Abnormal winding detection method of protective cover based on machine vision

Abstract

This invention relates to the field of abnormal winding detection of protective cases, and more particularly to a machine vision-based method for detecting abnormal winding of protective cases. The method includes: acquiring side views of a winding shaft assembly at different times, the winding shaft assembly including a winding shaft and a protective case; constructing a ray with the axis of the winding shaft as the origin, and taking the edge pixels where the ray intersects with the outer contour of the protective case as target pixels; calculating the edge height of the target pixels, the edge height representing the distance between the target pixel and the axis; calculating the standard height of the target pixels, the standard height representing the thickness of the protective case without misalignment; calculating the abnormal winding evaluation of the side view at each time, and issuing a signal of abnormal winding of the protective case in response to the abnormal winding evaluation being greater than a preset evaluation threshold; this invention has the effect of improving the accuracy of detecting abnormal winding of protective cases.

Description

Technical Field

[0001] The invention relates to the field of abnormal winding detection of a protective cover, and in particular to a method for detecting abnormal winding of a protective cover based on machine vision. Background Art

[0002] Protective covers usually refer to external coverings used to protect items, equipment or the surface of items from damage. They are commonly found in electronic products such as mobile phones, tablets, and laptops. Protective covers can prevent physical damage such as scratches, collisions, and falls. There are many types of protective covers, among which the most common is a double-layer plastic protective cover, which has a rectangular structure and is sealed on three sides to form an opening. Its structure is like a pocket and is used to hold electronic products to prevent the surface of the corresponding products from being damaged. During the production process of protective covers, adjacent protective covers are connected by tear lines and then wound onto the surface of the winding roller.

[0003] A Chinese patent with announcement number CN221644221U discloses a monitoring device for a composite yarn winding machine, including a support plate, a displacement sensor and a winding drum. The two mounting frames extend forward on the front side of the support plate, and the mounting frames can be movably mounted on the support plate. The transmitting end and the receiving end of the displacement sensor are respectively arranged on the mounting frames on the left and right sides. When the winding drum is abnormally wound, the winding drum will be unbalanced when rotating and winding, and abnormal jumping will occur on both sides, and the mounting frames on both sides will also jump abnormally, thereby generating jitter, which is detected by the displacement sensor and the abnormal situation is promptly warned.

[0004] It can be seen from the above technical solution that the abnormal winding of the protective cover can be detected by detecting the shaking of the mounting frame through the displacement sensor. However, in the process of detecting the abnormal winding of the protective cover in the prior art, the abnormal winding of the protective cover cannot be discovered in time. In other words, it cannot be discovered in time at the early stage of the abnormal winding to reduce the loss caused by the abnormal winding of the protective cover. Summary of the invention

[0005] In order to timely discover abnormal winding of a protective cover and reduce the loss caused in the production process of the protective cover, the present invention provides a method for detecting abnormal winding of the protective cover based on machine vision.

[0006] The present invention provides a method for detecting abnormal winding of a protective cover based on machine vision, which adopts the following technical solution:

[0007] A method for detecting abnormal winding of a protective cover based on machine vision comprises the following steps:

[0008] Obtaining side views of the winding shaft assembly at different times, the winding shaft assembly comprising a winding shaft and a protective sleeve;

[0009] A ray is constructed with the axis of the winding shaft as the origin, and the edge pixel point where the ray intersects with the outer contour of the protective cover is taken as the target pixel point;

[0010] Calculate the edge height of the target pixel, which represents the distance between the target pixel and the axis; calculate the standard height of the target pixel, which represents the thickness of the protective cover when no misalignment occurs;

[0011] Calculating the abnormal winding evaluation of the side view at each moment, and in response to the abnormal winding evaluation being greater than a preset evaluation threshold, issuing a signal of abnormal winding of the protective cover;

[0012] Among them, the expression for abnormal winding evaluation is: ; In the formula, represents the abnormal winding evaluation of the side view at the i-th moment, represents the edge height of the target pixel in the side view at the i-th moment, Represents the standard height of the target pixel in the side view at the i-th moment.

