A method, device and system for detecting film effect of folding screen

Abstract

The present invention relates to a method, device and system for detecting the film effect of a folding screen, and relates to the technical field of film effect detection, including: performing bubble detection on the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics; when the bubble distribution characteristics are different from the preset bubble distribution law, generating a folding screen film unqualified mark; when the bubble distribution characteristics are the same as the preset bubble distribution law, and the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified, generating a folding screen film qualified mark. The technical problem that the prior art cannot effectively identify the film state of the folding screen bending area and its surrounding areas, resulting in a large detection error rate, is solved. By judging whether the bubble distribution law of the folding screen bending mark area meets the preset distribution law, if it is different, the folding screen film is unqualified, and if it is the same, it is qualified, thereby achieving the technical effect of improving the accuracy of folding screen film detection.

Description

Technical Field

[0001] The present invention relates to the field of film sticking effect detection, and in particular to a method, device and system for detecting the film sticking effect of a folding screen. Background Art

[0002] The existing folding screen film effect recognition algorithms are inherited from the previous non-folding screen mobile phone production line. It mainly uses high-precision cameras to perform image recognition on the folding screen mobile phone with film applied, and judges the number and volume of bubbles in it to determine the quality of the film application effect.

[0003] However, the bending area of ​​the folding screen is generally concave or convex, so it is normal for there to be many larger bubbles in the bending area and the area near the bending. However, the existing folding screen film effect recognition algorithm will judge the folding screen products with such normal phenomena as unqualified products. In summary, the existing film effect recognition algorithm has the disadvantage of a large error rate in detecting the folding screen film effect. Summary of the invention

[0004] The present invention aims to solve the technical problem that the prior art cannot effectively identify the film status of the bending area of ​​the folding screen and its surrounding areas, resulting in a large detection error rate, and provides a folding screen film effect detection method, device and system to solve the problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for detecting the effect of a folding screen film, comprising: performing bubble detection on a bending mark area of ​​a folding screen film surface image to obtain a bubble distribution feature; when the bubble distribution feature is different from a preset bubble distribution law, generating a folding screen film unqualified mark; when the bubble distribution feature is the same as the preset bubble distribution law, and the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified, generating a folding screen film qualified mark;

[0007] Furthermore, bubble detection is performed on the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics, which also includes: obtaining the folding screen film surface image; performing bending area detection on the folding screen film surface image to obtain the bending area, wherein the bending area has a bending area width and a bending area center axis; based on the bending area center axis, the bending area width is enlarged to a preset multiple of the bending area width to obtain the bending mark area; the area outside the bending mark area of ​​the folding screen film surface image is set as the non-bending mark area.

[0008] Furthermore, the bending zone width is enlarged to a preset multiple of the bending zone width based on the central axis of the bending zone to obtain the bending mark area, including: when the bending zone width is greater than or equal to 0 mm and less than 5 mm, the preset multiple is 1.4~1.6; when the bending zone width is greater than or equal to 5 mm and less than 10 mm, the preset multiple is 1.7~2.1; when the bending zone width is greater than or equal to 10 mm, the preset multiple is 2.2~2.4.

[0009] Further, the bending area detection is performed on the folding screen film surface image to obtain the bending area, including:

[0010] The folding screen film surface image is detected by a bending zone detection model to obtain the bending zone; wherein the bending zone detection model construction step includes: collecting a folding screen film surface image set and a folding screen film surface bending zone identification image set; using the folding screen film surface bending zone identification image set as supervision and the folding screen film surface image set as input to train the bending zone detection model.

[0011] Furthermore, bubble detection is performed on the bend mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics, including: dividing the bend mark area into a first bend mark area and a second bend mark area along the central axis of the bend area; performing bubble detection on the first bend mark area to obtain a first bubble distribution characteristic; performing bubble detection on the second bend mark area to obtain a second bubble distribution characteristic; and adding the first bubble distribution characteristic and the second bubble distribution characteristic into the bubble distribution characteristic.

[0012] Furthermore, when the bubble distribution feature is different from the preset bubble distribution law, an unqualified mark for the folding screen film is generated, including: training a bubble trend detection model according to the preset bubble distribution law, wherein the bubble trend detection model includes a first bubble trend detection model and a second bubble trend detection model; the bubble distribution feature includes the first bubble distribution feature and the second bubble distribution feature; inputting the first bubble distribution feature into the first bubble trend detection model for trend detection to obtain a first trend detection result; inputting the second bubble distribution feature into the second bubble trend detection model for trend detection to obtain a second trend detection result; when any one of the first trend detection result and the second trend detection result is unqualified, an unqualified mark for the folding screen film is generated; wherein, training the bubble trend detection model according to the preset bubble distribution law includes: collecting a first bubble distribution feature record data set in the first bend mark area; marking the first bubble distribution feature record data set according to the preset bubble distribution law to obtain a first trend detection result identification data set; using the first trend detection result identification data set as supervision, using the second trend detection result identification data set as supervision, using the second trend detection result identification data set as supervision, using the second trend detection result identification data set as supervision, using the second trend detection result identification data set as supervision, using the second trend detection result identification data set as supervision, using the first trend detection result identification data set as supervision, using the first trend detection result identification data set as supervision, using the first trend detection result identification data set as supervision, using the first trend detection result identification data set as supervision, using the second trend detection result as supervision, using the first trend detection result identification data set ... A bubble distribution feature record data set is used as input to train the first bubble trend detection model; a second bubble distribution feature record data set is collected in the second bend marking area; the second bubble distribution feature record data set is marked according to the preset bubble distribution law to obtain a second trend detection result mark data set; the second bubble trend detection model is trained with the second trend detection result mark data set as supervision and the second bubble distribution feature record data set as input; wherein, the first bubble distribution feature record data set is marked according to the preset bubble distribution law to obtain a first trend detection result mark data set, including: obtaining the first bubble distribution feature record data of the first bubble distribution feature record data set; the preset bubble distribution law is that the number and area of ​​bubbles exceeding a preset number and proportion decrease successively along the direction away from the central axis of the bending area; when the first bubble distribution feature record data satisfies that the number and area of ​​bubbles decrease successively along the direction away from the central axis of the bending area, the first trend detection result of the first bubble distribution feature record data is marked as qualified; otherwise, the first trend detection result of the first bubble distribution feature record data is marked as unqualified.

