KR102429699B1 - Future car headlamp cover lens vision inspection device - Google Patents

Abstract

A futuristic vehicle headlamp cover lens vision inspection apparatus according to the present invention includes a lens vision inspection chamber; a base plate installed inside the lens vision inspection chamber; a lens holder installed on the base plate to hold and fix a lens to be inspected, which is an object of defect analysis, on the base plate; a camera mount installed on the base plate; a camera mounted on the camera mount to photograph a lens of an object to be inspected for defect analysis; a camera mount transfer means for reciprocating the camera mount in the longitudinal direction of the base plate so that the camera can photograph the entire section of the lens of the subject; a lighting means installed on the base plate and emitting light from the lower part of the lens toward the lens; and a center for controlling the camera mount transport means and lighting means, analyzing the image captured by the camera, automatically setting a lens defect inspection area in the captured image, and determining whether the lens to be inspected is defective in the lens defect inspection area control means.

Description

Future car headlamp cover lens vision inspection device

The present invention relates to a futuristic vehicle headlamp cover lens vision inspection apparatus, and is an apparatus for automatically detecting an appearance defect of a vehicle headlamp cover lens using the vision inspection apparatus.

In the current automobile parts industry, automobile headlamp cover lenses are made of polycarbonate, and the polycarbonate has the advantage of being as transparent as glass, moldable in various shapes, and light in weight.

Currently, the inspection method of the vehicle headlamp cover lens made of the polycarbonate depends on the operator's visual inspection. This visual inspection method not only wastes manpower for inspection, but also takes a long inspection time, and There was a problem that the inspection time varies according to the work ability.

In addition, the method for visually inspecting the vehicle headlamp cover lens has a problem that the quality reading deviation for each operator is severe, and thus process improvement is required.

On the other hand, as a prior art of the present invention, a "backlight unit inspection apparatus" of application number "10-2014-0143380" has been filed and registered, and the backlight unit inspection apparatus inspects foreign substances or damaged parts remaining in the backlight unit to detect defects. An invention relates to a backlight unit inspection apparatus for reading whether or not the backlight unit inspection apparatus is composed of a workbench having a flat upper portion, sidewalls formed perpendicular to both sides of the workbench, and a main body having a door at the front; a shuttle unit which is mounted to be rotated on the upper surface of the workbench of the main body and formed with a seating portion so that a plurality of backlight units are seated; a rotating part formed under the shuttle unit to rotate the shuttle unit; a dust collecting unit formed on the shuttle unit to suck and remove foreign substances from the surface of the backlight unit; It is installed on one side of the upper part of the main body, the inspection unit provided with a scan camera that scans the backlight unit seated on the shuttle unit while moving in a linear direction to inspect the defect.

Republic of Korea Patent Registration No. 10-1610316 (2016.04.07)

Therefore, in order to solve the above problem, the present invention is a future vehicle that enables a more accurate and quantitative inspection of the vehicle headlamp lens cover by performing a visual inspection of the exterior defect inspection of the vehicle headlamp lens cover through a vision system. It is an object of the present invention to provide a headlamp cover lens vision inspection device.

A futuristic vehicle headlamp cover lens vision inspection apparatus according to the present invention for achieving the above object includes: a lens vision inspection chamber; a base plate installed inside the lens vision inspection chamber; a lens holder installed on the base plate to hold and fix a lens to be inspected, which is an object of defect analysis, on the base plate; a camera mount installed on the base plate; a camera mounted on the camera mount to photograph a lens of an object to be inspected for defect analysis; a camera mount transfer means for reciprocating the camera mount in the longitudinal direction of the base plate so that the camera can photograph the entire section of the lens of the subject; a lighting means installed on the base plate and emitting light from the lower part of the lens toward the lens; and a center for controlling the camera mount transport means and lighting means, analyzing the image captured by the camera, automatically setting a lens defect inspection area in the captured image, and determining whether the lens to be inspected is defective in the lens defect inspection area control means.

As the effect of the futuristic vehicle headlamp cover lens vision inspection device according to the present invention having such a configuration, a technical ripple effect and an economic ripple effect can be presented.

First, as a technical ripple effect, we provide a visual inspection solution for transparent injection parts in addition to general mold parts or external inspection of injection parts.

In addition, it can detect both colored or transparent injection product defects, and it can provide an adaptive inspection solution for various structures of automobile headlamp cover lenses by combining the auto-focusing function that can adjust the depth and servo motor. .

In addition, a higher defect detection ability can be exhibited through image synthesis technology acquired by controlling pattern illumination.