[0013] The abnormal winding evaluation is obtained by calculating the standard height and edge height of the target pixel point. Through the abnormal winding evaluation, it is possible to timely determine whether the protective cover is wound, thereby improving the accuracy of detecting abnormal winding of the protective cover.

[0014] Preferably, the acquisition time expression of the side view is:

[0015]

[0016] In the formula, i represents the acquisition time of the nth side view, n is the number of times the side view is acquired, is the thickness of the protective sleeve on the winding shaft, is the angle between the ray and the horizontal direction, v is the moving linear velocity of the protective sleeve, is the radius of the winding axis.

[0017] By calculating the acquisition time each time, the winding axis can be in the same spatial posture each time the side view is acquired, so that the angle between the ray and the horizontal direction is the same, thereby improving the accuracy of calculating the edge height of the target pixel point.

[0018] Preferably, the angle between the ray and the horizontal direction is in the range of ( ).

[0019] Preferably, the edge height of the target pixel is calculated as follows:

[0020] A rectangular coordinate system is constructed with the center of the winding axis as the origin to calculate the edge height. The expression is:

[0021]

[0022] in, represents the edge height of the target pixel in the side view at the i-th moment, is the position coordinate of the target pixel, The position coordinates of the pixel point corresponding to the center of the winding axis.

[0023] The formula can accurately calculate the edge height of the target pixel, which improves the accuracy of calculating the edge pixel.

[0024] Preferably, the method further comprises: obtaining a top view of the winding shaft assembly at different times, the top view and the side view Figure 1 One by one, auxiliary lines are constructed in the top view, and the auxiliary lines coincide with the axis of the winding shaft.

[0025] Preferably, the top view is converted into a grayscale image, and the evaluation set is constructed using the grayscale values ​​of the pixels in the grayscale image corresponding to the auxiliary lines;

[0026] Calculate the eigenvalues ​​of the grayscale values ​​in the evaluation set so that the eigenvalues ​​of the grayscale values ​​in the evaluation set are 0 or 1;

[0027] The ratio of the number of gray values ​​with an eigenvalue of 0 to the number of gray values ​​with an eigenvalue of 1 is taken as the abnormal winding degree; when the abnormal winding degree is greater than the abnormal threshold, a signal that the protective cover is seriously abnormally wound is issued; wherein the expression of the eigenvalue is:

[0028]

[0029] In the formula, is the characteristic value of the grayscale value of the kth pixel in the evaluation set of the top view at the i-th moment, is the gray value of the kth pixel in the evaluation set of the top view at the i-th moment, It is the preset standard gray value when there are n layers of protective sleeves on the winding shaft. is the evaluation set of the top view at the i-th moment.

[0030] By calculating the abnormal winding degree, the abnormality of the winding of the protective sleeve on the surface of the winding shaft can be understood in time, so as to facilitate response and reduce the loss caused by abnormal winding.

[0031] Preferably, the method further comprises:

[0032] In the top view, a vector is constructed using the pixel values ​​of the pixel points corresponding to the auxiliary lines, and a vector set is constructed using the obtained multiple vectors;

[0033] Calculate the eigenvalue of each vector in the vector set so that the eigenvalues ​​in the vector set are 0 or 1;

[0034] The ratio of the number of pixels with a characteristic value of 0 to the number of pixels with a characteristic value of 1 is used as the abnormal winding degree; when the abnormal winding degree is greater than the abnormal threshold, a signal that the protective cover is seriously abnormally wound is issued;

[0035] The expression of the eigenvalue is:

[0036]

[0037] In the formula, is the eigenvalue of the pth vector in the vector set of the top view at the i-th moment, It is the cosine pixel degree between the pth vector in the vector set of the top view at the i-th moment and the preset standard vector n.

