Vehicle-mounted backlight assembly method and device

Abstract

The invention relates to the technical field of vehicle-mounted assembly, and provides a vehicle-mounted backlight assembly method and equipment, aiming at assembly materials with normal functions and partial appearance flaws, a virtual tolerance zone is preset to limit a fluctuation range of a product contour, and when the assembly materials which do not conform to an incoming material management and control standard are detected, the virtual tolerance zone is preset to limit the fluctuation range of the product contour. Image optimization is carried out to reduce the defect detection sensitivity, the similarity and the actual tolerance zone are further calculated through pattern recognition to ensure that the optimized product image is closer to the theoretical center, finally, the optimized image is judged again, if yes, fitting assembly is carried out, and if not, image optimization is carried out again. And closed-loop detection is realized, so that normal fitting of assembly materials with normal functions and weak appearance defects is ensured, the yield is improved, and the cost advantage is expanded.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-mounted assembly, and in particular to a vehicle-mounted backlight assembly method and equipment. Background Art

[0002] As an important component of the car display screen, the vehicle backlight module is generally tested for product qualification by mechanical vision in actual manufacturing. Usually, data is collected through visual sensors, and the collection objects mainly include two types: 1. Products in the assembly process, that is, the bonding process of the screen, screen mount and bottom shell; 2. Incoming materials on the production line, that is, the screen mount and bottom shell before installation. That is, the existing visual inspection mainly removes unqualified parts in advance through the inspection of incoming materials; or screens out unqualified products through the inspection of the bonding process.

[0003] However, in the existing material assembly process, defective product detection is mainly carried out on the bonding process of the backlight module and the bonded finished products. The material control accuracy is high during the detection process. Therefore, a large number of defective products that do not affect actual use (minor material defects, but qualified functions) will be screened out, resulting in serious waste of resources and high product costs. Summary of the invention

[0004] The present invention provides a vehicle-mounted backlight assembly method and device, which solves the technical problem that the existing vehicle-mounted assembly scheme cannot assemble defective products that meet the use standards, resulting in large product production losses and high production costs.

[0005] In order to solve the above technical problems, the present invention provides a vehicle-mounted backlight assembly method, comprising:

[0006] Create an image coordinate system and calibrate it, and perform area planning to obtain a virtual tolerance zone;

[0007] Acquire and identify product images to determine whether the current assembly materials meet the incoming material control standards. If not, optimize the image to obtain an optimized image.

[0008] Performing contour recognition on the optimized image to obtain the graphic contour of the assembly material;

[0009] Comparing the virtual tolerance zone with the graphic contour, and calculating the similarity;

[0010] Taking the graphic outline as a reference and combining it with the virtual tolerance zone, an offset measurement is performed to calculate the actual tolerance zone of the assembly material;

[0011] The similarity and the actual tolerance band are obtained to determine whether the assembly material meets the incoming material control standard. If so, fitting assembly is performed; otherwise, image optimization is performed again.

[0012] This basic solution aims at assembly materials with normal functions but partial defects in appearance. It pre-sets virtual tolerance bands to limit the fluctuation range of product contours. When assembly materials that do not meet the incoming material control standards are detected, image optimization is performed to reduce the sensitivity of defect detection, and further similarity and actual tolerance bands are calculated through graphic recognition to ensure that the optimized product image is closer to the theoretical center. Finally, the optimized image is judged again. If yes, it is assembled. If not, the image is optimized again to achieve a closed detection loop, so as to ensure that assembly materials with normal functions but slight defects in appearance are assembled normally, so as to improve the yield rate and expand cost advantages.

[0013] In a further embodiment, the creating and calibrating the image coordinate system comprises:

[0014] Fixing at least one image acquisition module according to a preset photographing position of the assembly material;

[0015] Calibrate at least one of the image acquisition modules in a unified image coordinate system;

[0016] The image coordinate system is calibrated so that its coordinate origin coincides with the geometric center of the product image.

[0017] This solution fixes at least one image acquisition module according to a preset photographing position of the assembly material, realizes multi-angle image acquisition based on the assembly material, and ensures the reliability of detection through sufficient image acquisition.

[0018] In a further implementation scheme, the execution area planning to obtain the virtual tolerance zone is specifically:

[0019] An image of a standard material is acquired, and the image coordinates of the standard material are determined according to the image coordinate system; an error range of the assembled material is acquired, the image coordinates are divided into regions, and a corresponding error range region is acquired as a virtual tolerance zone.