In addition, process capability technology can be improved by identifying the cause of defects through defect detection in the injection molding process, and technology can be applied to the production process of other parts and finished products by developing detection technology for defective injection molding products for the production of automobile parts. can be transferred

In addition, it has the advantage of being able to transfer technology to other processes by providing a PC-based control program that is flexible when adding a control unit.

Next, as an economic ripple effect, it is possible to reduce the defect rate of auto parts makers, thereby contributing to product quality improvement and reducing production costs.

In addition, it is possible to improve the production process by identifying the cause of the defect detection, and reduce the waste of raw materials required for injection by lowering the defect rate of the automobile headlamp cover lens.

In addition, there is an advantage that manpower and time can be invested in directly monitoring the injection molding manufacturing process for the automobile headlamp cover lens, and efficient manufacturing process innovation is possible through production automation.

In addition, by self-developing defect detection technology and system between manufacturing processes, it is possible to contribute to the limited budget operation of demanding companies through localization of technology, and to secure external competitiveness of demanding companies by contributing to quality improvement by reducing the incidence of product defects. can do.

1 is a conceptual diagram of the present invention;
2 is a front view of the present invention;
3 is a control block diagram of the present invention;
Figure 4 is a view for explaining the lighting means,
5 is a control block diagram of the central control means;
6 is a view for explaining the image processing procedure of the central control means to detect the location of the defect in the lens;
7 is a conceptual diagram for finding a defect location in a lens by summing the brightness values for each pixel coordinate calculated from the captured images for each of two or more pattern images for each pixel coordinate;
Figure 8a is a lens image taken through the camera when the first pattern is displayed through the lighting means,
Figure 8b is a lens image taken through the camera when the second pattern is displayed through the lighting means,
Figure 8c is a lens image taken through the camera when the third pattern is displayed through the lighting means,
Figure 8d is a lens image taken through the camera when the fourth pattern is displayed through the lighting means,
8E is a state diagram in which a defect location (red dotted line) in the lens is detected when the brightness values for each pixel coordinate of the lens image shown in FIGS. 8A to 8D are summed for each pixel coordinate and implemented as an image;
9 is a view for explaining an image processing process for including the lens defect region in the lens defect inspection region when the lens defect region is spread over the outer line of the lens defect inspection region;
Figure 10a is a view of converting a lens image taken through a camera into a black-and-white image;
10b is a diagram in which a lens defect inspection area (green line) is automatically extracted by applying a blob algorithm to the black-and-white image of FIG. 10a;
Figure 10c is a state diagram of a lens defect in the lens defect inspection area;
Figure 10d is a view in which the lens image captured by the camera is converted into a black-and-white image when the lens defect area is spread over the outer line of the lens defect inspection area;
10e and 10f are diagrams of a state in which the lens defect area is excluded from the lens defect inspection area when the blob algorithm is applied to the black and white image of FIG. 10d;
10g and 10h are diagrams for explaining the Convex Hull algorithm;
10i is a diagram showing a state in which the lens defect area is excluded from the lens defect inspection area when the blob algorithm is applied to the black and white image of FIG. 10d;
10j to 10M are diagrams for explaining a process of including a lens defect area in a lens defect inspection area by applying the Convex Hull algorithm to the image shown in FIG. 10i;

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

A futuristic vehicle headlamp cover lens vision inspection apparatus according to the present invention, as shown in FIGS. 1 to 3, includes a lens (L) vision inspection chamber (Chamber); a base plate (1) installed inside the lens (L) vision inspection chamber; a lens holder (2) installed on the base plate (1) and holding and fixing the target lens (L), which is a defect analysis target, on the base plate (1); a camera mount (3) installed on the base plate (1); a camera (4) mounted on the camera mounting table (3) to photograph the subject lens (L), which is an object of defect analysis; a camera mount transfer means (5) for reciprocating the camera mount (3) in the longitudinal direction of the base plate (1) so that the camera (4) can photograph the entire section of the subject lens (L); a lighting means (6) installed on the base plate (1) and emitting light in the direction of the lens (L) from the lower part of the lens (L); And by controlling the camera mount transport means (5) and the lighting means (6), and analyzing the image taken by the camera (4) to automatically set the lens (L) defect inspection area in the captured image, and the lens (L) and a central control means 7 for determining whether the lens L, which is an object to be inspected, is defective in the defect inspection area.

In addition, the camera mount transfer means 5 may reciprocate the lighting means 6 together with the camera mount 3 in the longitudinal direction of the base plate 1 .