[0038] Preferably, the expression of the standard height is:

[0039]

[0040] In the formula, It represents the standard height corresponding to the target point on the side view collected at the i-th moment, is the thickness of the single-layer protective cover on the winding shaft, is the radius of the winding axis, and n is the number of acquisitions.

[0041] By calculating the standard height corresponding to the target point, a reference standard is provided for judging whether the winding of the protective cover is abnormal, thereby improving the accuracy of the judgment result.

[0042] The present invention has the following technical effects:

[0043] 1. The abnormal winding evaluation is obtained by calculating the standard height and edge height of the target pixel point. The abnormal winding evaluation can timely determine whether the protective cover is wound, thereby improving the accuracy of the abnormal winding detection of the protective cover.

[0044] 2. By calculating the degree of abnormal winding, the abnormality of the protective sleeve winding on the surface of the winding shaft can be understood in time, so as to facilitate response and reduce the loss caused by abnormal winding. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0046] Figure 1 The present invention is a flow chart of a method for detecting abnormal winding of a protective cover based on machine vision.

[0047] Figure 2It is a side view of the winding shaft assembly of the present invention.

[0048] Figure 3 It is a schematic diagram of the state when the protective cover of the present invention is abnormally wound.

[0049] Figure 4 It is a top view schematic diagram of the winding shaft assembly of the present invention.

[0050] In the figure: 1. Winding shaft; 2. Protective cover. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0052] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0053] like Figure 2 As shown, during the production process of the protective cover, the protective cover 2 is wound on the winding shaft 1, and the protective cover 2 is wound on the winding shaft 1 at a fixed walking speed. The protective cover 2 and the winding shaft 1 together constitute a winding assembly, and the following is developed based on this background.

[0054] The embodiment of the present invention discloses a method for detecting abnormal winding of a protective cover based on machine vision, referring to Figure 1 , including the following steps, as follows:

[0055] S1: Obtain side views of the winding shaft assembly at different times, where the winding shaft assembly includes a winding shaft and a protective cover.

[0056] During the production process of the protective cover, the protective cover is wound on the surface of the winding shaft. It should be noted that during the winding process, the travel speed of the protective cover remains constant. As the thickness of the protective cover on the surface of the winding shaft gradually increases, the speed of the winding shaft gradually decreases to ensure that the protective cover can be smoothly wound on the surface of the winding shaft. An industrial camera is set on the left or right side of the winding shaft to obtain a side view of the winding shaft assembly.

[0057] S2: construct a ray with the axis of the winding shaft as the origin, and take the edge pixel point where the ray intersects with the outer contour of the protective cover as the target pixel point.

[0058] In the side view, the side view is converted into a grayscale image, and the Canny algorithm is used to detect the grayscale image to obtain the outer contour of the protective cover. The ray is constructed with the axis of the winding shaft as the origin, and the edge pixel point where the ray intersects with the outer contour of the protective cover is used as the target pixel point. When the protective cover is abnormally wound, the upper part of the contour in the side view will change, which is specifically manifested as uneven thickness of the protective cover above the winding shaft or partial loss of the contour line. In order to detect the winding abnormality in time, the value range of the angle between the ray and the horizontal direction is set to ( ), that is, the direction of the ray is upward or obliquely upward, so as to facilitate the timely discovery of abnormal winding phenomena.

[0059] It should be noted that during the winding process of the protective cover, the winding shaft is in a state of continuous rotation, and the rays also rotate with the winding shaft. In order to ensure the accuracy of the subsequent analysis results, when using the industrial camera to collect the side view, the collection time expression of the side view is:

[0060]

[0061] In the formula, i represents the acquisition time of the nth side view, n is the number of times the side view is acquired, is the thickness of the single-layer protective cover on the winding shaft, is the angle between the ray and the horizontal direction, v is the moving linear velocity of the protective sleeve, is the radius of the winding axis. The time for collecting the side view each time is calculated by the formula, so that the angle between the ray and the horizontal direction is kept consistent during each collection. It can also be understood that the side view is collected once every rotation of the winding axis.

