the straight line, and the target formula of the fitted straight line is ε 2 :

[0041] ε 2 =∑(ax i +by i +g) 2 (9)

[0042] Step 6.4, measure the lens according to the fitted profile and calculate the lens deviation.

[0043] The present invention provides a method for detecting machining errors of an eyepiece lens of a fundus camera, and the beneficial effects are as follows:

[0044] 1. The present invention can detect the eligibility of the eyepiece lens of the fundus camera lens, and the detection system has strong versatility;

[0045] 2. The present invention pre-processes the data by using the proposed smoothing filter and Laplace sharpening to improve the accuracy of the detection results;

[0046] 3. The present invention uses the Harris corner point extraction method to obtain the workpiece corner points, and finally fits the contour of the workpiece to be measured, and detects the workpiece processing error according to the fitting curve;

[0047] 4. The present invention provides an important technical means for detecting the eyepiece lens of a fundus camera, thereby improving the detection efficiency of the eyepiece lens of a fundus camera. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flow chart of the present invention;

[0049] Figure 2 is a cross-sectional view of a flat sheet of the present invention;

[0050] Figure 3 This is a 45-degree incident light field simulation of the present invention. DETAILED DESCRIPTION

[0051] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0052] The present invention proposes a method for detecting machining errors of an eyepiece lens of a fundus camera. The qualification of the eyepiece lens of the fundus camera is detected by building an error detection device for the eyepiece lens of the fundus camera. The data is preprocessed by using smoothing filtering and Laplace sharpening to improve the accuracy of the detection result. The corner points of the workpiece are obtained by using the Harris corner point extraction method. Finally, the contour of the workpiece to be tested is fitted, and the machining error of the workpiece is detected according to the fitting curve. Figure 1 Detailed description of the steps of the present invention is given below.

[0053] Step 1, build a fundus camera lens eyepiece lens error detection device, which consists of a lens rotation control module, an optical imaging module and an image processing module;

[0054] The lens rotation control module includes a turntable and tooling; the optical imaging module includes a dimension measuring instrument; the image processing module includes an industrial computer and dimension data detection model software;

[0055] During the inspection, the lens is placed on the tooling of the turntable and the turntable is rotated at a constant speed. The dimension measuring instrument collects images of the lens rotated clockwise by 0°, 45°, 90° and 135° respectively. The images are transmitted to the industrial computer via USB and PCIE bus. The dimension data detection model software is used in the industrial computer to detect the lens processing error. Once an unqualified lens is detected, the industrial computer will issue an alarm to remind the user.

[0056] Step 2, placing the fundus camera lens eyepiece lens to be tested on the tooling, placing the tooling on a turntable, and placing the turntable on the measuring position of the dimension measuring instrument;

[0057] Step 3: The dimension measuring instrument emits a parallel beam of light through the light source to illuminate the sample to be measured, and the linear array CCD collects the shadow image of the sample, and then uploads the image data to the industrial computer. The flat film cross-section of the sample is as follows: Figure 2 As shown;

[0058] The dimension measuring instrument irradiates the workpiece to be measured with parallel light emitted by the light source, and then forms an enlarged shadow image on the linear array CCD by the imaging lens and the aperture. The actual size of the workpiece to be measured can be obtained by dividing the measured value of the shadow image by the magnification factor; the CCD sensor uses the size of the discrete voltage signal corresponding to the intensity of the light received by the photosensitive element to represent the image signal corresponding to the Y-axis data in the image signal, and the timing of the signal output corresponds to the X-axis data in the image signal. The collected high-speed data acquisition card is converted into a digital signal through A / D conversion. The 45-degree incident light field simulation of the sample is as follows Figure 3 shown.

[0059] Step 4, preprocessing the image data collected by the dimension measuring instrument, including smoothing filtering and Laplace sharpening;

[0060] The image is sharpened by Laplace, and the formula is as follows:

[0061]

[0062] In the formula, f(x,y) is the original data of the image, c is the scale factor, is the Laplace operator of the original image data, and g(x,y) is the sharpened image data.

