WO2023038214A1

WO2023038214A1 - Surface defect detection device and method

Info

Publication number: WO2023038214A1
Application number: PCT/KR2022/004988
Authority: WO
Inventors: 윤기욱; 김태웅; 송정민; 김의석
Original assignee: 라온피플 주식회사
Priority date: 2021-09-10
Filing date: 2022-04-06
Publication date: 2023-03-16
Also published as: KR20230038078A

Abstract

Proposed are a surface defect detection device and method. The surface defect detection method performed by a surface defect detection device comprises the steps of: obtaining a captured image of an object to be inspected; and detecting a defect by using the obtained captured image. The step of obtaining a captured image comprises the steps of: irradiating, onto the surface of the object to be inspected, pattern light having stripe patterns at regular intervals; and obtaining a captured image by capturing reflective light reflected from the surface of the object to be inspected.

Description

Apparatus and method for detecting surface defects

Embodiments disclosed herein relate to a surface defect detection apparatus and method capable of detecting surface defects of products that are difficult to detect defects by photography due to surface gloss.

As one of the quality inspections of mass-produced products, there is a visual inspection. The product is inspected by optically analyzing the image taken of the exterior of the product, finding the part estimated to be defective, and classifying the defective product. In particular, the use of pre-learned artificial intelligence models in the exterior inspection process has improved inspection performance, and the inspection results have reached a reliable level even without human visual inspection. Particularly, by adopting the exterior inspection using machine vision, the inspection of mass-produced products is becoming more efficient in terms of time and cost.

However, there are products that are difficult to adopt the exterior inspection method that can maximize efficiency. Products with a glossy surface, such as automobiles, related parts, and home appliances that require glossy coating, reflect light of essential lighting for exterior photography, making it difficult to perform exterior inspection using machine vision. Even if there is a defect in the part where there is excessive reflected light due to the surface gloss in the photographed image of the product, it is difficult to identify it. As a result, these products had inefficiency in that a person had to find defects with the naked eye.

Korean Patent Registration No. 10-1611823 “Exterior Inspection Method” discloses an automatic exterior inspection device and method for detecting defects such as dents or stains on the surface of home appliances by photographing the exterior while transporting the appliance.

However, such prior literature does not suggest a problem or a method for improving it when gloss is included in a photographed image. Therefore, there is a need for a vision inspection method for inspecting the appearance of products having a glossy surface.

On the other hand, the above-mentioned background art is technical information that the inventor possessed for derivation of the present invention or acquired in the process of derivation of the present invention, and cannot necessarily be said to be known art disclosed to the general public prior to filing the present invention. .

Embodiments disclosed herein are aimed at presenting an apparatus and method for detecting surface defects of a product having a glossy surface.

Embodiments disclosed in this specification are aimed at presenting a device and method for detecting surface defects capable of improving defect detection performance of a product having a glossy surface.

Embodiments disclosed in this specification are intended to provide a surface defect detection device and method capable of detecting detailed locations of defects by selectively photographing a plurality of coating layers forming a glossy surface.

Embodiments disclosed in this specification are intended to provide a surface defect detection device and method capable of detecting all defects formed on a glossy surface, even if the locations or extension directions thereof vary.

As a technical means for achieving the above-described technical problem, according to an embodiment, a surface defect detection method performed by a surface defect detection apparatus includes obtaining a photographed image of an inspection target; The method includes: detecting a defect using an acquired captured image, wherein the acquiring of the captured image includes: irradiating pattern light having a stripe pattern at regular intervals on a surface of an object to be inspected; and obtaining a photographed image by photographing the reflected light reflected from the surface of the object to be inspected.

According to another embodiment, a surface defect detection apparatus includes an illumination unit for irradiating pattern light having a stripe pattern at regular intervals on a surface of an inspection target; a photographing unit for acquiring a photographed image by photographing reflected light reflected from the surface of the inspection target by the pattern light irradiated by the lighting unit; a storage unit for storing the photographed image acquired by the photographing unit; The control unit may include a controller configured to adjust characteristics of pattern light emitted from the lighting unit, control the photographing unit, and detect a defect of the inspection target using a photographed image stored in the storage unit.

