Apparatus for inspecting the appearance of resin parts and method for inspecting the appearance of resin parts

The resin part inspection apparatus and method address the challenge of quantifying low-contrast defects by setting specific angles for illumination and detection, removing specular reflection, and using image processing to efficiently detect and quantify appearance defects without machine learning, enhancing inspection accuracy and efficiency.

JP7864250B2Active Publication Date: 2026-05-22MITSUBISHI ELECTRIC CORP
View PDF 9 Cites 0 Cited by

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
MITSUBISHI ELECTRIC CORP
Filing Date
2024-02-01
Publication Date
2026-05-22

AI Technical Summary

Technical Problem

Existing resin part inspection methods, particularly those using machine learning and visual inspection, struggle to quantify appearance defects with low contrast and not caused by macroscopic irregularities, and require large amounts of training data.

Method used

A resin part appearance inspection apparatus and method that utilizes a camera, lighting, and image processing to detect defects by setting specific angles for illumination and detection, removing specular reflection, and extracting diffuse reflection for analysis, employing image processing steps to quantify defects without machine learning.

Benefits of technology

The method requires fewer images for inspection, can quantify appearance defects with low contrast, and accurately detects defects not caused by macroscopic irregularities, improving inspection efficiency and consistency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007864250000001
    Figure 0007864250000001
  • Figure 0007864250000002
    Figure 0007864250000002
  • Figure 0007864250000003
    Figure 0007864250000003
Patent Text Reader

Abstract

A resin component appearance inspection device (100) comprises: a component holding mechanism (2) for holding a resin component (1) to be inspected; a camera (3) for capturing an image of the resin component (1); an illumination (4) for illuminating the resin component (1); and an image processing device (5) for performing arithmetic processing on the image captured by the camera (3). The resin component (1) is opaque and glossy. The image processing device (5) performs the arithmetic processing on the image captured by the camera (3) to detect a defective portion (1F) on the surface of the resin component and quantify the quality of the appearance of the resin component (1).
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0005] , , ,

[0004] , , , , , , ,

[0001] The present disclosure relates to a resin part appearance inspection apparatus and a resin part appearance inspection method.

Background Art

[0002] Defects in resin parts include shape defects such as warping, twisting, and sink marks, defects due to thermal factors such as black spots and foreign matters, mold defects such as scratches and burrs, and appearance defects such as flow marks and silver streaks. Among these, visual inspection by an inspector is common for appearance inspection such as flow marks. This visual inspection is a subjective evaluation method that depends on the judgment of the inspector and cannot be quantified, so it is difficult to maintain a constant appearance quality.

[0003] For the purpose of depersonalization, a defect detection and classification system using machine learning is disclosed, which includes an acquisition unit for acquiring images and a detection and classification unit for determining the presence, position, and type of defects (for example, Patent Document 1). In addition, an inspection system is disclosed that includes an inspection camera for photographing an inspection area including a part of the inspection object, a detection camera for photographing a detection area wider than the inspection area, a stage on which the inspection object is placed, and an adjustment mechanism for adjusting the relative position between the stage and the inspection camera based on an image of the detection area (for example, Patent Document 2).

Prior Art Documents

Patent Documents

[0006] This disclosure aims to provide a resin part appearance inspection device and resin part appearance inspection method that can quantify appearance defects that require fewer images for inspection, have low contrast, and are not caused by macroscopic irregularities. [Means for solving the problem]

[0007] The resin part appearance inspection apparatus of this disclosure comprises a part holding mechanism for holding a resin part to be inspected, a camera for photographing the resin part, lighting for illuminating the resin part, and an image processing apparatus for processing the image captured by the camera, wherein the resin part is opaque and glossy, and the image processing apparatus processes the image captured by the camera to detect defective parts on the surface of the resin part and quantifies the appearance quality of the resin part. A resin parts appearance inspection device, The angle between the lighting and the defective portion of the resin part is defined as the incident angle, and the angle between the camera and the defective portion of the resin part is defined as the detection angle. When the lighting illuminates the glossy resin part, the reflection where the incident angle and reflection angle are equal is defined as specular reflection, the portion where the appearance of the lighting is reflected and visible on the surface of the resin part is defined as specular reflection, and the portion that reflects at an angle different from the incident angle because the surfaces of the good portion and the defective portion are not perfectly mirrored is defined as diffuse reflection. By utilizing the fact that the angular distribution of the light intensity of specular reflection and diffuse reflection differs between the good portion and the defective portion, the detection angle is set to have an angular difference from the incident angle. By trimming, the specular reflection portion of the specular reflection is removed, the diffuse reflection portion is extracted, and the appearance defect is detected. It was designed that way. The resin part appearance inspection method of this disclosure uses a part holding mechanism for holding the resin part to be inspected, a camera for photographing the resin part, a light source for illuminating the resin part, and an image processing device for processing the image captured by the camera, A resin part appearance inspection method is provided in which the angle between the illumination and the defective portion of the resin part is defined as the incident angle, the angle between the camera and the defective portion of the resin part is defined as the detection angle, when the illumination shines light onto the glossy resin part, the reflection where the incident angle and the reflection angle are equal is defined as specular reflection, the portion where the appearance of the illumination is reflected and visible on the surface of the resin part is defined as specular reflection, and the portion where reflection occurs at an angle different from the incident angle due to the surface of the resin part not being a perfect mirror is defined as diffuse reflection, and the detection angle is set to have an angle difference with respect to the incident angle, thereby removing the specular reflection portion from the specular reflection portion and extracting the diffuse reflection portion. Image acquisition step of taking an image captured by the camera into the image processing device, and the defective part on the surface of the resin part The aforementioned The system comprises a trimming step for performing trimming, a smoothing step for smoothing the image created in the trimming step, a difference image creation step for creating an image obtained by taking the difference between the image created in the trimming step and the image created in the smoothing step, a binarization step for binarizing the image created in the difference image creation step, a contour extraction step for extracting the contours of the defective parts, and a quantitative value calculation step for calculating and quantifying the pixel ratio of the defective parts.The present disclosure is a resin part appearance inspection method that performs an appearance quality inspection of a resin part using an injection molding apparatus that performs injection molding of a resin part to be inspected, and is performed using a camera for photographing the resin part, a light for illuminating the resin part installed in a mold or platen used for resin molding of the resin part, an image processing device for processing the image captured by the camera, a movable blackout curtain, and a removal machine for removing and moving the resin part from the mold, and includes a position adjustment step for adjusting the positions of the camera and the light, a continuous molding start step for starting continuous molding of the resin part, a resin filling step for filling the mold with resin, a first blackout curtain moving step for moving the blackout curtain to the vicinity of the mold, a mold opening step for opening the mold after the resin filling of the resin filling step is completed, and irradiating the resin part with light from the light and the resin part with the camera A shooting step of taking an image of the resin part; a second blackout curtain moving step of moving the blackout curtain to the outside of the mold; a chucking step of moving the removal machine to the vicinity of the resin part and chucking the resin part; an image acquisition step of importing the image of the resin part taken by the camera into the image processing device; an image processing step of performing image processing on the imported image and calculating a quantitative value of the appearance quality of the resin part; a quantitative value output step of outputting the quantitative value of the appearance quality of the resin part calculated in the image processing step; a quality determination step of comparing the appearance quality of the resin part output in the quantitative value output step with a pre-set required quality and determining whether the required quality is satisfied; and a good product processing step of moving the removal machine to a good product storage area and returning to the resin filling step if the quality determination step determines that the required quality is satisfied. If the quality determination step determines that the good faith required quality is not met, the removal machine includes a defective product processing step in which it moves the resin part to a defective product storage area; a defective frequency determination step in which it determines whether defective products have occurred continuously at a pre-set threshold or higher, and if they have not occurred continuously at a threshold or higher, it returns to the resin filling step, and if they have occurred continuously at a threshold or higher, it proceeds to the alert output step; and an alert output step in which it outputs a continuous defect alert and stops the continuous molding process. [Effects of the Invention]

[0008] According to the resin parts appearance inspection apparatus of this disclosure, a resin parts appearance inspection apparatus is obtained that requires fewer images for inspection, has low contrast, and can quantify appearance defects that are not caused by macroscopic irregularities. The resin part appearance inspection method disclosed herein provides a resin part appearance inspection method that requires fewer images for inspection, has low contrast, and can quantify appearance defects that are not caused by macroscopic irregularities. [Brief explanation of the drawing]

