A glass plane speed measurement method based on binocular telecentric lens

Abstract

The present invention discloses a method for measuring the speed of a glass plane based on a binocular telecentric lens, comprising the following steps: S1, placing the device in the direction of movement of the glass surface so that the lens is perpendicular to the side of the glass; S2, adjusting the distance between the lens and the side of the glass; S3, initializing two cameras and setting corresponding acquisition parameters. S4, starting the device so that the lower edge of the glass appears from the upper lens and the lower lens in sequence. The upper lens detects in real time whether the glass enters the field of view. If it enters the field of view, the upper and lower cameras take two photos respectively at a fixed time interval. S5, processing and calculating the two photos taken, obtaining the position of the lower edge of the glass in the picture, and calculating the accurate time interval between the two photos. S6, calculating the moving distance and time of the glass to obtain the moving speed of the glass. The present invention has higher measurement accuracy, lower environmental requirements, and can maintain high-precision measurement for a long time.

Description

Technical Field

[0001] The present invention relates to the technical field of machine vision, and specifically refers to a method for measuring the speed of a glass plane based on a binocular telecentric lens. Background Art

[0002] In actual industrial scenarios, it is necessary to measure the moving speed of transparent objects such as glass. Common methods for measuring speed use binocular vision, laser ranging, etc. to measure distance, and then calculate the required time to obtain the speed. Binocular ranging mainly uses two monocular ordinary lenses with a fixed distance to calculate the field-of-view difference of the same object in two images and then estimate the distance. However, the edges of transparent objects such as glass are often not obvious, and the field-of-view difference between the two images in binocular vision is very weak, making it difficult to calculate. To enhance the edge features of the glass, an auxiliary light source can be used to illuminate the glass to increase the brightness, but the imaging result is related to the ambient light. After the position of the glass changes, the imaging effect under the best illumination cannot be maintained. Therefore, the available moving range for measurement is small and the difficulty is high. Laser ranging mainly measures the distance by measuring the total time required for laser, ultrasonic wave, etc. to return to the origin after contacting the object surface, and its accuracy is very high. However, in an industrial scenario, the glass on the production line generally has a high temperature of about 300 degrees Celsius and a high environmental humidity. The above environmental factors will affect the reflection of laser and ultrasonic wave, making the ranging scheme based on the reflected light from the object surface inapplicable in industrial scenarios. Neither of the above two schemes can ensure the continuous stability of the measurement results for transparent objects such as glass in industrial scenarios with harsh environmental conditions, and they have high requirements for the environment and do not have universality. A telecentric lens is a camera lens with a special optical design, and its main feature is that within a certain object distance range, the obtained image magnification does not change. The technical features of the telecentric lens, such as no perspective error and almost zero distortion, enable visual measurement and detection to achieve high precision.

[0003] After using a telecentric lens to take multiple pictures of a moving glass plane and measuring the position of the lower surface of the glass, by calculating the resolution of the image and the field-of-view range of the telecentric lens, the moving distance of the glass between two adjacent shots can be obtained. The ultra-low distortion (the maximum image-side distortion does not exceed 0.1%) and high telecentricity of the telecentric lens ensure that the measurement results obtained from the images still maintain high precision. However, the field-of-view range of the telecentric lens is smaller than that of an ordinary lens. For example, the field-of-view size of the telecentric lens used in the present invention is about 5 cm * 3 cm. In scenarios with a fast moving speed and high error requirements, a single telecentric lens cannot meet the target requirements.

[0004] The present invention proposes a solution using a binocular telecentric lens to measure the speed of a glass plane. The interval between the two binocular lenses can be set to a fixed value by itself, and this value depends on the moving speed of the target object and the requirements of the actual deployment device. By increasing the interval time between two shootings, the measured speed error is also made smaller. Summary of the Invention

[0005] In view of the above technical problems existing in the prior art, the present invention provides a method for measuring the speed of a glass plane based on a binocular telecentric lens. Compared with general ranging solutions, the present invention has higher measurement accuracy, lower environmental requirements, and can maintain high-precision measurement for a long time.

