EP4681162A1

EP4681162A1 - Quantitative image analysis system for evaluating surface-based coating performance

Info

Publication number: EP4681162A1
Application number: EP24718658.8A
Authority: EP
Inventors: Yicheng Hu; Sun Hye Kim; Kevin J. Henderson; Jonathan P. Derocher; Michael W. LINSEN; Bo SHUANG; Nipun BISHT; Alicia LIEW; Matthew Nicholas BENEDICT
Original assignee: Dow Global Technologies LLC; Rohm and Haas Co
Current assignee: Dow Global Technologies LLC; Rohm and Haas Co
Priority date: 2023-03-14
Filing date: 2024-03-12
Publication date: 2026-01-21
Also published as: KR20250160985A; CN120937041A; JP2026509411A; WO2024192007A1

Abstract

A method for quantifying surface defects on a coated substrate, comprising: a) providing a system for acquiring and analyzing images; b) loading a coated substrate on the holder, wherein the coated substrate comprises a coating formed on a surface of a substrate; c) illuminating the coated substrate with a lighting system, wherein illuminating the coated substrate comprising illuminating the coated substrate at an angle of incidence of 15° to 85° relative to a plane parallel to a surface of the coated substrate to generate shadows of any defects on the surface of the coated substrate; d) acquiring at least one image of the coated substrate with an imaging system; e) transforming the at least one image of the coated substrate with an analysis unit to provide at least one transformed image and quantifying the surface defects on the coated substrate based on the at least one transformed image; and f) providing an output, wherein the output comprises a value identifying an amount or percentage of the defects on the coated substrate and/or a generated image illustrating the amount or percentage of the surface defects on the coated substrate.

Description

QUANTITATIVE IMAGE ANALYSIS SYSTEM FOR EVALUATING

SURFACE-BASED COATING PERFORMANCE

FIELD OF THE INVENTION

This invention relates generally to processes for quantitatively analyzing images for surface-based defects on coated substrates.

BACKGROUND

The appearance of a coated substrate is one of the most important performance evaluation metrics used by consumers and researchers. Defects in coated substrates can manifest themselves in many different ways, including color abnormalities, surface or texture differences, or other visible deviations. For example, the smoothness or leveling of a coated substrate may appear as a surface defect caused by shadows or variations in appearance. Resistance to defects caused by, for example, water, corrosion, dirt, grease, and efflorescence may appear as variations in the color of the coated substrate. Other defects in coated substrates may be more difficult to observe visually, such as the migration/secretion or extraction of components within the coated substrate.

Such defects in coated substrates are typically observed or measured by a human. Due to the difficulty of assessing many of the defects in coated substrates, the measurement of defects is often difficult to quantify accurately and/or reproducibly. Most measurements of defects are highly subjective and are typically measured on a simple scale, such as a numerical scale of 1 to 5, where the human observer assigns a value based on their interpretation of the surface of the coated substrate. Variations in the values assigned by one observer and the next are common, and quantification of the defects generally results in a rough estimate. Due to the subjective nature of the measurements, observations are typically normalized within each study. Therefore, data acquired by human observation from one study cannot be reliably combined with data from another study.

Attempts have been made to automate the defect detection process. U.S. Patent Application Publication No. US 2022/0082508 discloses a method for providing a coating composition-related prediction program which includes providing a database of qualitative and/or quantitative characterizations of coating surfaces and training a machine learning model to develop a composition-quality-prediction program for predicting the properties of a coating surface to be produced. However, the database of qualitative and/or quantitative characterizations are generated by manually identifying and labeling digital images, where the qualitative and/or quantitative characterization of the image is based on a scale with values assigned by a human observer. Therefore, the database is compiled with data based on human observations.

There is a need for a process that can more accurately and reproducibly detect defects in coated substrates to identify and quantify defects.