[0009] [Figure 1] This is an isometric view of the entire resin parts appearance inspection apparatus according to Embodiment 1. [Figure 2] This is a diagram showing the configuration of the resin parts appearance inspection apparatus according to Embodiment 1. [Figure 3] Figure 3A is a conceptual diagram of illumination and reflected light according to Embodiment 1. Figure 3B is an explanatory diagram of reflected light according to Embodiment 1. [Figure 4] This is an explanatory diagram of the incident angle of illumination and the detection angle of the camera according to Embodiment 1. [Figure 5] This is an explanatory diagram of the appropriate detection angle range according to Embodiment 1. [Figure 6] Figure 6A is an explanatory diagram of the measured results regarding the detection angle according to Embodiment 1. Figure 6B is an explanatory diagram of the measured results regarding the detection angle according to Embodiment 1. [Figure 7] This is an explanatory diagram illustrating an example of an image application of diffuse reflection according to Embodiment 1. [Figure 8] This is an explanatory diagram for determining the appropriateness of the detection angle according to Embodiment 1. [Figure 9] Figure 9A is an explanatory diagram of the luminance distribution characteristics of the lighting according to Embodiment 1. Figure 9B is an explanatory diagram of the luminance distribution characteristics of the lighting according to Embodiment 1. [Figure 10] This is an explanatory diagram illustrating the length requirements for the illumination light-emitting section according to Embodiment 1. [Figure 11]It is a flowchart of the resin part appearance inspection method according to Embodiment 1. [Figure 12] It is a flowchart of the image processing of the resin part appearance inspection according to Embodiment 1. [Figure 13] FIG. 13A is an example of a processed image of the resin part appearance inspection according to Embodiment 1. FIG. 13B is an example of a processed image of the resin part appearance inspection according to Embodiment 1. FIG. 13C is an example of a processed image of the resin part appearance inspection according to Embodiment 1. FIG. 13D is an example of a processed image of the resin part appearance inspection according to Embodiment 1. FIG. 13E is an example of a processed image of the resin part appearance inspection according to Embodiment 1. FIG. 13F is an example of a processed image of the resin part appearance inspection according to Embodiment 1. [Figure 14] It is an example of a processed image of a defective part in the resin part appearance inspection according to Embodiment 1. <00第11]] [Figure 15] It is an example of a processed image of a defective part in the resin part appearance inspection according to Embodiment 1. [Figure 16] It is a configuration diagram of the resin part appearance inspection apparatus according to Embodiment 2. [Figure 17] It is a flowchart of the resin part appearance inspection method according to Embodiment 2. [Figure 18] It is a configuration diagram of the resin part appearance inspection apparatus according to Embodiment 3. [Figure 19] It is a flowchart of the resin part appearance inspection method according to Embodiment 3. [Figure 20] It is a system configuration diagram of the resin part appearance inspection apparatus according to Embodiment 4. [Figure 21] It is a system configuration diagram of the resin part appearance inspection apparatus according to Embodiment 4. [Figure 22] It is a functional block diagram of each apparatus related to the injection molding process according to Embodiment 4. [Figure 23] FIG. 23A is an explanatory diagram of the injection molding process according to Embodiment 4. FIG. 23B is an explanatory diagram of the injection molding process according to Embodiment 4. [Figure 24] FIG. 24A is an explanatory diagram of the injection molding process according to Embodiment 4. FIG. 24B is an explanatory diagram of the injection molding process according to Embodiment 4. [Figure 25] This is a flowchart of the resin part appearance inspection method according to Embodiment 4. [Figure 26] This is a flowchart of the resin part appearance inspection method according to Embodiment 4. [Figure 27] This is an example of the hardware configuration of an image processing device and a control device. [Modes for carrying out the invention]

[0010] Embodiment 1. Embodiment 1 comprises a component holding mechanism for holding a resin component to be inspected, a camera for photographing the resin component, lighting for illuminating the resin component, and an image processing device for processing the image captured by the camera. The image processing device processes the image captured by the camera to detect defective parts on the surface of the resin component and quantifies the appearance quality of the resin component. This invention relates to a resin component appearance inspection device and a resin component appearance inspection method.

[0011] The resin parts appearance inspection apparatus and inspection method according to Embodiment 1 will be described below with reference to Figure 1, an isometric view of the entire resin parts appearance inspection apparatus; Figure 2, a configuration diagram of the resin parts appearance inspection apparatus; Figures 3A and 3B, conceptual diagrams of illumination and reflected light; Figure 4, an explanatory diagram of the incident angle of illumination and the detection angle of the camera; Figure 5, an explanatory diagram of the appropriate detection angle range; Figures 6A and 6B, explanatory diagrams of the measured results regarding the detection angle; Figure 7, an explanatory diagram of an example image of diffuse reflection; Figure 8, an explanatory diagram of the determination of appropriateness of the detection angle; Figures 9A and 9B, an explanatory diagram of the brightness distribution characteristics of illumination; Figure 10, an explanatory diagram of the length requirements of the illumination light-emitting part; Figure 11, a flowchart of the resin parts appearance inspection method; Figure 12, a flowchart of the image processing for resin parts appearance inspection; Figures 13A to 13F, which are examples of processed images for resin parts appearance inspection; and Figures 14 and 15, which are examples of processed images of defective parts for resin parts appearance inspection. In each figure, the same or corresponding parts are indicated by the same reference numeral.

[0012] First, the overall configuration of the resin parts appearance inspection apparatus 100 of Embodiment 1 will be explained based on Figure 1, which is an isometric view of the entire inspection apparatus, and Figure 2, which is a configuration diagram of the inspection apparatus. The resin part appearance inspection device 100 comprises, as its main components, a resin part 1 to be inspected, a part holding mechanism 2 to hold the resin part 1, a camera 3 and lighting 4 for photographing the resin part 1, an image processing device 5 for processing the captured images, a frame 10 which is the framework of the entire inspection device, and a blackout curtain 11 for blocking ambient light.

[0013] In Embodiment 1, it is assumed that the appearance of the resin part 1 is inspected in advance by visual inspection or other means to identify areas where cosmetic defects may occur, and that the presence or absence of these defective areas is checked. Therefore, when performing the appearance inspection, the distance and angle of the camera 3 and lighting 4 relative to the resin part 1 are adjusted so that the previously identified defective areas can be inspected with high accuracy.

[0014] Since the resin part 1 to be inspected is a part molded using a mold, if it is warped, the resin part appearance inspection described below cannot be performed accurately. Therefore, it is assumed that any warping will be corrected before inspection. The component holding mechanism 2 holds the resin component 1 suspended from the floor surface in order to inspect its appearance. Camera 3 captures images for visual inspection of the resin part 1, and lighting 4 emits the necessary illumination light for the camera 3 to capture images. Camera 3 and lighting 4 are mounted on the frame 10 and fixed to maintain an appropriate distance and angle from the resin part 1 to be inspected. Camera 3 and lighting 4 are attached to the frame 10, and a mounting device is needed to adjust the angle, but since this mounting device is publicly known, its explanation will be omitted. Furthermore, the camera 3 and lighting 4 are connected to the image processing device 5 by cable 51, and the image processing device 5 controls the turning on and off of the lighting 4 and the image capture of the resin part 1 by the camera 3. Note that the turning on and off of the lighting 4 may be controlled by a separate power supply or other device, or the lighting may be left on during inspection, rather than being controlled by the image processing device 5. Similarly, the image capture by the camera 3 may be controlled by a timer shooting function of a consumer compact digital camera or similar device, rather than being controlled by the image processing device 5. Note that in Figure 1, the cable connecting the image processing device 5 and the lighting 4 is omitted.

[0015] Next, we will explain each component in turn. First, let me explain resin part 1. Resin part 1 is a flat plate, and long parts with dimensions of approximately 300mm to 900mm in length and 20mm to 300mm in width are also expected to be subject to inspection. Resin part 1 is opaque and glossy. Furthermore, the system is designed to allow inspection even when ribs are present on the back surface of resin part 1.

[0016] Next, we will explain the component holding mechanism 2. The component holding mechanism 2 comprises an aluminum plate 21, a knurled bolt 22, and a magnet 23. The resin part 1 is secured by sandwiching it between the head of the knurled bolt 22 and the magnet 23.

[0017] A tapped hole (or screw insert) is drilled in the aluminum plate 21 to secure the knurled bolt 22, and the iron knurled bolt 22 is inserted and secured. The resin part 1 is placed on top of the knurled bolt 22, and the magnet 23 is placed on top of the resin part 1 and sandwiched between the head of the knurled bolt 22 to secure the resin part 1. This structure holds the resin part 1 suspended above the floor surface. The position of the knurled bolt 22 is set so that it does not interfere with the rib of the resin part 1. By configuring the component holding mechanism 2 as described above, it is possible to ensure reproducibility of the holding position for resin parts 1 that have warping and resin parts 1 that have protrusions including ribs.

[0018] When dealing with multiple resin parts 1, multiple aluminum plates 21 may be made, one for each resin part 1, or they may be consolidated into a single aluminum plate 21. When consolidating into a single aluminum plate 21, positions are set so that the ribs of each resin part 1 do not interfere with each other, and then tapped holes for all the target resin parts 1 are drilled in the single aluminum plate 21.

[0019] The reason for using knurled bolts 22 instead of dowel pins is that if dowel pins are installed using interference fit or intermediate fit, it is difficult to insert and remove the dowel pins. Also, if they are installed using clearance fit, there is a risk that the dowel pins may come out of the holes in the aluminum plate 21 if the resin part 1 is warped.

[0020] First, before describing specific visual inspection methods, we will define the terms used in this specification and explain the basic concepts of illumination and reflected light. The following provides general definitions of basic optics terms. (1) Specular reflection: This is a type of reflection that occurs at the interface between two materials, where the angle of incidence and the angle of reflection are equal. (2) Specular reflection This is a type of specular reflection that occurs on a surface with almost no irregularities. It is characterized by the appearance of a mirror image, a virtual image. (3) Irregular reflection: Specular reflection that occurs on a rough surface with uneven surfaces. (4) Diffuse reflection The opposite of specular reflection, it is characterized by emitting a similar level of luminosity in various directions, regardless of the angle of reflection. It is reflection due to multiple reflections, transmission, and scattering that occur slightly inside the interface of a material.

[0021] Next, we will explain "specular reflection" and "diffuse reflection," which are important concepts in the following explanation, using specific examples. When a light source is shone on a glossy resin molded product, the appearance of the light source is reflected onto the surface of the molded product. This phenomenon is referred to as specular reflection in this disclosure. On the other hand, since the surface of a molded product is not a perfect mirror surface, neither the good-looking parts nor the defective parts are perfectly mirror-like, it also reflects light at angles slightly different from the angle of incidence. This is referred to as diffuse reflection in this specification. Figures 3A and 3B illustrate the case where a glossy resin molded product with cosmetic defects is illuminated and photographed with a camera. In Figure 3A, the symbol LT represents light, the symbol SR represents specular reflection, the symbol IR represents diffuse reflection, and the symbol DL represents diffuse reflection. In Figure 3B, symbol A indicates the reflection of illumination due to specular reflection. Symbol B indicates the area shifted from the reflection of the light due to diffuse reflection, i.e., the area around the reflection. It is dimly bright and becomes darker as you move away from the reflection of the light. Symbol C indicates a part with a cosmetic defect. It appears slightly brighter in the area that is slightly off from the reflection of the lighting indicated by symbol B. The phenomenon described above is due to the difference in the angular distribution of specular and diffuse reflected light intensity between the parts of the product that are in good condition and those that are defective.