According to another embodiment, in a computer readable recording medium on which a program for performing a method for detecting surface defects is recorded, the method for detecting surface defects includes acquiring a photographed image of an object to be inspected; The method includes: detecting a defect using an acquired captured image, wherein the acquiring of the captured image includes: irradiating pattern light having a stripe pattern at regular intervals on a surface of an object to be inspected; A step of acquiring a photographed image by photographing reflected light reflected from the surface of the inspection target may be included.

Furthermore, according to another embodiment, the surface defect detection method in a computer program stored in a medium to perform the surface defect detection method, which is performed by a surface defect detection device, includes acquiring a photographed image of an inspection target; The method includes: detecting a defect using an acquired captured image, wherein the acquiring of the captured image includes: irradiating pattern light having a stripe pattern at regular intervals on a surface of an object to be inspected; A step of acquiring a photographed image by photographing reflected light reflected from the surface of the inspection target may be included.

According to any one of the above-mentioned problem solving means, it is possible to present a method for detecting surface defects of a product having a glossy surface and an apparatus for performing the same.

According to any one of the above-mentioned problem solving means, it is possible to present a method for detecting surface defects capable of improving the detection performance of defects of a product having a glossy surface and an apparatus for performing the same.

According to any one of the above-described problem solving means, a surface defect detection method capable of detecting a detailed location of a defect by selectively photographing a plurality of paint layers forming a glossy surface and a device for performing the same are presented. can

According to any one of the above-described problem solving means, it is possible to propose a surface defect detection method and a device for performing the same, which can detect all of the defects formed on the glossy surface even if the locations or extension directions thereof vary.

Effects obtainable from the disclosed embodiments are not limited to those mentioned above, and other effects not mentioned are clear to those skilled in the art from the description below to which the disclosed embodiments belong. will be understandable.

1 is a block diagram schematically showing a functional configuration of a surface defect detection apparatus according to an embodiment.

2 is a conceptual diagram showing some configurations of a surface defect detection apparatus according to an embodiment.

3 is a diagram showing a method of detecting surface defects performed by a surface defect detection apparatus according to an embodiment step by step.

4 is an exemplary diagram showing types of defects occurring in a coating layer of a product to be inspected having a glossy surface.

5 is an exemplary view showing an example of defects detected by a surface defect detection method according to an embodiment.

Hereinafter, various embodiments will be described in detail with reference to the accompanying drawings. Embodiments described below may be modified and implemented in various different forms. In order to more clearly describe the characteristics of the embodiments, detailed descriptions of matters widely known to those skilled in the art to which the following embodiments belong are omitted. And, in the drawings, parts irrelevant to the description of the embodiments are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a component is said to be "connected" to another component, this includes not only the case of being 'directly connected', but also the case of being 'connected with another component in between'. In addition, when a certain component "includes" a certain component, this means that other components may be further included without excluding other components unless otherwise specified.

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

1 is a block diagram schematically showing a functional configuration of a surface defect detection device according to an embodiment, and FIG. 2 is a conceptual diagram showing some configurations of a surface defect detection device according to an embodiment.

The surface defect detection apparatus 100 is a device that acquires a photographed image of an inspection target and detects defects based on the obtained photographed image in order to detect defects on the surface of the inspection target having a glossy surface. The surface defect detection apparatus 100 may be implemented as a general user terminal or a server-client system composed of a user terminal and a server.

Here, the user terminal may be implemented as a computer or portable terminal capable of accessing a remote server through the network N or connecting to other terminals and servers. Here, the computer includes, for example, a laptop, desktop, or laptop equipped with a web browser, and the portable terminal is, for example, a wireless communication device that ensures portability and mobility. , PCS(Personal Communication System), PDC(Personal Digital Cellular), PHS(Personal Handyphone System), PDA(Personal Digital Assistant), GSM(Global System for Mobile communications), IMT(International Mobile Telecommunication)-2000, CDMA(Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet), Smart Phone, Mobile WiMAX (Mobile Worldwide Interoperability for Microwave Access), etc. (Handheld)-based wireless communication device may be included.

In this case, a light irradiation unit or a camera corresponding to a lighting unit or a photographing unit, which will be described later, may be integrally configured or connected to the user terminal in the form of a module.

Meanwhile, at this time, the server implementing the server-client system may be implemented as a computer capable of communicating with a user terminal on which an application for interaction with a user or a web browser is installed through a network, or may be implemented as a cloud computing server. In addition, the server may include a storage device capable of storing data or may store data through a third server.