As shown in FIG. 2 , the camera mount 3 has an arc shape.

As shown in FIG. 1 , a camera moving means 41 for reciprocating the camera 4 along the outer line of the camera mounting table 3 is mounted on the camera mount 3 .

The camera moving means 41 may use a servo motor for moving the camera 4 by generating rotational power using electric energy.

The camera moving means 41 is integrally coupled with the camera 4 , and a camera tilting means 42 for tilting the camera 4 is mounted on the camera moving means 41 .

The camera 4 installed on the camera mount 3 is composed of three or more, and the peripheral surface of the lens L is divided and photographed.

As shown in FIG. 2 , the lens holder 2 adjusts the height of the lens holder 2 according to the control signal of the central control means 7 to adjust the height between the camera 4 and the lens L. It further includes a lens height adjustment means 21 for adjusting the distance.

As shown in FIG. 4, the lighting means 6 includes an LCD panel 61 (LCD Panel) that outputs a specific pattern image according to a control signal of the central control means 7, and the LCD panel 61 The backlight unit 62 (Back Light) that emits light from the bottom so that the image output from the LCD panel 61 can be identified, and the backlight unit 62 are controlled to adjust the illuminance of the backlight unit 62 and a backlight control unit 63 that does.

In addition, as shown in FIG. 5, the central control means 7 includes a pattern image storage unit 7A1 in which pattern image data is stored.

In addition, the lighting means 6 controls the LCD panel 61 according to the control signal of the central control means 7 to display the pattern image stored in the pattern image storage unit 7A1 to the LCD panel 61 . and LCD panel control means (64) for outputting through it.

An HDMI cable or a DP cable is communication-connected between the central control means 7 and the LCD panel control means 64 to transmit and receive control signals and image data.

In addition, the backlight control unit 63 and the central control unit 7 may send and receive a control signal for adjusting backlight illuminance through the RS-232C communication protocol.

Preferably, a heat dissipation unit for dissipating heat generated from the backlight unit 62 is mounted on the backlight unit 62 .

The lens holder 2 is an actuator that grips the lens L using the pneumatic pressure of the compressed air. The lens holder 2 controls the flow of compressed air so that the lens holder 2 bites the lens L. A solenoid valve portion for holding or releasing may be additionally equipped.

The central control means 7 may control the solenoid valve unit to hold the lens L through the lens holder 2 or release the lens L.

The lens (L) vision inspection chamber (Chamber) for the lens (L) vision inspection chamber is additionally equipped with a light blocking means for blocking the light flowing in from the outside, the light blocking means as shown in FIG. likewise, a blackout curtain unit installed in the lens (L) vision inspection chamber to block light entering the lens (L) vision inspection chamber when it is unfolded; and blackout curtain installation means (CI) for unfolding or folding the blackout curtain according to the control signal of the central control means (7).

In addition, the present invention uses a liquid lens (L) as the lens (L) of the camera (4), and controls the liquid lens (L) to automatically adjust the depth and focus of the camera (4).

As shown in Fig. 4, the lighting means 6 has an LCD panel 61 for displaying a specific pattern image according to a control signal from the central control means 7, and is displayed on the LCD panel 61. The image pattern to be used may be composed of a stripe pattern in which two lines having different colors are alternately displayed.

The central control means 7 outputs two or more pattern images through the LCD panel 61 when inspecting one subject lens L, and coordinates of line positions of specific colors in the two or more pattern images. is different for each pattern image.

The central control means 7 controls the camera 4 whenever the pattern image displayed from the LCD panel 61 provided in the lighting means 6 changes, as shown in FIGS. 6 and 8e. A camera control unit 7A2 for photographing the lens L, which is an object to be inspected, a photographed image storage unit 7A3 for storing photographed images for each two or more pattern images photographed by the camera 4, and the photographed image storage unit A grayscale conversion unit 7A4 that converts the captured image for each pattern image stored in 7A3 to grayscale, and the pixel coordinates in the captured image for each pattern image converted to grayscale by the grayscale conversion unit 7A4 A brightness calculation unit 7A5 for each pixel coordinate for calculating a brightness value for each pixel, a brightness summing unit 7A6 for each pixel coordinate for summing the brightness values for each pixel coordinate calculated from the captured images for each of two or more pattern images for each pixel coordinate, 7A6, the pixel a brightness comparison unit 7A7 for each pixel coordinate that compares the summed brightness values for each pixel coordinate added by the brightness summing unit 7A6 for each coordinate for each pixel coordinate; and a defect position estimation unit 7A8 in the image estimating that there is a lens (L) defect in the pixel coordinate portion where the summed brightness value for each pixel coordinate is the minimum value as a result of the comparison of the brightness comparison unit 7A7 for each pixel coordinate. .