[0022] Next, the relationship between the appropriate incident angle of illumination and the detection angle of the camera will be explained based on Figure 4. In Figure 4, the dotted arrows indicate "specular reflected light with an angular distribution." The angle between the light source 4 and the defective area 1F on the surface of the resin part 1 is defined as the incident angle θ1, and the incident angle θ1 is set to 25° to 35°. The angle between the camera 3 and the defective area 1F on the surface of the resin part 1 is defined as the detection angle θ2, and the detection angle θ2 is set to be approximately 3° to 10° above or below the incident angle θ1. By setting a difference between the incident angle θ1 and the detection angle θ2, detection can be performed at the angle where the difference in reflected light intensity between the good part and the defective area 1F appears in the specularly reflected light, which has an angular distribution. The reason for setting a difference between the incident angle θ1 and the detection angle θ2 will be explained later. The distance between the light source 4 and the defective portion 1F on the surface of the resin part 1 (d in Figure 4) can be set to 500 mm or more to ensure the linearity of the light beam. Note that "defective part 1F" refers to the external defects of the resin part 1 being inspected, and does not include shape defects or mold defects. Additionally, "defective parts on the surface of resin parts" should be appropriately described as "defective parts of resin parts."

[0023] Next, the reason for the difference between the incident angle and the detection angle will be explained based on Figure 5, which is an explanatory diagram of the appropriate detection angle range. In Figure 5, the symbol GD corresponding to the solid line represents the "good part," and the symbol BD corresponding to the dotted line represents the "defective part." The function curve in Figure 5 is a conceptual diagram of the bidirectional scattering distribution function (BSDF) obtained by measuring the good and defective parts 1F of resin part 1 using a scattering detector. Figure 5 is a diagram that emphasizes the concept to make the difference easier to understand, but the actual measurement results are shown in Figures 6A and 6B. Figure 6A shows the measured results corresponding to Figure 5. Figure 6B is an enlarged view of section A in Figure 6A. As can be seen in Figure 6B, a slight difference in light intensity occurs at certain angles.

[0024] The reflected light intensity is maximum at an angle equal to the incident angle θ1 in both the good and defective parts 1F, and the reflected light intensity decreases as the angle moves away from this incident angle θ1. The defective part 1F has a wider angular distribution of reflected light intensity than the good part, and in the area shown as "Setting range of detection angle θ2" in the figure, the reflected light intensity is stronger in the defective part 1F. On the other hand, near the incident angle θ1, the reflected light intensity is stronger in the good part, but the angular range is narrower than "Setting range of detection angle θ2". Therefore, it is difficult to adjust the position of camera 3 near the incident angle θ1. Therefore, the detection angle θ2 of camera 3 is set to the "setting range of detection angle θ2" where the reflected light intensity is stronger in the defective area 1F. For this reason, it is necessary to set the detection angle θ2 of camera 3 to either the upper or lower side of the incident angle θ1 by approximately 3° to 10°.

[0025] If there is no difference between the detection angle and the incident angle, that is, if the reflected portion of the illumination (specular reflection) is cropped during image processing, it is not possible to detect appearance defects that do not have high contrast and are not caused by macroscopic irregularities, such as flow marks and tiger stripe defects on resin parts. A real-world example is illustrated in Figure 7. Figure 7 shows the results of defect detection processing for "flow marks" and "tiger stripes," which are surface defects on a glossy surface, using specular reflection (the part where the light is reflected) and diffuse reflection (the part shifted from the reflected light). In the processed image, "○" indicates detectable, and "×" indicates undetectable. In the case of specular reflection, neither flow mark defects nor tiger stripe defects were detected, but they were detected in the case of diffuse reflection. In the case of surface defects in resin parts, such as scratches, chips, and dirt, defects can be detected by the occurrence of non-specular reflection areas (dark areas) within specular reflection light. However, specular reflection light cannot detect surface defects that are not caused by macroscopic irregularities, such as flow marks and tiger stripe defects.

[0026] As explained above, by utilizing the fact that the angular distribution of specular and diffuse light intensity differs between good and defective parts, and by setting the incident angle to 35 to 55 degrees and the detection angle to have an angular difference of 3 to 10 degrees relative to the incident angle, the specular reflection portion of the specularly reflected light is removed by cropping in image processing, which will be explained later, and the diffuse reflection portion is extracted. This makes it possible to detect flow mark defects and tiger stripe defects in resin part 1, which are appearance defects with low contrast and not caused by macroscopic irregularities.

[0027] Next, we will explain how to determine whether the detection angle is appropriate, based on Figure 8, which is an explanatory diagram for determining the appropriateness of the detection angle. In Figure 8, symbol A represents "the external defect was not captured in the photograph," symbol B represents "the external defect was captured in the photograph," symbol C represents "the external defect was not captured in the photograph," and symbol D represents "reflection of lighting." In Figure 8, the image captured for installation pattern (a) (incident angle θ1 = 30°, detection angle θ2 = 30°) is determined to be insufficient to capture the external defect because the reflection of illumination 4 overlaps with the defective part of the resin component. The image captured using installation pattern (c) (incident angle θ1=30°, detection angle θ2=20°) is determined to be insufficient to capture the external defect because the reflection of illumination 4 is far from the defective part of the resin component. In contrast, in the installation pattern (b) where the detection angle θ2 is intermediate (incident angle θ1 = 30°, detection angle θ2 = 25°), it can be confirmed that the defective part of the resin component is captured, and therefore it can be determined that the external defect has been captured. If the above determination method indicates that the installation pattern is (a) or (c), the positional relationship between the lighting 4 and the resin part 1, and the positional relationship between the camera 3 and the resin part 1 are adjusted so that the installation pattern becomes (b).

[0028] Next, the requirements for lighting 4 will be explained with reference to Figures 9A to 9C, which are explanatory diagrams of the luminance distribution characteristics of the lighting. Figure 9A represents luminance from 0% to 100%, Figure 9B shows an example of a uniform luminance distribution, and Figure 9C shows an example of a non-uniform luminance distribution. Lighting fixture 4 will be a bar type, with a diffuse beam pattern and a surface-emitting type that prevents point light source reflections. Furthermore, it will have high uniformity in its luminance distribution characteristics along the long axis, and the length of the light-emitting section will be at least twice the length of the defective portion 1F of resin part 1. Figures 9B and 9C show examples of lighting with high and low uniformity in the luminance distribution characteristics along the long axis. In Figure 9C, the brightness is high only near the LED element. In contrast, in Figure 9B, the brightness is uniform across the entire width of the LED along its long axis, except for the ends.

[0029] Next, the relationship between the length of the light-emitting part of the lighting fixture 4 and the length of the defective part 1F of the resin component 1 will be explained based on Figure 10, which is an explanatory diagram of the length requirements for the light-emitting part of the lighting fixture. In order to illuminate the entire defective portion 1F of the resin part 1 uniformly, the length (L) of the light-emitting part of the lighting fixture 4 is set to be at least twice the length (w) of the defective portion 1F of the resin part 1. If the intensity is less than 2x, the incident light intensity weakens only at the edge of the defective portion 1F of resin part 1, making it impossible to detect the defective portion 1F.

[0030] Next, I will explain the requirements for camera 3. Camera 3 is equipped with an area sensor, has a focal length capable of capturing the defective part 1F of the resin part 1 and its surroundings, and uses an effective pixel count of 1 million pixels (1000 pixels x 1000 pixels) or more. Furthermore, the camera is designed so that the focus can be manually adjusted, and the F-number, shutter speed, and ISO sensitivity can be set arbitrarily.

[0031] Next, I will explain the requirements for the blackout curtain 11. When photographing the resin part 1 with the camera 3 in an environment with ambient light, a blackout curtain 11 is used to block the ambient light. To block the ambient light, the curtain is placed over the entire resin part appearance inspection device 100. Figure 1 shows a portion of the blackout curtain 11 as an example.

[0032] The blackout curtain 11 must satisfy the following specifications regarding light blocking, tear strength, and flame retardancy. (1) Light-blocking properties: The product must meet the light-blocking class 1 (light-blocking rate of 99.99% or more) according to JIS L 1055. (2) Tear strength: 15N or more according to JIS K 6328, or 15N or more according to JIS L 1096 Method D. (3) Flame retardancy: The product must have obtained a flame retardancy performance test number issued by the Japan Fire Retardant Association. Light-shielding properties are related to the performance of visual inspection by the resin parts visual inspection device 100. While tear strength and flame retardancy are not related to the performance of visual inspection, it is desirable that these specifications be met as a fire prevention measure in practical operation.

[0033] A resin part appearance inspection method for inspecting the appearance of a resin part 1 using a resin part appearance inspection device 100 will be explained based on Figure 11, which is a flowchart of the resin part appearance inspection method, Figure 12, which is a flowchart of the image processing, and Figures 13A to 13F, which are examples of processed images for resin part appearance inspection. First, the overall process flow of the resin parts visual inspection method will be explained in the flowchart in Figure 11. Using the resin part appearance inspection device 100, the resin part 1 is placed in the part holding mechanism 2, the resin part 1 is illuminated with light from the lighting device 4, and an image of the resin part 1 is captured by the camera 3. By processing this captured image with the image processing device 5, a quantitative value of the appearance quality of the resin part 1 is calculated and output. Note that in Figure 11, the "part holding mechanism" is referred to as the "holding mechanism." The term "holding mechanism" will also be used as appropriate in subsequent drawings.