The above-described user terminal and server may cooperate with each other to configure the surface defect detection device 100 implemented as one server-client system. In this case, the surface defect detection device 100 interfaces with the user through the user terminal. While doing so, it is possible to operate in such a way that a photographed image of an inspection target is acquired from a user terminal, and a server processes it to detect a defect.

Meanwhile, as shown in FIG. 1 , the surface defect detection apparatus 100 includes a controller 110. The control unit 110 controls the overall operation of the surface defect detection apparatus 100 and may include a processor such as a CPU that controls various components included in the surface defect detection apparatus 100 to be described later. The control unit 110 may execute a program stored in the storage unit 140 to be described later or calculate data using an algorithm or a machine learning model stored in the storage unit 140. Also, the control unit 110 may store the processed data in the storage unit 140 again.

A specific operation of the control unit 110 will be elaborated below.

Meanwhile, the surface defect detection apparatus 100 may include a lighting unit 120 for irradiating pattern light to the inspection target in order to photograph the inspection target. At this time, the pattern light is the light of the stripe pattern P having straightness as shown in FIG. 2, and at this time the stripe patterns are formed with a certain width and may be arranged in parallel at regular intervals. For example, the lighting unit 120 may include a lamp in which a plurality of LED elements are arranged, and such LED lamps may be arranged in a stripe pattern.

In addition, the lighting unit 120 may be configured to change characteristics of pattern light, such as color temperature, width of stripe patterns, interval between stripe patterns, direction in which stripe patterns extend, and the like. The lighting unit 120 may adjust the color temperature of the pattern light to have two or more different values selected from, for example, 3000K (Kelvin) to 7000K.

For example, the lighting unit 120 may include two or more types of LED elements emitting light of different color temperatures, and the color temperature may be adjusted while controlling only LED elements having a specific color temperature to emit light at a time. .

As another example, the lighting unit 120 may be configured to be rotatable so that the direction of the stripe pattern may be adjusted.

As another example, the lighting unit 120 includes a plurality of LED elements arranged in a matrix so that the interval or width of the stripe pattern can be adjusted, and while the LED elements selectively emit light, the light of the stripe pattern having a desired width and interval is emitted. It may also be configured to emit light.

As another example, in the lighting unit 120, LED elements arranged in a matrix form emit light at the rear, and an LCD panel is configured at the front. When the LCD panel outputs a desired pattern, the LED elements at the rear emit light and emit pattern light. It may be configured to form. According to this, the characteristics of the pattern light can be controlled by outputting a two-dimensional image of the pattern light to be irradiated by the LCD panel. Furthermore, the color temperature of light emitted from the LED device can be controlled using an image output from the LCD panel.

In this way, since the lighting unit 120 is configured to be able to adjust the characteristics of the pattern light, the control unit 110 can drive the lighting unit 120 according to the characteristics of one or more pattern lights required according to the inspection target and obtain a photographed image. there is.

For example, with respect to an inspection target including a plurality of paint layers, the control unit 110 stores color temperature information of two or more pattern lights set so that the pattern light can reach each paint layer in the storage unit 140. The lighting unit 120 may sequentially irradiate pattern light having different characteristics to the same inspection target according to the set color temperature.

Alternatively, if there is a possibility that a defect cannot be easily detected by the pattern light in one specific direction depending on the location of the defect or the extension direction of the defect, the control unit 110 sets two directions of movement of the stripe pattern forming the pattern light. After the preset is set and stored in the storage unit 140, the lighting unit 120 may obtain two or more captured images while sequentially changing and irradiating pattern light according to two or more set directions.

The surface defect detection device 100 may include a photographing unit 130 . The photographing unit 130 is provided with an optical means such as a conventional camera module, and can photograph the surface of the inspection target while the lighting unit 120 radiates pattern light to the inspection target. Accordingly, the photographed image obtained by the photographing unit 130 may include an image of reflected light reflected from the glossy surface of the inspection target to which the pattern light is irradiated.

Meanwhile, as shown in FIG. 2 , the photographed image I obtained by the photographing unit 130 may have a substantially similar shape corresponding to the stripe pattern P forming the pattern light. However, if there is a curve on the glossy surface of the inspection target, it may be formed into a shape that is bent according to the curve. Meanwhile, if there is a defect other than the intended curve, the stripe pattern may be distorted by the defect. Accordingly, the photographed image I obtained by the photographing unit 130 may include distortion due to defects.