[0020] This solution uses the error range of the assembly material as a benchmark and divides the contour range of the assembly material into regions, thereby digitally limiting the fluctuation range of the product contour, which is conducive to controlling the size and angle of the product.

[0021] In a further embodiment, the obtaining and identifying the product image, determining whether the current assembly material meets the incoming material control standard, and if not, optimizing the image to obtain the optimized image includes:

[0022] Capturing images of the current assembly material from multiple angles to obtain at least one product image;

[0023] Identify the product image and determine whether the assembly material meets the incoming material control standard. If yes, perform bonding assembly. If not, proceed to the next step to perform image optimization.

[0024] The product image is compared with a registered template, and pixel value optimization is performed on the product image to obtain an optimized image.

[0025] When this solution detects assembly materials and identifies that they do not meet the incoming material control standards, it optimizes the pixel assignment of the product image based on the registered template. By optimizing the product image, the sensitivity of incoming material detection is reduced, thereby assisting assembly materials with normal functions and slight appearance defects to fit normally, thereby ensuring product quality and avoiding material waste.

[0026] In a further embodiment, the step of performing contour recognition on the optimized image to obtain the graphic contour of the assembled material comprises:

[0027] Performing contour recognition on at least one of the optimized images to capture product contour features;

[0028] Performing image stitching according to the product contour features to construct a geometric figure corresponding to the assembly material;

[0029] After controlling the graphic center of the geometric figure to coincide with the geometric center of the theoretical center of the virtual tolerance zone, the edge coordinates of the geometric figure are obtained as the image contour.

[0030] This solution directly captures product contour features from the optimized image, and after image stitching and center overlap, provides digitized image contour data for subsequent processing and calculation to assist in calculation and detection.

[0031] In a further embodiment, comparing the virtual tolerance zone and the graphic profile and calculating the similarity comprises:

[0032] Obtaining the theoretical center of the virtual tolerance zone;

[0033] Determine the coordinates of all pixel points on any edge of the graphic outline, and make a difference with the corresponding theoretical center to obtain a difference set, and calculate the similarity based on the difference set;

[0034] Alternatively, the coordinates of all pixel points on any edge of the graphic outline are obtained, the offset angle of the image outline is calculated according to the theoretical center, and the similarity is calculated according to all the offset angles.

[0035] This solution starts from the image contour edge of the assembled material, sets up two different similarity calculation schemes: difference calculation scheme and offset angle calculation scheme, and calculates the similarity between the image contour in the optimized image and the theoretical value, thereby providing a basis for judging whether the optimized image is close to the theoretical center, and assisting in improving the process efficiency of image optimization to reduce detection sensitivity.

[0036] In a further embodiment, the actual tolerance zone of the assembly material is calculated by taking the graphic profile as a reference and combining the virtual tolerance zone to perform offset measurement, including:

[0037] Taking the graphic outline as a reference, the upper limit of the virtual tolerance zone is translated inwardly until the upper limit is tangent to the graphic outline, and the outer edge line of the upper limit at this time is obtained;

[0038] Taking the graphic contour as a reference, the lower limit of the virtual tolerance zone is translated outward until the lower limit is tangent to the graphic contour, and the inner edge line of the lower limit at this time is obtained;

[0039] The relative distance between each of the outer edge lines and the corresponding inner edge line is calculated as the actual tolerance zone of the assembly material.

[0040] This solution uses the image contour edge of the assembled material as a reference and shrinks the virtual tolerance zone to make it tangent to the virtual tolerance zone, thereby determining the actual error range of the assembled material (i.e., the actual tolerance zone), and then providing a secondary judgment basis for whether the optimized image is close to the theoretical center, thereby assisting in improving the process efficiency of image optimization to reduce detection sensitivity.

[0041] In a further embodiment, obtaining the similarity and the actual tolerance zone, determining whether the assembly material meets the incoming material control standard, and if so, performing fitting assembly, and if not, re-acquiring the optimized image includes:

[0042] Determine whether the similarity is greater than a first threshold, if so, proceed to the next step, if not, reacquire the optimized image;

[0043] Determine whether the actual tolerance zone is less than a second threshold, if so, proceed to the next step, if not, reacquire the optimized image;

[0044] The assembly material is photographed twice to obtain a verification image, and the verification image is identified to determine whether it meets the incoming material control standard. If so, assembly is performed, and if not, an optimized image is obtained again.