[0013] Furthermore, when the bubble distribution characteristics are different from the preset distribution pattern of bubbles, an unqualified mark for the folding screen film is generated, including: counting the number of first bubbles whose bubble area is greater than or equal to a first bubble area threshold according to the bubble distribution characteristics; when the first number of bubbles is greater than or equal to a first bubble number threshold, or / and when the bubble distribution characteristics are different from the preset distribution pattern of bubbles, an unqualified mark for the folding screen film is generated.

[0014] Furthermore, when the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified, it includes: performing bubble detection on the non-bending mark area to obtain the bubble distribution characteristics of the non-bending mark area; according to the bubble distribution characteristics of the non-bending mark area, counting the number of second bubbles whose bubble area is greater than or equal to a second bubble area threshold, wherein the second bubble area threshold is less than or equal to the first bubble area threshold; when the second bubble number is greater than or equal to the second bubble number threshold, generating the folding screen film unqualified mark; when the second bubble number is less than the second bubble number threshold, it is deemed that the film effect detection of the non-bending mark area is qualified.

[0015] In a second aspect, the present invention provides a folding screen film effect detection device, which is used to implement a folding screen film effect detection method described in any one of the first aspects.

[0016] In a third aspect, the present invention provides a folding screen film effect detection system, comprising: a bubble detection module, used to perform bubble detection on the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics; a first decision module, used to generate an unqualified folding screen film mark when the bubble distribution characteristics are different from the preset bubble distribution pattern; a second decision module, used to generate a qualified folding screen film mark when the bubble distribution characteristics are the same as the preset bubble distribution pattern and the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified.

[0017] The beneficial effects of the present invention are:

[0018] A method for detecting the effect of a folding screen film is provided, specifically, bubble detection is performed on the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics; when the bubble distribution characteristics are different from the preset bubble distribution pattern, an unqualified folding screen film mark is generated; when the bubble distribution characteristics are the same as the preset bubble distribution pattern, and the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified, a technical solution of generating a qualified folding screen film mark is provided. Since the film bubbles in the folding screen bending mark area have a certain distribution pattern, it is judged whether the bubble distribution pattern in the folding screen bending mark area conforms to the preset distribution pattern. If they are different, the folding screen film is unqualified, and if they are the same, the folding screen film is qualified. Compared with the traditional detection method that only focuses on the volume, number and area of ​​bubbles, the detection accuracy is higher, thereby achieving the technical effect of improving the detection accuracy of folding screen films. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of a method for detecting the effect of folding screen film sticking in the prior art;

[0020] Figure 2A schematic diagram of the structure of the folding screen bending mark area of ​​a folding screen film sticking effect detection method provided by the present invention;

[0021] Figure 3 A schematic diagram of the partition structure of the folding screen bend identification area of ​​a folding screen film effect detection method provided by the present invention.

[0022] Description of reference numerals:

[0023] L: the center axis of the bending zone, b: the width of the bending zone, B: the width of the bending mark area, A: the length of the bending mark area, S1: the direction of the first bending mark area away from the center axis, S2: the direction of the second bending mark area away from the center axis. DETAILED DESCRIPTION

[0024] 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 only 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.

[0025] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0026] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0027] Embodiment 1:

[0028] like Figure 1 As shown, an embodiment of the present invention provides a method for detecting the effect of a folding screen film, comprising the steps of:

[0029] S10: performing bubble detection on the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics;

[0030] S20: When the bubble distribution feature is different from the preset bubble distribution rule, generating a folding screen film unqualified mark;

[0031] S30: When the bubble distribution feature is the same as the preset bubble distribution rule, and the film pasting effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified, a folding screen film passing mark is generated.

[0032] Specifically, the folding screen of the embodiment of the present application can be a product that uses a folding screen, such as a mobile phone, a tablet, and a computer, and the image of the film surface of the folding screen refers to the image of the film surface of the folding screen.

[0033] The bending mark area includes the bending area of ​​the folding screen, also known as the keel structure and the area near the bending area. The non-bending mark area is the area other than the bending mark area. Preferably, any folding screen film surface image has preset markings of the bending mark area and the non-bending mark area, so as to facilitate the subsequent use of a different folding screen film effect detection algorithm for different areas. The specific marking method can be manual marking or machine vision automated marking.

[0034] In a preferred embodiment, the film effect detection algorithm for the non-bending mark area can be detected by using the conventional film effect detection algorithm for ordinary mobile phones, that is, a high-precision camera is used to perform image recognition on the folding screen mobile phone with the film, and the quality of the film effect is determined by judging the number and volume of bubbles therein. When the number of bubbles exceeding the area threshold or the volume threshold is more than a certain number, it is regarded as unqualified, otherwise it is regarded as qualified. Since the film effect detection algorithm for ordinary mobile phones is relatively mature, it will not be elaborated here.

