KR20200046149A - Area-based vision testing device - Google Patents

Abstract

The present invention relates to an area-based vision inspection apparatus capable of inspecting the omission of a component occurring on the exterior in regard to an assembled finished product. The area-based vision inspection apparatus includes: a fixing plate (10) helping a variety of assembled finished products to be fixed to a placement jig; the placement jig (20) fixing the fixing plate to prevent the fixing plate from being moved when rotated; a rotary stage (30) for the rotation of the assembled finished products; a Z stage (40) for obtaining an image of the upper surface of the assembled finished products; XZ stages (50, 51) for obtaining images of the front and rear sides and left and right sides of the assembled finished products; lighting devices (60, 61, 62) emitting light required for obtaining images; a stage controller (80) controlling Z, XY stages; an image processing part (90) decoding whether a component is omitted based on a specific area by obtaining the images of the assembled finished products; and a vision inspection monitor (100) displaying a component omission decoding result of the assembled finished products decoded by the image processing part.

Description

Area-based vision testing device

The present invention relates to an area-based vision inspection device, and more specifically, in order to automatically inspect parts of various types of assembled finished products, an area to be inspected is displayed on an image input through an imaging device, and learning is performed. It relates to a device for automatically determining whether a part is missing.

Since the present invention relates to a device for inspecting parts of various types of assembled finished products, more specifically, it is a vision inspection device based on a specific area for detecting omission of various parts including screws that appear on the appearance of the assembled finished product.

Of the parts shown on the exterior, screws and labels, among other things, cannot be detected by the functional inspection. Until now, these tests are mostly determined by the human eye, and it is difficult to always make the correct judgment due to the afterimage effect.

Therefore, in the prior art, in the process of inspecting the finished assembly, a marker is used on the screw to determine whether parts appearing on the exterior are missing. This method has a problem in that it is inefficient in terms of detection accuracy and work efficiency. Due to the afterimage effect of the human naked eye, it is possible to confirm marking even on the missing parts, which may result in missing parts in the inspection process. Due to missing parts found in sample inspections when shipping an assembled finished product, delays in delivery and economic loss due to a drop in reliability can cause serious problems. To solve these problems, recently developed machine learning technology was applied. In the image of the finished assembly, the part area to be inspected was displayed, the image of the displayed area was learned, and the learned result was used to judge missing parts.

An object of the present invention is to provide an apparatus for detecting a missing part based on a specific area, which receives and receives an image of a computer for a specific part in the exterior from various assembly finished products, and automatically analyzes and detects the missing part. .

The present invention for achieving the above object is a vision inspection device that displays a specific area for parts appearing on the exterior of the assembled finished product and determines whether parts are missing for this area. The configuration of the present invention is a fixing plate (10) that helps to ensure that the various assembled finished products are fixed to the seating jig, a seating jig (20) that fixes the fixing plate so that it does not move when rotating, a rotating stage (30) for rotating the assembled product ) Z-stage 40 for acquiring the image of the top surface of the assembled finished product XZ stage 50,51 for acquiring the image of the left and right and front and back surfaces of the assembled finished product 60 , 61,62) Z, XY stage controller 80 that controls the stage. 80. Image processing unit 90 that acquires the image of the assembled finished product and reads out whether parts are missing based on a specific area. The assembled unit read from the image processing unit. A vision inspection monitor 100 is provided to display the result of reading whether the finished product is missing.

At this time, on the seating jig on which the assembled finished product is seated, there are various fixing plates to help various assembled finished products be fixed on the seating jig, so that the assembled finished product can be easily fixed.

In addition, it is possible to automatically rotate the finished product through the above rotating stage, and through the stage controller that controls the position of the Z and XY stages, the 5 sides (top, tightened, right, front, and back) images of the finished product are easily assembled. Can be captured. The area to be inspected is displayed on the obtained five-sided image (top, side, right, front, back) and learning data for the displayed area is generated. Learning is performed on the generated learning data. Based on the learned results, the missing parts are inspected for the assembled finished product to be inspected. In addition, the execution result displays a missing result on the assembly completed image through the monitor.

The region-based vision inspection apparatus according to an embodiment of the present invention can quickly determine whether omission of a visible part is omitted by applying a vision inspection to an assembled finished product as a result of an assembly process.

In addition, the region-based vision inspection apparatus according to an embodiment of the present invention learns an image of a specific area for each assembled product in order to detect defects in various assembled finished products. With the learned result, the missing parts are inspected for the assembled finished product to be inspected, and the missing results are displayed on the assembled finished product image through the monitor, so that it is easy to judge the missing parts appearing on the exterior.

In addition, the region-based vision inspection apparatus according to an embodiment of the present invention, only by learning the input image for the various types of assembly finished products, without software modification, only the fixed plate to help the assembly completed to be fixed to the seating jig If you change it, you can do it.