[0063] Step 5, segment the image to be tested according to the negative film in the database, and detect the corner points of the segmentation contour;

[0064] The window is slid on the sharpened image, and the change in the grayscale of the image pixels during sliding is compared. The grayscale change E of the image pixels during sliding of the window is described as follows:

[0065]

[0066] Where u and v are the offsets of the sliding window, x and y are the coordinate positions of the pixel points of the image data, ω(x, y) is the window function, I(x, y) is the pixel value of a point in the window before translation, and I(x+u, y+v) is the pixel value of the corresponding point after the window is translated;

[0067] Taylor expansion of the grayscale part in the grayscale change description formula:

[0068]

[0069] In the formula, I x is the gradient coordinate of the image data on the x-axis, I y is the gradient coordinate of the image data on the y-axis, where M can be expressed as:

[0070]

[0071] In the formula, I x (x,y) is the gradient coordinate of the image data point (x,y) on the x-axis, I y (x,y) is the gradient coordinate of the image data point (x,y) on the y-axis;

[0072] Calculate the corner point response R based on M:

[0073] R=det|M|-k·T r (5)

[0074] k is between 0.04 and 0.06, det means the determinant, when R is greater than the threshold, the pixel is considered to be a corner point, T r Represents the trace of the matrix.

[0075] Step 6, fitting the contour corner points, determining the boundary of the image to be measured, and calculating the error of the eyepiece lens of the fundus camera lens;

[0076] Step 6.1, sort the detected corner points according to the clockwise contour of the image, and match the sample corner points according to the position of the first corner point;

[0077] Step 6.2, fit the matched corner points and perform ellipse fitting according to the fitting formula. The ellipse fitting formula is as follows:

[0078] x 2 +Axy+By 2 +Cx+Dy+E=0 (6)

[0079] In the formula, x represents the x-axis data of the ellipse, y represents the y-axis data of the ellipse, A, B, C, D, E represent the parameters of the ellipse, and the target formula of the fitted ellipse is ε 1 :

[0080]

[0081] In the formula, (x i ,y i ) is the i measurement point on the ellipse contour, x i is the horizontal coordinate value of the detected corner point, y i is the ordinate value of the detected corner point;

[0082] Step 6.3, determine whether the corner point is an arc point set. If the aspect ratio of the fitted ellipse is greater than the threshold, it is determined to be a straight line point set, otherwise it is determined to be an arc point set. The fitting formula of the straight line point set is as follows:

[0083] ax+by+g=0 (8)

[0084] In the formula, a, b, g represent the parameters of the straight line, and the target formula of the fitted straight line is ε 2 :

[0085] ε 2 =∑(ax i +by i +g) 2 (9)

[0086] Step 6.4, measure the lens according to the fitted profile and calculate the lens deviation.

[0087] The above description is only a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent change made based on the technical essence of the present invention still falls within the scope of protection required by the present invention.