[0035] In a preferred embodiment, the film effect detection algorithm for the bend mark area adopts the detection algorithm customized in the embodiment of the present application for detection, and the specific process is as follows:

[0036] Since the keel structure of the folding screen is raised or sunken, the bubbles in the normal film show a certain distribution pattern from the central axis to the two sides of the folding screen, that is, the height difference between the film and the screen caused by the keel structure becomes smaller and smaller from the central axis to the two sides of the folding screen, so the number of bubbles and the area of ​​the bubbles are also getting smaller and smaller. Therefore, the pattern of the number of bubbles and the area of ​​the bubbles decreasing from the central axis to the two sides of the folding screen can be quantified as a preset distribution pattern of bubbles, that is, the number of bubbles decreases from the central axis to the two sides of the folding screen, and the area of ​​the bubbles decreases from the central axis to the two sides of the folding screen.

[0037] By machine vision, the bubbles in the bending mark area are identified, the distribution data of each bubble is determined, and stored as the bubble distribution characteristics. Preferably, the algorithm flow for determining the bubble distribution characteristics from the central axis to both sides of the folding screen is as follows:

[0038] Step 1: Divide the bubble distribution characteristics into first-side bubble distribution characteristics and second-side bubble distribution characteristics along the central axis;

[0039] Step 2: Then, according to the bubble distribution characteristics on the first side, the distance between any two bubbles on the first side in the direction perpendicular to the central axis is calculated. When the distance is less than or equal to the distance threshold preset by the user, they are considered to belong to a straight line parallel to the central axis.

[0040] Step 3: Repeat clustering to gather all the bubbles on the first side onto multiple straight lines parallel to the central axis.

[0041] Step 4: Count the number of bubbles on each straight line parallel to the central axis from near to far, and obtain the distribution sequence of the number of bubbles on the first side. Use the same method to determine the distribution sequence of the number of bubbles on the second side.

[0042] Step 5: Count the average area of ​​bubbles on each straight line from the distance to the central axis to obtain the bubble area distribution sequence on the first side, and determine the bubble area distribution sequence on the second side in the same way.

[0043] Furthermore, it is determined whether the bubbles conform to the rule that the number of bubbles decreases from the central axis to the two sides of the folding screen, and the area of ​​bubbles decreases from the central axis to the two sides of the folding screen. If so, the bubble distribution characteristics are the same as the preset bubble distribution pattern. If the film effect test of the non-bending mark area of ​​the folding screen film surface image is qualified, a folding screen film qualified mark is generated. When the bubble distribution characteristics are different from the preset bubble distribution pattern, an unqualified folding screen film mark is generated.

[0044] Furthermore, the embodiment of the present application performs film pasting effect detection on the bending mark area, which can be performed simultaneously with the film pasting effect detection on the non-bending mark area; or the film pasting effect detection on the bending mark area can be performed first, and then the film pasting effect detection on the non-bending mark area; or the film pasting effect detection on the non-bending mark area can be performed first, and then the film pasting effect detection on the bending mark area. The embodiment of the present application does not impose many restrictions on this, and it can be selected according to actual conditions during specific implementation.

[0045] Further, such as Figure 2 As shown, the folding screen film surface image is subjected to bubble detection in the bending mark area to obtain bubble distribution characteristics, and the step S40 is also included before: Step S40 includes the steps of:

[0046] S410: Obtaining the folding screen film surface image;

[0047] S420: Perform bending zone detection on the folding screen film surface image to obtain the bending zone, wherein the bending zone has a bending zone width and a bending zone center axis;

[0048] S430: Enlarging the width of the bending zone to a preset multiple of the width of the bending zone based on the central axis of the bending zone to obtain the bending mark area;

[0049] S440: setting the area outside the bending mark area of ​​the folding screen film surface image as the non-bending mark area.

[0050] Specifically, after the product has been film-laminated on the film production line, it will be transported to the designated area for positioning, and then a photo will be taken of one side of the product screen with the film laminated on it to obtain an initial picture. The initial picture will then be cut according to the positioning area of ​​the product screen to obtain an image of the film surface of the folding screen.

[0051] The bending area is the area where the keel of the folding screen belongs, which is generally a concave area or a convex area; the central axis of the bending area refers to the axis connecting the two folded screens, such as Figure 2 As shown, L is the center axis of the bending area, and the width of the bending area is the width of the concave area or the convex area. Figure 2 The b shown is the width of the bending zone. Further, the length of the bending zone is the length of the screen, that is, Figure 2 The A in is also the length of the subsequent bend mark area.

[0052] The bending zone is easy to determine, but since the width of the bending zone of different products is different, it is difficult to determine the width of the corresponding area near the bending zone affected by the bending zone. Therefore, the embodiment of the present application sets an adjustable width, that is, along the central axis of the bending zone, the bending zone width is enlarged to a preset multiple of the bending zone width, and the area is set as the bending mark area; the preset multiple of the bending zone width is the adjustable width, and the user can customize the setting according to the actual application scenario, such as Figure 2 B in the figure is the area width obtained by enlarging the bending zone width to a preset multiple of the bending zone width along the central axis of the bending zone.

[0053] Furthermore, the area outside the bending mark area of ​​the folding screen film surface image is set as the non-bending mark area.

[0054] Furthermore, the bend mark area of ​​the present application is for ease of understanding. Figure 2 It is shown as a rectangle, but it can actually be other shapes and can be adjusted according to actual needs.

