Automobile armrest quality rapid evaluation method and device

Abstract

The application relates to the technical field of material detection, in particular to a method and device for rapidly evaluating the quality of an automobile handrail, which comprises the following steps: setting a camera to face the surface of the handrail to be detected, determining the position of the component in the image and the effective illumination angle; the light source is inclined at 45 degrees and rotates around the center of the handrail, and the camera takes an image every fixed step length delta theta; the image is subjected to gray uniformization processing; the shadow change area in the front and back two images is calculated, and the shadow change area is corresponded to each component; all the corresponding shadow change areas of each component within the effective illumination angle are screened out, then the shadow areas are combined to form an equivalent centroid, the perpendicularity and distance of the connecting line between the center of each component and the equivalent centroid are calculated, and the final evaluation score is calculated; if the evaluation score is greater than a set threshold value, the automobile handrail quality is determined to be unqualified. The application can effectively solve the problems of low detection efficiency and poor detection accuracy in the prior art.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of material detection, and particularly relates to a method and device for rapidly evaluating the quality of an automobile handrail. BACKGROUND

[0002] With the rapid development of automobiles, the functions of automobiles are gradually increasing, and various operation buttons, operation interfaces or additional components such as decorations are increasingly arranged in the vehicle, especially in places such as automobile handrails that can be quickly contacted by personnel. With the increase in components, the probability of missing or misplacing components on the production line or loosening or falling components during the carrying process will increase, so it is necessary to finally detect the quality of the automobile handrail to ensure accurate installation of all components. The existing quality detection method usually adopts a personnel observation method, which is not only low in efficiency, but also prone to missing or misidentifying some non-obvious quality problems, and has the problems of low detection efficiency and poor detection accuracy. SUMMARY

[0003] The present application provides a method and device for rapidly evaluating the quality of an automobile handrail, which can effectively solve the problems of low detection efficiency and poor detection accuracy in the background art.

[0004] The present application provides a method for rapidly evaluating the quality of an automobile handrail, which comprises the following steps:

[0005] A camera is arranged to face the surface of the handrail to be detected, the positions of the components on the surface of the handrail in the images captured by the camera are determined, and the effective illumination angles of the components are determined;

[0006] The light source is arranged to illuminate the surface of the handrail from top to bottom at an inclination angle of 45°, and the light source rotates uniformly around the center of the handrail in the horizontal direction, and the camera captures an image every fixed step length △θ;

[0007] The images are subjected to gray scale unification processing to make the gray scales of the non-shaded areas in the images consistent;

[0008] The shaded areas in the front and rear images are calculated, and the shaded areas are matched with the components;

[0009] All corresponding shaded areas within the effective illumination angle of each component are screened out through the effective illumination angle, and then the shaded areas are combined to form a shaded area, and the equivalent centroid of the shaded area is calculated;

[0010] The perpendicularity and distance of the connecting line between the center of each component and the equivalent centroid are calculated, and the final evaluation score is calculated according to the perpendicularity and distance;

[0011] The evaluation scores of all components are compared with the set threshold value, and if there is an evaluation score greater than the set threshold value, it is determined that the quality of the automobile handrail is unqualified.

[0012] Further, the specific calculation process of △θ is as follows:

[0013] The direction of the front of the handrail is taken as 0°, and the handrail circumferential direction is numbered from 0° to 360°;

[0014] The effective illumination angles of each component are determined to be [a1min, a1max], [a2min, a2max]…[anmin, anmax] respectively;

[0015] The middle angles of each component are calculated as aimid=(aimin+aimax) / 2, i taking 1, 2…n;

[0016] The set S={a1min, a1mid, a1max, a2min, a2mid, a2max, …, anmin, anmid, anmax} is constructed;

[0017] The repeated angles in the set S are removed, and then the angles in the set S are sorted in ascending order s1, s2, …, sm;

[0018] The adjacent angle difference values are calculated, and the greatest common divisor of the difference values is △θ.