Meanwhile, the controller 110 may control the photographing unit 130 . The controller 110 may change settings of the photographing unit 130 according to the characteristics of the pattern light emitted from the lighting unit 120 . For example, when the control unit 110 changes the characteristics of the pattern light of the lighting unit 120, the controller 110 changes the settings of the photographing unit 130 according to the characteristics of the pattern light so that a captured image can be obtained clearly. can do. For example, when the color temperature of pattern light emitted by the lighting unit 120 is changed, a white balance set value of the photographing unit 130 may be adjusted to acquire a photographed image according to the changed color temperature.

Meanwhile, the control unit 110 changes the characteristics of the pattern light emitted from the lighting unit 120 for the same inspection target by a preset number of times, and at the same time, the photographing unit 130 photographs the inspection target whenever the characteristics of the pattern light are changed. By repeating this, it is possible to acquire a plurality of captured images of one inspection target.

At this time, a plurality of photographed images are obtained by photographing a specific region of the inspection target while changing only the characteristics of the irradiated pattern light without movement of the photographing unit 130 or the inspection target. Accordingly, a plurality of captured images of an inspection target described below are images captured while only changing illumination of the same part of the inspection target.

Meanwhile, the surface defect detection apparatus 100 may include a storage unit 140 . Various types of data such as files or programs may be installed and stored in the storage unit 140 . The controller 110 may access and use data stored in the storage unit 140 or may store new data in the storage unit 140 . Also, the controller 110 may execute a program installed in the storage unit 140 .

The storage unit 140 may store information on characteristics of pattern light to be irradiated by the lighting unit 120 according to the type of inspection target. For example, when it is necessary to acquire three captured images while irradiating pattern light having three different color temperatures on one inspection target, the storage unit 140 sets the color temperature to be adjusted by the lighting unit 120. Values can be stored.

As such, information on a plurality of pattern light characteristics including information on the characteristics of the pattern light, that is, color temperature, width or spacing of the stripe pattern, direction of the stripe pattern, and the like, may be previously stored in the storage unit 140 .

In addition, the storage unit 140 may store a plurality of captured images for each examination target. Of course, only one photographed image may be acquired and stored according to the type of inspection target.

At this time, when a plurality of captured images of one inspection target are stored in the storage unit 140, each captured image included in the plurality of captured images contains information on pattern light characteristics when the corresponding captured image is acquired. can be associated with and stored. For example, when a plurality of captured images are acquired while sequentially changing the color temperature of pattern light to 4000K, 5500K, and 7000K, 1, 2, and 3 may be associated with each of the plurality of captured images as reference values. At this time, for example, A reference value of '1' may represent an image captured while irradiating pattern light having a color temperature of 4000K.

In addition, various data or program modules such as parameters related to the pre-learned defect detection model may be stored in the storage unit 140 as one or more packages. At this time, the defect detection model is a machine learning model that receives a photographed image and outputs information on whether a defect exists or a defect grade, and has been pre-learned before being installed in the storage unit 140 or stored in the storage unit 140. This state may be a state in which learning continues.

The control unit 110 may determine whether or not there is a defect on the surface of the inspection target, the location of the defect, the type of defect, and the grade of the defect by using the defect detection model. That is, the defect detection model is configured to determine only whether or not there is a defect, or, as described above, calculates the grade of the product to be inspected according to the location of the defect, for example, the coating layer where the defect is formed, the type of defect, the number or size of the defect, etc. It can be designed in advance to output it.

Meanwhile, in detecting a defect using a plurality of captured images obtained for one inspection target, the defect detection model receives a plurality of captured images one by one and is trained to output whether or not the corresponding captured image has a defect. can In this case, the controller 110 may check whether each of the plurality of captured images has a defect, and accordingly, if a defect is detected in one of the plurality of captured images, it may be determined that the inspection target has a defect.

In addition, for example, when the color temperature is set so that the wavelengths of light according to the color temperature reach different paint layers, three captured images a, b, and c are obtained while irradiating pattern light of three different color temperatures. When the detection result of the defect detection model for each captured image is output as defect for captured image a, defect for captured image b, and no defect for captured image c, the controller 110 outputs the first It can be determined that only the second coating layer and the second coating layer are defective.