[0006] To solve the above technical problems, the technical solution of the present invention is as follows:

[0007] A method for measuring the speed of a glass plane based on a binocular telecentric lens, comprising the following steps:

[0008] S1. Set up an image shooting device, the image shooting device includes two cameras, the telecentric lenses of the two cameras are respectively denoted as α and β, the two cameras are located on one side of the glass plane to be measured, and the telecentric lenses of the two cameras are perpendicular to the side of the glass plane to be measured;

[0009] S2. Set the shooting time interval T of the two cameras according to the moving speed of the glass plane to be measured and the distance between the two cameras. The distance d between the two telecentric lenses is a fixed known value;

[0010] S3. Initialize the two cameras and set the acquisition parameters;

[0011] S4. Let α take a picture M α , and record the time stamp time corresponding to this photo α . After an interval time T, let β take a picture M β , and record the time stamp time corresponding to this photo β ;

[0012] When shooting, start the device so that the lower edge of the glass appears in the upper lens and the lower lens in turn. The upper lens detects in real time whether the glass enters the field of view. If it enters the field of view, the upper and lower cameras respectively take two pictures at a fixed time interval.

[0013] The distance d between the two telecentric lenses should be set in combination with the moving speed range of the object. For example, if the two lenses are spaced 1 s for shooting, it should be ensured that the object appears in the field of view of the upper lens at the moment before 1 s starts, and the object appears in the field of view of the lower lens at the moment when 1 s ends. At the same time, the situation where the object can be obtained only in the upper lens and not photographed in the lower lens should also be avoided.

[0014] Preferably, the value of the object space working distance is related to the optical parameters of the telecentric lens and should be adjusted in combination with the specific lens used. It should be ensured as much as possible that the distance between the object and the lens is within this distance, and the error should not exceed 3 mm.

[0015] Preferably, before the measurement starts, it should be ensured that the lower edge of the glass is above the upper lens, that is, the lower edge of the glass plane has not entered the field of view of the upper lens.

[0016] S5. Locate the lower edge of the glass in the two pictures

[0017] S5-1. First, process the picture horizontally using the Sobel operator;

[0018] S5-2. Perform binarization processing on the picture after horizontal processing;

[0019] S6. Detect the position of the boundary of the picture after locating the lower edge of the glass

[0020] S6-1. For the binarized picture dst w*h , sum all the points in each column to obtain a one-dimensional vector P w , where each value in the vector is calculated as follows:

[0021]

[0022] Traverse the values in the vector from front to back. If the value of a certain point is greater than 255 * Thickness, it can be considered that the coordinate of this point is the abscissa value x of the lower edge of the glass α , x β ,

[0023] where 255 is the pixel value representing white, and Thickness is the pixel length of the glass thickness.

[0024] S6-2. In the picture dst w*h , draw the detected lower edge straight line;

[0025] S7. Calculate the average moving speed of the glass based on the time interval I between the movements of the glass planes corresponding to the two photos and the distance the glass moves.

[0026] Preferably, the specific method of step S2 is as follows:

[0027] Estimate the change interval [v 0 , v 1 of the moving speed of the target glass plane, adjust the distance d (mm) between the two telecentric lenses, and calculate the adjacent shooting time interval T:

[0028]

[0029] Preferably, in step S2, it further includes adjusting the distance between the telecentric lens and the side of the glass.

[0030] Preferably, the telecentric lenses α and β are used to take pictures at a shooting frame rate of 50 Hz.

[0031] Preferably, in step S4, it further includes image preprocessing:

[0032] Convert the color RGB image into a grayscale image and then continue the detection:

[0033] Gray = Red * 0.3 + Green * 0.59 + Blue * 0.11

[0034] Obtain the resulting grayscale two-dimensional image matrix M w*h , where each value in the matrix represents the grayscale value at that point, and the value range is [0, 255]. The field of view of the telecentric lens is that the camera rotates 90 degrees counterclockwise when placed to obtain a larger viewing distance. Sum all the points in each column of matrix M to obtain a one-dimensional vector N w , where the calculation method of each value in the vector is as follows:

[0035]

[0036] Traverse the values in the vector from back to front. If the value of a point exceeds the predetermined width Thickness = 40 pixel, it indicates that the glass appears in the field of view. Here, 40 pixels refers to the thickness of the glass in the image.

[0037]

[0038] If the glass appears, proceed to the next step; if not, repeat step 3.