As shown in FIG. 9 , the central control means 7 automatically extracts the lens L defect inspection region from the image of the subject photographed by the camera 4 , a region of interest (ROI). and an automatic extraction unit 7B.

As shown in FIG. 9 , the automatic region-of-interest extraction unit 7B automatically extracts the lens L defect region in the image of the subject by automatically extracting the lens L defect by the automatic region-of-interest extraction unit 7B. and an automatic defect area accommodating part 7B1 for including the lens (L) defect area in the lens (L) defect examination area when it spans the outer line of the extracted lens (L) defect inspection area.

As shown in FIG. 9 , the defect region automatic accommodation unit 7B1 includes an image pre-processing unit 7B11 that converts the subject image photographed by the camera 4 into a black-and-white image, and the image pre-processing unit 7B11 ) includes a primary lens defect inspection area extraction unit 7B12 that automatically extracts the lens (L) defect inspection area set in the central control means 7 by applying a blob algorithm to the black and white image converted by .

In addition, as shown in FIG. 9 , the automatic defect area receiving unit 7B1 includes a lens (L) on the outer line of the defect inspection area automatically extracted by the primary lens defect inspection area extracting unit 7B12. L) Convex hull on a black and white image designated as a lens defect inspection region through the primary lens defect inspection region extraction unit 7B12 to solve the problem that the lens defect region is excluded from the lens defect inspection region when the defect region is over and a defect area accommodating part 7B13 for accommodating the lens defect area extending over the outer line of the lens defect examination area in the lens defect examination area by applying a (Convex Hull) algorithm.

As shown in FIG. 10A , the image preprocessor 7B11 converts the image of the subject photographed by the camera 4 into a black-and-white image.

10B is a diagram in which a lens defect inspection area is automatically extracted by the primary lens defect inspection area extraction unit 7B12, and a green line in FIG. 10B is an outer line of the lens defect inspection area.

At this time, as shown in FIG. 10B , if no foreign material or defect is detected on the outer line of the lens defect inspection region, the lens defect inspection region is successfully extracted.

In this case, as shown in FIG. 10C , when there is a defect component in the lens defect inspection area, it is possible to detect a defect.

On the other hand, as shown in FIG. 10D , when a defect part spans the outer line of the lens defect inspection area, the primary lens defect inspection area extraction unit 7B12 determines that the defective part is within the lens defect inspection area. , as shown in FIGS. 10E and 10F, the defective portion is excluded from the lens defect inspection area.

In this case, since the defective part is completely excluded from the lens defect inspection area, no matter how much the defective part is detected in the lens coupling inspection area, it is impossible to know whether the defect is there, so there is a problem in that a defective lens is judged as a good product.

In order to solve such a problem, the defect region accommodating unit 7B13 uses a Convex Hull algorithm, which is a convex hull algorithm.

As shown in FIGS. 10g and 10h, the Convex Hull algorithm creates a convex polygon by using some of the points located outside among several points when there are several points on a two-dimensional plane, and all points within the convex polygon are say to include

When there is a set of points, there are Javis's March algorithm and Graham's Scan algorithm in the Convex Hull algorithm that can include all points.

The application results of the Javis's March algorithm and the Graham's Scan algorithm are the same, and the Javis's March algorithm is more intuitive than the Graham's Scan algorithm, but takes a long time to process because it is more intuitive.

Returning to the main point, the defective area accommodating unit 7B13 uses the Convex Hull algorithm selected from either Javis's March algorithm or Graham's Scan algorithm as shown in FIGS. can be included in

1. Base plate L. Lens
2. Lens holder 21. Lens height adjustment means
3. Camera Mount 4. Camera
41. Camera moving means 42. Camera tilting means
5. Camera mount transport means 6. Lighting means
61. LCD panel 62. Backlight part
63. Backlight control 64. LCD panel control means
7. Central control means 7A1. pattern image storage
7A2. Camera Control 7A3. recording video storage
7A4. Grayscale conversion unit 7A5. Brightness calculator by pixel coordinates
7A6. Brightness summing unit for each pixel coordinate 7A7. Brightness comparison unit by pixel coordinates
7A8. Defect location estimation unit in the image 7B. region of interest automatic extraction unit
7B1. Defect area automatic receptacle 7B11. image preprocessor
7B12. Primary lens defect inspection area extraction unit 7B13. Defect area receptacle