SUMMARY OF THE INVENTION

The invention relates to a method for quantifying surface defects on a coated substrate, comprising: a) providing a system for acquiring and analyzing images, comprising i) an imaging system for acquiring one or more images of the coated substrate; ii) a lighting system comprising a light source for illuminating the coated substrate; iii) a holder for holding the coated substrate in a position to be illuminated by the light source; iv) an analysis unit configured to transform the one or more images and quantitatively analyze the one or more transformed images for defects on the coated substrate, b) loading a coated substrate on the holder, wherein the coated substrate comprises a coating formed on a surface of a substrate; c) illuminating the coated substrate with the lighting system, wherein illuminating the coated substrate comprises illuminating the coated substrate at an angle of incidence of 15° to 85° relative to a plane parallel to the surface of the coated substrate to generate shadows of any defects on the surface of the coated substrate; d) acquiring at least one image of the coated substrate with the imaging system; e) transforming the at least one image of the coated substrate with the analysis unit, wherein transforming the at least one image of the coated substrate comprises processing the at least one image of the coated substrate with an algorithm selected from the group consisting of image thresholding, wavelet transformation, morphological transformation, color detection, pattern detection, clustering, and combinations thereof, to provide at least one transformed image and quantifying the surface defects on the coated substrate based on the at least one transformed image; and f) providing an output, wherein the output comprises a value identifying an amount or percentage of the defects on the coated substrate and/or a generated image illustrating the amount or percentage of the surface defects on the coated substrate.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 is a schematic drawing of a system for acquiring and analyzing images according to an embodiment of the present invention. Figure 2 is a schematic drawing of the position of a light source with respect to a coated substrate according to an embodiment of the invention.

Figure 3 is a transformed color image from a rollout charted used for the smoothness test analyzed by an image analysis system according to an embodiment of the present invention.

Figure 4 is a transformed texture image from a rollout charted used for the smoothness test analyzed by an image analysis system according to an embodiment of the present invention

Figure 5 is a black and white image generated by illumination by white light from the top of the sample from a leveling test method analyzed by an image analysis system according to an embodiment of the present invention.

Figure 6 is a transformed texture image from a leveling test method analyzed by an image analysis system according to an embodiment of the present invention.

DETAILED DESCRIPTION

The inventors have found a process for reproducibly and accurately identifying and quantifying defects in coated substrates.

As used herein, the term “coated substrate” refers to a substrate comprising a coating on a surface thereof, such as, for example, a paint coating on a metal or paper substrate. The coating preferably has a thickness of less than 500 pm, more preferably less than 300 pm, and even more preferably less than 200 pm, and a thickness preferably greater than 50 nm, more preferably greater than 100 nm, and even more preferably greater than 250 nm. Multilayer coatings may have a greater thickness. The coated substrate may also comprise multiple layers, including for example, a primer or base coat.

Preferably, the coated substrate comprises a coating selected from polyurethane coatings, epoxy coatings, acrylic coatings (including, for example, acrylic coatings, vinyl- acrylic coatings, and styrene-acrylic coatings), alkyd coatings, and zinc-rich coatings. More preferably, the coated substrate comprises a paint. The substrate may comprise a metal, plastic, wood, glass, composite, fiberglass, paper, fabric, leather, or other substrate. For testing, it is preferred that the substrate has a flat or planar surface.

Defects in or on the coated substrate may be caused by a variety of issues. Examples of defects include, but are not limited to, color defects, surface or texture defects, and migration/secretion or extraction defects.

Color defects may include stains from dirt, household stains (e.g., wine, pencil, lipstick, crayons, ink, markers, etc.), adhesion defects in which the coating does not adhere adequately to the substrate, hiding defects in which the coating allows the underlying substrate or sublayers to show through, grease resistance which evaluates the penetration of oil through the coating, early rain resistance in which a coating is tested for its ability to resist rain washout shortly after the coating is applied and cured, and efflorescence resistance in which the coatings ability to resist formation of efflorescence and alkali burnout is tested.

Texture or surface defects may be indicated by smoothness or leveling issues in which the coating may have an irregular or imperfect surface. Other texture or surface issues may include, for example, cracking or flaking.