[0045] This solution sets the corresponding first and second thresholds for similarity and actual tolerance, and detects and judges the optimized image to determine whether the optimized image is closer to the theoretical center. If not, it is optimized again. If so, it is re-determined whether the incoming material control standards are met and the fitting test is performed. This helps to improve the inspection of incoming assembly materials and avoid serious defective materials being bonded, which may lead to defective whole machines and cause greater waste.

[0046] The present invention provides a vehicle-mounted backlight assembly device, comprising:

[0047] one or more processors;

[0048] A memory, used to store one or more programs, which, when executed by the one or more processors, enable the one or more processors to perform the method described above;

[0049] The robot is used to receive instructions sent by the processor, transport assembly materials to the photographing station and / or perform bonding assembly.

[0050] In a further embodiment, at least one image acquisition module is included, and at least one of the image acquisition modules is respectively installed and fixed at different preset photographing positions and surrounds the photographing station.

[0051] This solution sets up multiple image acquisition modules fixed at different preset photographing positions to obtain comprehensive images of assembly materials, thereby further optimizing assembly efficiency by improving the precision of collected data. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a workflow diagram of a vehicle-mounted backlight assembly method provided by an embodiment of the present invention;

[0053] Figure 2 is an operation flow chart of a vehicle backlight assembly method provided by an embodiment of the present invention;

[0054] Figure 3 is a schematic diagram of an image coordinate system provided by an embodiment of the present invention;

[0055] Figure 4 Schematic diagram of similarity difference symbols provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following specifically illustrates the implementation mode of the present invention in conjunction with the accompanying drawings. The embodiments are provided for illustrative purposes only and are not to be construed as limitations of the present invention. The accompanying drawings are provided for reference and illustration only and do not constitute limitations on the scope of patent protection of the present invention, because many changes may be made to the present invention without departing from the spirit and scope of the present invention.

[0057] Example 1

[0058] An embodiment of the present invention provides a vehicle-mounted backlight assembly method, such as Figure 1 to Figure 4 As shown, in this embodiment, it includes:

[0059] S0, create an image coordinate system and calibrate it, perform regional planning to obtain a virtual tolerance zone;

[0060] S01, fixing at least one image acquisition module according to a preset photographing position of the assembly material;

[0061] S02, calibrating at least one of the image acquisition modules in a unified image coordinate system;

[0062] S03. Calibrate the image coordinate system so that its coordinate origin coincides with the geometric center of the product image.

[0063] Taking a four-cornered rectangular assembly material as an example, an image acquisition module (such as a camera or a video camera) is installed at each of the four diagonal corners of the assembly material, and the four cameras are calibrated in a unified coordinate system (image coordinate system) using a calibration plate or a calibration block. For example, the origin of the image coordinate system forms the geometric center of the figure at the origin of the physical coordinate system of the four camera images.

[0064] In this embodiment, at least one image acquisition module is fixed according to a preset photographing position of the assembly material, so as to realize multi-angle image acquisition based on the assembly material, and the reliability of detection is ensured through sufficient image acquisition.

[0065] S04, acquiring an image of the standard material, and determining the image coordinates of the standard material according to the image coordinate system;

[0066] S05. Obtain the error range of the assembled material, divide the image coordinates into regions, and obtain corresponding error range regions as virtual tolerance zones.

[0067] like Figure 3 As shown; taking the origin of the image coordinate system as the center, in the image coordinate system, referring to the size range allowed by the product drawing, calculate the virtual distribution area of ​​the product size that meets the drawing requirements, and obtain the virtual tolerance zone.

[0068] This embodiment divides the contour range of the assembly material into regions based on the error range of the assembly material, thereby limiting the fluctuation range of the product contour in a digital manner, which is beneficial for controlling the size and angle of the product.

[0069] S1. Obtain and identify product images, determine whether the current assembly materials meet the incoming material control standards, and if not, optimize the image to obtain an optimized image, including:

[0070] S11, acquiring images of the current assembly material from multiple angles to obtain at least one product image;

[0071] S12, identifying the product image, determining whether the assembly material meets the incoming material control standard, if yes, performing bonding assembly, if no, proceeding to S13 to perform image optimization;

[0072] S13: Compare the product image with a registration template (generated by a registration picture), optimize pixel assignment of the product image, and obtain an optimized image.

[0073] Among them, the pixel point assignment optimization is used to repair / pass pixel data, which is a conventional technology in this field (such as Gaussian filtering and other technologies) and will not be repeated in this embodiment.

