Automobile interior parts defect detection method and system

Abstract

The present invention belongs to the field of image processing technology, and specifically relates to a method and system for detecting defects in automotive interior parts. The method comprises: obtaining an image to be detected and a standard image of the interior part containing depth information; obtaining the contour deviation of the target pixel based on its position in the image to be detected and the distance between the target pixel and its nearest edge pixel; obtaining the importance of defect detection based on the contour deviation of the target pixel, the depth value of the pixel in the region to which the target pixel belongs, and the number of characteristic value types; obtaining a defect index for the target pixel based on the depth difference between the target pixel and its reference point and the importance of defect detection for the target pixel; and performing interior part defect detection based on the magnitude of the defect index. The present invention introduces image depth information and analyzes the characteristics of interior part three-dimensional structural defects to detect subtle interior part three-dimensional structural defects.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and more particularly to a method and system for detecting defects in automotive interior trim parts. Background Art

[0002] Automobile interior parts are usually produced through injection molding. During the production process, there are usually certain defects on the automobile interior parts. If the defects on the automobile interior parts are not detected, it will affect the subsequent covering and installation of the automobile interior parts. Therefore, high-precision automobile interior part defect detection is particularly important for the quality of automobile interior parts.

[0003] In the related technology, for example, the Chinese patent document with authorization announcement number CN118334019B discloses a method and system for detecting the injection molding quality of injection molded parts, including obtaining a grayscale image of the injection molded part, determining the edge texture and grayscale area; determining matching pixel points based on slope characteristics; screening to obtain wavy edges based on texture morphological characteristics, grayscale distribution characteristics and distance characteristics; screening to obtain a light spot area based on the morphological characteristics and grayscale values ​​of the grayscale area; fitting the overall edge based on the edge texture of the intersection of the light spot area and the wave edge; and then determining the wave area, and detecting the injection molding quality of the injection molded part based on the wave area.

[0004] In the related art, injection molded part defects are detected through grayscale images of injection molded parts, but defects in the three-dimensional structure of the injection molded parts are ignored; in the processing of automotive interior parts, due to factors such as the quality of raw materials and inherent defects in the processing process, the produced automotive interior parts sometimes have structural defects such as warping and deformation, resulting in the inability to accurately install the automotive interior parts on the corresponding components during assembly; and warped automotive interior parts have similar color characteristics to normal automotive interior parts. Relying solely on the color characteristics of the image cannot effectively identify the three-dimensional structural deviation of the interior parts. Sometimes it is difficult to detect the warping of automotive interior parts through the color characteristics of the image. Therefore, there are certain limitations to defect detection of automotive interior parts using existing technologies. Summary of the Invention

[0005] In order to solve the technical problem that the defect detection of the above-mentioned automobile interior parts lacks the detection of three-dimensional structural defects of injection molded parts, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for detecting defects in automotive interior parts, comprising: obtaining an image to be detected and a standard image of an interior part containing depth information; taking any pixel point in the image to be detected as a target pixel point, and obtaining a contour deviation of the target pixel point based on the relative position of the target pixel point in the image to be detected and the distance between the target pixel point and the nearest edge pixel point; obtaining several areas in the image to be detected by a clustering algorithm; obtaining the defect detection importance of the target pixel point based on the contour deviation of the target pixel point, the standard deviation of the depth value of the pixel point in the area to which the target pixel point belongs, and the number of types of local binary pattern eigenvalues ​​of the pixel point in the neighborhood range of the target pixel point; obtaining the depth difference between the target pixel point and its reference point based on the difference in depth value mean and the difference in depth value standard deviation between the pixel point in the neighborhood range of the target pixel point and the pixel point in the neighborhood range of the reference point of the target pixel point, wherein the reference point of the target pixel point is a pixel point in the standard image with the same coordinates as the target pixel point; obtaining a defect index of the target pixel point based on the depth difference and the defect detection importance; and performing interior part defect detection based on the size of the defect index.