[0034] The resin part appearance inspection method consists of steps 01 (S01) to 06 (S06). In the resin part installation step (S01), the resin part 1 is installed in the part holding mechanism 2. In the position adjustment step (S02), the distance and angle of the camera 3 and lighting 4 relative to the resin part 1 are adjusted. In the shooting step (S03), the lighting 4 illuminates the resin part 1 with light, and the camera 3 captures an image of the resin part 1. In the image acquisition step (S04), the image of the resin part 1 captured by the camera 3 is imported into the image processing device 5. In the image processing step (S05), the captured image is subjected to image processing, which will be explained in detail below, to calculate a quantitative value of the appearance quality of the resin part 1. In the quantitative value output step (S06), the quantitative value of the appearance quality of the resin part 1, calculated by processing the image in the image processing step (S05), is output. The image processing device 5 stores this data in its internal memory and displays it on the display device. At this time, it is also possible to output it to an external higher-level system, such as a data acquisition device.

[0035] Next, the specific processing details of the image processing step (S05) will be explained using the flowchart in Figure 12. The image processing steps (S05) from the time the image processing device 5 acquires an image in the image acquisition step (S04) until it outputs a quantitative value of the appearance quality in the quantitative value output step (S06) consist of steps 51 (S51) to 56 (S56).

[0036] In the trimming step (S51), since the image also shows parts of the resin part 1 that are not subject to visual inspection, as well as a part of the resin part visual inspection device 100, the defective part 1F of the resin part 1 is trimmed.

[0037] From the smoothing step (S52) to the difference image creation step 53 (S53), a process is performed to correct the luminance distribution according to the distance from the reflection of illumination 4 (hereinafter referred to as the luminance distribution due to illumination 4).

[0038] In the smoothing step (S52), since the brightness distribution due to the defective part 1F of the resin part 1 occurs in a narrower area than the brightness distribution due to illumination 4, the image is smoothed to temporarily eliminate the brightness distribution due to the defective part 1F of the resin part 1, leaving only the brightness distribution due to illumination 4.

[0039] In the difference image creation step (S53), the difference between the image from the trimming step (S51) and the image from the smoothing step (S52) is taken, thereby eliminating only the brightness distribution due to illumination 4 and leaving the brightness distribution due to the defective part 1F of the resin part 1. When calculating the difference, since brightness is represented by a numerical value within the range of 256 levels from 0 to 255, we use equation (1) to keep the brightness value within the range of 0 to 255. [(Image from Step 51 - Image from Step 52) / 2] + 128... Formula (1) Here, since the image from the trimming step (S51) and the image from the smoothing step (S52) have similar brightness levels, the average brightness of the image from the difference image creation step (S53) is approximately 128.

[0040] In the binarization step (S54), the brightness distribution due to illumination 4 is corrected and then binarized to reduce the brightness gradation and make it easier to derive quantitative values. The binarization threshold is set to 128, which is around the average brightness of the image in the difference image creation step (S53). Because the defective portion 1F has a higher brightness than the good portion, the binarized image shows white pixels in the defective portion 1F joined together, forming a white pixel region with a large outline. On the other hand, in the good portion, when the difference is taken in the difference image creation step (S53), the regions that slightly exceed a brightness of 128 and the regions that do not exceed it are processed by binarization in the binarization step (S54), thus forming a white pixel region with a small outline.

[0041] Alternatively, instead of performing the smoothing step (S52), the difference image creation step (S53), and the binarization step (S54) consecutively, the process can be combined into a single step by using "adaptive thresholding," which determines a threshold according to the brightness value within the filter of a given pixel while performing the binarization process. Adaptive thresholding has the advantage of reducing computational load by grouping steps. However, it has the disadvantage of being prone to arbitrary setting of filter sizes because there are no clear criteria for determining the filter size. On the other hand, the method of dividing the process into a smoothing step (S52), a difference image creation step (S53), and a binarization step (S54) has the advantage that the criteria for determining the filter size are clear, as the filter size is set to eliminate any imperfections in the appearance of the output image by looking at the output image from the smoothing step (S52), making it less likely to be set arbitrarily.

[0042] In the contour extraction step (S55), the contour of the defective portion 1F is detected, and only contours with an area greater than or equal to a reference area are extracted, thereby extracting only the defective portion 1F.

[0043] In the quantitative value calculation step (S56), since the defective parts 1F were converted to white pixels and the good parts to black pixels in the processing up to the contour extraction step (S55), the white pixel ratio, that is, the area ratio of the defective parts to the entire image, is calculated using equation (2), and this is taken as a quantitative value representing the appearance quality of the resin part 1. Percentage of white pixels = Number of white pixels / Total number of pixels ... Equation (2)

[0044] By calculating a quantitative value of the appearance quality of resin part 1, it is possible to separate the characteristic that the area around the reflection of lighting 4 becomes darker as it moves away from the reflection of lighting 4, and the characteristic that the defective part 1F is brighter than the area around the defective part 1F.

[0045] Next, examples of processed images at each step of the image processing performed by the image processing device 5 are explained in Figures 13A to 13F. The image in Figure 13A is the image before image processing, captured by the image processing device 5 in the image acquisition step (S04). The image in Figure 13B is the image processed in the trimming step (S51). The image in Figure 13C is the image processed in the smoothing step (S52). The image in Figure 13D is the image processed in the difference image creation step (S53). The image in Figure 13E is the image processed in the binarization step (S54). The image in Figure 13F is the image processed in the contour extraction step (S55).

[0046] Next, two types of cosmetic defects are explained in Figures 14 and 15, showing examples of images before and after image processing by the image processing device 5. Figures 14 and 15 show the original image before and after image processing, as well as quantitative values ​​of appearance quality, for resin part 1 when it is a good product and when it is a defective product. In the original image, the dotted square in the upper right corner indicates the area that was trimmed in the trimming step (S51).

[0047] Figure 14 shows a surface defect in injection molding called "tiger stripes," which consists of multiple streaks. Figure 15 shows a cosmetic defect called a flow mark in the field of injection molding, which is an arc-shaped or fan-shaped pattern that occurs around the resin inlet. In both Figures 14 and 15, defective areas can be detected as white pixels, and the quantitative value (percentage of white pixels) is higher for defective products. By setting a threshold between the quantitative values ​​of defective and good products, it is possible to determine whether a product is good or bad.

[0048] This section summarizes the resin part appearance inspection method using the resin part appearance inspection device 100, focusing on the image processing performed by the image processing device 5. This resin part appearance inspection method is performed using a part holding mechanism 2 that holds the resin part 1 to be inspected, a camera 3 that takes an image of the resin part 1, an illumination device 4 that irradiates the resin part 1 with illumination light, and an image processing device 5 that processes the image.

[0049] The resin part appearance inspection method consists of steps 04 (S04) and steps 51 (S51) to 56 (S56). In the image acquisition step (S04), the image captured by camera 3 is imported into the image processing device 5. In the trimming step (S51), the defective portion 1F of the resin part 1 is trimmed. In the smoothing step (S52), the image is smoothed. In the difference image creation step (S53), an image is created by taking the difference between the image from the trimming step and the image from the smoothing step. In the binarization step (S54), the image from the difference image creation step is binarized. In the contour extraction step (S55), only the contour of the defective portion 1F is extracted. In the quantitative value calculation step (S56), the pixel ratio of the defective part 1F, i.e., the white pixel ratio, is calculated.

[0050] In the resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 1, quantitative values ​​are calculated by using image processing such as smoothing and contour detection to make defective parts white pixels and good parts black pixels, without using machine learning. For this reason, only a minimum of two images (limit samples of good and defective products) are required for resin part appearance inspection.

[0051] Furthermore, the resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 1 utilize the difference in the angular distribution of reflected light intensity between defective and good parts, making it possible to detect flow marks, tiger stripes, etc., as differences in brightness. This allows for the quantification of appearance defects that are not caused by macroscopic irregularities.

[0052] As described above, the resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 1 require a small number of images for inspection, have low contrast, and can quantify appearance defects that are not caused by macroscopic irregularities.

[0053] Embodiment 2. The resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 2 are equipped with a drive device that moves the part holding mechanism, so that even long resin parts can be moved and the appearance of the entire resin part can be inspected.

[0054] The resin parts appearance inspection apparatus and resin parts appearance inspection method of Embodiment 2 will be described, focusing on the differences from Embodiment 1, based on Figure 16, which is a configuration diagram of the resin parts appearance inspection apparatus, and Figure 17, which is a flowchart of the resin parts appearance inspection method. In the drawings of Embodiment 2, parts that are the same as or equivalent to those in Embodiment 1 are denoted by the same reference numerals. To distinguish it from Embodiment 1, it is referred to as the resin parts appearance inspection device 200.

[0055] The overall configuration of the resin parts appearance inspection apparatus 200 of Embodiment 2 will be described based on Figure 16, which is a configuration diagram of the inspection apparatus. The resin part appearance inspection device 200 comprises, as its main components, a resin part 1 to be inspected, a part holding mechanism 2 to hold the resin part 1, a camera 3 and lighting 4 for photographing the resin part 1, an image processing device 5 for processing the captured images, a frame 10 which is the framework of the entire inspection device, and a drive device 25 for moving the part holding mechanism 2. The difference from the resin part appearance inspection apparatus 100 of Embodiment 1 is the addition of a drive device 25 equipped with a motor, etc., for horizontally moving the part holding mechanism 2 on which the resin part 1 is placed. Note that the image processing device 5 and the blackout curtain 11 are omitted in Figure 16.

[0056] The resin part 1 to be inspected is flat, just like in Embodiment 1. The structure of the component holding mechanism 2, the camera 3, and the specifications of the lighting 4 are the same as in Embodiment 1. The method for setting the relative positions of resin part 1, camera 3, and lighting 4 is the same as in Embodiment 1.

[0057] In Embodiment 1, it was assumed that the appearance of the resin part 1 would be inspected in advance by visual inspection to identify areas where cosmetic defects might occur, and then inspected for the presence or absence of these defective areas. Therefore, when performing the appearance inspection, the position of the resin part 1 relative to the lighting 4 and camera 3 was adjusted so that the previously identified defective areas could be inspected with high accuracy. In the resin part appearance inspection apparatus 200 of Embodiment 2, such prior appearance inspection is unnecessary, and even for long resin parts 1, the entire appearance of the resin part 1 can be inspected by moving the resin part 1 and taking multiple images. However, adjustment of the position and angle of the lighting 4 and camera 3, as well as the cropping position, is necessary, and the entire appearance of the resin part 1 is achieved by moving only the resin part 1 horizontally while maintaining these positional relationships.