In this way, in order to design and learn a defect detection model to receive each of a plurality of captured images and output defects, the defect detection model identifies the surface of a product having surface defects as a product of the same kind as the inspection target, and the surface defects are on the image. It is possible to pre-learn each of the learning images obtained by photographing while irradiating pattern light whose characteristics are adjusted to be expressed in the image. That is, the defect detection model may pre-learn a learning image in which a reflected light pattern distorted due to the surface defect is captured by photographing the surface of a product having surface defects as a product of the same kind as the inspection target.

For example, even if an object to be inspected includes a defect, distortion due to the defect may not be expressed in a photographed image according to an arrangement direction of a stripe pattern of pattern light. Therefore, the learning image is an image taken in a state in which the characteristics of pattern light, such as color temperature, stripe pattern width, spacing, and arrangement direction, are adjusted so that distortion due to defects of the same defective product as the actual inspection target is expressed on the image. By configuring the defect detection model, it is possible to learn the distortion of the patterned light due to defects.

Of course, at this time, the defect detection model may learn a plurality of training images including distortion obtained from a plurality of defective products.

In addition, the defect detection model receives a plurality of captured images obtained for one inspection object as a set and outputs whether or not there are defects or the number, location, size, type or grade of the product to be inspected due to the defect. may be learned.

To this end, the defect detection model sets a plurality of learning images obtained by photographing the surface of a product having surface defects as a product of the same kind as the inspection target while changing the characteristics of pattern light as one learning image set, A plurality of training image sets of defective products may be trained.

In this case, the defect detection model, even if a plurality of training images included in one training image set includes a training image in which distortion due to a defect is not expressed, if distortion due to a defect is expressed in another training image, it is comprehensively It is possible to learn the distortion that the defects of the product show according to the characteristics of the patterned light.

For example, a defect generated in a specific paint layer by a wavelength that varies according to a color temperature may not be expressed in one learning image in one learning image set but may be expressed in another learning image. At this time, a specific type of defect is included in the learning image set In the case of learning a defect detection model by labeling information on the paint layer with defects or by labeling information on the paint layer with defects, the defect detection model receives a plurality of images taken for the actual inspection target and detects defects as well as It can also be learned to classify the type of defect.

Meanwhile, the defect detection result for each inspection target calculated by the controller 110 may be stored in the storage unit 140 of the surface defect detection apparatus 100 .

Furthermore, the surface defect detection device 100 may include an input/output unit 150 . Specifically, the input/output unit 120 may include an input unit for receiving an input from a user and an output unit for displaying information such as a job performance result or a state of the surface defect detection device 100 . For example, the input/output unit 120 may include an operation panel for receiving a user input and a display panel for displaying a screen.

Specifically, the input unit may include devices capable of receiving various types of user inputs, such as a keyboard, a physical button, a touch screen, or a microphone. Also, the output unit may include a display panel or a speaker. However, the input/output unit 120 is not limited thereto and may include a configuration supporting various input/output.

The input/output unit 150 may receive specific setting values for the lighting unit 120 or the camera unit 130 from a user and transmit them to the control unit 110 or store them in the storage unit 140 . Alternatively, the input/output unit 150 may receive a type of examination target selected by a user, and the controller 110 may adjust settings of the lighting unit 120 or the camera unit 130 according to the selected examination target type.

Furthermore, the input/output unit 150 may output defect detection results for inspection targets. For example, when the control unit 110 determines that the grade of a specific inspection target is unqualified, the input/output unit 150 may generate an alarm so that the user can exclude the corresponding inspection target product from the inspection line.

Meanwhile, the surface defect detection device 100 may include a communication unit 160 . The communication unit 160 is a means for mediating data exchange between other devices and the surface defect detection apparatus 100, and may perform wired/wireless communication with other devices or networks. To this end, the communication unit 160 may include a communication module supporting at least one of various wired/wireless communication methods. For example, the communication module may be implemented in the form of a chipset.

The wireless communication supported by the communication unit 160 may be, for example, Wireless Fidelity (Wi-Fi), Wi-Fi Direct, Bluetooth, Ultra Wide Band (UWB), or Near Field Communication (NFC). In addition, wired communication supported by the communication unit 140 may be, for example, USB or High Definition Multimedia Interface (HDMI).

The communication unit 160 may transmit a defect detection result from the surface defect detection device 100 to another device, for example, a separate user device.