[0007] The present invention addresses the limitations of the prior art that relies on grayscale images and ignores the three-dimensional structural defects of interior trim parts. It introduces depth information to reflect the three-dimensional structural characteristics of automobile interior trim parts, thereby providing a reliable data basis for detecting three-dimensional structural defects of interior trim parts. The present invention obtains the importance of defect detection through contour deflection, combined with the distribution characteristics of pixel depth values ​​and the number of characteristic value types, so that the pixels in the defective area receive more attention. Furthermore, the defect index obtained according to the depth difference and the importance of defect detection can provide an important basis for defect detection of interior trim parts. The present invention can identify three-dimensional structural defects that are difficult to capture in the prior art, thereby timely eliminating interior trim parts with three-dimensional structural defects in the production process, effectively solving the shortcomings of the prior art in defect detection of automobile interior trim parts.

[0008] Preferably, the method for obtaining the image to be inspected and the standard image containing depth information is as follows: placing a defect-free automobile interior part on the inspection table, and taking a depth image of the defect-free automobile interior part from a top-down perspective using a depth camera as the standard image; placing the automobile interior part to be inspected at the same angle on the inspection table, and taking a depth image of the automobile interior part to be inspected from a top-down perspective using a depth camera as the image to be inspected.

[0009] Preferably, the profile deflection satisfies the relationship: Where, is the contour deviation of the target pixel point, is the distance between the target pixel and the nearest edge pixel, is the difference between the horizontal coordinate of the target pixel point and the horizontal coordinate of the image center point, is the difference between the ordinate of the target pixel and the ordinate of the image center point, and are half the height and half the width of the image to be detected, respectively. is an exponential function with a natural constant as base, is the maximum value function.

[0010] The present invention utilizes the contour deflection to make the pixel points closer to the contour of the interior decoration parts receive more attention. Since the three-dimensional structural defects of the automobile interior decoration parts are relatively obvious at the contour of the interior decoration parts, the present invention can make the defect detection of the automobile interior decoration parts more sensitive to the three-dimensional structural defects of the automobile interior decoration parts, and make the detection of the three-dimensional structural defects of the automobile interior decoration parts more accurate.

[0011] Preferably, the method for acquiring the neighborhood range of the target pixel point includes: taking the target pixel point as the center and a circular range with the distance between the target pixel point and the nearest edge pixel point as the radius as the neighborhood range of the target pixel point.

[0012] Preferably, the defect detection importance satisfies the relationship: Where, is the importance of defect detection for the target pixel, is the contour deviation of the target pixel point, is the standard deviation of the depth values ​​of all pixels in the area where the target pixel belongs, is the number of types of pixel feature values ​​within the neighborhood of the target pixel, is the number of types of feature values ​​of all pixels in the image to be detected, is a linear normalization function.

[0013] The present invention obtains the importance of defect detection by combining the contour deflection with the depth information and texture information of the image, so that the pixels that can reflect the structural defects of the interior parts receive more attention during the defect detection of automobile interior parts, and the influence of the image background area and the area in the image that cannot obviously reflect the three-dimensional structural defects of the interior parts on the defect detection of automobile interior parts is weakened, thereby reducing the interference with the defect detection of automobile interior parts and making the defect detection more accurate.

[0014] Preferably, obtaining the depth difference between the target pixel and its reference point includes: obtaining the absolute value of the difference in the mean depth value between the pixels within the neighborhood of the target pixel and the pixels within the neighborhood of the reference point of the target pixel, and the absolute value of the difference in the standard deviation of the depth value, and obtaining the depth difference between the target pixel and its reference point based on the product of the absolute values ​​of the two differences.

[0015] The depth difference between the target pixel point and its reference point in the present invention can well reflect the difference in depth information between the target pixel point and its reference point within the neighborhood range. Therefore, when there are defects in the automobile interior parts in the image to be detected, the depth difference can well reflect the difference in three-dimensional structure between the automobile interior parts in the standard image and the automobile interior parts in the image to be detected, providing an important basis for defect detection of automobile interior parts and making the defect detection results more reliable.