[0074] When the present embodiment detects assembly materials, if it is identified that they do not meet the incoming material control standards, the pixel assignment of the product image is optimized based on the registration template. By optimizing the product image, the sensitivity of incoming material detection is reduced by optimizing the image, thereby assisting the normal fitting of assembly materials with normal functions and slight poor appearance, thereby ensuring product quality and avoiding material waste.

[0075] S2, performing contour recognition on the optimized image to obtain the graphic contour of the assembly material, including:

[0076] Performing contour recognition on at least one of the optimized images to capture product contour features;

[0077] Performing image stitching according to the product contour features to construct a geometric figure corresponding to the assembly material;

[0078] See attached Figure 3 , 4 cameras take 4 images respectively, capture the product contour features of the four optimized images after optimization, get four intersection points, and use the four intersection points to construct geometric figures.

[0079] After controlling the graphic center of the geometric figure to coincide with the geometric center of the theoretical center of the virtual tolerance zone, the edge coordinates of the geometric figure are obtained as the image contour.

[0080] This embodiment directly captures the product contour features from the optimized image, and after image splicing and center overlap, provides digitized image contour data for subsequent processing and calculation, to assist in calculation and detection.

[0081] S3, comparing the virtual tolerance zone and the graphic contour to calculate similarity, including:

[0082] Obtaining the theoretical center of the virtual tolerance zone;

[0083] The similarity is used to measure the image coordinates of the optimized image and the theoretical center of the virtual tolerance zone (i.e., the edge contour of the standard material, such as Figure 3 The similarity between L0 and L1 can be expressed in many forms. The methods for calculating similarity include but are not limited to the following:

[0084] (1) determining the coordinates of all pixel points on any edge of the graphic outline, and subtracting them from the corresponding theoretical center to obtain a difference set, and calculating the similarity based on the difference set;

[0085] like Figure 4 As shown, the inner rectangle represents the graphic outline formed after the product image is optimized, and the dotted box represents the field of view of the four cameras; a single positive direction is fixed in each direction, and the horizontal axis is used for difference, and the vertical axis is used for difference, and the sign of the difference is taken.

[0086] I. Specifically, the product of the differences / the average of the absolute values ​​of the differences can be obtained, and the similarity level can be set according to the value range of the "product of the differences" / "the average of the absolute values ​​of the differences". After obtaining the difference set, the product / average value is calculated and compared with the similarity level to obtain the similarity.

[0087] II. Determine the positive and negative of the difference. When the difference signs in the same direction (the difference of pixels on the same edge) are the same or the product is greater than 0, the direction can be considered to have acceptable similarity; vice versa, when the difference signs in the same direction are different or the product is less than 0, the direction can be considered to have unacceptable similarity. For example, if the difference signs are all the same, the optimized image is included in the theoretical center range, or the theoretical center is included in the optimized image, and they are similar here, then the similarity is 1, otherwise the similarity is 0.

[0088] For images with high similarity, the corresponding products are of the same size as the theoretical values, which is conducive to fitting.

[0089] (2) Obtaining the coordinates of all pixel points on any edge of the graphic contour, calculating the offset angle of the image contour based on the theoretical center, and calculating the similarity based on all the offset angles.

[0090] Specifically, the angles (ie, offset angles) between each edge and the corresponding line segment in the theoretical center may be calculated, the average of the four angles may be obtained, the similarity levels may be set according to the angles, and the similarity may be obtained by comparison.

[0091] This embodiment starts from the image contour edge of the assembled material, sets two different similarity calculation schemes, namely the difference calculation scheme and the offset angle calculation scheme, and performs similarity calculation on the image contour in the optimized image and the theoretical value, thereby providing a basis for judging whether the optimized image is close to the theoretical center, and assisting in improving the process efficiency of image optimization to reduce detection sensitivity.

[0092] S4, taking the graphic outline as a reference and combining the virtual tolerance zone to perform offset measurement to calculate the actual tolerance zone of the assembly material, including:

[0093] Taking the graphic outline as a reference, the upper limit L1 of the virtual tolerance zone is translated inwardly until the upper limit is tangent to the graphic outline, and the outer edge line of the upper limit at this time is obtained;

[0094] Taking the graphic contour as a reference, the lower limit L2 of the virtual tolerance zone is translated outward until the lower limit is tangent to the graphic contour, and the inner edge line of the lower limit at this time is obtained;

[0095] The relative distance between each of the outer edge lines and the corresponding inner edge line is calculated as the actual tolerance zone of the assembly material.