[0062] Exemplary, combined Figure 2 As shown, in the initial state, the angle between the ray and the horizontal direction is 0°, the winding shaft starts to rotate, and the time required when the angle between the ray and the horizontal direction is 145° is calculated. The required time is the first acquisition time. If the first acquisition time is calculated to be 1.0 second, the side view of the winding shaft is first acquired at 1.0 second; then, the second acquisition time is calculated to be 3.0 seconds, and the side view of the winding shaft is acquired at 4.0 seconds; the third acquisition time is calculated to be 3.2 seconds, and the side view of the winding shaft is acquired at 7.2 seconds. In each acquisition, the angle between the ray and the horizontal direction is 145°. It can be understood that for each acquired side view, the target pixel points are all pixel points on the outermost contour.

[0063] S3: Calculate the edge height of the target pixel, where the edge height represents the distance between the target pixel and the axis; calculate the standard height of the target pixel, where the standard height represents the thickness of the protective cover when no misalignment occurs.

[0064] A rectangular coordinate system is constructed with the center of the winding axis as the origin to calculate the edge height. The expression is:

[0065]

[0066] in, It represents the edge height of the target pixel in the side view at the i-th moment, which can also be understood as the edge height of the target pixel in the side view collected at the n-th time. is the position coordinate of the target pixel, is the position coordinate of the pixel point corresponding to the axis of the winding axis. The distance between the target pixel point and the axis can be calculated by the formula.

[0067] The expression of the standard height of the target pixel is:

[0068]

[0069] In the formula, It represents the standard height corresponding to the target point on the side view collected at the i-th moment, is the thickness of the single-layer protective cover on the winding shaft, is the radius of the winding axis.

[0070] S4: Calculate the abnormal winding evaluation of the side view at each moment, and in response to the abnormal winding evaluation being greater than a preset evaluation threshold, send a signal indicating that the protective cover is abnormally wound.

[0071] The expression for abnormal winding evaluation is:

[0072]

[0073] In the formula, represents the abnormal winding evaluation of the side view at the i-th moment, represents the edge height of the target pixel in the side view at the i-th moment, Represents the standard height of the target pixel in the side view at the i-th moment.

[0074] like Figure 3As shown, when the protective cover is abnormally wound, the protective cover is wound on the surface of the winding shaft in an irregular winding manner. Therefore, the edge height of the target pixel point in the side view at the corresponding moment is greatly different from the standard height, so that the edge height of the target pixel point in the side view is greater than or less than the standard height. The abnormal winding evaluation can be used to determine whether the protective cover is in an abnormal state. Specifically, the larger the value of the abnormal winding evaluation is, the greater the possibility that the protective cover is in an abnormal winding state. Conversely, the smaller the value of the abnormal winding evaluation is, the smaller the possibility that the protective cover is in an abnormal winding state. The evaluation threshold is artificially set according to the actual situation. Exemplarily, the evaluation threshold is 0.3.

[0075] S5: Calculate the abnormal winding degree. When the abnormal winding degree is greater than the abnormal threshold, send a signal that the protective cover is seriously abnormally wound.