Claims (4)

Lens (L) and vision inspection chamber (Chamber);
a base plate (1) installed inside the lens (L) vision inspection chamber;
a lens holder (2) installed on the base plate (1) and holding and fixing the target lens (L), which is a defect analysis target, on the base plate (1);
a camera mount (3) installed on the base plate (1);
a camera (4) mounted on the camera mount (3) to photograph the subject lens (L), which is an object of defect analysis;
a camera mount transfer means (5) for reciprocating the camera mount (3) in the longitudinal direction of the base plate (1) so that the camera (4) can photograph the entire section of the subject lens (L);
a lighting means (6) installed on the base plate (1) and emitting light in the direction of the lens (L) from the lower part of the lens (L);
And by controlling the camera mount transport means (5) and the lighting means (6), and analyzing the image taken by the camera (4) to automatically set the lens (L) defect inspection area in the captured image, and the lens (L) and a central control means (7) for determining whether the lens (L), which is the object to be inspected, is defective in the defect inspection area;
The central control means 7 includes an automatic region-of-interest extraction unit 7B for automatically extracting a lens (L) defect inspection region from the image to be inspected taken by the camera 4,
The region of interest automatic extraction unit 7B inspects the lens (L) defect in which the lens L defect region in the image of the subject is automatically extracted by the region of interest automatic extraction unit 7B. and a defect area automatic accommodating part 7B1 for including the lens (L) defect area in the lens (L) defect inspection area when it spans the outer line of the area,
The defect region automatic accommodating unit 7B1 includes an image preprocessing unit 7B11 for converting the subject image photographed by the camera 4 into a black and white image,
A primary lens defect inspection area extraction unit that automatically extracts a lens (L) defect inspection area set in the central control means 7 by applying a blob algorithm to the black and white image converted by the image preprocessing unit 7B11 (7B12);
The defect area automatic accommodating part 7B1 is the lens (L) defect area over the outer line of the lens (L) defect inspection area automatically extracted by the primary lens defect inspection area extraction unit 7B12. In order to solve the problem that the defective area is excluded from the lens defect inspection area, the convex Hull algorithm is applied to the black-and-white image designated as the lens defect inspection area through the primary lens defect inspection area extractor 7B12 to solve the problem that the lens and a defect area accommodating part (7B13) for accommodating the lens defect area spanning the outer line of the defect examination area in the lens defect examination area.
The method of claim 1,
The lens holder 2 is a lens height adjusting means for adjusting the height gap between the camera 4 and the lens L by adjusting the height of the lens holder 2 according to the control signal of the central control means 7 . (21) Future vehicle headlamp cover lens vision inspection device, characterized in that it further comprises.
The method of claim 1,
The lighting means (6) has an LCD panel (61) that displays a specific pattern image according to the control signal of the central control means (7),
The image pattern displayed on the LCD panel 61 consists of a stripe pattern in which two lines having different colors are alternately displayed,
The central control means 7 outputs two or more pattern images through the LCD panel 61 when examining one subject lens L,
The coordinates of the line position of a specific color in the two or more pattern images are different for each pattern image,
The central control means 7 controls the camera 4 whenever the pattern image displayed from the LCD panel 61 provided in the lighting means 6 changes to photograph the lens L as an object to be inspected. a control unit 7A2;
A photographed image storage unit (7A3) for storing photographed images for each of two or more pattern images photographed by the camera (4);
a grayscale conversion unit 7A4 for converting the captured images for each pattern image stored in the captured image storage unit 7A3 into grayscale;
a brightness calculation unit 7A5 for each pixel coordinate that calculates a brightness value for each pixel coordinate in the captured image for each pattern image converted to grayscale by the grayscale conversion unit 7A4;
a brightness summing unit 7A6 for each pixel coordinate that sums the brightness values for each pixel coordinate calculated from the captured images for each of two or more pattern images for each pixel coordinate;
a brightness comparison unit for each pixel coordinate (7A7) for comparing the summed brightness values for each pixel coordinate added by the brightness summing unit (7A6) for each pixel coordinate for each pixel coordinate;
and a defect location estimation unit 7A8 in the image estimating that there is a lens (L) defect in the pixel coordinate portion where the summed brightness value for each pixel coordinate is the minimum value as a result of the comparison of the brightness comparison unit 7A7 for each pixel coordinate. Future vehicle headlamp cover lens vision inspection device, characterized in that.
delete

Images