Meanwhile, a surface defect detection method performed by the above-described surface defect detection apparatus 100 will be described step by step with reference to FIGS. 3 to 5 . 3 is a diagram showing a surface defect detection method step by step according to an embodiment, and FIG. 4 is an exemplary view showing types of defects occurring in a coating layer of a product to be inspected having a glossy surface. 5 is an exemplary view showing an example of defects detected by a surface defect detection method according to an embodiment.

Meanwhile, the surface defect detection method according to the embodiment shown in FIG. 3 includes steps processed time-sequentially in the surface defect detection apparatus 100 shown in FIGS. 1 and 2 . Therefore, even if the contents are omitted below, the above description of the surface defect detection apparatus 100 shown in FIGS. 1 and 2 can also be applied to the surface defect detection method according to the embodiment shown in FIG. 3 .

First, the surface defect detection apparatus 100 acquires a photographed image of an object to be inspected, and detects a defect using the obtained photographed image.

Specifically, as shown in FIG. 3 , steps S301 to S306 may be performed to obtain one or more captured images of the object to be inspected.

First, the surface defect detection apparatus 100 irradiates pattern light having preset characteristics to an inspection target (S301). In this case, two or more characteristics of pattern light may be set for one inspection target, and thus, two or more captured images of the inspection target may be obtained. For example, three different pattern light characteristics for obtaining three photographed images may be preset. Accordingly, the surface defect detection apparatus 100 may sequentially change pattern light according to preset characteristics.

Meanwhile, in a state in which the pattern light is being irradiated onto the surface of the inspection target in step S301, the surface defect detection apparatus 100 may capture the reflected light formed by reflecting the pattern light on the surface of the inspection target (S302). Subsequently, the surface defect detection apparatus 100 stores the photographed image (S303), and then determines whether or not the photographing has been completed a preset number of times, that is, a number corresponding to the preset number of pattern light characteristics (S304). ). For example, when two different pattern light characteristics, for example, a vertical pattern light characteristic and a horizontal pattern light characteristic in which the arrangement direction of a stripe pattern of pattern light is stored in advance, the surface defect detection device 100 In step S304, it is possible to check whether two times of image capturing according to each characteristic have been completed.

When the photographing is completed, the surface defect detection apparatus 100 may detect defects of the inspection target based on one or more stored captured images (S307).

On the other hand, when it is determined that the set number of times of shooting has not been completed in step S304, the surface defect detection apparatus 100 sequentially changes the characteristics of the pattern light according to a preset one (S305), and irradiates the pattern light with the changed characteristics to capture the captured image. A series of processes (S301 to S304) of obtaining may be repeated.

At this time, the surface defect detection apparatus 100 can selectively change the setting value of the camera, that is, the photographing unit according to the characteristics of the pattern light set by changing in step S305 (S306). For example, you can change the camera's white balance settings.

By repeating this process, the surface defect detection apparatus 100 may acquire one or more, depending on embodiments, a plurality of captured images of the inspection target.

Also, in detecting defects using the photographed image (S307), the surface defect detection apparatus 100 may use a previously learned defect detection model.

At this time, as described above, the defect detection model may learn learning images obtained by photographing a defective product as a product of the same type as the inspection target in image units. In this case, it is preferable to train the defect detection model using only images in which distortion due to defects is expressed as training images.

Meanwhile, the defect detection model may be trained to determine defects by using a plurality of training images acquired while changing the characteristics of pattern light identically to the actual inspection environment for one defective product as one unit set.

In this case, the defect detection model is labeled with the location or type of the defect, the shape, size, depth, etc. of the defect, and learns the characteristics of distortion that appear or do not appear in a plurality of learning images sequentially photographed with different characteristics of pattern light. By doing so, it may be learned to distinguish the type of defect.

For example, as shown in FIG. 4 , various types of defects may be distributed on a glossy surface having a plurality of paint layers, such as an automobile. For example, when the undercoating layer (L2), the color painting layer (L3), and the transparent painting layer (L4) are sequentially formed on the body panel (L1) of the vehicle, foreign matter 401, color aggregation 402, Defects such as poor gloss 403, bubbles 404, and scratches 405 and 406 may be distributed.

In this case, the surface defect detection apparatus 100 may calculate the number, size, type, etc. of defects from a plurality of captured images by photographing the surface of the vehicle while irradiating pattern light of different color temperatures in various directions.

In particular, by inputting a plurality of captured images as one set to the defect detection model, the defect detection model comprehensively analyzes the plurality of captured images to further determine the type, depth, distribution, and the like of defects.