[0016] Preferably, obtaining the defect index of the target pixel includes: calculating the depth difference between the target pixel and its reference point, and obtaining the defect index of the target pixel according to the product of the depth difference and the defect detection importance of the target pixel.

[0017] The present invention obtains the defect index of the target pixel point based on the depth difference and the defect detection importance of the target pixel point, making the defect detection more sensitive to the depth value change at the outline of the automobile interior decoration part in the image, so that the depth difference in the area that is not important for defect detection can be identified as a defect only when it is large enough, suppressing the response of the pixel points that cannot reflect the defects of the interior decoration part to the defect, and improving the reliability of detection.

[0018] Preferably, the interior parts defect detection according to the size of the defect index includes: in response to the average value of the defect index of all pixel points within the neighborhood range of the target pixel point being greater than a preset threshold, the neighborhood range of the target pixel point is the defect area of ​​the automobile interior parts to be detected, and the automobile interior parts to be detected have defects.

[0019] Preferably, the interior parts defect detection according to the size of the defect index includes: in response to the number of pixel points whose defect index is greater than a preset threshold within the neighborhood range of the target pixel point exceeding a preset number, the neighborhood range of the target pixel point is the defect area of ​​the automobile interior parts to be detected, and the automobile interior parts to be detected have defects.

[0020] In a second aspect, the present invention provides an automobile interior part defect detection system, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned automobile interior part defect detection method is implemented.

[0021] By adopting the above technical solution, the above-mentioned automobile interior defect detection method is generated into a computer program and stored in a memory to be loaded and executed by a processor, so that a terminal device is made based on the memory and the processor for easy use.

[0022] The beneficial effects of the present invention are as follows: in response to the deficiency of the prior art in detecting three-dimensional structural defects of automotive interior parts, the present invention introduces depth information, and supplements the limitation of using only grayscale images to detect three-dimensional structural defects such as warping and local deformation; the contour deflection is calculated by the distance from the pixel to the nearest edge and the relative position of the pixel, so that the defect detection has a higher sensitivity to areas such as warping and deformation of the outer contour of the interior part and the protrusion of the three-dimensional structure inside the interior part, so that the defect detection is focused on the area where deformation is most likely to occur, providing an important basis for calculating the importance of defect detection; the depth value of the pixel point in the local range of the pixel point and the number of local binary pattern feature value types and the contour deflection are combined to obtain the importance of defect detection, so that the defect detection is further focused on the area that can reflect the three-dimensional structural defects of the automotive interior part, and the reliability of defect detection is further improved; the defect index is obtained by combining the depth difference and the defect detection importance, which suppresses the interference of background areas that are not important for defect detection, makes the defect detection more sensitive to three-dimensional structural defects such as warping and dents in the automotive interior part, and makes the defect detection of automotive interior parts more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart schematically illustrating a method for detecting defects in automotive interior parts according to the present invention;

[0024] Figure 2 Schematically shows an image to be detected;

[0025] Figure 3 It is a schematic diagram schematically showing the distribution of pixel feature values. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work shall fall within the scope of protection of the present invention.

[0027] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] The embodiment of the present invention discloses a method for detecting defects in automobile interior parts, referring to Figure 1 , including steps S1 to S5:

[0029] S1. Acquire an interior part image containing depth information through a depth camera.

[0030] Specifically, an image of an interior part of an automobile of any model is taken as the target model, and an angle facing the center point of the interior part is taken as the shooting angle. The image of the interior part of the target model without defects is obtained by a depth camera, and the image is used as the standard image; an image of the interior part to be inspected of the target model is obtained, and the image is used as the image to be inspected; wherein the image to be inspected is an RGB image with depth information, and each pixel in the image has values ​​of four channels R, G, B, and D. R, G, and B are the values ​​of the red, green, and blue channels respectively, representing the color characteristics of the pixel, and D is the depth value, representing the distance between the sensor of the depth camera and the position of the interior part corresponding to the pixel in the image; exemplarily, Figure 2 is the image to be detected.