1 shows a configuration diagram of a region-based vision inspection apparatus according to an embodiment of the present invention.
2 is a view showing various types of fixed plates in a region-based vision inspection apparatus according to an embodiment of the present invention.
Referring to Figure 2, the fixing plate has two holes, which are fixed on the seating jig so that the rotating plate rotates, so that the fixing plate does not shake. In addition, the shape of the top plate of the fixing plate is manufactured in a form that the assembled finished product can be easily fixed.
3 is an operation flowchart showing an operation sequence of a region-based vision inspection apparatus according to an embodiment of the present invention.
Referring to FIG. 3, first, the assembled finished product to be inspected is fixed to the fixing plate on the seating jig. When the assembled finished product is fixed to the fixed plate, the operator enters the model name of the assembled finished product and selects a learning step and a test step.
When the operator selects the learning step, the Z stage and the XZ stage position are set on the 5 sides of the finished assembly.
When the operator completes the Z-stage and XZ-stage positioning operation, an image of the finished assembly is captured. After setting the area for the part to be checked for missing parts in the captured image, learning data for the set area is generated.
When the learning data for the area of the part is generated for the five sides of the finished assembly, learning is performed.
On the other hand, when the operator selects the test in the step selection in Figure 3 captures the image of the five sides of the finished assembly to be inspected.
A missing part test is performed on the captured image. When the test is completed, the image of the assembled product inspected with the result and the inspection result are shown as pictures.
4 shows inspection results of a region-based vision inspection apparatus according to an embodiment of the present invention.
In the figure, the green area indicates that the part is attached, and the red area indicates that the part is missing.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, the following embodiments are provided to help understanding of the present invention, and the scope of the present invention is not limited to the following embodiments. In addition, the following embodiments are provided to more fully describe the present invention to those of ordinary skill in the art, and detailed descriptions of well-known configurations determined to unnecessarily obscure the technical gist of the present invention Each component illustrated in the drawings may be exaggerated for convenience of description.

1 is a front view showing a region-based vision inspection apparatus according to an embodiment of the present invention. Referring to Figure 1, the region-based vision inspection apparatus 1 according to an embodiment of the present invention is fixed plate 10, seating jig 20, rotating stage 30, Z stage 40, XZ stage ( 50, 51), lighting devices (60,61,62), a stage controller (80), an image processing unit (90), and a vision inspection monitor (100) may be included.

The seating jig 20 may be formed in the form of a plate installed at a predetermined height on the bottom surface. The seating jig 20 is located between the fixing plate and the rotating stage, and when the rotating plate rotates, the fixing plate is fixed so that the position of the fixing plate is not changed. The fixing plate 10 is to be fixed to the assembly jig 20 various assembly finished products Make it possible.

The rotating stage 30 rotates the fixing plate to obtain images of the front and back sides of the assembled finished product.

The Z stage 40 is movable in the Z direction to obtain an image of the top surface of the assembled finished product.

The XZ stages 50 and 51 are movable in the Z- and X-directions to obtain images of the front and back and left and right sides of the assembled finished product.

The stage controller 80 controls the rotating stage 30, the Z stage 40, and the XZ stages 50 and 51 to acquire an image of the assembled finished product.

The cameras 70, 71 and 72 and the stage controller 80 capture an image for learning and discriminating a specific area image of each assembled product. The stage controller 80 moves the camera to a preset position in order to obtain a clear image according to each assembled product.

The lighting devices 60, 61, and 62 provide lighting to obtain a clear image according to each assembled product.

The image processing unit 90 processes the learning with the image of the area captured for learning, and determines whether the component is missing from the captured image to determine the missing part. For learning, a learning image is obtained by displaying an area for learning on the images captured through the Z stage 40 and the XZ stages 50 and 51. After creating various types of learning images using the acquired learning images, learning is performed using them. After capturing the image of the assembled finished product, the omission detection determines whether the displayed area is defective during learning.

Claims (1)

Fixing plate (10) and fixing plate to help the finished product to be fixed to the seating jig in a device that can quickly determine whether the parts that appear on the exterior are missing by applying vision inspection to the assembled finished product that is the result of the assembly process The seating jig 20, the rotating plate stage 30 for rotating the assembled finished product, the Z stage 40 for acquiring the image of the upper surface of the assembled finished product, the left and right sides of the assembled finished product XZ stage (50, 51), Z, X, Y, stage controller (80) for controlling the rotating plate stage, image processing unit (90) that acquires the image of the finished product and reads out whether parts are missing based on a specific area, assembly A device consisting of a vision inspection monitor (100) that displays the results of reading the missing parts of the finished product.

Images