[0055] Specifically, in a preferred embodiment, the preset multiple can be set based on the following parameters:

[0056] Further, based on the central axis of the bending zone, the width of the bending zone is enlarged to a preset multiple of the width of the bending zone to obtain the bending mark area, and step S430 includes the steps of:

[0057] S431: When the width of the bending zone is greater than or equal to 0 mm and less than 5 mm, the preset multiple is 1.4 to 1.6, preferably 1.5 times;

[0058] S432: When the width of the bending zone is greater than or equal to 5 mm and less than 10 mm, the preset multiple is 1.7 to 2.1, preferably 2 times:

[0059] S433: When the width of the bending area is greater than or equal to 10 mm, the preset multiple is 2.2~2.4. For mobile phones with a width of more than 10 mm, since there is sufficient space to place the keel structure, its height is not high. Therefore, a width magnification multiple of 2.3 times is sufficient, preferably 2.3 times.

[0060] If there is no customized preset multiple, the preset multiple is configured according to the above range by default, which ensures the intelligence of the marking in the bend mark area.

[0061] Further, the bending area detection is performed on the folding screen film surface image to obtain the bending area, and step S420 includes the steps of:

[0062] S421: Detecting the folding screen film surface image using a bending zone detection model to obtain the bending zone;

[0063] S4211: wherein the step of constructing the bending zone detection model comprises:

[0064] S42111: collecting a set of images of the folding screen film surface and a set of images of markings of the bending area of ​​the folding screen film surface;

[0065] S42112: Using the folding screen film surface bending area identification image set as supervision and the folding screen film surface image set as input, training the bending area detection model.

[0066] Specifically, the film images of the folding screen products are collected and stored as a folding screen film surface image set, and then the bending areas on the film images of the folding screen products are marked to obtain a folding screen film surface bending area identification image set, which is set as a bending area detection model construction data set.

[0067] Preferably, the bending zone detection model construction data set is divided into a ratio of 8:2, wherein the bending zone detection model construction data set with a ratio of 8 is set as the bending zone detection model training data set, and the bending zone detection model construction data set with a ratio of 2 is set as the bending zone detection model verification data set.

[0068] During training, the convolutional neural network is trained with the image set of the bending area identification on the folding screen film surface as supervision and the image set of the folding screen film surface as input, and the ratio of the non-intersection area area and the union area area of ​​the output bending area and the bending area identification is used as the loss function. When the loss function is less than or equal to the loss function threshold for a consecutive preset number of times, the bending area detection model verification data set is retrieved for verification. When the loss function is less than or equal to the loss function threshold for a consecutive preset number of times, it indicates that the model has converged, and the output is the bending area detection model.

[0069] The neural network model can efficiently detect the bending area, ensuring detection efficiency while ensuring detection accuracy.

[0070] Further, such as Figure 3 As shown, the folding mark area of ​​the folding screen film surface image is subjected to bubble detection to obtain bubble distribution characteristics. Step S10 includes the following steps:

[0071] S11: dividing the bending mark area into a first bending mark area and a second bending mark area along the central axis of the bending area;

[0072] S12: performing bubble detection on the first bending mark area to obtain a first bubble distribution feature;

[0073] S13: performing bubble detection on the second bending mark area to obtain a second bubble distribution feature;

[0074] S14: adding the first bubble distribution feature and the second bubble distribution feature into the bubble distribution feature.

[0075] Specifically, the bending mark area can be divided into left and right bending mark areas along the central axis of the bending area, one side of the bending mark area is set as the first bending mark area, and the other side of the bending mark area is set as the second bending mark area.

[0076] For example, Figure 3 L is the central axis, the left side of the central axis is the first bending mark area, and the right side of the central axis is the second bending mark area; S1 represents the direction of the first bending mark area away from the central axis, and S2 represents the direction of the second bending mark area away from the central axis.

[0077] Perform bubble detection on the first bend mark area to obtain the first bubble distribution feature, that is, detect the distribution coordinates of the bubbles in the first bend mark area; perform bubble detection on the second bend mark area to obtain the second bubble distribution feature, that is, detect the distribution coordinates of the bubbles in the second bend mark area, and add the first bubble distribution feature and the second bubble distribution feature into the bubble distribution feature. Preferably, the embodiment of the present application detects the distribution characteristics of bubbles through a machine vision model. It should be explained that the bubble distribution coordinates of the embodiment of the present application are not one coordinate point for each bubble. One bubble corresponds to a central coordinate and an edge coordinate. The central coordinate can facilitate the analysis of the number of bubbles, and the edge coordinate can be used to analyze the area of ​​the bubbles. Since the specific feature detection of bubbles is already relatively mature in the existing technology, it will not be elaborated here.

[0078] Further, when the bubble distribution feature is different from the preset bubble distribution law, a folding screen film unqualified mark is generated, and step S20 includes the steps of:

[0079] S21: training a bubble trend detection model according to the preset bubble distribution law, wherein the bubble trend detection model includes a first bubble trend detection model and a second bubble trend detection model;

[0080] S22: the bubble distribution feature includes the first bubble distribution feature and the second bubble distribution feature;

[0081] S23: inputting the first bubble distribution feature into the first bubble trend detection model to perform trend detection to obtain a first trend detection result;

[0082] S24: inputting the second bubble distribution feature into the second bubble trend detection model to perform trend detection to obtain a second trend detection result;

[0083] S25: When any one of the first trend detection result and the second trend detection result is unqualified, generating an unqualified mark of the folding screen film.