[0039] Preferably, the specific method of step S5-1 is:

[0040] Use the matrix in the X direction of the Sobel operator to perform a convolution operation on M α，β to obtain the preliminary edge position:

[0041]

[0042] Preferably, the specific method of step S5-2 is:

[0043] Using binarization processing to enhance the effect of the boundary and obtaining the processed result image dst after binarization. The value of each pixel point on dst is calculated from the value of the corresponding point on SobelX, so the size of the dst matrix is also w*h. The calculation method for each point in dst is as follows:

[0044]

[0045] Preferably, the specific method of step S6 is as follows:

[0046] Next, the position of the glass boundary is detected. Similarly, for the binarized image dst, the sum of all points in each column is calculated to obtain a one-dimensional vector P, and the vector has h values. The calculation method for each value in the vector is as follows:

[0047]

[0048] Among them, i represents the i-th column in the image M and also represents the i-th value in the vector P, and the value range is [1, w]. j represents the row number, and the value range is [1, h].

[0049] Traverse the values in the vector from front to back. If the value of a certain point is greater than 255*Thickness, it can be considered that the coordinate of this point is the abscissa value x of the lower edge of the glass α , x β , where 255 is the pixel value represented by white, and Thickness is the pixel length of the glass thickness.

[0050] Preferably, in step S7, the time interval I (ms) for the movement of the glass planes corresponding to the two photos is:

[0051] I = (time β - time α ) / 10 9 ,

[0052] The calculation method for the moving distance of the glass is:

[0053] The moving distance of the glass is divided into three parts: the distance from the glass edge in M α to the lower boundary of α; the distance from the glass edge in M β to the upper boundary of β; the distance from the lower boundary of α to the upper boundary of β, that is, the distance d between the two lenses α and β,

[0054] Then the moving distance of the glass is:

[0055] Distance = (x α + w - x β ) * K + d

[0056] Among them, K is the proportionality coefficient for converting the pixel length into the actual length.

[0057] Preferably, the calculation of the proportionality coefficient K is as follows:

[0058]

[0059] where len w is the horizontal field of view of the telecentric lens, with the unit of mm. image w is the horizontal resolution of the picture, with the unit of pixel.

[0060] The present invention has the following characteristics and beneficial effects:

[0061] Adopting the above technical solution, the present invention proposes a solution using a binocular telecentric lens to measure the speed of a glass plane, which is simple and convenient to deploy and use. No preset parameters are required and it can be directly used for measurement. By adjusting the distance between the cameras and the detection time interval, a wide range of speeds can be measured. The present invention uses a combination of a telecentric lens and a global exposure camera, and there will be no blurred imaging effect when photographing high-speed moving objects. The telecentric lens is used to photograph the object, and its imaging characteristics with ultra-low distortion enable high-precision ranging. At the same time, the telecentric lens has low requirements for the light in the environment, and there is no need to additionally illuminate the glass. Just avoid direct light shining on the telecentric lens. The entire device has low requirements for the environment and can meet the high-precision speed measurement requirements of transparent objects in various scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0063] Figure 1 is the flowchart of the embodiment of the present invention.

[0064] Figure 2 is the schematic diagram of the position of the glass and the camera of the present invention

[0065] Figure 3 is the grayscale imaging diagram obtained by the camera shooting.

[0066] Figure 4 is the result diagram after processing using the Sobel operator.

[0067] Figure 5 is the result diagram after binary processing.

[0068] Figure 6To draw the result graph of the detected position of the lower edge of the glass. Detailed implementation mode

[0069] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0070] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first", "second", etc. may explicitly or implicitly include one or more of such features. In the description of the present invention, unless otherwise specified, the meaning of "a plurality" is two or more.

[0071] In the description of the present invention, it should be noted that unless otherwise clearly specified and limited, the terms "installation", "connection", "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood through specific situations.

[0072] In the following steps, the model of the telecentric lens used is: Daheng Image - ZXCM118 - 64H - AL, and the model of the camera used is: Daheng Image - MER2 - 302 - 56U3M. The present invention has no fixed requirements for the models and brands of the telecentric lens and the camera. It should be noted that during the implementation, calculations and implementations should be carried out in combination with the optical parameters of specific equipment.

[0073] The present invention provides a solution for measuring the speed of a glass plane based on a binocular telecentric lens. The specific flowchart is referred to Figure 1 The glass plane speed measurement solution based on the binocular telecentric lens of the present invention includes the following steps:

[0074] Step 1: Make the lens perpendicular to the side of the glass plane. The line between the two lenses is the direction of the plane movement. Estimate the change interval [v 0 , v 1, adjust the distance d (mm) between the two telecentric lenses. Calculate the adjacent shooting time interval T.