[0031] S2. Obtain the contour deviation of the target pixel point according to the relative position of the target pixel point in the image to be detected and the distance between the target pixel point and the edge pixel point closest to it.

[0032] It should be noted that since the warping of the automobile interior parts to be inspected will manifest as the shape of the interior parts partially deviating from the designed shape, and usually when the interior parts have three-dimensional structural defects, the warping parts at the outer contour of the interior parts and the contour formed by the three-dimensional structure of the interior parts are more obvious, in order to better detect whether there are three-dimensional structural defects on the injection-molded parts to be inspected, the present invention obtains the contour deviation of each pixel point through the position of each pixel point in the image to be inspected.

[0033] Specifically, an edge detection algorithm is used to obtain edges in the image to be detected. Any pixel in the image to be detected is used as the target pixel. The distance between the target pixel and the nearest edge pixel is obtained, where the nearest edge pixel is the edge pixel closest to the target pixel. The contour deviation of the target pixel is obtained based on the distance between the target pixel and its nearest edge pixel and the pixel's coordinates.

[0034] In one embodiment, the edge detection algorithm is a Canny edge detection algorithm.

[0035] Specifically, the contour deflection of the target pixel satisfies the relationship:

[0036] ;

[0037] Where, is the contour deviation of the target pixel point, is the distance between the target pixel and the nearest edge pixel, is the difference between the horizontal coordinate of the target pixel point and the horizontal coordinate of the image center point, is the difference between the ordinate of the target pixel and the ordinate of the image center point, and are half the height and half the width of the image to be detected, respectively. is an exponential function with a natural constant as base, is the maximum value function.

[0038] in, It represents the horizontal relative distance between the target pixel point and the center point of the image to be detected. The larger the value, the closer the target pixel point is to the boundary of the image to be detected, the closer the target pixel point is to the outer contour of the car interior, and the greater the contour deviation of the target pixel point; the smaller the value, the farther the target pixel point is from the boundary of the image to be detected, the farther the target pixel point is from the outer contour of the car interior, and the smaller the contour deviation of the target pixel point. Represents the vertical relative distance between the target pixel and the center point of the image to be detected. The principle and Similar, I will not go into details here.

[0039] Represents the relative distance between the target pixel and its nearest edge pixel. Since the contour of the car interior parts will form an edge in the image, the larger the value, the farther the target pixel is from the contour of the interior parts, and the smaller the contour deviation of the target pixel; the smaller the value, the closer the target pixel is to the contour of the interior parts, and the larger the contour deviation of the target pixel; where, and Represent half of the height and half of the width of the image to be detected, respectively. Used for Perform negative correlation normalization, Divide by Used to prevent Larger, resulting in The result always approaches 0.

[0040] S3. Obtain the defect detection importance of the target pixel according to the contour deviation of the target pixel, the standard deviation of the pixel depth values ​​in the area to which the target pixel belongs, and the number of types of local binary pattern eigenvalues ​​of the pixel within the neighborhood of the target pixel.

[0041] It should be noted that in addition to the interior parts, there is also a background in the interior parts image. The pixels in the image background are not important for the defect detection of the interior parts. Different models of interior parts are made of different raw materials and have different colors. Therefore, it is not possible to directly measure whether the pixel point is on the car interior parts by the grayscale value of the pixel point; in order to avoid the influence of the part of the image background that is close to the outline of the interior parts on the defect detection of the interior parts, the present invention obtains the defect detection importance of each pixel point based on the outline deviation of the target pixel point and the pixel value of the pixel point.