[0084] Specifically, the bubble trend detection model is an intelligent model used to analyze whether the bubble trend conforms to the preset distribution law of bubbles. In basic applications, bubble trend detection is performed through a custom algorithm. Here, a preferred implementation method is provided to realize bubble trend detection through a machine learning model, which can improve detection efficiency and stability. The details are as follows:

[0085] The bubble trend detection model includes a first bubble trend detection model and a second bubble trend detection model. The first bubble trend detection model is used to detect the bubble trend on one side of the central axis, and the second bubble trend detection model is used to detect the bubble trend on the other side of the central axis. The training process of the first bubble trend detection model and the second bubble trend detection model is the same, as follows:

[0086] Furthermore, the bubble trend detection model is trained according to the preset distribution law of the bubbles, including:

[0087] Collecting a first bubble distribution feature record data set in a first bend mark area;

[0088] The first bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a first trend detection result mark data set;

[0089] Using the first trend detection result identification data set as supervision and the first bubble distribution feature record data set as input, training the first bubble trend detection model;

[0090] Collect a second bubble distribution feature record data set in the second bend mark area;

[0091] The second bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a second trend detection result mark data set;

[0092] The second trend detection result identification data set is used as supervision and the second bubble distribution feature record data set is used as input to train the second bubble trend detection model.

[0093] Specifically, a first bubble distribution feature record data set of the first bend identification area in the historical folding screen image set is collected; the first bubble distribution feature record data set is identified according to a preset bubble distribution law to obtain a first trend detection result identification data set; the convolution classifier is trained with the first trend detection result identification data set as supervision and the first bubble distribution feature record data set as input to obtain a first bubble trend detection model, which is a binary classification model, and the output is qualified or unqualified.

[0094] Similarly, a second bubble distribution feature record data set of the second bend mark area in the historical folding screen image set is collected; the second bubble distribution feature record data set is marked according to the preset bubble distribution law to obtain a second trend detection result mark data set; the second trend detection result mark data set is used as supervision and the second bubble distribution feature record data set is used as input to train the convolution classifier to obtain a second bubble trend detection model, which is a two-class model, and the output is qualified or unqualified.

[0095] Furthermore, the specific model training process is relatively conventional and will not be described in detail. In order to enable the bubble trend detection model to achieve intelligent detection that conforms to the preset distribution law of bubbles, special data identification is required. Taking the first trend detection result identification data set as an example, the detailed process is as follows:

[0096] Further, the first bubble distribution feature record data set is marked according to the preset bubble distribution law to obtain a first trend detection result mark data set, including:

[0097] Obtaining first bubble distribution feature record data of the first bubble distribution feature record data set;

[0098] The preset distribution rule of the bubbles is that the number and area of ​​the bubbles exceeding the preset number and proportion decrease in sequence along the direction away from the central axis of the bending zone;

[0099] When the first bubble distribution characteristic record data satisfies that the number and area of ​​bubbles decrease in sequence along a direction away from the central axis of the bending zone, the first trend detection result of the first bubble distribution characteristic record data is marked as qualified;

[0100] Otherwise, the first trend detection result of the first bubble distribution characteristic record data is marked as unqualified.

[0101] The first bubble distribution feature record data of the first bubble distribution feature record data set is randomly extracted. Preferably, the preset distribution rule of bubbles is that the number and area of ​​bubbles exceeding the preset number and proportion decrease successively in the direction away from the central axis of the bending zone. When the first bubble distribution feature record data satisfies the condition that the number and area of ​​bubbles decrease successively in the direction away from the central axis of the bending zone, the first trend detection result of the first bubble distribution feature record data is marked as qualified; otherwise, the first trend detection result of the first bubble distribution feature record data is marked as unqualified. In the embodiment of the present application, by setting the preset number of bubbles, it does not mean that all bubbles need to satisfy the condition that the number and area decrease successively in the direction away from the central axis of the bending zone. That is to say, it means that the bubbles as a whole only need to meet the rule that the number and area of ​​bubbles decrease successively in the direction away from the central axis of the bending zone. Occasionally, there are deviations in the area and number of some bubbles, which can be regarded as fault tolerance, and such folding screen film will not be classified as abnormal.

[0102] Further, when the bubble distribution feature is different from the preset bubble distribution law, generating a folding screen film unqualified mark includes step S50:

[0103] S51: according to the bubble distribution characteristics, counting the number of first bubbles whose bubble areas are greater than or equal to a first bubble area threshold;

[0104] S52: When the first bubble quantity is greater than or equal to a first bubble quantity threshold, or / and when the bubble distribution feature is different from a preset bubble distribution pattern, a mark indicating that the folding screen film is unqualified is generated.

[0105] Specifically, when judging the bubble distribution characteristics of the bend mark area and the preset bubble distribution law, it is also necessary to judge whether the bubble area of ​​the bend mark area is larger. If it exists, it is also unqualified. Therefore, the bubble area can be counted according to the edge characteristics of the bubbles of the bubble distribution characteristics, and then the number of first bubbles whose bubble area is greater than or equal to the first bubble area threshold is calculated. The first bubble area threshold is the bubble area threshold preset by the user. When the first number of bubbles is greater than or equal to the first bubble number threshold, preferably, the first bubble number threshold is the maximum number of bubbles exceeding the first bubble area threshold that can be tolerated by the user. By default, the first bubble number threshold is 1. Therefore, when the first number of bubbles is greater than or equal to the first bubble number threshold, or / and when the bubble distribution characteristics are different from the preset bubble distribution law, the folding screen film unqualified mark is generated.