[0019] Further, the specific gray scale unification processing of the image is as follows:

[0020] A reference region is set in the image shot by the camera, so that the reference region is always located on the handrail; a protection region is further set in the reference region, and each protection region center corresponds to a component and its shadow in the protection region;

[0021] The gray scale average value μref of all images in the reference region except the protection region part is calculated, and then the gray scale average value μq of each image in the reference region except the protection region part is calculated, q being the image number;

[0022] The gray scale of the pixel point Gq(x, y) in the qth image is recorded, and the gray scale of the pixel point Gqnew(x, y) in the new qth image is calculated as Gqnew(x, y)=Gq(x, y)·μref / μq, to obtain a gray scale consistent image.

[0023] Further, before obtaining the gray scale consistent image, the standard deviation of the gray scale average value μq of all images is calculated, and if the standard deviation value is greater than a set threshold, the image corresponding to the gray scale average value μq with the largest outlier value is determined, and if the image does not belong to the image within the effective illumination angle, the image is removed and μref and μq are updated;

[0024] If the image is within the effective illumination angle, then the contrast of the image is adjusted, and μref and μq are updated after the adjustment.

[0025] Furthermore, the specific calculation of the shadow change area in the two images is as follows:

[0026] For the q-th image, calculate the difference in gray values ​​of the pixels in the x-th row and y-th column between the two images, ΔGq(x,y)=Gqnew(x,y)-G(q-1)new(x,y);

[0027] If △Gq(x, y) is greater than the set threshold, then the corresponding pixel in the x-th row and y-th column is recorded as the changed pixel.

[0028] By combining all the changing pixels, the area of ​​shadow change can be obtained.

[0029] Furthermore, after calculating all the changed pixels, these pixels are divided into regions, so that each changed pixel in different regions must be adjacent to at least one other changed pixel.

[0030] After the regions are divided, the number of changing pixels in each region is counted, and regions with a number less than a set threshold are removed. The remaining regions are the shadow change regions.

[0031] Furthermore, the correspondence between the shadow variation areas and each component is as follows:

[0032] Calculate the minimum bounding box of each shadow change region, determine the center point of the component that falls within the minimum bounding box, and then the component is the component corresponding to the shadow change region.

[0033] Furthermore, the equivalent centroid of the shadowed region of each component is calculated as follows:

[0034] Calculate the minimum bounding box for each shaded region; the center of the minimum bounding box is the equivalent centroid.

[0035] Furthermore, the evaluation score is calculated as follows:

[0036] Let kj be the perpendicularity of the line connecting the center of the j-th component to its equivalent centroid, and dj be the distance between them. Then, the evaluation score sj of the j-th component is calculated as follows:

[0037] sj=1 / (1+a1|kj|+a2|dj-[dj]|);

[0038] Where a1 and a2 are both weighting coefficients, and [dj] is the standard distance between the center of the j-th component and the equivalent centroid.

[0039] This invention also provides a rapid quality assessment device for car handrails, comprising:

[0040] The testing platform is equipped with a circular track;

[0041] The camera is fixedly mounted on the testing platform;

[0042] The light source slides on a circular track;

[0043] A limiting platform, set on the detection platform, is used to limit the position of the handrail;

[0044] The processor is used to implement the aforementioned method for rapid quality assessment of automotive handrails.

[0045] The technical solution of this invention can achieve the following technical effects:

[0046] This method combines a special light source movement method with a camera shooting method, and designs a special algorithm adapted to the aforementioned hardware settings. It can quickly and accurately identify the shadow areas of each component, and use the shadow areas to calculate the installation qualification of each component, and quantify the installation qualification into an evaluation score. This enables quantitative evaluation of the handrail. The evaluation process can be automated and processed quickly, effectively reducing the influence of personnel, thereby improving the efficiency and accuracy of handrail quality inspection. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart illustrating the rapid quality assessment method for car handrails in this invention.