[0076] In one embodiment, a top view of the winding shaft assembly at different times is obtained, and the top view is as follows: Figure 4 As shown, top view and side view Figure 1 One by one, construct auxiliary lines in the top view, the auxiliary lines coincide with the axis of the winding shaft, that is, the auxiliary lines are located in the middle of the winding shaft assembly. The top view is an RGB format image, the top view is converted into a grayscale image, and the grayscale values ​​of the pixels in the grayscale image corresponding to the auxiliary lines are used to construct an evaluation set; calculate the characteristic values ​​of the grayscale values ​​in the evaluation set, so that the characteristic values ​​of the grayscale values ​​in the evaluation set are 0 or 1; the ratio of the number of grayscale values ​​with a characteristic value of 0 to the number of grayscale values ​​with a characteristic value of 1 is used as the abnormal winding degree; when the abnormal winding degree is greater than the abnormal threshold, a signal that the protective cover is seriously abnormally wound is issued;

[0077] The expression of the eigenvalue is:

[0078]

[0079] In the formula, is the characteristic value of the grayscale value of the kth pixel in the evaluation set of the top view at the i-th moment, is the gray value of the kth pixel in the evaluation set of the top view at the i-th moment, It is the preset standard gray value when there are n layers of protective sleeves on the winding shaft. The standard gray value is obtained according to the experiment. is the evaluation set of the top view at the i-th moment.

[0080] When the protective cover is not abnormally wound, the grayscale values ​​of the pixels in the grayscale image corresponding to the auxiliary lines are the same, and their grayscale values ​​are standard grayscale values; when abnormal winding occurs, the protective cover is wound on the surface of the winding shaft in an irregular manner, resulting in a decrease in the grayscale values ​​of some pixels. When the ratio of the number of grayscale values ​​with a characteristic value of 0 to the number of grayscale values ​​with a characteristic value of 1 in the evaluation set is greater than the preset abnormal threshold, it indicates that the protective cover is seriously abnormally wound, and the winding operation needs to be stopped at this time.

[0081] In one embodiment, a top view of the winding shaft assembly at different times is obtained, and the top view and the side view are Figure 1 One-to-one correspondence, construct auxiliary lines in the top view, and the auxiliary lines are parallel to the axis of the winding shaft. The top view is an RGB format image. In the top view, the pixel values ​​of the pixels corresponding to the auxiliary lines are used to construct vectors, and the obtained multiple vectors are used to construct a vector set; the eigenvalue of each vector in the vector set is calculated, so that the eigenvalue in the vector set is 0 or 1; the ratio of the number of pixels with an eigenvalue of 0 to the number of pixels with an eigenvalue of 1 is used as the abnormal winding degree; when the abnormal winding degree is greater than the abnormal threshold, a signal that the protective cover is seriously abnormally wound is issued;

[0082] The expression of the eigenvalue is:

[0083]

[0084] In the formula, is the eigenvalue of the pth vector in the vector set of the top view at the i-th moment, It is the cosine pixel degree between the pth vector in the vector set of the top view at the i-th moment and the preset standard vector n.

[0085] When the protective cover is not abnormally wound, the vectors of the pixels in the grayscale image corresponding to the auxiliary lines are the same. When abnormal winding occurs, the protective cover is wrapped around the surface of the winding shaft in an irregular manner, causing the vectors of some pixels to change. Therefore, the degree of pixel change is calculated by cosine similarity. When the ratio of the number of grayscale values ​​with an eigenvalue of 0 to the number of grayscale values ​​with an eigenvalue of 1 in the evaluation set is greater than the preset abnormal threshold, it indicates that the protective cover is seriously abnormally wound, and the winding operation needs to be stopped.

[0086] For example, for a pixel point in the top view, the values ​​of r, g, and b are 150, 120, and 200 respectively, then the vector of the pixel point is (150, 120, 200), and the standard vector is (160, 170, 180), so the cosine similarity between the vector (150, 120, 200) and the standard vector (160, 170, 180) is calculated. The method for calculating the cosine similarity is prior art, and the specific steps are not repeated here.