[0075]

[0076] When the target moving speed is between 85 mm / s and 90 mm / s, the calculated time interval T should be made greater than 1 second as much as possible, so that the calculated speed error is lower. Therefore, on the premise that the target moving speed interval is fixed, make d > v 1 *1 s is enough. Here, take d = 100 mm, and the calculated T ≈ 1.14 s. The interval time here does not need to be very accurate, so it can be set to 1.1 s.

[0077] Step 2: Adjust the distance between the telecentric lens and the glass plane, that is, the object-side working distance is within the range of 158 mm ± 3 mm. It should be noted that the object-side working distance is related to the optical parameters of the telecentric lens, and the distance length should be determined according to the specific model used. The relative position information of the device and the object can be referred to Figure 2 .

[0078] Step 3: Initialize the two cameras to ensure that the two cameras start up normally and can obtain images. Set the corresponding acquisition parameters. The specific parameters implemented in the present invention are shown in the following table:

[0079] Parameter Value Exposure time 5000 μs Gain 12 dB Trigger mode On Trigger source Soft trigger Acquisition mode Single frame mode

[0080] It should be set in combination with the specific actual application scenario to ensure that the lens is completely black when there is no object. It should be noted that the exposure time has a great influence on the brightness of the photographed photo. The longer the exposure time, the stronger the brightness. However, it should not be too strong to avoid affecting the accuracy of the time interval. The telecentric lens has a strong ability to capture objects within the field of view. Without additional supplementary lighting enhancement, the transparent glass surface can still be seen in the field of view. At the same time, shooting the glass from the side can further enhance the imaging effect of the transparent glass.

[0081] Step 4: Before the glass enters the field of view, start the detection program. Let the upper telecentric lens be α and the lower telecentric lens be β. Let α perform shooting detection at a shooting frame rate of 50 Hz. For each acquired frame of image, detect the content of the shot.

[0082] If a color camera is used for shooting, the color RGB image should be converted into a grayscale image before continuing the detection:

[0083] Gray = Red * 0.3 + Green * 0.59 + Blue * 0.11 (2)

[0084] Obtain the grayscale two-dimensional image matrix M w*h , such as Figure 3As shown. Each value in the matrix represents the grayscale value at that point, and the value range is [0, 255]. The field of view of the telecentric lens is that the camera rotates 90 degrees counterclockwise when placed to obtain a larger viewing distance. Therefore, from the picture, the glass moves from right to left. Sum all the points in each column of the matrix M to obtain a one-dimensional vector N w , where each value in the vector is calculated as follows:

[0085]

[0086] Traverse the values in the vector from back to front. If the value of a point exceeds the predetermined width Thickness = 40 pixels, it indicates that the glass appears in the field of view. Here, 40 pixels refers to the thickness of the glass in the picture.

[0087]

[0088] If the glass appears, go to the next step; if not, repeat step 3.

[0089] Step 5: Let α take a photo M α , and record the timestamp time α corresponding to this photo. After an interval of time T, let β take a photo M β , and record the timestamp time β corresponding to this photo. The timestamp is provided by the internal clock of the camera, and the unit is nanoseconds.

[0090] Locate the lower edge of the glass for the above two pictures. First, process the picture horizontally using the Sobel operator, that is, perform a convolution operation on M α，β with the matrix in the X direction of the Sobel operator to obtain the preliminary edge position:

[0091]

[0092] The result of the processing is as shown in Figure 4 .

[0093] Use binary processing to enhance the boundary effect. The value of each point on the picture is calculated as follows:

[0094]

[0095] The threshold value should be set according to the brightness obtained by the Sobel operator. In this method, it is set to 175. The picture obtained after binary processing is as shown in Figure 5 .