[0049] Figure 2 This is a schematic diagram showing the shadow changes of the protruding component in this invention;

[0050] Figure 3 This is a schematic diagram of the shaded area of ​​the protruding component in this invention;

[0051] Figure 4 This is a schematic diagram of the shadow changes of the recessed component in this invention;

[0052] Figure 5 This is a schematic diagram of the shaded area of ​​the recessed component in this invention. Detailed Implementation

[0053] The basic principles and main features of the technical solution of the present invention will be described below with reference to the accompanying drawings of the embodiments of the present invention. The following description will use one or more embodiments for a more intuitive understanding. These embodiments are merely some, not all, of the embodiments of the present invention.

[0054] In the description of this invention, the terms indicating orientation or positional relationship (such as up, down, left, right, etc.) are based on the orientation shown in the drawings or some conventional positional relationships, and are only for the convenience of describing this invention and simplifying the description, and are not intended to indicate or imply that the features referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting this invention.

[0055] A rapid method for assessing the quality of car handrails, such as Figure 1 As shown, the main steps include image acquisition, image unification, shadow region calculation, and score calculation. The specific process of each step is as follows:

[0056] Image acquisition: Set the camera to face the handrail to take grayscale images. The main surfaces to be measured are the front, left and right, back and top surfaces of the handrail. The components to be measured will be on the handrail's surface to be measured. The number of cameras corresponds to the number of surfaces to be measured. Each camera takes a picture of one surface to be measured. In subsequent calculations, each camera is calculated separately.

[0057] Each camera captures an image with a corresponding coordinate system, reflecting the number of rows and columns of each pixel in the image. Before obtaining the final inspection image, it's necessary to first capture an image with the camera, or determine the position of each component on the handrail surface within the camera's captured image based on the dimensions of each component in the design. This is the coordinate region of each component within the camera's captured image. After determining the position of each component, the effective illumination angle for each component must also be determined. The effective illumination angle is an angle range within which the component casts a shadow that is easily observable when the light source shines on the handrail. This effective illumination angle range is symmetrical about the angle of the component on the handrail. For example, if the front of the handrail is designated as 0°, and the angles are numbered clockwise from 0° to 360°, and the angle of a button component is located at 30°, then the effective illumination angle range needs to be symmetrical about 30°, such as 0° to 60°.

[0058] The light source illuminates the handrail surface from above at a 45° angle, ensuring that the light source casts shadows on the components regardless of its position. After the handrail is positioned, the light source is driven to rotate horizontally at a uniform speed around the center of the handrail. An image is captured by the camera every fixed step length △θ. The setting of △θ is primarily for controlling the camera's shooting interval. For example, if the angular velocity of the light source is ω = 360° / min, and we want to capture an image every 10°, then △θ = 10°. This allows us to obtain the camera's shooting speed vph = ω / △θ = 36 frames / min, thus facilitating the control and adjustment of the camera's shooting speed.

[0059] Image unification: After a camera captures images of a light source moving to various angles, the movement of the light source in the images may cause differences in the grayscale of the handrail in each image, which is not conducive to the comparison between subsequent images. Therefore, it is necessary to perform grayscale unification processing on the images to make the grayscale of the non-shaded areas in each image consistent.

[0060] Shadow area calculation: After unification, the two images obtained before and after can be compared pairwise. The shadow will move and change with the movement of the light source. Therefore, the parts with different gray levels in the two images are identified. These different parts are the non-overlapping parts of the shadow areas of the two images, which are the shadow change areas in the two images. A camera may capture images of multiple parts at the same time, so there will be multiple shadow change areas. It is necessary to match the shadow change areas with each part.

[0061] Each component uses its effective illumination angle to filter out all corresponding shadow change areas within that angle, and then combines them to form a shadow area. For example, in the button component mentioned above, the effective illumination angle is 0°~60°. The camera takes an image every time the light source passes through Δθ=10°. Therefore, for this button component, the shadow area is formed by combining the six shadow change areas obtained from the before-and-after comparison calculations of the images when the light source is at the positions of 0°, 10°, 20°, 30°, 40°, 50°, and 60°.