In addition, the surface defect detection apparatus 100 may classify the product grade of the inspection target according to the size, number, type, and depth of the detected defects.

In this way, by detecting defects by photographing the inspection target while variously adjusting the characteristics of the pattern light, the detection performance of defects that are difficult to be detected can be greatly increased.

For example, as shown at the top of FIG. 5, the outermost layer of the glossy surface of the inspection object, for example, by irradiating the inspection object with patterned light having a color temperature corresponding to the wavelength reflected from the transparent coating layer of FIG. In the case of photographing a specularly reflected image, the detection performance does not deteriorate according to the color of the product to be inspected by using the glossy characteristics of the painted surface, and defects found on the surface can be easily detected.

In addition, as shown at the bottom of FIG. 5, the lower layer, for example, the undercoating layer (L2 of FIG. ) or an image reflected from the color coating layer L3, etc., it is possible to improve the accuracy of defect detection by maximizing the visibility of defects generated inside the coating.

The term '~unit' used in the above embodiments means software or a hardware component such as a field programmable gate array (FPGA) or ASIC, and '~unit' performs certain roles. However, '~ part' is not limited to software or hardware. '~bu' may be configured to be in an addressable storage medium and may be configured to reproduce one or more processors. Therefore, as an example, '~unit' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, and procedures. , subroutines, segments of program patent code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables.

Functions provided within components and '~units' may be combined into smaller numbers of components and '~units' or separated from additional components and '~units'.

In addition, components and '~units' may be implemented to play one or more CPUs in a device or a secure multimedia card.

The surface defect detection method according to the embodiment described with reference to FIG. 3 may be implemented in the form of a computer-readable medium storing instructions and data executable by a computer. In this case, instructions and data may be stored in the form of program codes, and when executed by a processor, a predetermined program module may be generated to perform a predetermined operation. Also, computer-readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, a computer-readable medium may be a computer recording medium, which is a volatile and non-volatile memory implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules, or other data. It may include both volatile, removable and non-removable media. For example, computer storage media include magnetic storage media such as HDDs and SSDs, optical media such as CDs, DVDs and Blu-ray discs, or media accessible through a network. It may be a memory included in the server.

In addition, the surface defect detection method according to the embodiment described with reference to FIG. 3 may be implemented as a computer program (or computer program product) including instructions executable by a computer. A computer program includes programmable machine instructions processed by a processor and may be implemented in a high-level programming language, object-oriented programming language, assembly language, or machine language. . Also, the computer program may be recorded on a tangible computer-readable recording medium (eg, a memory, a hard disk, a magnetic/optical medium, or a solid-state drive (SSD)).

Therefore, the surface defect detection method according to the embodiment described with reference to FIG. 3 can be implemented by executing the computer program as described above by a computing device. A computing device may include at least some of a processor, a memory, a storage device, a high-speed interface connected to the memory and a high-speed expansion port, and a low-speed interface connected to a low-speed bus and a storage device. Each of these components are connected to each other using various buses and may be mounted on a common motherboard or mounted in any other suitable manner.

Here, the processor may process commands within the computing device, for example, to display graphic information for providing a GUI (Graphic User Interface) on an external input/output device, such as a display connected to a high-speed interface. Examples include instructions stored in memory or storage devices. As another example, multiple processors and/or multiple buses may be used along with multiple memories and memory types as appropriate. Also, the processor may be implemented as a chipset comprising chips including a plurality of independent analog and/or digital processors.

Memory also stores information within the computing device. In one example, the memory may consist of a volatile memory unit or a collection thereof. As another example, the memory may be composed of a non-volatile memory unit or a collection thereof. Memory may also be another form of computer readable medium, such as, for example, a magnetic or optical disk.

Also, the storage device may provide a large amount of storage space to the computing device. A storage device may be a computer-readable medium or a component that includes such a medium, and may include, for example, devices in a storage area network (SAN) or other components, such as a floppy disk device, a hard disk device, an optical disk device, or a tape device, flash memory, or other semiconductor memory device or device array of the like.

The above-described embodiments are for illustrative purposes, and those skilled in the art to which the above-described embodiments belong can easily transform into other specific forms without changing the technical spirit or essential features of the above-described embodiments. You will understand. Therefore, it should be understood that the above-described embodiments are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

The scope to be protected through this specification is indicated by the following claims rather than the detailed description above, and should be construed to include all changes or modifications derived from the meaning and scope of the claims and equivalent concepts thereof. .