[0096] This embodiment uses the image contour edge of the assembled material as a reference and shrinks the virtual tolerance zone to make it tangent to the virtual tolerance zone, thereby determining the actual error range of the assembled material (i.e., the actual tolerance zone), and then providing a secondary judgment basis for whether the optimized image is close to the theoretical center, thereby assisting in improving the process efficiency of image optimization to reduce detection sensitivity.

[0097] S5, obtaining the similarity and the actual tolerance band, determining whether the assembly material meets the incoming material control standard, and if so, performing fitting assembly, and if not, performing image optimization again, including:

[0098] S51, determining whether the similarity is greater than a first threshold A, if so, proceeding to S52, if not, reacquiring an optimized image;

[0099] S52, determining whether the actual tolerance band is less than a second threshold value B, if so, proceeding to S53, if not, reacquiring an optimized image;

[0100] S53, taking a second photo of the assembly material to obtain a verification image, identifying the verification image to determine whether it meets the incoming material control standard, if so, performing bonding assembly, if not, returning to step S13 to re-acquire the optimized image.

[0101] Among them, the common purpose of similarity and actual tolerance zone is to ensure that the optimized image is closer to the theoretical center. Therefore, both can be flexibly set according to product requirements and debugging requirements. The first threshold A and the second threshold B can also be set according to actual conditions.

[0102] In this embodiment, a first threshold value A and a second threshold value B are set corresponding to the similarity and the actual tolerance, and the optimized image is detected and judged to determine whether the optimized image is closer to the theoretical center. If not, the optimization is performed again. If so, it is re-judged whether the incoming material control standards are met and the fitting detection is performed. This helps to improve the inspection of incoming assembly materials and avoid serious defective materials being bonded, which may cause the whole machine to be defective and cause greater waste.

[0103] In the embodiment of the present invention, for assembly materials with normal functions but partial defects in appearance, a virtual tolerance band is pre-set to limit the fluctuation range of the product contour. When assembly materials that do not meet the incoming material control standards are detected, image optimization is performed to reduce the sensitivity of defect detection, and the similarity and actual tolerance band are further calculated through graphic recognition to ensure that the optimized product image is closer to the theoretical center. Finally, the optimized image is judged again after optimization. If yes, assembly is performed. If not, image optimization is performed again to realize a closed detection loop, so as to ensure that assembly materials with normal functions but slight defects in appearance are normally fitted, so as to improve the yield rate and expand the cost advantage.

[0104] Example 2

[0105] An embodiment of the present invention provides a vehicle-mounted backlight assembly device, comprising:

[0106] one or more processors;

[0107] A memory, used to store one or more programs, which, when executed by the one or more processors, causes the one or more processors to execute the method according to embodiment 1;

[0108] The robot is used to receive instructions sent by the processor, transport assembly materials to the photographing station and / or perform bonding assembly.

[0109] In this embodiment, at least one image acquisition module (such as Figure 3 CAM), at least one of the image acquisition modules is respectively installed and fixed at different preset photographing positions and surrounds the photographing station.

[0110] In this embodiment, a plurality of image acquisition modules are arranged and fixed at different preset photographing positions to acquire comprehensive images of the assembly materials, thereby further optimizing the assembly efficiency by improving the precision of the collected data.

[0111] The present invention forms a closed loop for detecting defective materials (normal functions but slight appearance defects) to achieve normal lamination of defective materials, and the beneficial effects achieved are as follows:

[0112] (1) For material manufacturers who manufacture laminated products, this can be equivalent to lowering the appearance standards, which is beneficial to improving product yield and further enhancing cost advantages.

[0113] (2) For manufacturers of bonding products, it helps to improve the efficiency of incoming material inspection and avoid serious defective materials being bonded, which may lead to machine failure and further waste.

[0114] (3) In terms of assembly application, there is no need to make major changes to the existing hardware architecture. It is only necessary to introduce the assembly method in Example 1 based on the existing bonding principle, which has strong applicability.

[0115] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A vehicle-mounted backlight assembly method, characterized in that: include: Create an image coordinate system and calibrate it, and perform area planning to obtain a virtual tolerance zone; Acquire and identify product images to determine whether the current assembly materials meet the incoming material control standards. If not, optimize the image to obtain an optimized image. Performing contour recognition on the optimized image to obtain the graphic contour of the assembly material; Comparing the virtual tolerance zone with the graphic contour, and calculating the similarity; Taking the graphic outline as a reference and combining it with the virtual tolerance zone, an offset measurement is performed to calculate the actual tolerance zone of the assembly material; The similarity and the actual tolerance band are obtained to determine whether the assembly material meets the incoming material control standard. If so, assembly is performed. If not, image optimization is performed again based on the similarity.