[0062] After the shadow area is formed, the equivalent centroid of the shadow area is calculated. The equivalent centroid can be regarded as the center of the shadow formed when the light source is located at the corresponding angular position of the component. The shadow formed by the light source at the corresponding angular position of the component best reflects the installation status of the component. However, if only one image is captured at this time, the shadow part needs to be accurately distinguished in a single image. This not only requires high accuracy of the discrimination algorithm, but also requires frequent adjustment of the algorithm parameters due to background light and other factors. To avoid this problem, this method uses a before-and-after image comparison to obtain the shadow area simply and accurately. Then, the calculated equivalent centroid, based on its position and distance relationship with the center of the component, is used to replace the requirement of finding the center of the shadow formed when the light source is located at the corresponding angular position of the component.

[0063] For example: Figure 2 and Figure 4 As shown, Figure 2 For protruding parts such as buttons that protrude from the surface to be measured, Figure 4 For recessed components such as cup grooves that are recessed into the surface to be measured, after the light source moves once, the shadow area changes from ①+② to ②+③. Therefore, the first calculated shadow change area is ①+③. After the light source moves again, the shadow area changes from ④+⑤ to ⑤+⑥. Therefore, the second calculated shadow change area is ④+⑥. Adding these multiple shadow change areas together yields the result as shown below. Figure 3 and Figure 5 The shaded area shown.

[0064] Score Calculation: After obtaining the equivalent centroid of each component, if the component is installed accurately, the equivalent centroid should be aligned with the component center in the height direction (when the inspection surface is a vertical surface) or in the frontal direction (when the inspection surface is a top surface). That is, the line connecting the equivalent centroid and the component center is perpendicular to the image. Since the tilt angle of the light source is 45°, the distance between the equivalent centroid and the component center is equal to the height of the component's protrusion or depression. Therefore, it is only necessary to calculate the perpendicularity and distance between the center of each component and the equivalent centroid. The final evaluation score of the corresponding component can be calculated based on the perpendicularity and distance. The smaller the evaluation score, the closer the perpendicularity and distance are to the aforementioned standard values, and the higher the qualification of the component installation.

[0065] The evaluation scores of all components are compared with the set thresholds. If any evaluation score is greater than the set threshold, the quality of the car armrest is deemed unqualified.

[0066] As can be seen, this method can quickly and accurately identify the shadow areas of each component, calculate the installation qualification of each component using the shadow areas, and quantify the installation qualification into an evaluation score, thereby enabling quantitative evaluation of the handrail. The evaluation process can be automated and processed quickly, effectively reducing the influence of personnel, thus improving the efficiency and accuracy of handrail quality inspection.

[0067] △θ can be set manually or calculated automatically using an algorithm. The value of △θ should be maximized as much as possible while covering all required angles. This minimizes the camera's shooting speed (the number of shots taken per unit time). A lower shooting speed reduces the burden and demands on the camera, ensuring the continuous operation of this method. This method provides a specific calculation method for △θ, the process of which is as follows:

[0068] Take the front of the handrail as 0°, and number the horizontal circumference of the handrail from 0° to 360° in a clockwise direction (counterclockwise is also possible, but the parameters of the corresponding parts need to be adjusted accordingly in the calculation).

[0069] Determine the effective illumination angles of each component as [a1min, a1max], [a2min, a2max], ..., [anmin, anmax]. For example, there are three components: [0°, 60°], [90°, 150°], and [210°, 270°].

[0070] Calculate the intermediate angle aimid = (aimin + aimmax) / 2 for each component, where i takes the values ​​1, 2, ..., n. This intermediate angle is also the angle corresponding to the angular position of the component, such as 30°, 120°, and 240° in the example above.