[0058] A resin part 1 is placed in the part holding mechanism 2, illumination light is shone on the resin part 1 by the lighting device 4, and an image of the resin part 1 is taken by the camera 3. After the part holding mechanism 2 with the resin part 1 placed in it is moved a predetermined distance by the drive device 25, the next part of the resin part 1 is photographed by the camera 3. By repeating this operation, an image of the entire resin part 1 is taken, and the image processing device 5 performs the image processing described in Figure 11 on each of the captured images. As a result, the appearance of the entire resin part 1 can be inspected.

[0059] The resin part appearance inspection method of Embodiment 2 consists of steps 21 (S21) to 28 (S28). In the resin part installation step (S21), the resin part 1 is installed in the part holding mechanism 2. In the position adjustment step (S22), the distance and angle of the camera 3 and lighting 4 relative to the resin part 1 are adjusted. In the shooting step (S23), the illumination light 4 is used to illuminate the resin part 1, and the camera 3 is used to photograph the resin part 1. In the completion determination step (S24), it is determined whether the entire resin part 1 has been photographed. If the entire resin part 1 has been photographed, the process proceeds to the next image acquisition step (S26); otherwise, the process proceeds to the resin part movement step (S25). In the resin part movement step (S25), the drive unit 25 moves the part holding mechanism 2 (i.e., the resin part 1) a predetermined distance. After the resin part movement step (S25), the process returns to the shooting step (S23). In the image acquisition step (S26), the image of the resin part 1 captured by the camera 3 is imported into the image processing device 5. In the image processing step (S27), the image processing described in detail earlier is performed on each captured image to calculate a quantitative value of the appearance quality of the resin part 1. In the quantitative value output step (S28), the quantitative value of the appearance quality of the resin part 1, which was calculated by processing each image in the image processing step (S27), is output. The image processing device 5 stores this data in its internal memory and also displays it on the display device. At this time, it is also possible to output it to an external higher-level system, such as a data acquisition device.

[0060] The predetermined distance over which the resin part 1 is moved in the resin part movement step (S24) is approximately the same as the length in the sliding direction of the defective part 1F that can be captured in a single image. This sliding distance (i.e., the distance over which the defective part 1F can be detected in a single image) is approximately 10 cm to 15 cm. Furthermore, by labeling each image when it is taken at various locations on the resin part 1, it becomes easier to identify the location of the defect if the image processing device 5 determines that the appearance quality inspection result is defective. Furthermore, by setting a threshold value between the quantitative values ​​of a defective product and a good product for each position (i.e., each image) of the resin part 1 in which the image was taken, appropriate quality determination can be made. For example, if there are three defective parts 1F on the resin part 1, an appropriate threshold value can be set for each defective part.

[0061] As described above, the resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 2 are equipped with a drive device that moves the part holding mechanism, so that even long resin parts can be moved and the appearance of the entire resin part can be inspected. Therefore, the resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 2 require fewer images for inspection, have low contrast, and can quantify appearance defects that are not caused by macroscopic irregularities. Furthermore, it can inspect the entire resin part.

[0062] Embodiment 3. The resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 3 are equipped with multiple cameras and multiple lights, and by switching between the multiple cameras and lights, the overall appearance inspection of long resin parts can be performed.

[0063] The resin parts appearance inspection apparatus and resin parts appearance inspection method of Embodiment 3 will be described, focusing on the differences from Embodiment 1, based on Figure 18, which is a configuration diagram of the resin parts appearance inspection apparatus, and Figure 19, which is a flowchart of the resin parts appearance inspection method. In the drawings of Embodiment 3, parts that are the same as or corresponding to those in Embodiment 1 are denoted by the same reference numerals. To distinguish it from Embodiment 1, it is referred to as the resin parts appearance inspection device 300.

[0064] The overall configuration of the resin parts appearance inspection apparatus 300 of Embodiment 3 will be described based on Figure 18, which is a configuration diagram of the inspection apparatus. The resin part appearance inspection device 300 comprises, as its main components, a resin part 1 to be inspected, a part holding mechanism 2 to hold the resin part 1, two cameras 31 and 32 for capturing images of the resin part 1, two lights 41 and 42, an image processing device 5 for processing the captured images, and a frame 10 which is the framework of the entire inspection device. The difference from the resin parts appearance inspection apparatus 100 of Embodiment 1 is that it is equipped with two cameras 31 and 32 and two lights 41 and 42. Note that the blackout curtain 11 is omitted in Figure 18. In Embodiment 3, an example is described in which two cameras and two lights are provided, but three or more cameras and lights may be provided as appropriate to match the dimensions of the resin part 1.

[0065] The resin part 1 to be inspected is flat, just like in Embodiment 1. The structure of the component holding mechanism 2, the specifications of the camera 3, and the lighting 4 are the same as in Embodiment 1. Furthermore, the method for setting the relative positions of the resin component 1, camera 3, and lighting 4 is also the same as in Embodiment 1.

[0066] In Embodiment 1, it was assumed that the appearance of the resin part 1 would be inspected in advance by visual inspection to identify areas where cosmetic defects might occur, and then inspected for the presence or absence of these defective areas. Therefore, when performing the appearance inspection, the position of the resin part 1 relative to the lighting 4 and camera 3 was adjusted so that the previously identified defective areas could be inspected with high accuracy. In the resin part appearance inspection apparatus 300 of Embodiment 3, such prior appearance inspection is unnecessary. Even for long resin parts 1, the entire appearance of the resin part 1 can be inspected by switching between multiple cameras and multiple lights according to the dimensions of the resin part 1 and taking multiple images. However, adjustment of the position and angle of the lights and cameras, as well as the cropping position, is necessary.

[0067] The resin part 1 is placed in the part holding mechanism 2, the first part of the resin part 1 is illuminated with light from the lighting 41, and an image of the first part of the resin part 1 is taken with the camera 31. Next, the camera is switched to camera 32 and the lighting to lighting 42, and the next part of the resin part 1 is illuminated with light from lighting 41, and an image of the next part of the resin part 1 is taken with camera 32. If there are three or more cameras and lights, this operation is repeated to capture images of the entire resin part 1, and the image processing device 5 performs the image processing described in Figure 11 on each captured image. As a result, the appearance of the entire resin part 1 can be inspected.

[0068] The resin part appearance inspection method of Embodiment 3 consists of steps 31 (S31) to 38 (S38). In the resin part installation step (S31), the resin part 1 is installed in the part holding mechanism 2. In the position adjustment step (S32), the distance and angle of the cameras 31, 32 and lights 41, 42 relative to the resin part 1 are adjusted. In the shooting step (S33), the resin part 1 is illuminated with light from the lighting 41, and an image of the resin part 1 is taken with the camera 31. In the completion determination step (S34), it is determined whether the entire resin part 1 has been photographed. If the entire resin part 1 has been photographed, the process proceeds to the next image acquisition step (S36); otherwise, the process proceeds to the camera and lighting switching step (S35). In the camera and lighting switching step (S35), the camera is switched to camera 32 and the lighting to lighting 42. After the camera and lighting switching step (S35), the process returns to the shooting step (S33). In the image acquisition step (S36), images of the resin part 1 captured by cameras 31 and 32 are imported into the image processing device 5. In the image processing step (S37), the image processing described in detail earlier is performed on each captured image to calculate a quantitative value of the appearance quality of the resin part 1. In the quantitative value output step (S38), the quantitative value of the appearance quality of the resin part 1, which was calculated by processing each image in the image processing step (S37), is output. The image processing device 5 stores this data in its internal memory and displays it on the display device. At this time, it is also possible to output it to an external higher-level system, such as a data acquisition device.

[0069] For example, the distance that can be captured in a single image using camera 31 and lighting 41 (i.e., the distance at which a defective portion 1F can be detected in a single shot) is approximately 10cm to 15cm. By labeling each image when photographing the resin part 1, it becomes easier to identify the location of the defect if the image processing device 5 determines that the appearance quality inspection result is defective. Furthermore, by setting a threshold value between the quantitative values ​​of a defective product and a good product for each position (i.e., each image) of the resin part 1 in which the image was taken, it is possible to make an appropriate judgment of whether the product is good or bad.

[0070] As described above, the resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 3 are equipped with multiple cameras and multiple lights, and by switching between the multiple cameras and lights, the overall appearance inspection of long resin parts can be performed. Therefore, the resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 3 require fewer images for inspection, have low contrast, and can quantify appearance defects that are not caused by macroscopic irregularities. Furthermore, it can inspect the entire resin part.

[0071] Embodiment 4. The resin part appearance inspection apparatus and resin part appearance inspection method of Embodiment 4 use a mold of an injection molding apparatus instead of the part holding mechanism used in the resin part appearance inspection apparatus of Embodiments 1 to 3.

[0072] The resin parts appearance inspection apparatus and resin parts appearance inspection method of Embodiment 4 will be explained, focusing on the differences from Embodiment 1, based on Figures 20 and 21, which are system configuration diagrams of the resin parts appearance inspection apparatus; Figure 22, which is a functional block diagram of each device related to the injection molding process; Figures 23A, 23B, 24A, and 24B, which are explanatory diagrams of the injection molding process; and Figures 25 and 26, which are flowcharts of the resin parts appearance inspection method. In the drawings of Embodiment 4, parts that are the same as or equivalent to those in Embodiment 1 are denoted by the same reference numerals. However, to distinguish it from Embodiment 1, the resin part appearance inspection device 400, resin part 401, camera 403, lighting 404, image processing device 405, and blackout curtain 411 are used. Furthermore, in Embodiments 1 to 3, the object of inspection, which is a molded product formed by an injection molding machine, was described as a resin part. Although parts formed by the mold of an injection molding machine are sometimes referred to as molded products, in order to maintain consistency with the explanations in Embodiments 1 to 3, the term "resin part" will also be used in Embodiment 4.