[0042] Specifically, the values ​​of the R, G, and B channels of the image to be detected are used to convert the image to be detected into a grayscale image, and the pixels in the image to be detected are clustered according to the grayscale values, and the pixels in the image to be detected are divided into several categories. The area composed of all pixels of the same category as the target pixel is the area to which the target pixel belongs; the standard deviation of the depth values ​​of the pixels in the area to which the target pixel belongs is obtained; the characteristic value of each pixel in the image to be detected is obtained by the local binary pattern algorithm; the circular range with the target pixel as the center and the distance between the target pixel and its nearest edge pixel as the radius is the neighborhood range of the target pixel, and the number of types of pixel characteristic values ​​in the neighborhood range of the target pixel is counted, and the importance of defect detection of the target pixel is obtained according to the contour deviation of the target pixel, the standard deviation of the depth values ​​of all pixels in the area to which the target pixel belongs, and the number of types of pixel characteristic values ​​in the neighborhood range of the target pixel.

[0043] In one embodiment, the clustering algorithm is a mean-shift clustering algorithm.

[0044] Specifically, the defect detection importance of the target pixel satisfies the relationship:

[0045] ;

[0046] Where, is the importance of defect detection for the target pixel, is the contour deviation of the target pixel point, is the standard deviation of the depth values ​​of all pixels in the area where the target pixel belongs, is the number of types of pixel feature values ​​within the neighborhood of the target pixel, is the number of types of feature values ​​of all pixels in the image to be detected, is a linear normalization function.

[0047] in, Represents the discrete degree of pixel depth value in the area where the target pixel belongs. Since the car interior has a complex three-dimensional structure and the background in the image is flat, the depth value distribution of the pixel points in the area corresponding to the interior in the image is more discrete than that in the background area. The larger the value, the more likely the target pixel is in the automotive interior area of ​​the image, and the greater the importance of defect detection for the target pixel. The smaller it is, the more likely the target pixel is in the background area of ​​the image, and the less important the defect detection of the target pixel is; Represents the local texture complexity of the target pixel, Figure 3 is a schematic diagram of the distribution of pixel feature values, such as Figure 3As shown in the figure, since the surface of the car interior parts has a certain texture, the eigenvalue distribution of the interior parts in the image is complex. The number of types of pixel eigenvalues ​​in the local range of the pixel points is relatively large compared to the number of types of pixel eigenvalues ​​in the background area. However, the image background usually does not have a texture structure, and the number of types of eigenvalues ​​in the local range of the pixel points is relatively small. The larger the value, the more likely the target pixel is in the automotive interior area of ​​the image, and the greater the importance of defect detection for the target pixel. The smaller the value, the more likely the target pixel is in the background area of ​​the image, and the less important the defect detection of the target pixel is. Degree of deviation of the contour of the target pixel Further corrections are made so that the pixels in the interior part area of ​​the image that are closer to the interior part outline have a higher defect detection importance, and the pixels in the interior part area of ​​the image that are farther away from the interior part outline and the pixels in the background area have a lower defect detection importance.

[0048] S4. Obtain a defect index of the target pixel according to the defect detection importance of the target pixel and the depth difference between the target pixel and its reference point.

[0049] It should be noted that when a three-dimensional structural defect appears in the automobile interior parts in the image to be inspected, there will be a certain difference between the image to be inspected and the standard image; at the same time, the contour within the automobile interior parts area in the image has a prominent expression of the three-dimensional structural defects of the automobile interior parts. Therefore, the present invention obtains the defect index of the target pixel point based on the importance of defect detection of the target pixel point and the depth difference between the target pixel point and its reference point.

[0050] Specifically, the pixel point with the same coordinates as the target pixel point in the standard image is used as the reference point of the target pixel point; the circular range with the reference point of the target pixel point as the center and the size of the radius of the neighborhood range of the target pixel point as the radius is used as the neighborhood range of the reference point; the mean and standard deviation of the depth values ​​of the pixels in the neighborhood range of the target pixel point are obtained; the mean and standard deviation of the depth values ​​of the pixels in the neighborhood range of the reference point of the target pixel point are obtained; the defect index of the target pixel point is obtained based on the difference in the mean of the depth values ​​between the target pixel point and its reference point in the neighborhood range, the difference in the standard deviation of the depth values ​​between the target pixel point and its reference point in the neighborhood range, and the importance of defect detection of the target pixel point.