Migration/secretion and/or extraction occurs when a component separates from other components or travels through the coating. Migration/secretion and/or extraction may occur as a result of materials contacting the coating or incompatibilities of components within the coating. For example, surfactant leaching may occur when water-soluble materials leach to the surface of the coating when it is contacted with water. Other examples include the migration of binders or additives when influenced by time, temperature, or environmental conditions. While migration/secretion and/or extraction may lead to visible defects, oftentimes such defects are not easily visible. However, the present inventors have surprisingly found that Migration/secretion and/or extraction defects may be identified and analyzed using the infrared or ultraviolet spectra. The species that migrate/secrete or may be extracted may have different properties that are observable in the infrared or ultraviolet spectra. For example, when analyzing for surfactant leaching, the inventors have found that the migrating components have a different thermal conductivity than the remaining components, making those defects readily identifiable in the infrared spectrum.

To analyze the coated substrate, a system for analyzing and quantifying defects is provided. A schematic drawing of a system for analyzing and quantifying defects 100 is shown in FIG. 1. The system 100 comprises an imaging system 10, a lighting system 20, a holder 30 for holding a coated substrate 35, and an analysis unit 40.

Imaging system 10 is configured to acquire one or more images of the coated substrate 35. The imaging system 10 may comprise, for example, a camera or an image sensor. The imaging system 10 may further comprise a filter for preferentially or selectively transmitting or blocking predetermined wavelengths of light, such as, for example, at least one channel of a predetermined wavelength.

The lighting system 20 comprises a light source for illuminating the coated substrate 35. The lighting system 20 is preferably configured to emit radiation in the visible light spectrum. The light source preferably comprises a single light source so that surface defects are visible as shadows when illuminated. Preferably, the light source is a point light source or a directional light source. Preferably, the lighting system 20 is configured to allow for adjustments to the intensity of the light, the angle of incidence on the coated substrate 35, or wavelength of light emitted. To reduce potential influence by outside lighting, the system 100 may be covered or enclosed (not shown) such that only light from the lighting system 20 is used to acquire the images. The lighting system 20 is configured to illuminate the coated substrate at an angle of incidence, a as shown in FIG. 2, of 15° to 85°, and preferably from 20° to 70°.

The substrate holder 30 is used to hold the coated substrate 35 for imaging by the imaging system 10. The holder 30 is configured to hold the coated substrate in a position to be illuminated by the lighting system 20 as it is imaged. The holder 30 may be configured to hold a single coated substrate 35 or a plurality of coated substrates. The holder 30 may be stationary or adapted to allow samples to be automatically loaded/unloaded.

Preferably, at least one of the imaging system 10 and the holder 30 is adjustable so that the position of the coated substrate 35 can be changed relative to the imaging system 10. For example, as shown in FIG. 1, the imaging system 10 may be mounted on an arm 101 attached to a vertical support 102. The arm 101 may be configured to be adjustable such that the distance between the imaging system 10 and the holder 30 can be selected. Alternatively, the arm 101 may be moveable between more than one position. In another alternative, the holder 30 may be adjustable to raise or lower the holder 30 using base 103, or the angle of the holder with respect to a fixed position may be adjusted.

Preferably, at least one of a the lighting system 20 and the holder 30 is adjustable relative to each other such that the position of the coated substrate 35 can be changed relative to the lighting system 20. For example, the lighting system 20 may be height or angle adjustable relative to the holder 30 to change the angle of incidence, a, of the light 25 as shown in FIG. 2. For example, the lighting system 20 may be adjustable to allow for a shallower or steeper angle of incidence on the coated substrate 35 to control the size of the shadows generated. For example, a coated substrate with small surface defects may be illuminated at a greater angle of incidence to increase the size of the shadows, whereas a coated substrate with larger surface defects may be illuminated at a smaller angle of incidence to decrease the size of the shadows. By controlling the angle of incidence, the size of the shadows can be controlled to minimize the amount of overlap between adjacent surface defects.

Additionally, the lighting system 20 may be adjustable to allow for rotation of the lighting system 20 around the coated substrate 35 such that the angle of incidence, a, is the same, but the light is directed on the coated substrate 35 from a different angle, e.g., from the side rather than the front of the coated substrate 35. By allowing for different lighting angles, multiple images may be acquired to identify defects that may be otherwise unaccounted for in a single image. For example, a small surface defect aligned with a larger surface defect and the source of light may fall within the shadow of the larger surface defect. By rotating the light source relative to the coated substrate 35, multiple images may be acquired and compared by the analysis unit 40, described below, to identify all of the surface defects on the coated substrate. Additionally, acquiring multiple images at different angles may allow for more accurate quantification of defects. For example, a wide but narrow defect may generate a large shadow at one angle, but a small shadow at another angle. Acquisition and analysis of multiple images therefore allows for more accurate analysis of surface defects. Such analysis would be very difficult for a human observer attempting to make a similar analysis.