[0073] The system configuration of the resin parts appearance inspection apparatus 400 of Embodiment 4 will be described with reference to Figures 20 and 21. The resin parts appearance inspection device 400 is an injection molding machine 60 to which necessary equipment and devices for visual inspection of resin parts 401, such as a camera 403, lighting 404, a blackout curtain unit 82, and an image processing device 405, are arranged. Figure 20 is a side view of the resin parts visual inspection device 400, where the left-right direction in the figure is the X direction and the up-down direction is the Z direction. The Z-direction + side (upper side of the figure) is the top side and the Z-direction - side (lower side of the figure) is the ground side. Figure 21 is a top view of the resin parts visual inspection device 400, where the left-right direction in the figure is the X direction and the up-down direction is the Y direction. The Y-direction - side (bottom of the figure) is the operating side, and the Y-direction + side (top of the figure) is the non-operating side. In Figure 20, symbol K1 indicates the movement of the removal machine 70 as injection molding and visual inspection of the resin part 401 progresses, and symbol K2 indicates the movement of the blackout curtain unit 82. Furthermore, in Figures 20 and 21, the clamping unit holding section 65 is referred to as the unit holding section 65, and the clamping unit wall surface 75 is referred to as the unit wall surface 75.

[0074] First, the configuration and function of the injection molding apparatus 60 will be explained. The injection molding apparatus 60 consists of a resin injection unit 61, a clamping unit 62, and a mold 63. These devices are mounted on a base 68. The injection molding apparatus 60 also includes a removal machine 70 for removing the molded resin part 401. The resin injection device 61 is a device for heating and melting plastic material and injecting it into the mold 63 during the injection molding process. One of the main components of the resin injection device 61 is the cylinder 61C, which controls the supply and injection of plastic material. The resin injection device 61 is equipped with a hopper 61H for material input at its top. The mold clamping unit 62 is a device for holding and opening / closing the mold 63 used for molding in the injection molding process, and consists of a fixed platen 64F, a movable platen 64M, a mold clamping unit holding section 65, a crosshead 66, and a tie bar 67. When referring to both the fixed-side platen 64F and the movable-side platen 64M collectively, they are simply referred to as platen 64. The platen 64 is a component that holds the mold 63, which consists of a fixed mold 63F and a movable mold 63M, and transmits the clamping force. The platen 64 is often manufactured by casting and is designed not to deform under the clamping force. When referring to both the fixed mold 63F and the movable mold 63M collectively, they are simply referred to as mold 63. As the crosshead 66 moves, the mold 63 opens and closes, and the tie bar 67 elastically deforms, generating a high-pressure clamping force. The clamping unit holding part 65 holds the tie bar 67 and holds and drives the crosshead 66.

[0075] Safety devices include a safety door 73, a safety cover 74, and a clamping unit wall 75. The safety door 73 is a device for protecting the working area of ​​the injection molding machine 60. Because the inside of the injection molding machine 60 is subject to high temperature and pressure, the safety door 73 is installed to ensure the safety of the operator. A transparent material is used in part of the safety door 73, allowing observation of the mold 63. The safety cover 74 is installed below the mold 63 to protect the working area of ​​the injection molding machine 60. The safety cover 74 is generally made of sheet metal and does not allow the contents inside to be seen. The clamping unit wall 75 is installed for safety and dust protection and is often made of sheet metal.

[0076] The control panel 71 is a device for operating and controlling the injection molding apparatus 60, and performs temperature control, injection control, and mold control. An extraction machine 70 is an essential piece of equipment for the injection molding process. The extraction machine 70 removes the resin parts 401 after visual inspection and moves them to designated locations (good parts storage area, defective parts storage area). The blackout curtain unit 82 is equipped with a moving device 83 because it needs to move in accordance with the progress of the injection molding process.

[0077] Next, we will explain the resin part appearance inspection device 400, focusing on the method for inspecting the appearance of the resin part 401. The camera 403 and lighting 404 are fixed to either the fixed mold 63F or the movable mold 63M. Figures 20 and 21 show an example where the ejection mechanism is installed in the movable mold 63M (i.e., the resin part 401 is ejected on the movable side), and the camera 403 and lighting 404 are fixed to the fixed mold 63F. When the resin part 401 is ejected on the fixed side, it is preferable to fix it to the movable mold 63M. Furthermore, Figures 20 and 21 show an example where the cooling pipes 69 of the mold 63 extend in the operating and non-operating directions, and the camera 3 and lighting 4 are fixed to the top and bottom sides of the mold 63. When the cooling pipes 69 extend in the top and bottom directions, it is desirable to fix the camera 3 and lighting 4 to the operating and non-operating sides of the mold 63. Furthermore, while Figures 20 and 21 illustrate the case where the camera 403 and lighting 404 are fixed to either the fixed mold 63F or the movable mold 63M, the camera 403 and lighting 404 can also be installed on either the fixed platen 64F or the movable platen 64M.

[0078] The blackout curtain unit 82 consists of a blackout curtain 411 made up of light-blocking parts and support parts, a moving device 83 made up of casters, a motor, etc., and a control device 443 that controls the moving device 83. The blackout curtain 411 blocks ambient light in the same way as the blackout curtain 11 in Figure 1 in Embodiment 1. In Embodiment 4, in order to avoid interference with the movement of the resin part removal machine 70 for the resin part 401, the blackout curtain unit 82 is moved out of the way before the removal machine 70 approaches the inspected resin part 401.

[0079] The functional blocks of the resin parts appearance inspection apparatus 400 of Embodiment 4 will be described based on Figure 22. Note that Figure 22 represents the functions of each device of the resin parts appearance inspection apparatus 400 as blocks and does not correspond one-to-one with the hardware in the system configuration diagrams of Figures 20 and 21. In Figure 22, symbol C1 represents the "mold opening completion signal," symbol C2 represents the "injection completion signal," symbol C3 represents "good / defective product information," symbol C4 represents "image data," and symbol C5 represents the "shooting completion signal."

[0080] The molding apparatus 420 consists of a mold clamping device 421, an injection device 422, and a control device 423. The clamping device 421 opens and closes the mold 63, and the injection device 422 heats and melts the resin material and injects it into the mold 63. The control device 423 controls the mold clamping device 421 and the injection device 422.

[0081] The extraction device 430 consists of a chucking device 431, a moving device 432, and a control device 433. The chucking device 431 chucking the resin part 401 by air suction or the like. The moving device 432 moves the removal machine 70 to the vicinity of the mold 63 after the mold 63 has been opened, or moves the chucked resin parts 401 to the good / defective product storage area. When a mold opening completion signal (C1) is received from the molding apparatus 420, the resin part 401 is moved to either the good product area or the defective product area based on the good / defective product information (C3) from the image processing apparatus 405. The control device 433 controls the chucking device 431 and the moving device 432.

[0082] The blackout curtain unit device 440 consists of a moving device 441, a blackout curtain 411, and a control device 443. The blackout curtain 411 blocks ambient light in the same way as the blackout curtain 11 in Figure 1 in Embodiment 1. When the injection start signal (C2) is input from the molding apparatus 420, the moving device 441 moves the blackout curtain 411 near the mold 63 based on the control of the control device 443. Furthermore, when a shooting completion signal (C5) is received from the shooting device 450, the moving device 441 moves the blackout curtain 411 outside the mold 63 based on the control of the control device 443.

[0083] The imaging device 450 consists of a camera 403, a lighting 404, and a control device 453. The functions of the camera 403 and lighting 404 are the same as in Embodiment 1. The control device 453 controls the camera 403 and the lighting 404. When the mold opening signal (C1) is input from the molding apparatus 420, the control device 453 turns on the lighting 404, takes off the shutter of the camera 403, outputs image data (C4) to the image processing apparatus 405, and turns off the lighting 404.

[0084] The image processing device 405 has the same functions as the image processing device 5 of Embodiment 1. When the image processing device 405 receives image data (C4) from the imaging device 450, it performs image processing in the same manner as in Embodiment 1 and outputs good / defective product information (C3) to the molding device 420 and the removal device 430.

[0085] Next, the operation of each device in Embodiment 4, that is, the operation of each device from resin filling to molded product removal, will be explained based on Figures 23A, 23B, 24A, and 24B. In Figures 23A, 23B, 24A, and 24B, the cooling pipes 69 are installed in the vertical direction, with the camera 403 installed on the non-operating side of the fixed mold 63F and the lighting 404 installed on the operating side of the fixed mold 63F. (1) In Figure 23A, when resin filling begins, the blackout curtain 411 is moved near the mold 63 to prepare it so that ambient light can be blocked when taking pictures with the camera 403. (2) In Figure 23B, once the resin filling is complete, the mold 63 is opened, the necessary illumination light is shone on the resin part 401 with the light 4, and a photograph is taken with the camera 403. (3) In Figure 24A, after the photograph is taken, the blackout curtain 411 is moved to the outside of the mold 63 to prevent interference when the resin part 401 is removed by the removal machine 70. (4) In Figure 24B, after the blackout curtain 411 is retracted, the resin part 401 is removed by the removal machine 70. By repeating the above steps (1) through (4), continuous molding can be performed while inspecting the appearance quality of all resin parts 401. Figure 24B shows the extraction machine 70 moving in the direction opposite to the operation side of the fixed mold 63F, but it also includes a configuration in which it moves upwards relative to the fixed mold 63F.

[0086] Next, a flowchart of the resin part appearance inspection method of Embodiment 4 will be explained based on Figures 25 and 26. In practice, the resin part appearance inspection method of Embodiment 4 is a resin part appearance inspection method that uses an injection molding apparatus 60 to perform injection molding of the resin part 401 to be inspected to inspect the appearance quality of the resin part 401, and is performed using a camera 403 to photograph the resin part 401 and a light 404 to illuminate the resin part 401 installed in the mold 63 or platen 64 used for resin molding of the resin part 401, an image processing device 405 that performs calculation processing on the image captured by the camera 403, a movable blackout curtain 411 and a removal machine 70 that removes and moves the resin part 401 from the mold 63.