[0051] Specifically, the defect index of the target pixel satisfies the relationship:

[0052] ;

[0053] Where, is the defect index of the target pixel, is the importance of defect detection for the target pixel, and are the mean and standard deviation of the depth values ​​of the pixels in the neighborhood of the target pixel, and are the mean and standard deviation of the depth values ​​of the pixels in the neighborhood of the reference point of the target pixel, is a linear normalization function.

[0054] in, Represents the difference in depth value between the target pixel and its reference point in the local range. The larger the value, the greater the difference in depth value between the pixels in the neighborhood of the target pixel and the pixels in the neighborhood of its reference point. The more likely the part of the automotive interior corresponding to the neighborhood of the target pixel is to have a three-dimensional structural defect such as warping of the interior, and the larger the defect index of the target pixel. The smaller the value, the smaller the difference in depth value between the pixels in the neighborhood of the target pixel and the pixels in the neighborhood of the reference point. The more likely it is that the part of the automotive interior corresponding to the neighborhood of the target pixel does not have a three-dimensional structural defect, and the smaller the defect index of the target pixel. It represents the difference between the depth value dispersion of the target pixel and its reference point in the local range. The larger the value, the greater the difference in depth value distribution between the pixels in the neighborhood of the target pixel and the pixels in the neighborhood of the reference point. The more likely the automobile interior part corresponding to the neighborhood of the target pixel has a three-dimensional structural defect, and the larger the defect index of the target pixel. The smaller the value, the smaller the difference in depth value distribution between the pixels in the neighborhood of the target pixel and the pixels in the neighborhood of the reference point. The more likely the automobile interior part corresponding to the neighborhood of the target pixel has no three-dimensional structural defect, and the smaller the defect index of the target pixel. It is the depth difference between the target pixel and its reference point. Since the contour of the automobile interior parts area in the image has a prominent expression of the three-dimensional structural defects of the automobile interior parts, in order to reduce the influence of the pixels that are not important for automobile interior parts defect detection in the image on automobile interior parts defect detection, the interior parts defect detection should be more sensitive to the depth difference corresponding to the pixels with greater defect detection importance, and less sensitive to the pixels with smaller defect detection importance. Therefore, the defect detection importance of the target pixel is used to determine the depth difference of the automobile interior parts. Depth Difference Make further corrections, The smaller it is, the smaller the defect index of the pixel points that are not important for interior defect detection is, making the pixels with less importance for defect detection less sensitive. The defects can only be identified when the depth difference of the pixel points that are not important for interior defect detection is large enough. The larger it is, the greater the defect index of the pixel points that are not important for interior defect detection, the more sensitive the pixel points that are more important for defect detection are, and the easier it is to identify defects when there are defects at the pixel points that are important for interior defect detection.

[0055] S5. Perform defect detection on interior parts according to the size of the defect index.

[0056] In one embodiment, in response to the average defect index of all pixels within the neighborhood of the target pixel being greater than a preset threshold, the neighborhood of the target pixel is a defective area of ​​the automotive interior to be inspected, and the automotive interior to be inspected has defects.

[0057] In another embodiment, in response to the number of pixels in the neighborhood of the target pixel point having a defect index greater than a preset threshold exceeding a preset number, the neighborhood of the target pixel point is a defective area of ​​the automobile interior part to be inspected, and the automobile interior part to be inspected has defects.

[0058] The preset threshold and preset number are set by the implementer according to the actual implementation situation. For example, the preset threshold is 0.5, and the preset number is 0.5 times the number of pixels in the neighborhood of the target pixel.

[0059] An embodiment of the present invention further discloses an automobile interior defect detection system, comprising a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, an automobile interior defect detection method according to the present invention is implemented.