The system 100 further comprises an analysis unit 40 configured to transform images acquired by the imaging system 10. The analysis unit 40 further quantitatively analyzes the transformed images to identify and/or quantify the amount or percentage of defects in or on the coating of the coated substrate 35. The analysis unit 40 may comprise, for example, a computer, workstation, notebook computer, tablet computer, or smartphone. The analysis unit 40 may comprise an application or program adapted to transform and analyze the images from the imaging system 10. Information obtained and/or generated by the system 100 may be stored locally within the analysis unit 40, a server, cloud storage, or media storage device. The analysis unit 40 is preferably configured to transform acquired images by processing the acquired images with an algorithm selected from image thresholding, wavelet transformation, morphological transformation, color detection, pattern detection, contrast detection, clustering, and combinations therefore. The transformed images may then be analyzed by the analysis unit 40 to identify and/or quantify defects in or on the coating of the coated substrate 35 and to provide an output of the analysis. Preferably, the output comprises a value signifying the quantity/percentage of defects and/or an image or data set identifying the location, size and/or quantity/percentage of defects.

Preferably, the analysis unit 40 comprises or is connected to a display comprising a graphical user interface (GUI). The GUI is preferably configured to display the output of the analysis unit 40. For example, the GUI may display a value quantifying the amount or percentage of defects present in the coated substrate 35. Alternatively, the GUI may display a transformed image identifying the location, size, and/or quantity/percentage of defects.

The method of identifying and quantifying defects in a coated substrate according to the present invention comprises providing the system for acquiring and analyzing images, loading a substrate on the holder, illuminating the coated substrate with the lighting system, acquiring at least one image of the coated substrate with the imaging system, transforming the at least one image of the coated substrate with the analysis unit to provide at least one transformed image and identifying and quantifying defects on the coated substrate based on the at least one transformed image, and providing the output.

To identify and quantify defects based on migration/secretion and/or extraction of one or more components in the coating of the coated substrate, illuminating the coated substrate with the lighting system comprises illuminating the coated substrate with radiation in the infrared spectrum, the ultraviolet spectrum or both. The images acquired by the imaging system comprise infrared and/or ultraviolet images. Migration/secretion and/or extraction can be determined by the transformed images, which show different regions based on the migrated/secreted or extracted components different absorbance or reflectance of wavelengths in the infrared or ultraviolet spectra.

Examples

A system having a similar arrangement as shown in FIG. 1 was prepared using a 5 MP camera as the imaging system, an 8 channel multi-spectrum light ring as the lighting system, and a customizable sample holder to hold a coated substrate for imaging and analysis. The 8 channel multi-spectrum light ring was configured to emit in channels consisting of ultraviolet, blue, green, yellow, red, far red, infrared, and white light. The camera was configured to have the ability to acquire images in each of the channels emitted by the lighting system.

All or a subset of the acquired images were then transformed using an image analysis algorithm. The image analysis algorithm identified and quantified the defects in the coated surface.

Smoothness Test

A coated substrate was prepared by applying a coating of paint to a substrate with a roller. The coated substrate was tested by a human tester and an image analysis system according to the present invention. For the human tester, smoothness was estimated based on observations of the roller pattern, the uniformity of the surface, and the surface roughness. The human tester provided a subjective score based on the observations on a scale of 1 to 5.

The coated substrate was also analyzed by the inventive process. An image was acquired using the system described above. For one test, a color image was reconstituted from the images from each of the light channels resulting in FIG. 3. In another test, a texture image was reconstituted from four images illuminated by white light from the top, left, right, and bottom of the sample, respectively, as shown in FIG. 4. In both tests, the acquired image was processed and transformed by applying a wavelet transform to locate defects and quantify smoothing by constructing a linear regression between pixel variance. A quantifiable and reproducible value for the amount of surface defects was provided by the image analysis process of the present invention.