[0087] The method for inspecting the appearance of resin parts consists of steps 61 (S61) to 76 (S76). In the position adjustment step (S61), the positions of the camera 403 and the lighting 404 are adjusted. In the continuous molding start step (S62), the continuous molding of the resin part 401 is started. In the resin filling step (S63), resin is filled into the mold 63 of the resin injection device 61. In the first blackout curtain movement step (S64), the blackout curtain 411 is moved to the vicinity of the mold 63. In the mold opening step (S65), the mold 63 is opened after the resin filling step (S63) is completed. In the shooting step (S66), the illumination 404 is used to illuminate the resin part 401, and the camera 403 is used to capture an image of the resin part 401. In the second blackout curtain movement step (S67), the blackout curtain 411 is moved to the outside of the mold 63. In the chucking step (S68), the removal machine 70 is moved to the vicinity of the resin part 401, and the resin part 401 is chucking. In the image acquisition step (S69), the image of the resin part 401 captured by the camera 403 is imported into the image processing device 405. In the image processing step (S70), the image processing described in Embodiment 1 is performed on the captured image to calculate a quantitative value of the appearance quality of the resin part 401. In the quantitative value output step (S71), the quantitative value of the appearance quality of the resin part 401, which was calculated in the image processing step (S70), is output. The image processing device 405 stores this data in its internal memory and also displays it on the display device. In the quality determination step (S72), the appearance quality of the resin part 401 output in the quantitative value output step (S71) is compared with the pre-set required quality. It is determined that the required quality is satisfied. Note that other quality items and different evaluation methods may be used during the quality determination and included in the pass / fail determination. In the good product processing step (S73), if the required quality is determined to be met in the quality judgment step (S72), the removal machine 70 moves the resin part 401 to the good product storage area. After that, the process returns to the resin filling step (S63) to enter the next molding cycle. In the defective product handling step (S74), if it is determined in the quality determination step (S72) that the required quality is not met, the removal machine 70 moves the resin part 401 to the defective product storage area. In the defect frequency determination step (S75), it is determined whether defective products have occurred consecutively for a predetermined threshold number of shots (N shots) or more. If no defective products have occurred consecutively for N shots or more, the process returns to the resin filling step (S63) to proceed to the next molding cycle. If N or more consecutive shots occur, the process proceeds to the alert output step (S76). In the alert output step (S76), a continuous defect alert is output, and the continuous molding process is stopped.

[0088] By processing the resin parts visually inspected using the method described above, consisting of steps 61 (S61) to 76 (S76), continuous molding can be performed while inspecting the visual quality of all resin parts 401. If a defective product is produced, it is discarded in the defective product area, and if defective products are produced continuously, continuous molding can be stopped. If continuous molding of resin parts is stopped due to a series of defects, the molding conditions are adjusted so that the resin parts 401 become good products. In addition, it is checked whether the lot of the resin material has changed, the condition of the mold 63 is checked, and changes in environmental factors such as ambient temperature and humidity are checked. By taking the necessary measures for the cause of the defects, continuous molding of resin parts 401 can be resumed.

[0089] The effects of the resin parts appearance inspection apparatus and molded product appearance inspection method of Embodiment 4 will be described. In Embodiment 4, the mold of the injection molding apparatus can be used as a resin part appearance inspection device, eliminating the need to prepare a separate resin part appearance inspection device. Furthermore, it is possible to ensure the reproducibility of the holding position for resin parts 401 with warping and resin parts 401 with protrusions including ribs. Furthermore, in the overall flowchart shown in Figure 11 of Embodiment 1, step S01 "Installing resin parts in the holding mechanism" becomes unnecessary, reducing the man-hours required for visual inspection. In addition, the appearance of all molded products during continuous molding can be automatically inspected in-line, allowing defective products to be sorted and automatically discarded by a removal machine. Moreover, by issuing an alert and stopping continuous molding when defects occur consecutively, monitoring can be automated in the mass production of injection molded products.

[0090] In the resin parts appearance inspection apparatus and molded product appearance inspection method of Embodiment 4, it is assumed that the areas where appearance defects may occur have been identified in advance. However, by installing multiple lights and cameras as in Embodiment 3, it becomes possible to detect defects at multiple locations. Furthermore, by controlling the angles of the camera 403 and the lighting 404 relative to the resin part 401, detection accuracy can be ensured within an acceptable range, allowing for visual inspection of the entire resin part 401.

[0091] As described above, the resin part appearance inspection apparatus and molded product appearance inspection method of Embodiment 4 use a mold of an injection molding apparatus instead of the part holding mechanism used in the resin part appearance inspection apparatus of Embodiments 1 to 3. Therefore, the resin part appearance inspection apparatus and molded product appearance inspection method of Embodiment 4 require fewer images for inspection, have low contrast, and can quantify appearance defects that are not caused by macroscopic irregularities. Furthermore, there is no need to prepare a separate resin part appearance inspection apparatus.

[0092] Here, an example of the hardware of the image processing apparatus 5, the control device 423 of the molding apparatus 420, the control device 433 of the extraction apparatus 430, the control device 443 of the blackout curtain unit apparatus 440, the control device 453 of the imaging apparatus 450, and the image processing apparatus 405 from Embodiments 1 to 4 will be described based on Figure 27. The image processing device 5, control devices 423, 433, 443, 453, and image processing device 405 are composed of a processor 1000 and a storage device 1001. The storage device 1001 includes a volatile storage device such as random access memory (not shown) and a non-volatile auxiliary storage device such as flash memory. Alternatively, a hard disk may be provided as an auxiliary storage device instead of flash memory. The processor 1000 executes a program input from the storage device 1001. In this case, the program is input to the processor 1000 from the auxiliary storage device via a volatile storage device. The processor 1000 may also output data such as calculation results to the volatile storage device of the storage device 1001, or it may save the data to the auxiliary storage device via the volatile storage device.

[0093] While this disclosure describes various exemplary embodiments and examples, the various features, aspects, and functions described in one or more embodiments are not limited to the application of a particular embodiment and can be applied individually or in various combinations to the embodiments. Accordingly, countless variations not illustrated are conceivable within the scope of the art disclosed in this specification. These include, for example, modifying, adding or omitting at least one component, or even extracting at least one component and combining it with components of other embodiments.

[0094] The various aspects of this disclosure are summarized below as an appendix.

[0095] (Note 1) The system comprises a component holding mechanism for holding the resin component to be inspected, a camera for photographing the resin component, a light source for illuminating the resin component, and an image processing device for performing calculations on the image captured by the camera. The image processing device is a resin part appearance inspection device that performs calculations on images captured by the camera to detect defective parts on the surface of the resin part and quantifies the appearance quality of the resin part. (Note 2) The image processing apparatus is a resin part appearance inspection apparatus as described in Appendix 1, which performs smoothing and difference acquisition, binarization and contour extraction of the defective portion, calculation of the pixel ratio of the defective portion, and calculation of a quantitative value of the appearance quality on an image captured by the camera. (Note 3) When the angle between the illumination and the defective portion of the resin part is defined as the incident angle, and the angle between the camera and the defective portion of the resin part is defined as the detection angle, The resin part appearance inspection apparatus according to Appendix 1 or Appendix 2, wherein the incident angle is 35 to 55 degrees, and the detection angle has an angular difference of 3 to 10 degrees with respect to the incident angle. (Note 4) The resin part appearance inspection apparatus according to any one of the appendices 1 to 3, wherein the illumination is bar-type in shape, the beam spread is diffuse, it is a surface-emitting type, it has a highly uniform brightness distribution characteristic in the long axis direction, and the length of the light-emitting part is at least twice the length of the defective part of the resin part. (Note 5) The aforementioned camera is equipped with an area sensor, has an effective pixel count of 1 million pixels or more, and the focus, F-number, shutter speed, and ISO sensitivity are adjustable. This is a resin part appearance inspection apparatus as described in any one of the appendices 1 to 4. (Note 6) Furthermore, it is equipped with a drive device for moving the component holding mechanism, A resin part appearance inspection apparatus according to any one of Appendix 1 to Appendix 5, which moves the resin part with the drive device and repeatedly photographs the resin part with the camera to perform an overall appearance inspection of the resin part. (Note 7) The system includes multiple cameras and multiple lights. A resin part appearance inspection apparatus according to any one of Appendix 1 to Appendix 5, which switches between the multiple cameras and multiple lights, repeatedly photographs the resin part with the cameras, and performs an overall appearance inspection of the resin part with the cameras. (Note 8) The resin part appearance inspection apparatus according to any one of the appendices 1 to 7, wherein if the resin part has warping during manufacturing, the warping has been corrected by the time of appearance inspection. (Note 9) Furthermore, the resin parts appearance inspection apparatus according to any one of the appendices 1 to 8, further comprising a blackout curtain for shielding the camera and the lighting from ambient light. (Note 10) Using a component holding mechanism for holding a resin component to be inspected, a camera for photographing the resin component, a light source for illuminating the resin component, and an image processing device for performing calculations on the image captured by the camera, An image acquisition step in which an image captured by the camera is taken into the image processing device, A trimming step in which defective portions are trimmed from the surface of the resin part, A smoothing step is performed to smooth the image created in the trimming step, A difference image creation step is to create an image obtained by taking the difference between the image created in the trimming step and the image created in the smoothing step, A binarization step in which the image created in the difference image creation step is binarized, A contour extraction step for extracting the contour of the defective portion, A quantitative value calculation step involves calculating and quantifying the pixel ratio of the defective portion. A method for inspecting the appearance of resin parts, comprising the following features. (Note 11) Furthermore, it is equipped with a drive device for moving the component holding mechanism, A resin part moving step for moving the aforementioned resin part, A completion determination step to determine whether the overall inspection of the resin parts has been completed, A method for inspecting the appearance of resin parts as described in Appendix 10, comprising the features described herein. (Note 12) The camera and lighting are provided in multiple units, The camera that switches between the multiple cameras and multiple lights, the lighting switching step, A completion determination step to determine whether the overall inspection of the resin parts has been completed, A method for inspecting the appearance of resin parts as described in Appendix 10, comprising the features described herein. [Explanation of symbols]