Claims

1. A method for detecting processing errors of eyepiece lenses of fundus camera lenses, comprising the following specific steps, characterized in that: Step 1, build a fundus camera lens eyepiece lens error detection device, which consists of a lens rotation control module, an optical imaging module and an image processing module; The process of the fundus camera lens eyepiece lens error detection device in step 1 is expressed as: The lens rotation control module includes a turntable and tooling; the optical imaging module includes a dimension measuring instrument; the image processing module includes an industrial computer and dimension data detection model software; During the inspection, the lens is placed on the tooling of the turntable and the turntable is rotated at a constant speed. The dimension measuring instrument collects images of the lens rotated clockwise at 0°, 45°, 90° and 135° respectively. The images are transmitted to the industrial computer via USB and PCIE bus. The industrial computer uses the dimension data detection model software to detect the lens processing error. Once an unqualified lens is detected, the industrial computer will issue an alarm to remind the user. Step 2, placing the fundus camera lens eyepiece lens to be tested on the tooling, placing the tooling on a turntable, and placing the turntable on the measuring position of the dimension measuring instrument; Step 3, the dimension measuring instrument emits a parallel light beam through the light source to illuminate the sample to be measured, and the linear array CCD collects the shadow image of the sample, and then uploads the image data to the industrial computer; The process of the detection method in step 3 can be expressed as follows: The dimension measuring instrument irradiates the workpiece to be measured with parallel light emitted by the light source, and then forms an enlarged shadow image on the linear array CCD by the imaging lens and the aperture. The actual size of the workpiece to be measured can be obtained by dividing the measured value of the shadow image by the magnification factor; the CCD sensor uses the size of the discrete voltage signal corresponding to the intensity of the light received by the photosensitive element to represent the image signal corresponding to the Y-axis data in the image signal, and the timing of the signal output corresponds to the X-axis data in the image signal. The collected high-speed data acquisition card performs A / D conversion and converts it into a digital signal; Step 4, preprocessing the image data collected by the dimension measuring instrument, including smoothing filtering and Laplace sharpening; In step 4, the process of Laplace sharpening is expressed as follows: The image is sharpened by Laplace, and the formula is as follows: In the formula, f(x,y) is the original data of the image, c is the scale factor, is the Laplace operator of the original image data, and g(x,y) is the sharpened image data; Step 5, segment the image to be tested according to the negative film in the database, and detect the corner points of the segmentation contour; In step 5, the process of detecting the corner points of the segmentation contour is expressed as follows: The window is slid on the sharpened image, and the change in the grayscale of the image pixels during sliding is compared. The grayscale change E of the image pixels during sliding of the window is described as follows: Where u and v are the offsets of the sliding window, x and y are the coordinate positions of the pixel points of the image data, ω(x, y) is the window function, I(x, y) is the pixel value of a point in the window before translation, and I(x+u, y+v) is the pixel value of the corresponding point after the window is translated; Taylor expansion of the grayscale part in the grayscale change description formula: In the formula, I x is the gradient coordinate of the image data on the x-axis, I y is the gradient coordinate of the image data on the y-axis, where M can be expressed as: In the formula, I x (x,y) is the gradient coordinate of the image data point (x,y) on the x-axis, I y (x,y) is the gradient coordinate of the image data point (x,y) on the y-axis; Calculate the corner point response R based on M: R=det|M|-k·T r (5) k is between 0.04 and 0.06, det means the determinant, when R is greater than the threshold, the pixel is considered to be a corner point, T r represents the trace of a matrix; Step 6, fitting the contour corner points, determining the boundary of the image to be measured, and calculating the error of the eyepiece lens of the fundus camera lens; In step 6, the process of fitting the contour corner points is as follows: Step 6.1, sort the detected corner points according to the clockwise contour of the image, and match the sample corner points according to the position of the first corner point; Step 6.2, fit the matched corner points and perform ellipse fitting according to the fitting formula. The ellipse fitting formula is as follows: x 2 +Axy+By 2 +Cx+Dy+E=0 (6) In the formula, x represents the x-axis data of the ellipse, y represents the y-axis data of the ellipse, A, B, C, D, E represent the parameters of the ellipse, and the target formula ε1 of the fitted ellipse is: In the formula, (x i ,y i ) is the i measurement point on the ellipse contour, x i is the horizontal coordinate value of the detected corner point, y i is the ordinate value of the detected corner point; Step 6.3, determine whether the corner point is an arc point set. If the aspect ratio of the fitted ellipse is greater than the threshold, it is determined to be a straight line point set, otherwise it is determined to be an arc point set. The fitting formula of the straight line point set is as follows: ax+by+g=0 (8) Where a, b, and g represent the straight line parameters, and the target formula for fitting the straight line is ε2: ε2=∑(ax i +by i +g) 2 (9) Step 6.4, measure the lens according to the fitted profile and calculate the lens deviation.

Images