[0096] Next, detect the position of the glass boundary. Similarly, for the picture dst after binary processingw*h Sum all the points in each column to obtain a one-dimensional vector P w where each value in the vector is calculated as follows:

[0097]

[0098] Traverse the values in the vector from front to back. If the value of a certain point is greater than 255 * Thickness, the coordinate of this point can be considered as the abscissa value x of the lower edge of the glass α x β where 255 is the pixel value represented by white, and Thickness is the pixel length of the glass thickness

[0099] In M, draw the detected lower edge line, as shown in Figure 6 For M α M β Do the same processing to obtain the positions x of the glass boundaries in the two photos α x β

[0100] The calculation of the proportionality coefficient K for converting pixel length to actual length is as follows:

[0101]

[0102] For the telecentric lens and camera used in this patent, it can be calculated that K = 0.02568359375 mm / pixel. Step 6: From the values obtained in the above steps, the time interval I (ms) for the glass plane to move corresponding to the two photos is:

[0103] I=(time β -time α ) / 10 9 (9)

[0104] The moving distance of the glass is divided into three parts: the distance from the glass edge in M to the lower boundary of α; the distance from the glass edge in M to the upper boundary of β; the distance from the lower boundary of α to the upper boundary of β, that is, the distance d between the two lenses α and β α in M β in M

[0105] Then the moving distance of the glass is:

[0106] Distance=(x α +w - x β )*K + d (10)

[0107] The average moving speed of the glass is:

[0108] ​

[0109] Based on the above steps, conduct experiments multiple times and count the data. Use the robotic arm to ensure that the glass moves at a specified speed. The measured speed and speed error are retained to 4 decimal places, as shown in the following table:

[0110]

[0111] As can be seen from the data in the table, the speed measurement method provided by this patent has high precision and can still ensure sufficiently high-precision speed measurement results in the case of relatively fast speeds.

[0112] The above method provided by this application has the following advantages compared with the prior art:

[0113] This algorithm is a glass plane speed measurement method based on a binocular telecentric lens, with strong applicability and can calculate objects in various scenarios including transparent and high-temperature objects. It does not require calculation of preset parameters, and is convenient for deployment and use. The accuracy value of this algorithm depends on the calculated K value, which is related to the field of view of the telecentric lens and the imaging element of the camera. In the above experiment, a high-precision level of about 0.02 mm / pixel can be achieved. If a telecentric lens with a larger field of view and a camera with a higher resolution are selected, the accuracy can be further improved. For the time interval I of the glass movement, due to using the built-in clock of the camera, the accuracy can reach the nanosecond level. And when calculating the time interval, the moment corresponding to the photo returned by the camera is used, and the time calculation includes the time overhead of optical imaging processes such as exposure time, so the calculation of the time interval is accurate enough. In the detection of the lower edge of the glass, different from common edge detection algorithms, starting from the imaging characteristics of the telecentric lens, this method does not use image processing methods such as Gaussian filtering that may blur the imaging content, and retains as much as possible the part of the glass in the image, making the measurement results as accurate as possible. Judging from the experimental data, the method proposed in this patent has the characteristics of wide applicability, high recognition rate and high measurement accuracy.

[0114] The above description of the embodiments is to enable those of ordinary skill in the art to understand and apply the present invention. It is obvious that those who are familiar with the technology in this field can easily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative labor. Therefore, the present invention is not limited to the above embodiments, and all improvements and modifications made by those skilled in the art according to the disclosure of the present invention should be within the protection scope of the present invention.

Claims

1. A glass plane speed measurement method based on binocular telecentric lens, characterized in that: The steps include: S1. Build an image capture device, wherein the image capture device includes two cameras, the telecentric lenses of the two cameras are respectively denoted as α and β, the two cameras are located on one side of the glass plane to be measured, and the telecentric lenses of the two cameras are perpendicular to the side of the glass plane to be measured; S2, setting the shooting time interval T of the two cameras according to the moving speed of the glass plane to be tested and the distance between the two cameras; S3, initialize two cameras and set acquisition parameters; S4, let α take a picture M α , and record the timestamp corresponding to the photo α , after an interval of time T, let β take a picture M β , and record the timestamp corresponding to the photo β ; S5. Position the bottom edge of the glass for the two images S5-1, first use the Sobel operator to process the image horizontally; S5-2, binarize the horizontally processed image to obtain image dst; S6. Detect the position of the border of the image after the bottom edge of the glass is located S6-1. For the binary image dst, sum all the points in each column to obtain a one-dimensional vector P, where each value in the vector is calculated as follows: Where i represents the i-th column in the image M, and also represents the i-th value in the vector P, with a value range of [1, w]; j represents the row, with a value range of [1, h]; Traverse the values ​​in the vector from front to back. If the value of a point is greater than 255*Thickness, the coordinate of the point can be considered to be the horizontal coordinate value x of the lower edge of the glass. α , x β , Among them, 255 is the pixel value represented by white, and Thickness is the pixel length of the glass thickness; S6-2, in the image dst, draw a straight line of the detected lower edge; S7. Calculate the average moving speed of the glass according to the time interval I of the glass plane movement corresponding to the two photos and the distance the glass moves.