[0106] Further, when the film pasting effect detection of the non-bending mark area of ​​the folding screen film pasting surface image is qualified, step S30 includes the steps of:

[0107] S31: performing bubble detection on the non-bending mark area to obtain bubble distribution characteristics of the non-bending mark area;

[0108] S32: according to the bubble distribution characteristics of the non-bending mark area, counting the number of second bubbles whose bubble area is greater than or equal to a second bubble area threshold, wherein the second bubble area threshold is less than or equal to the first bubble area threshold;

[0109] S33: when the second bubble quantity is greater than or equal to a second bubble quantity threshold, generating an unqualified mark for the folding screen film;

[0110] S34: When the second bubble number is less than a second bubble number threshold, it is deemed that the film pasting effect detection of the non-bending mark area is qualified.

[0111] Specifically, the bubble distribution characteristics of the non-bending mark area refer to the results of bubble detection of the non-bending mark area through machine vision. According to the bubble distribution characteristics of the non-bending mark area, the number of second bubbles whose bubble area is greater than or equal to the second bubble area threshold is counted. Since the fault tolerance of the bending mark area needs to be higher than that of the non-bending mark area, it is preferred that the second bubble area threshold is less than or equal to the first bubble area threshold; when the second bubble number is greater than or equal to the second bubble number threshold, the folding screen film unqualified mark is generated; when the second bubble number is less than the second bubble number threshold, the film effect detection of the non-bending mark area is deemed to be qualified.

[0112] A folding screen film effect detection method provided by an embodiment of the present invention has at least the following technical effects:

[0113] A method for detecting the effect of a folding screen film is provided, specifically, bubble detection is performed on the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics; when the bubble distribution characteristics are different from the preset bubble distribution pattern, an unqualified folding screen film mark is generated; when the bubble distribution characteristics are the same as the preset bubble distribution pattern, and the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified, a technical solution of generating a qualified folding screen film mark is provided. Since the film bubbles in the folding screen bending mark area have a certain distribution pattern, it is judged whether the bubble distribution pattern in the folding screen bending mark area conforms to the preset distribution pattern. If they are different, the folding screen film is unqualified, and if they are the same, the folding screen film is qualified. Compared with the traditional detection method that only focuses on the volume, number and area of ​​bubbles, the detection accuracy is higher, thereby achieving the technical effect of improving the detection accuracy of folding screen films.

[0114] Embodiment 2:

[0115] Based on the same inventive concept of a folding screen film effect detection method provided in Example 1, the present invention also provides a folding screen film effect detection device, which is used to implement a folding screen film effect detection method as described in any one of Example 1.

[0116] Embodiment 3:

[0117] Based on the same inventive concept as the folding screen film effect detection method provided in Example 1, the present invention also provides a folding screen film effect detection system, including:

[0118] The bubble detection module is used to detect bubbles in the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics;

[0119] The first decision module is used to generate a folding screen film unqualified mark when the bubble distribution feature is different from the preset bubble distribution law;

[0120] The second decision module is used to generate a folding screen film qualified mark when the bubble distribution feature is the same as the preset bubble distribution law and the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified.

[0121] Furthermore, the bubble detection module executes the steps further comprising:

[0122] Obtaining the folding screen film surface image;

[0123] Performing bending zone detection on the folding screen film surface image to obtain the bending zone, wherein the bending zone has a bending zone width and a bending zone center axis;

[0124] Based on the central axis of the bending zone, the width of the bending zone is enlarged to a preset multiple of the width of the bending zone to obtain the bending mark area;

[0125] The area outside the bending mark area of ​​the folding screen film surface image is set as the non-bending mark area.

[0126] Furthermore, the bubble detection module executes the steps further comprising:

[0127] When the width of the bending zone is greater than or equal to 0 mm and less than 5 mm, the preset multiple is 1.4 to 1.6;

[0128] When the width of the bending zone is greater than or equal to 5 mm and less than 10 mm, the preset multiple is 1.7 to 2.1;

[0129] When the width of the bending zone is greater than or equal to 10 mm, the preset multiple is 2.2-2.4.

[0130] Furthermore, the bubble detection module executes the steps further comprising:

[0131] The bending area detection model is used to detect the folding screen film surface image to obtain the bending area;

[0132] The step of constructing the bending zone detection model includes:

[0133] Collecting a set of images of the folding screen film surface and a set of images of markings of the bending area of ​​the folding screen film surface;

[0134] The bending area detection model is trained by taking the bending area identification image set of the folding screen film surface as supervision and the bending area identification image set of the folding screen film surface as input.

[0135] Furthermore, the bubble detection module executes the steps further comprising:

[0136] Dividing the bending mark area into a first bending mark area and a second bending mark area along the central axis of the bending area;

[0137] Performing bubble detection on the first bending mark area to obtain a first bubble distribution feature;

[0138] Performing bubble detection on the second bending mark area to obtain a second bubble distribution feature;

[0139] The first bubble distribution feature and the second bubble distribution feature are added to the bubble distribution feature.