2. A vehicle-mounted backlight assembly method as claimed in claim 1, characterized in that: The creating and calibrating of the image coordinate system comprises: Fixing at least one image acquisition module according to a preset photographing position of the assembly material; Calibrate at least one of the image acquisition modules in a unified image coordinate system; The image coordinate system is calibrated so that its coordinate origin coincides with the geometric center of the product image.

3. A vehicle-mounted backlight assembly method as claimed in claim 2, characterized in that: The virtual tolerance zone obtained by executing regional planning is specifically: An image of a standard material is acquired, and the image coordinates of the standard material are determined according to the image coordinate system; an error range of the assembled material is acquired, the image coordinates are divided into regions, and a corresponding error range region is acquired as a virtual tolerance zone.

4. A vehicle-mounted backlight assembly method as claimed in claim 2, characterized in that: The obtaining and identifying of the product image, determining whether the current assembly material meets the incoming material control standard, and if not, optimizing the image to obtain the optimized image includes: Capturing images of the current assembly material from multiple angles to obtain at least one product image; Identify the product image and determine whether the assembly material meets the incoming material control standard. If yes, perform bonding assembly. If not, proceed to the next step to perform image optimization. The product image is compared with a registered template, and pixel value optimization is performed on the product image to obtain an optimized image.

5. The vehicle-mounted backlight assembly method according to claim 1, characterized in that: The step of performing contour recognition on the optimized image to obtain the graphic contour of the assembly material includes: Performing contour recognition on at least one of the optimized images to capture product contour features; Performing image stitching according to the product contour features to construct a geometric figure corresponding to the assembly material; After controlling the graphic center of the geometric figure to coincide with the geometric center of the theoretical center of the virtual tolerance zone, the edge coordinates of the geometric figure are obtained as the image contour.

6. The vehicle-mounted backlight assembly method according to claim 1, characterized in that: The comparing the virtual tolerance zone and the graphic contour and calculating the similarity comprises: Obtaining the theoretical center of the virtual tolerance zone; Determine the coordinates of all pixel points on any edge of the graphic outline, and make a difference with the corresponding theoretical center to obtain a difference set, and calculate the similarity based on the difference set; Alternatively, the coordinates of all pixel points on any edge of the graphic outline are obtained, the offset angle of the image outline is calculated according to the theoretical center, and the similarity is calculated according to all the offset angles.

7. The vehicle-mounted backlight assembly method according to claim 1, characterized in that: Taking the graphic outline as a reference and combining the virtual tolerance zone to perform offset measurement, the actual tolerance zone of the assembly material is calculated, including: Taking the graphic outline as a reference, the upper limit of the virtual tolerance zone is translated inwardly until the upper limit is tangent to the graphic outline, and the outer edge line of the upper limit at this time is obtained; Taking the graphic contour as a reference, the lower limit of the virtual tolerance zone is translated outward until the lower limit is tangent to the graphic contour, and the inner edge line of the lower limit at this time is obtained; The relative distance between each of the outer edge lines and the corresponding inner edge line is calculated as the actual tolerance zone of the assembly material.

8. The vehicle-mounted backlight assembly method according to claim 1, characterized in that: The obtaining of the similarity and the actual tolerance band, determining whether the assembly material meets the incoming material control standard, and if so, performing fitting assembly, and if not, re-obtaining the optimized image comprises: Determine whether the similarity is greater than a first threshold, if so, proceed to the next step, if not, reacquire the optimized image; Determine whether the actual tolerance zone is less than a second threshold, if so, proceed to the next step, if not, reacquire the optimized image; The assembly material is photographed twice to obtain a verification image, and the verification image is identified to determine whether it meets the incoming material control standard. If so, assembly is performed, and if not, an optimized image is obtained again.

9. A vehicle-mounted backlight assembly device, characterized in that: include: one or more processors; A memory, for storing one or more programs, which, when executed by the one or more processors, causes the one or more processors to execute the method according to any one of claims 1 to 8; The robot is used to receive instructions sent by the processor, transport assembly materials to the photographing station and / or perform bonding assembly.

10. The vehicle-mounted backlight assembly device according to claim 9, characterized in that: It comprises at least one image acquisition module, and at least one of the image acquisition modules is respectively installed and fixed on different preset photographing positions and surrounds the photographing station.

Images