2. The glass plane velocity measurement method based on binocular telecentric lens according to claim 1, characterized in that: The specific method of step S2 is: Estimate the change range of the moving speed of the target glass plane [v0, v1], adjust the distance d (mm) between the two telecentric lenses, and calculate the time interval T between adjacent shots: Where d is the distance between the two telecentric lenses.

3. The glass plane velocity measurement method based on binocular telecentric lens according to claim 1, characterized in that: The step S2 also includes adjusting the distance between the telecentric lens and the side surface of the glass.

4. The glass plane velocity measurement method based on binocular telecentric lens according to claim 1, characterized in that: The telecentric lenses α and β take pictures at a frame rate of 50 Hz.

5. The glass plane velocity measurement method based on binocular telecentric lens according to claim 2, characterized in that: The step S4 also includes image preprocessing: After converting the color RGB image into a grayscale image, continue the detection: Gray=Red*0.3+Green*0.59+Blue*0.11 The obtained grayscale two-dimensional image matrix M w*h , each value in the matrix represents the grayscale value at that point, and the value range is [0, 255]. The field of view of the telecentric lens is that the camera is rotated 90 degrees counterclockwise when placed to obtain a larger field of view distance. Sum all the points on each column of the matrix M to obtain a one-dimensional vector N w , where each value in the vector is calculated as follows: Traverse the values ​​in the vector from back to front. If a point's value exceeds the preset width Thickness = 40 pixels, it means that glass appears in the field of view. Here 40 pixels refers to the thickness of the glass in the picture. If the glass appears, proceed to the next step; if not, repeat step S3.

6. The glass plane velocity measurement method based on binocular telecentric lens according to claim 1, characterized in that: The specific method of step S5-1 is: Use the matrix in the X direction of the Sobel operator to M α，β Perform a convolution operation to get the preliminary edge position:

7. The glass plane velocity measurement method based on binocular telecentric lens according to claim 6, characterized in that: In the step S5-2: The value of each point on the image dst is calculated by the conversion between the corresponding point values ​​on the SobelX image as follows: Where (x, y) represents the horizontal and vertical coordinates of a point, x∈[1,w], y∈[1,h].

8. The glass plane velocity measurement method based on binocular telecentric lens according to claim 7, characterized in that: The specific method of step S6 is: For the binarized image dst, the resolution is w*h, the unit is pixel, and all the points on each column are summed to obtain a one-dimensional vector P, where the number of values ​​in the vector is w, and each value is calculated as follows: Where i represents the i-th column in the image M, and also represents the i-th value in the vector P, with a value range of [1, w]; j represents the row, with a value range of [1, h]; Traverse the values ​​in the vector from front to back. If the value of a point is greater than 255*Thickness, the coordinate of the point can be considered to be the horizontal coordinate value x of the lower edge of the glass. α , x β , where 255 is the pixel value represented by white, and Thickness is the pixel length of the glass thickness.

9. The glass plane velocity measurement method based on binocular telecentric lens according to claim 8, characterized in that: In step S7, the time interval I (ms) between the movement of the glass plane corresponding to the two photos is: I=(time β -time α ) / 10 9 , The distance the glass moves is calculated as: The moving distance of the glass is divided into three parts: M α The distance between the edge of the glass and the lower boundary of α; M β The distance between the edge of the glass in and the upper boundary of β; the distance between the lower boundary of α and the upper boundary of β, that is, the distance d between the two lenses of α and β, The distance the glass moves is: Distance=(x a +w-x β )*K+d Wherein, K is the ratio coefficient of converting pixel length to actual length.

10. The method for measuring speed of a glass plane based on a binocular telecentric lens according to claim 9, characterized in that: The proportionality coefficient K is calculated as follows: where len w is the lateral field of view of the telecentric lens, in mm, image w It is the horizontal resolution of the image in pixels.

Images