Leveling Test

A coated substrate was prepared by applying paint at a set spread rate using a threaded steel rod to produce a coating with parallel ridges and valleys to simulate brush marks. The coated substrate was analyzed by a human tester and the process according to the present invention. The human tester observed the coated substrate to identify the degree to which the applied coating flowed to form a uniform film with a smooth surface based on the test method of ASTM D 4062. A rating on a scale of 1 to 10 was provided.

The coated substrate was then analyzed using the process according to the present invention. For one test, a black and white image was generated using white light illuminating the image from the top of the sample as shown in FIG. 5. In another test, a texture image was reconstituted from four images illuminated by white light from the top, left, right, and bottom of the sample, respectively, as shown in FIG. 6. In both test, the acquired image was analyzed by algorithms including edge detection to detect defect areas and wavelet transformation to provide a quantifiable leveling score.

Claims

WHAT IS CLAIMED IS:

1. A method for quantifying surface defects on a coated substrate, comprising: a) providing a system for acquiring and analyzing images, comprising i) an imaging system for acquiring one or more images of the coated substrate; ii) a lighting system comprising a light source for illuminating the coated substrate; iii) a holder for holding the coated substrate in a position to be illuminated by the light source; iv) an analysis unit configured to transform the one or more images and quantitatively analyze the one or more transformed images for defects on the coated substrate, b) loading a coated substrate on the holder, wherein the coated substrate comprises a coating formed on a surface of a substrate; c) illuminating the coated substrate with the lighting system, wherein illuminating the coated substrate comprising illuminating the coated substrate at an angle of incidence of 15° to 85° relative to a plane parallel to a surface of the coated substrate to generate shadows of any defects on the surface of the coated substrate; d) acquiring at least one image of the coated substrate with the imaging system; e) transforming the at least one image of the coated substrate with the analysis unit, wherein transforming the at least one image of the coated substrate comprises processing the at least one image of the coated substrate with an algorithm selected from the group consisting of image thresholding, wavelet transformation, morphological transformation, color detection, pattern detection, clustering, and combinations thereof, to provide at least one transformed image and quantifying the surface defects on the coated substrate based on the at least one transformed image; and f) providing an output, wherein the output comprises a value identifying an amount or percentage of the defects on the coated substrate and/or a generated image illustrating the amount or percentage of the surface defects on the coated substrate.

2. The method according to claim 1 , wherein the light source is selected from a point light source and a directional light source.

3. The method according any one of the preceding claims, wherein at least one of the imaging system and the holder is adjustable to vary at least one parameter selected from an angle between the imaging system and the holder, a distance between the imaging system and the holder, and a relative position between the imaging system and the holder, and acquiring at least one image of the coated substrate with the imaging system comprises adjusting a relative position between the imaging system and the holder to acquire at least two images of the coated substrate at different positions.

4. The method according to any one of the preceding claims, wherein at least one of the lighting system and the holder is adjustable to vary at least one parameter selected from an angle between the lighting system and the holder, a distance between the lighting system and the holder, and a relative position between the lighting system and the holder, and acquiring at least one image of the coated substrate with the imaging system comprises adjusting a relative position between the lighting system and the holder to acquire at least two images of the coated substrate at different positions.

5. The method according to claim 4, wherein the step of illuminating the coated substrate comprises illuminating the coated substrate at multiple angles and acquiring at least one image of the coated substrate comprises acquiring at least one image at each of the multiple angles.

6. The method according to claim 5, wherein the multiple angles are different angles of incidence between 15° and 85° relative to a plane parallel to the surface of the coated substrate.

7. The method according to claim 5, wherein the multiple angles are different angles formed by rotating the holder relative to the light source, wherein the holder is rotated within the plane parallel to the surface of the coated substrate.

8. The method according to any one of the preceding claims, wherein the surface defects in the coated substrate are selected from smoothness defects and leveling defects.

9. The method according to any one of the preceding claims, further comprising displaying the output on a graphical user interface (GUI).

10. The method according to any one of the preceding claims, wherein the coating comprises paint.