[0140] Furthermore, the first decision module execution step further includes:

[0141] Training a bubble trend detection model according to the preset bubble distribution law, wherein the bubble trend detection model includes a first bubble trend detection model and a second bubble trend detection model;

[0142] The bubble distribution feature includes the first bubble distribution feature and the second bubble distribution feature;

[0143] Inputting the first bubble distribution feature into the first bubble trend detection model to perform trend detection to obtain a first trend detection result;

[0144] Inputting the second bubble distribution feature into the second bubble trend detection model to perform trend detection to obtain a second trend detection result;

[0145] When any one of the first trend detection result and the second trend detection result is unqualified, generating an unqualified mark of the folding screen film;

[0146] Wherein, training a bubble trend detection model according to the preset bubble distribution law includes:

[0147] Collecting a first bubble distribution feature record data set in a first bend mark area;

[0148] The first bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a first trend detection result mark data set;

[0149] Using the first trend detection result identification data set as supervision and the first bubble distribution feature record data set as input, training the first bubble trend detection model;

[0150] Collect a second bubble distribution feature record data set in the second bend mark area;

[0151] The second bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a second trend detection result mark data set;

[0152] Using the second trend detection result identification data set as supervision and the second bubble distribution feature record data set as input, training the second bubble trend detection model;

[0153] The step of marking the first bubble distribution feature record data set according to the preset bubble distribution law to obtain a first trend detection result marking data set includes:

[0154] Obtaining first bubble distribution feature record data of the first bubble distribution feature record data set;

[0155] The preset distribution rule of the bubbles is that the number and area of ​​the bubbles exceeding the preset number and proportion decrease in sequence along the direction away from the central axis of the bending zone;

[0156] When the first bubble distribution characteristic record data satisfies that the number and area of ​​bubbles decrease in sequence along a direction away from the central axis of the bending zone, the first trend detection result of the first bubble distribution characteristic record data is marked as qualified;

[0157] Otherwise, the first trend detection result of the first bubble distribution characteristic record data is marked as unqualified.

[0158] Furthermore, the first decision module execution step further includes:

[0159] According to the bubble distribution characteristics, counting the number of first bubbles whose bubble areas are greater than or equal to a first bubble area threshold;

[0160] When the first number of bubbles is greater than or equal to a first bubble number threshold, or / and when the bubble distribution feature is different from a preset bubble distribution pattern, an unqualified mark for the folding screen film is generated.

[0161] Furthermore, the second decision module execution step further includes:

[0162] Performing bubble detection on the non-bending mark area to obtain bubble distribution characteristics of the non-bending mark area;

[0163] According to the bubble distribution characteristics of the non-bending mark area, counting the number of second bubbles whose bubble area is greater than or equal to a second bubble area threshold, wherein the second bubble area threshold is less than or equal to the first bubble area threshold;

[0164] When the second bubble quantity is greater than or equal to a second bubble quantity threshold, generating an unqualified mark for the folding screen film;

[0165] When the second bubble number is less than the second bubble number threshold, it is considered that the film pasting effect detection of the non-bending mark area is qualified.

[0166] The present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0167] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or multiple boxes.

[0168] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction system, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0169] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0170] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0171] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technology, the present invention is also intended to include these changes and variations.

Claims

1. A method for detecting the effect of film application on a folding screen, characterized in that: include: Perform bubble detection on the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics; When the bubble distribution feature is different from the preset bubble distribution law, an unqualified mark of the folding screen film is generated; When the bubble distribution feature is the same as the preset bubble distribution rule, and the film pasting effect detection of the non-bending mark area of ​​the folding screen film pasting surface image is qualified, a folding screen film pasting qualified mark is generated; Among them, the bubble detection is performed on the bending mark area of ​​the folding screen film surface image to obtain the bubble distribution characteristics, including: Dividing the bending mark area into a first bending mark area and a second bending mark area along the central axis of the bending area; Performing bubble detection on the first bending mark area to obtain a first bubble distribution feature; Performing bubble detection on the second bending mark area to obtain a second bubble distribution feature; adding the first bubble distribution feature and the second bubble distribution feature into the bubble distribution feature; Wherein, when the bubble distribution feature is different from the preset bubble distribution law, generating a folding screen film unqualified mark includes: Training a bubble trend detection model according to the preset bubble distribution law, wherein the bubble trend detection model includes a first bubble trend detection model and a second bubble trend detection model; The bubble distribution feature includes the first bubble distribution feature and the second bubble distribution feature; Inputting the first bubble distribution feature into the first bubble trend detection model to perform trend detection to obtain a first trend detection result; Inputting the second bubble distribution feature into the second bubble trend detection model to perform trend detection to obtain a second trend detection result; When any one of the first trend detection result and the second trend detection result is unqualified, generating an unqualified mark of the folding screen film; Wherein, training a bubble trend detection model according to the preset bubble distribution law includes: Collecting a first bubble distribution feature record data set in a first bend mark area; The first bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a first trend detection result mark data set; Using the first trend detection result identification data set as supervision and the first bubble distribution feature record data set as input, training the first bubble trend detection model; Collect a second bubble distribution feature record data set in the second bend mark area; The second bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a second trend detection result mark data set; Using the second trend detection result identification data set as supervision and the second bubble distribution feature record data set as input, training the second bubble trend detection model; The step of marking the first bubble distribution feature record data set according to the preset bubble distribution law to obtain a first trend detection result marking data set includes: Obtaining first bubble distribution feature record data of the first bubble distribution feature record data set; The preset distribution rule of the bubbles is that the number and area of ​​the bubbles exceeding the preset number and proportion decrease in sequence along the direction away from the central axis of the bending zone; When the first bubble distribution characteristic record data satisfies that the number and area of ​​bubbles decrease in sequence along a direction away from the central axis of the bending zone, the first trend detection result of the first bubble distribution characteristic record data is marked as qualified; Otherwise, the first trend detection result of the first bubble distribution characteristic record data is marked as unqualified.