[0060] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

Claims

1. A method for detecting defects in automotive interior parts, characterized in that: include: Obtain an image to be inspected and a standard image of the interior part containing depth information; Taking any pixel in the image to be detected as the target pixel, the contour deviation of the target pixel is obtained according to the relative position of the target pixel in the image to be detected and the distance between the target pixel and the nearest edge pixel. A clustering algorithm is used to obtain several regions in the image to be inspected. The importance of defect detection of the target pixel is obtained based on the contour deviation of the target pixel, the standard deviation of the depth value of the pixel in the region to which the target pixel belongs, and the number of types of local binary pattern eigenvalues ​​of the pixel in the neighborhood of the target pixel. The importance of defect detection satisfies the relationship: Where, is the importance of defect detection for the target pixel, is the contour deviation of the target pixel point, is the standard deviation of the depth values ​​of all pixels in the area where the target pixel belongs, is the number of types of pixel feature values ​​within the neighborhood of the target pixel, is the number of types of feature values ​​of all pixels in the image to be detected, is a linear normalization function; Obtaining a depth difference between a target pixel and its reference point based on a difference in depth mean values ​​and a difference in depth standard deviation between pixels within a neighborhood of the target pixel and pixels within a neighborhood of a reference point of the target pixel, where the reference point of the target pixel is a pixel with the same coordinates as the target pixel in the standard image; Obtain the defect index of the target pixel point based on the depth difference and the importance of defect detection; Perform interior parts defect detection based on the size of the defect index.

2. The method for detecting defects in automotive interior parts according to claim 1, wherein: The method for obtaining the image to be detected and the standard image containing depth information is as follows: Place a defect-free automobile interior part on the inspection table, and use the depth camera to take a depth image of the defect-free automobile interior part from a top-down perspective as the standard image; place the automobile interior part to be inspected on the inspection table at the same angle, and use the depth camera to take a depth image of the automobile interior part to be inspected from a top-down perspective as the image to be inspected.

3. The method for detecting defects in automotive interior parts according to claim 1, wherein: The profile deflection satisfies the relationship: ； Where, is the contour deviation of the target pixel point, is the distance between the target pixel and the nearest edge pixel, is the difference between the horizontal coordinate of the target pixel point and the horizontal coordinate of the image center point, is the difference between the ordinate of the target pixel and the ordinate of the image center point, and are half the height and half the width of the image to be detected, respectively. is an exponential function with a natural constant as base, is the maximum value function.

4. The method for detecting defects in automotive interior parts according to claim 1, wherein: The method for obtaining the neighborhood range of the target pixel point includes: The neighborhood range of the target pixel is a circular range with the target pixel as the center and the distance between the target pixel and the nearest edge pixel as the radius.

5. The method for detecting defects in automotive interior parts according to claim 1, wherein: Obtaining the depth difference between the target pixel and the reference point includes: Obtain the absolute value of the difference in depth mean between the pixels in the neighborhood of the target pixel and the pixels in the neighborhood of the reference point of the target pixel, as well as the absolute value of the difference in depth standard deviation. The depth difference between the target pixel and its reference point is obtained based on the product of the absolute values ​​of the two differences.

6. The method for detecting defects in automotive interior parts according to claim 1, wherein: The step of obtaining the defect index of the target pixel includes: The depth difference between the target pixel and its reference point is calculated, and the defect index of the target pixel is obtained based on the product of the depth difference and the defect detection importance of the target pixel.

7. The method for detecting defects in automotive interior parts according to claim 1, wherein: The interior decoration defect detection according to the size of the defect index includes: In response to the average defect index of all pixels within the neighborhood of the target pixel being greater than a preset threshold, the neighborhood of the target pixel is a defective area of ​​the automotive interior component to be inspected, and the automotive interior component to be inspected has defects.

8. The method for detecting defects in automotive interior parts according to claim 1, wherein: The interior decoration defect detection according to the size of the defect index includes: In response to the number of pixels in the neighborhood of the target pixel having a defect index greater than a preset threshold exceeding a preset number, the neighborhood of the target pixel is a defective area of ​​the automotive interior part to be inspected, and the automotive interior part to be inspected has defects.

9. An automotive interior parts defect detection system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an automobile interior component defect detection method according to any one of claims 1 to 8 is implemented.

Images