[0096] 1,401 Resin parts, 2 Part holding mechanism, 3,31,32,403 Camera, 4,41,42,404 Lighting, 5,405 Image processing device, 10 Frame, 11,411 Blackout curtain, 21 Aluminum plate, 22 Knurled bolt, 23 Magnet, 25 Drive unit, 51 Cable, 100,200,300,400 Resin parts appearance inspection device, 60 Injection molding device, 61 Resin injection device, 61C Cylinder, 61H Hopper, 62 Clamping unit, 63 Mold, 63F Fixed side mold, 63M Movable side mold, 64 Platen, 64F Fixed side platen, 64M Movable side platen, 65 Clamping unit holder, 66 Crosshead, 67 Tie bar, 68 Base, 69 Cooling piping, 70 Take-out machine, 71 Control panel, 73 Safety door, 74 Safety cover, 75 Clamping unit wall, 82 Blackout curtain unit, 83 Moving device, 420 Molding device, 421 Clamping device, 422 Injection device, 423, 433, 443, 453 Control device, 430 Extraction device, 431 Chucking device, 432 Moving device, 440 Blackout curtain unit device, 441 Moving device, 450 Imaging device, 1000 Processor, 1001 Memory device.

Claims

1. A component holding mechanism for holding the resin component to be inspected, and a camera for photographing the resin component, A light source for illuminating the aforementioned resin component, and an image processing device for performing calculations on the image captured by the camera. Equipped with, The aforementioned resin part is opaque and glossy. The image processing device is a resin part appearance inspection device that performs calculation processing on an image captured by the camera to detect defective parts on the surface of the resin part and quantifies the appearance quality of the resin part, The angle between the illumination and the defective portion of the resin part is defined as the incident angle, and the angle between the camera and the defective portion of the resin part is defined as the detection angle. When the illumination irradiates the glossy resin component with illumination light, the reflection where the angle of incidence and the angle of reflection are equal is defined as specular reflection, and the portion where the appearance of the illumination is reflected and visible on the surface of the resin component is defined as specular reflection. If the surfaces of the good parts and the defective parts are not perfectly mirror-like, and the parts that reflect at angles different from the angle of incidence are considered to be diffuse reflection, By taking advantage of the fact that the angular distribution of specular reflection light intensity and diffuse reflection light intensity differs between the good part and the defective part, The detection angle is set to have an angle difference with respect to the incident angle, thereby removing the specularly reflected portion from the specularly reflected light by trimming, extracting the diffusely reflected portion, and detecting defects in appearance. Apparatus for inspecting the appearance of resin parts.

2. The image processing device performs smoothing and difference acquisition, binarization and contour extraction of the defective portion, calculation of the pixel ratio of the defective portion, and calculation of a quantitative value of the appearance quality on the image captured by the camera, In the area surrounding the reflection of the aforementioned lighting, The aforementioned characteristic is that it becomes darker as you move away from the reflection of the light, The defective portion separates features that are brighter than the surrounding area. The resin part appearance inspection apparatus according to claim 1.

3. The detection angle is set to have an angle difference of 3 to 10 degrees with respect to the incident angle, thereby extracting the diffusely reflected light portion and detecting flow mark defects and tiger stripe defects of the resin part. The resin part appearance inspection apparatus according to claim 1 or claim 2.

4. The resin part appearance inspection apparatus according to claim 1 or claim 2, wherein the illumination is bar-type in shape, the light beam spread is diffuse, it is a surface-emitting type, it has a highly uniform brightness distribution characteristic in the long axis direction, and the length of the light-emitting part is twice or more the length of the defective part of the resin part.

5. The resin part appearance inspection apparatus according to claim 1 or claim 2, wherein the camera is equipped with an area sensor, has an effective pixel count of 1 million pixels or more, and the focus, F-number, shutter speed, and ISO sensitivity are adjustable.

6. Furthermore, it is equipped with a drive device for moving the component holding mechanism, The resin part appearance inspection apparatus according to claim 1 or claim 2, which moves the resin part with the drive device and repeatedly photographs the resin part with the camera to perform an overall appearance inspection of the resin part.

7. The system includes multiple cameras and multiple lights. The resin part appearance inspection apparatus according to claim 1 or claim 2, which switches between the multiple cameras and multiple lights, repeatedly takes pictures of the resin part with the cameras, and performs an overall appearance inspection of the resin part with the cameras.

8. The resin part appearance inspection apparatus according to claim 1 or claim 2, wherein if the resin part has warping during manufacturing, the warping has been corrected by the time of appearance inspection.

9. Furthermore, the resin part appearance inspection apparatus according to claim 1 or claim 2 further comprises a blackout curtain for shielding the camera and the lighting from ambient light.

10. The aforementioned component holding mechanism is It consists of an aluminum plate, a knurled bolt, and a magnet. The resin part is fixed by sandwiching it between the head of the knurled bolt and the magnet, The aforementioned resin component is held suspended from the floor surface, This configuration ensures reproducibility of the holding position for the aforementioned resin part that has warping and the aforementioned resin part that has protrusions including ribs. The resin part appearance inspection apparatus according to claim 1 or claim 2.

11. The aforementioned component holding mechanism is Among the molds used in the resin molding process of the aforementioned resin parts, the fixed side or the movable side where the ejection mechanism exists, This configuration ensures reproducibility of the holding position for the aforementioned resin part that has warping and the aforementioned resin part that has protrusions including ribs. The resin part appearance inspection apparatus according to claim 1 or claim 2.

12. Using a component holding mechanism for holding a resin component to be inspected, a camera for photographing the resin component, a light source for illuminating the resin component, and an image processing device for performing calculations on the image captured by the camera, The angle between the illumination and the defective portion of the resin part is defined as the incident angle, and the angle between the camera and the defective portion of the resin part is defined as the detection angle. When the illumination irradiates the glossy resin component with illumination light, the reflection where the angle of incidence and the angle of reflection are equal is defined as specular reflection, and the portion where the appearance of the illumination is reflected and visible on the surface of the resin component is defined as specular reflection. If the surface of the aforementioned resin part is not a perfect mirror surface, and the portion that reflects at an angle different from the angle of incidence is considered diffuse reflection, The aforementioned detection angle is set to have an angular difference with respect to the incident angle, thereby removing the specularly reflected portion of the specularly reflected light and extracting the diffusely reflected portion from the specularly reflected light by trimming, in a resin part appearance inspection method, An image acquisition step in which an image captured by the camera is taken into the image processing device, A trimming step in which defective portions of the surface of the resin part are trimmed, A smoothing step is performed to smooth the image created in the trimming step, The difference between the image created in the trimming step and the image created in the smoothing step is The steps for creating a difference image by creating the captured image, A binarization step in which the image created in the difference image creation step is binarized, A contour extraction step for extracting the contour of the defective portion, A quantitative value calculation step involves calculating and quantifying the pixel ratio of the defective portion. A method for inspecting the appearance of resin parts, comprising the following features.

13. Furthermore, it is equipped with a drive device for moving the component holding mechanism, A resin part moving step for moving the aforementioned resin part, A completion determination step to determine whether the overall inspection of the resin parts has been completed, A method for inspecting the appearance of a resin part according to claim 12, comprising:

14. The camera and lighting are provided in multiple units, The camera that switches between the multiple cameras and multiple lights, the lighting switching step, A completion determination step to determine whether the overall inspection of the resin parts has been completed, A method for inspecting the appearance of a resin part according to claim 12, comprising:

15. A resin part appearance inspection method is a resin part appearance inspection method that uses an injection molding apparatus to perform injection molding of a resin part to be inspected to inspect the appearance quality of the resin part, This method involves installing a camera for photographing the resin part and a light source for illuminating the resin part in a mold or platen used for resin molding the resin part, using an image processing device for processing the image captured by the camera, a movable blackout curtain, and a removal machine for removing and moving the resin part from the mold. A position adjustment step to adjust the positions of the camera and the lighting, A continuous molding start step in which the continuous molding of the resin part is initiated, A resin filling step of filling the mold with resin, A first blackout curtain moving step involves moving the blackout curtain to the vicinity of the mold, After the resin filling step is completed, the mold is opened in a mold opening step, A shooting step in which the resin part is illuminated with the aforementioned lighting and an image of the resin part is taken with the aforementioned camera, A second blackout curtain moving step involves moving the blackout curtain to the outside of the mold, A chucking step involves moving the extraction machine to the vicinity of the resin part and chucking the resin part, An image acquisition step of taking an image of the resin part captured by the camera and importing it into the image processing device, Image processing step: Performs image processing on the captured image to calculate a quantitative value of the appearance quality of the resin part. A quantitative value output step that outputs a quantitative value of the appearance quality of the resin part calculated in the image processing step, A quality determination step in which the appearance quality of the resin part output in the quantitative value output step is compared with a pre-set required quality and it is determined whether the required quality is satisfied, If the quality determination step determines that the required quality is satisfied, the removal machine moves the resin part to the good product storage area and returns to the resin filling step in a good product processing step. If the quality determination step determines that the good-faith required quality is not met, the removal machine moves the resin part to the defective product storage area in a defective product processing step. A defect frequency determination step that determines whether defective products have occurred consecutively at a pre-set threshold or higher, returns to the resin filling step if they have not occurred consecutively at a threshold or higher, and proceeds to the alert output step if they have occurred consecutively at a threshold or higher, An alert output step that outputs a continuous defect alert and stops the continuous molding process, A method for inspecting the appearance of resin parts, comprising the following features.