[0071] Construct a set S = {a1min, a1mid, a1max, a2min, a2mid, a2max, ..., anmin, anmid, anmax}, for example: S = {0°, 30°, 60°, 90°, 120°, 150°, 210°, 240°, 270°};

[0072] Remove duplicate angles from set S, and then sort the angles in set S in ascending order as s1, s2, ..., sm;

[0073] Calculate the difference between adjacent angles, i.e., calculate △s2=s2-s1, △s3=s3-s2……△sm=sm-s(m-1), and calculate the greatest common divisor of all differences △s2~△sm, which is △θ. For example, in the above example, the differences between adjacent angles are: 30°-0°=30°, 60°-30°=30°, 90°-60°=30°, 120°-90°=30°, 150°-120°=30°, 210°-150°=60°, 240°-210°=30°, 270°-240°=30°, and the greatest common divisor is 30°, i.e., △θ=30°. Normally, the angular position of the component and the effective illumination angle will be relatively regular integers, such as the values ​​in the example above. Therefore, the calculated Δθ will be more suitable. If Δθ is not suitable (such as too many decimals, or cannot be calculated), an alarm can be issued and a value can be given manually.

[0074] Preferably, the grayscale unification process for the image is specifically as follows:

[0075] In the image captured by the camera, a reference area is set so that the reference area is always located on the armrest; a protection area is then set within the reference area. The protection area is a certain pixel area that is extended outward from the component area, so that the component and its shadow corresponding to the center of each protection area are within the protection area.

[0076] Calculate the average gray value μref of the reference region excluding the protected region in all images, and then calculate the average gray value μq of the reference region excluding the protected region in each image, where q is the image number, usually numbered 1, 2, 3... according to the image generation time.

[0077] Let Gq(x, y) be the gray level of the pixel in the x-th row and y-th column of the q-th image. Calculate the gray level of the pixel in the x-th row and y-th column of the new q-th image as Gqnew(x, y) = Gq(x, y)·μref / μq to obtain an image with consistent gray levels.

[0078] It is important to note that when calculating μref and μq, pixels within the protected region of each image are not included in the calculation; however, when calculating Gqnew(x, y), the calculation is performed on all pixels in each image (including the protected region). Pixels outside the reference region can be retained or removed as needed.

[0079] Preferably, before obtaining a uniform grayscale image, the standard deviation of the grayscale average value μq of all images is calculated. If the standard deviation value is greater than a set threshold, the image corresponding to the grayscale average value μq with the largest outlier is judged. If the image does not belong to the effective illumination angle, the image will not affect the calculation of the shadow area, and the image can be directly removed and μref and μq updated.

[0080] If the image is within the effective illumination angle, it must be retained to avoid affecting the calculation of the shadow area. Therefore, only the contrast of the image can be adjusted. After adjustment, μref and μq are updated. If the standard deviation value is still greater than the set threshold after adjustment, the image is adjusted again until the standard deviation value enters the threshold.

[0081] Since the environment may be affected by sudden changes in external light sources, causing the overall grayscale of a certain image to be too bright or too dark compared to other images, if the above calculation is performed directly, it will cause a large difference between μref and all images, resulting in significant changes to all images and potentially causing all details in all images to disappear. Therefore, this method first identifies the overly bright or dark image and processes it separately to minimize damage to other images.

[0082] Preferably, the calculation of the shadow change area in the two images is specifically as follows:

[0083] Since the images were taken by the same camera, each pixel in the image corresponds to the other. Therefore, for the q-th image, calculate the difference in gray values ​​of the pixels in the x-th row and y-th column between the two images, ΔGq(x,y)=Gqnew(x,y)-G(q-1)new(x,y).

[0084] If △Gq(x,y) is greater than the set threshold, it means that the gray level at the pixel (x,y) in the two images has changed significantly. It is very likely that there is a shadow in the previous image but the shadow moves away in the next image, or there is no shadow in the previous image but the shadow moves over in the next image. The corresponding pixel in the xth row and yth column is recorded as the changed pixel.