2. The method according to claim 1, characterized in that The bubble detection is performed on the bending mark area of ​​the folding screen film surface image to obtain the bubble distribution characteristics, which also includes: Obtaining the folding screen film surface image; Performing bending zone detection on the folding screen film surface image to obtain the bending zone, wherein the bending zone has a bending zone width and a bending zone center axis; Based on the central axis of the bending zone, the width of the bending zone is enlarged to a preset multiple of the width of the bending zone to obtain the bending mark area; The area outside the bending mark area of ​​the folding screen film surface image is set as the non-bending mark area.

3. The method according to claim 2, characterized in that The bending zone width is enlarged to a preset multiple of the bending zone width based on the central axis of the bending zone to obtain the bending mark area, including: When the width of the bending zone is greater than or equal to 0 mm and less than 5 mm, the preset multiple is 1.4 to 1.6; When the width of the bending zone is greater than or equal to 5 mm and less than 10 mm, the preset multiple is 1.7 to 2.1; When the width of the bending zone is greater than or equal to 10 mm, the preset multiple is 2.2-2.

4.

4. The method according to claim 2, characterized in that Performing bending area detection on the folding screen film surface image to obtain the bending area includes: The bending area detection model is used to detect the folding screen film surface image to obtain the bending area; The step of constructing the bending zone detection model includes: Collecting a set of images of the folding screen film surface and a set of images of markings of the bending area of ​​the folding screen film surface; The bending area detection model is trained by taking the bending area identification image set of the folding screen film surface as supervision and the bending area identification image set of the folding screen film surface as input.

5. The method according to any one of claims 1 to 4, characterized in that: When the bubble distribution feature is different from the preset bubble distribution law, a folding screen film unqualified mark is generated, including: According to the bubble distribution characteristics, counting the number of first bubbles whose bubble areas are greater than or equal to a first bubble area threshold; When the first number of bubbles is greater than or equal to a first bubble number threshold, or / and when the bubble distribution feature is different from a preset bubble distribution pattern, an unqualified mark for the folding screen film is generated.

6. The method according to claim 5, characterized in that When the film effect detection of the non-bending mark area of ​​the folding screen film surface image is qualified, it includes: Performing bubble detection on the non-bending mark area to obtain bubble distribution characteristics of the non-bending mark area; According to the bubble distribution characteristics of the non-bending mark area, counting the number of second bubbles whose bubble area is greater than or equal to a second bubble area threshold, wherein the second bubble area threshold is less than or equal to the first bubble area threshold; When the second bubble quantity is greater than or equal to a second bubble quantity threshold, generating an unqualified mark for the folding screen film; When the second bubble number is less than the second bubble number threshold, it is considered that the film pasting effect detection of the non-bending mark area is qualified.

7. A folding screen film effect detection device, characterized in that: Used to implement a folding screen film effect detection method as described in any one of claims 1 to 6.

8. A folding screen film effect detection system, characterized in that: include: The bubble detection module is used to detect bubbles in the bending mark area of ​​the folding screen film surface image to obtain bubble distribution characteristics; The first decision module is used to generate a folding screen film unqualified mark when the bubble distribution feature is different from the preset bubble distribution law; The second decision module is used to generate a folding screen film-sticking qualified mark when the bubble distribution feature is the same as the preset bubble distribution law and the film-sticking effect detection of the non-bending mark area of ​​the folding screen film-sticking surface image is qualified; Among them, the bubble detection is performed on the bending mark area of ​​the folding screen film surface image to obtain the bubble distribution characteristics, including: Dividing the bending mark area into a first bending mark area and a second bending mark area along the central axis of the bending area; Performing bubble detection on the first bending mark area to obtain a first bubble distribution feature; Performing bubble detection on the second bending mark area to obtain a second bubble distribution feature; adding the first bubble distribution feature and the second bubble distribution feature into the bubble distribution feature; Wherein, when the bubble distribution feature is different from the preset bubble distribution law, generating a folding screen film unqualified mark includes: Training a bubble trend detection model according to the preset bubble distribution law, wherein the bubble trend detection model includes a first bubble trend detection model and a second bubble trend detection model; The bubble distribution feature includes the first bubble distribution feature and the second bubble distribution feature; Inputting the first bubble distribution feature into the first bubble trend detection model to perform trend detection to obtain a first trend detection result; Inputting the second bubble distribution feature into the second bubble trend detection model to perform trend detection to obtain a second trend detection result; When any one of the first trend detection result and the second trend detection result is unqualified, generating an unqualified mark of the folding screen film; Wherein, training a bubble trend detection model according to the preset bubble distribution law includes: Collecting a first bubble distribution feature record data set in a first bend mark area; The first bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a first trend detection result mark data set; Using the first trend detection result identification data set as supervision and the first bubble distribution feature record data set as input, training the first bubble trend detection model; Collect a second bubble distribution feature record data set in the second bend mark area; The second bubble distribution feature record data set is marked according to the preset bubble distribution rule to obtain a second trend detection result mark data set; Using the second trend detection result identification data set as supervision and the second bubble distribution feature record data set as input, training the second bubble trend detection model; The step of marking the first bubble distribution feature record data set according to the preset bubble distribution law to obtain a first trend detection result marking data set includes: Obtaining first bubble distribution feature record data of the first bubble distribution feature record data set; The preset distribution rule of the bubbles is that the number and area of ​​the bubbles exceeding the preset number and proportion decrease in sequence along the direction away from the central axis of the bending zone; When the first bubble distribution characteristic record data satisfies that the number and area of ​​bubbles decrease in sequence along a direction away from the central axis of the bending zone, the first trend detection result of the first bubble distribution characteristic record data is marked as qualified; Otherwise, the first trend detection result of the first bubble distribution characteristic record data is marked as unqualified.

Images