Claims

carried out by a surface defect detection device of an inspection object having a glossy surface,

obtaining a photographed image of the inspection target;

Including the step of detecting a defect using the acquired photographed image,

Acquiring the photographed image is,

irradiating pattern light having a stripe pattern at regular intervals on a surface of an object to be inspected;

And acquiring a photographed image by photographing the reflected light reflected from the surface of the inspection target.
According to claim 1,

Acquiring the photographed image is,

The surface defect detection method of claim 1, further comprising adjusting characteristics of the patterned light including at least one of a color temperature of the patterned light, a distance between stripe patterns, a width of the stripe pattern, and an arrangement direction of the stripe pattern.
According to claim 2,

Acquiring the photographed image is,

It is repeatedly performed a plurality of times while differently adjusting the characteristics of the pattern light according to preset conditions,

The step of detecting the defect is,

A surface defect detection method of detecting surface defects of the inspection target using a plurality of captured images obtained by repeatedly performing the acquiring of the captured images.
According to claim 3,

Acquiring the photographed image is,

The surface defect detection method of claim 1, further comprising adjusting settings of a camera for acquiring the photographed image when the characteristics of the patterned light are changed.
According to claim 3,

The surface defect detection method,

Acquiring a learning image including a reflected light pattern distorted due to the surface defect by photographing the surface of a product having a surface defect as a product of the same kind as the inspection target; and

Further comprising the step of learning a defect detection model using each acquired training image,

The step of detecting the defect is,

Surface defect detection method, wherein each of the plurality of captured images is input to the defect detection model, and when a defect is detected in at least one of the plurality of captured images, it is determined that the surface of the inspection target has a defect.
According to claim 3,

The surface defect detection method,

Obtaining a learning image set including a plurality of learning images obtained by photographing a surface of a product having surface defects as a product of the same kind as the inspection target while changing characteristics of the pattern light; and

Further comprising the step of training a defect detection model using the acquired training image set,

The step of detecting the defect is,

The surface defect detection method of detecting a defect of the inspection target by inputting the plurality of captured images as one set to the defect detection model.
In the surface defect detection device,

a lighting unit that irradiates pattern light having a stripe pattern at regular intervals on the surface of the inspection target;

a photographing unit for acquiring a photographed image by photographing reflected light reflected from the surface of the inspection target by the pattern light irradiated by the lighting unit;

a storage unit for storing the photographed image obtained by the photographing unit; and

A surface defect detection apparatus comprising a control unit for adjusting characteristics of pattern light emitted from the lighting unit, controlling the photographing unit, and detecting defects of the inspection target using a photographed image stored in the storage unit.
According to claim 7,

The control unit,

A surface defect detection device that adjusts characteristics of the patterned light including at least one of a color temperature of the patterned light emitted by the lighting unit, a distance between stripe patterns, a width of the stripe pattern, and an arrangement direction of the stripe pattern according to the inspection target.
According to claim 8,

The control unit,

A plurality of photographed images are obtained by controlling the photographing unit to obtain a photographed image whenever the characteristic of the pattern light is changed while differently adjusting the characteristics of the pattern light irradiated by the lighting unit. A surface defect detection device that detects surface defects of an inspection target.
According to claim 9,

The control unit,

A surface defect detection device that adjusts settings of the imaging unit when changing characteristics of the patterned light.
According to claim 9,

The control unit,

Each of the plurality of captured images is input to a defect detection model, and when a defect is detected in at least one of the plurality of captured images, a defect is detected by determining that the surface of the inspection target has a defect,

The defect detection model,

Surface defect detection device, which is a machine learning model obtained by pre-learning each learning image including a distorted reflected light pattern due to a surface defect, obtained by photographing the surface of a product having surface defects as a product of the same kind as the inspection target.
According to claim 9,

The control unit,

The plurality of captured images are input as one image set to a defect detection model to detect defects in the inspection target;

The defect detection model,

Surface defect detection, which is a machine learning model pre-learned from a learning image set including a plurality of learning images obtained by photographing the surface of a product having surface defects as the same product as the inspection target while changing the characteristics of the pattern light. Device.
A computer-readable recording medium in which a program for performing the method according to claim 1 is recorded.