[0085] After combining all the changing pixels, they will form a large area, which is the shadow variation area.

[0086] It should be noted that some impurities, light spots, etc., may cause the calculation of some pixel changes that are not in the shadow area. Therefore, these noise points need to be removed. Specifically:

[0087] After calculating all the changed pixels, these pixels are divided into regions, so that each changed pixel in different regions must be adjacent to at least one other changed pixel.

[0088] After the region is divided, the number of changing pixels in each region is counted. If the number is too small, it means that the area of ​​the region is too small, which is noise. Regions with a number less than the set threshold are removed, and the remaining regions are the shadow change regions.

[0089] Preferably, the correspondence between the shadow variation area and each component is as follows:

[0090] Calculate the minimum bounding box for each shadow change region. The length of the minimum bounding box is parallel to the x-axis in the camera, and the width is parallel to the y-axis in the camera. Determine the center point of the part that falls within the minimum bounding box. Then, the part is the part corresponding to the shadow change region.

[0091] Preferably, the calculation of the equivalent centroid of the shadow region of each component is specifically as follows:

[0092] Calculate the minimum bounding box for each shadow variation region. The length of the minimum bounding box is parallel to the x-axis in the camera, and the width is parallel to the y-axis in the camera. The center of the minimum bounding box is the equivalent centroid.

[0093] Preferably, the evaluation score is calculated as follows:

[0094] Let the center of the j-th component be P(xp, yp), and its equivalent centroid be Q(xq, yq). Then the perpendicularity of the connecting line is kj = (yq - yp) / (xq - xp), and the distance is dj = ;

[0095] The evaluation score sj for the j-th component is calculated as follows:

[0096] sj=1 / (1+a1|kj|+a2|dj-[dj]|);

[0097] Where a1 and a2 are both weighting coefficients, a1 is 0.05 and a2 is 0.1. The values ​​of a1 and a2 can be adjusted according to actual needs; [dj] is the standard distance between the center of the j-th component and the equivalent centroid, which can be obtained by drawing in CAD.

[0098] A rapid quality assessment device for car handrails, comprising:

[0099] The testing platform is equipped with a circular track, and the handrail is located below the circular track during testing.

[0100] The camera is fixedly mounted on the testing platform; the number of cameras is the same as the number of surfaces to be tested on the handrail.

[0101] The light source slides on a circular track. Parallel light sources are preferred to minimize shadow distortion.

[0102] A limiting stage is set on the detection stage. The limiting stage is equipped with clamps to limit the position of the handrails and ensure that the positions of each handrail to be tested are consistent in the camera.

[0103] The processor, which can be a computer, PLC, or other computing device, is used to control the movement of the aforementioned components and to acquire data from them for calculation, thereby realizing the aforementioned method for rapid quality assessment of car handrails.