[0087] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0088] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting abnormal winding of a protective cover based on machine vision, characterized in that: Includes steps: Obtaining a side view and a top view of the winding shaft assembly at different times, wherein the top view corresponds to the side view one by one, and the winding shaft assembly includes a winding shaft and a protective cover; A ray is constructed with the axis of the winding shaft as the origin, and the edge pixel point where the ray intersects with the outer contour of the protective cover is taken as the target pixel point; Calculate the edge height of the target pixel, which represents the distance between the target pixel and the axis; calculate the standard height of the target pixel, which represents the thickness of the protective cover when no misalignment occurs; Calculating the abnormal winding evaluation of the side view at each moment, and in response to the abnormal winding evaluation being greater than a preset evaluation threshold, issuing a signal of abnormal winding of the protective cover; Among them, the expression for abnormal winding evaluation is: ; In the formula, represents the abnormal winding evaluation of the side view at the i-th moment, represents the edge height of the target pixel in the side view at the i-th moment, represents the standard height of the target pixel in the side view at the i-th moment; Construct auxiliary lines in the top view, and the auxiliary lines coincide with the axis of the winding shaft; The top view is converted into a grayscale image, and the evaluation set is constructed using the grayscale values ​​of the pixels in the grayscale image corresponding to the auxiliary lines; Calculate the eigenvalues ​​of the grayscale values ​​in the evaluation set so that the eigenvalues ​​of the grayscale values ​​in the evaluation set are 0 or 1; The ratio of the number of gray values ​​with an eigenvalue of 0 to the number of gray values ​​with an eigenvalue of 1 is taken as the abnormal winding degree; when the abnormal winding degree is greater than the abnormal threshold, a signal that the protective cover is seriously abnormally wound is issued; wherein the expression of the eigenvalue is: In the formula, is the characteristic value of the grayscale value of the kth pixel in the evaluation set of the top view at the i-th moment, is the gray value of the kth pixel in the evaluation set of the top view at the i-th moment, It is the preset standard gray value when there are n layers of protective sleeves on the winding shaft. is the evaluation set of the top view at the i-th moment.

2. The method for detecting abnormal winding of a protective cover based on machine vision according to claim 1, characterized in that: The acquisition time expression of the side view is: In the formula, i represents the acquisition time of the nth side view, n is the number of times the side view is acquired, is the thickness of the protective sleeve on the winding shaft, is the angle between the ray and the horizontal direction, v is the moving linear velocity of the protective sleeve, is the radius of the winding axis.

3. The method for detecting abnormal winding of a protective cover based on machine vision according to claim 1, characterized in that: The range of the angle between the ray and the horizontal direction is ( ).

4. The method for detecting abnormal winding of a protective cover based on machine vision according to claim 1, characterized in that: The calculation method of the edge height of the target pixel is: A rectangular coordinate system is constructed with the center of the winding axis as the origin to calculate the edge height. The expression is: in, represents the edge height of the target pixel in the side view at the i-th moment, is the position coordinate of the target pixel, The position coordinates of the pixel point corresponding to the center of the winding axis.

5. The method for detecting abnormal winding of a protective cover based on machine vision according to claim 1, characterized in that: The method also includes: In the top view, a vector is constructed using the pixel values ​​of the pixel points corresponding to the auxiliary lines, and a vector set is constructed using the obtained multiple vectors; Calculate the eigenvalue of each vector in the vector set so that the eigenvalues ​​in the vector set are 0 or 1; The ratio of the number of pixels with a characteristic value of 0 to the number of pixels with a characteristic value of 1 is used as the abnormal winding degree; when the abnormal winding degree is greater than the abnormal threshold, a signal that the protective cover is seriously abnormally wound is issued; The expression of the eigenvalue is: In the formula, is the eigenvalue of the pth vector in the vector set of the top view at the i-th moment, It is the cosine pixel degree between the pth vector in the vector set of the top view at the i-th moment and the preset standard vector n.

6. The method for detecting abnormal winding of a protective cover based on machine vision according to claim 1, characterized in that: The expression for standard height is: In the formula, represents the standard height corresponding to the target point on the side view collected at the i-th moment, is the thickness of the single-layer protective cover on the winding shaft, is the radius of the winding axis, and n is the number of acquisitions.

Images