[0104] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for rapid quality assessment of automotive handrails, characterized in that, include: Set the camera to face the handrail surface to be measured, determine the position of each component on the handrail surface in the image captured by the camera, and the effective illumination angle of each component; The light source illuminates the handrail surface from above at a 45° tilt angle, and rotates at a constant speed in the horizontal direction around the center of the handrail. The camera takes an image once every time it passes a fixed step length △θ. Perform grayscale unification processing on the images to ensure that the grayscale of non-shaded areas in each image is consistent. Calculate the shadow change areas in the two images and associate the shadow change areas with each component; Each component filters out all corresponding shadow change areas within the effective illumination angle through the effective illumination angle, then combines them to form a shadow area, and calculates the equivalent centroid of the shadow area; Calculate the perpendicularity and distance between the center of each component and the line connecting the center of the component and the equivalent centroid, and calculate the final evaluation score based on the perpendicularity and distance. The evaluation scores of all components are compared with the set thresholds. If any evaluation score is greater than the set threshold, the quality of the car armrest is deemed unqualified. Specifically, the grayscale unification process for the image involves: Set a reference area in the image captured by the camera, ensuring that the reference area is always located on the armrest; then set a protection area within the reference area, with each component and its shadow corresponding to the center of the protection area being within the protection area; Calculate the average gray value μref of the reference region excluding the protected region in all images, and then calculate the average gray value μq of the reference region excluding the protected region in each image, where q is the image number; Let Gq(x, y) be the gray level of the pixel in the x-th row and y-th column of the q-th image. Calculate the gray level of the pixel in the x-th row and y-th column of the new q-th image as Gqnew(x, y) = Gq(x, y)·μref / μq to obtain an image with consistent gray levels. The specific steps for calculating the change in shadow area between the two images are as follows: For the q-th image, calculate the difference in gray values ​​of the pixels in the x-th row and y-th column between the two images, ΔGq(x,y)=Gqnew(x,y)-G(q-1)new(x,y); If △Gq(x, y) is greater than the set threshold, then the corresponding pixel in the x-th row and y-th column is recorded as the changed pixel. By combining all the changing pixels, the area of ​​shadow change can be obtained; The specific mapping of shadow variation areas to individual components is as follows: Calculate the minimum bounding box of each shadow change region, determine the center point of the component that falls within the minimum bounding box, and then the component is the component corresponding to the shadow change region; The equivalent centroid of the shadowed region of each component is calculated as follows: Calculate the minimum bounding box for each shaded region; the center of the minimum bounding box is the equivalent centroid.

2. The method for rapid quality assessment of automotive handrails according to claim 1, characterized in that, The specific calculation process for △θ is as follows: The direction of the front of the handrail is taken as 0°, and the circumferential direction of the handrail is numbered from 0° to 360°. The effective illumination angles of each component are determined to be [a1min, a1max], [a2min, a2max], ..., [anmin, anmax]; Calculate the midpoint angle of each component: aimid = (aimin + aimmax) / 2, where i takes the values ​​1, 2, ..., n; Construct a set S = {a1min, a1mid, a1max, a2min, a2mid, a2max, ..., anmin, anmid, anmax}; Remove duplicate angles from set S, and then sort the angles in set S in ascending order as s1, s2, ..., sm; The difference between adjacent angles is calculated, and the greatest common divisor of the difference is Δθ.

3. The method for rapid quality assessment of automotive handrails according to claim 1, characterized in that, Before obtaining a uniform grayscale image, the standard deviation of the average grayscale value μq of all images is calculated. If the standard deviation value is greater than the set threshold, the image corresponding to the average grayscale value μq with the largest outlier is judged. If the image does not belong to the effective illumination angle, the image is removed and μref and μq are updated. If the image is within the effective illumination angle, then the contrast of the image is adjusted, and μref and μq are updated after the adjustment.

4. The method for rapid quality assessment of automotive handrails according to claim 1, characterized in that, After calculating all the changed pixels, these pixels are divided into regions, so that each changed pixel in different regions must be adjacent to at least one other changed pixel. After the regions are divided, the number of changing pixels in each region is counted, and regions with a number less than a set threshold are removed. The remaining regions are the shadow change regions.

5. The method for rapid quality assessment of automotive handrails according to claim 1, characterized in that, The specific calculation of the evaluation score is as follows: Let kj be the perpendicularity of the line connecting the center of the j-th component to its equivalent centroid, and dj be the distance between them. Then, the evaluation score sj of the j-th component is calculated as follows: sj=1 / (1+a1|kj|+a2|dj-[dj]|); Where a1 and a2 are both weighting coefficients, and [dj] is the standard distance between the center of the j-th component and the equivalent centroid.

6. A rapid quality assessment device for automobile handrails, characterized in that, include: The testing platform is equipped with a circular track; The camera is fixedly mounted on the detection platform; The light source slides on the circular track; A limiting platform, installed on the detection platform, is used to limit the position of the handrail; A processor for implementing the rapid quality assessment method for automotive handrails as described in any one of claims 1 to 5.

Images