Method and device for evaluating rough surface treatment of metal surfaces

Abstract

Execution: the step S2 of acquiring image data of the surface of the inspection object on which the roughening treatment is performed, the step S7 of calculating the NDSI value in each pixel of the image data with respect to light of two wavelengths selected in advance based on the image data, and the steps S8 to S12 of performing evaluation related to the roughening treatment of the inspection object based on the NDSI value.

Description

Technical Field

[0001] This disclosure relates to a method for evaluating roughening treatments applied to the surface of metals such as steel, and to an evaluation apparatus capable of performing the method. Background Technology

[0002] For the surfaces of metals such as steel, roughening treatments such as shot peening are sometimes performed to remove rust or improve the corrosion resistance of coatings. Especially in shipbuilding, roughening treatment is mandatory for the steel components that make up the hull, depending on the location. For example, areas with intact rust-preventive coatings undergo a lighter roughening treatment called sweep peening, while areas with incomplete coatings require a more severe roughening process. Inspectors then conduct on-site inspections to determine whether the treated steel surfaces meet the necessary requirements for roughness, rust removal, and cleanliness. However, these on-site inspections are almost always subjective, relying on visual evaluation. Therefore, the evaluation depends heavily on the inspector's skill and experience, making it prone to bias.

[0003] To mitigate evaluation biases arising from such evaluation methods, photographs of metal surfaces with varying degrees of roughening are sometimes used as benchmarks, and these are compared to the actual objects for evaluation. However, optical conditions at inspection sites vary, and the metals being inspected may sometimes show signs of discoloration over the years, making it difficult to achieve highly reproducible and stable evaluations even when using photographs.

[0004] Therefore, various apparatuses and methods for objectively measuring the degree of roughness treatment in metal surfaces have been invented and put into practical use (for example, see Patent Documents 1 and 2 below).

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: Japanese Patent Application Publication No. 2019-158820

[0008] Patent Document 2: Japanese Patent Application Publication No. 2019-168353 Summary of the Invention

[0009] The problem that the invention aims to solve

[0010] However, even when using the techniques described in Patent Documents 1 and 2 above, surface roughness can be evaluated, but rust removal and cleanliness cannot be evaluated. These techniques cannot replace the evaluation performed by the inspector's visual inspection. In addition to these techniques for evaluating the roughness of metal surfaces, various optical techniques and methods have been developed, but they have drawbacks such as a significantly narrow range that can be evaluated at one time. Therefore, they may not be considered sufficient as evaluation techniques for roughness.

[0011] Therefore, in view of such actual circumstances, this disclosure describes an evaluation method and apparatus for roughening the surface of a metal surface, which can easily and appropriately evaluate the roughening treatment of a metal surface.

[0012] Solution for solving the problem

[0013] This disclosure relates to an evaluation method for rough surface treatment of metal surfaces, wherein the following steps are performed:

[0014] The steps for obtaining image data of the surface of an object to be inspected after roughening the surface;

[0015] Based on the image data, the steps of calculating the NDSI value of each pixel in the image data for two pre-selected wavelengths of light; and

[0016] The steps for evaluating the surface roughness of the inspected object based on the NDSI value.

[0017] In the above-mentioned evaluation method for rough surface treatment of metal surfaces, the wavelength of light used in the calculation of NDSI value can be selected based on the following conditions.

[0018] Condition 1) The magnitude of reflectivity is positively correlated with the magnitude of surface roughness.

[0019] Condition 2) On the basis of satisfying condition 1, in the two wavelengths of reflected light, the amplitude of the reflectivity caused by the surface roughness of the object being inspected should be as small as possible, and the amplitude of the reflectivity should be as large as possible.

[0020] In the above-mentioned evaluation method for rough surface treatment of metal surfaces, the average value of the NDSI value of the pixels of the object can be calculated from the acquired image data, and the evaluation related to surface roughness can be performed based on the average value.

[0021] In the above-mentioned evaluation method for rough surface treatment of metal surfaces, the NDSI value of each pixel of the object can be compared with a preset threshold in the acquired image data to perform an evaluation related to rust.

[0022] In the above-mentioned evaluation method for rough surface treatment of metal surfaces, a range can be selected as the evaluation object from the acquired image data, and an evaluation related to rough surface treatment can be performed within the selected range.

[0023] Furthermore, this disclosure relates to an evaluation apparatus for rough surface treatment of a metal surface, wherein the apparatus is configured to perform the evaluation method for rough surface treatment of the metal surface, the evaluation apparatus comprising: an image generation unit for generating image data based on at least two pre-selected wavelengths of light; and a resolution unit for performing NDSI resolution based on the image data.

[0024] Invention Effects

[0025] The evaluation method and apparatus for rough surface treatment of metal surfaces according to the present invention can achieve excellent results in easily and appropriately evaluating the rough surface treatment of metal surfaces. Attached Figure Description

[0026] Figure 1 This is a block diagram illustrating an example of the structure of an evaluation apparatus for rough surface treatment of a metal surface according to an embodiment of the present disclosure.

[0027] Figure 2 It is a graph showing the relationship between the wavelength of light and reflectivity on a steel surface that has undergone roughening treatment.

[0028] Figure 3 It is Figure 2 A graph representing the reflectance of each wavelength relative to the reflectance of a specific wavelength, in a normalized manner.

[0029] Figure 4 This is a graph illustrating an example of the relationship between the calculated NDSI value and surface roughness.

[0030] Figure 5 This is a flowchart illustrating an example of the operational sequence of an evaluation method for rough surface treatment of a metal surface according to an embodiment of the present disclosure.

[0031] Figure 6 This is a diagram showing an example of a screen displayed on the display unit in an embodiment of this disclosure, showing an image of a measurement object based on visible light and its selection range.

[0032] Figure 7 This is a diagram showing another example of a screen displayed on the display unit in an embodiment of this disclosure, showing what an image with the selection range color-differentiated according to NDSI values ​​looks like.

[0033] Figure 8This is a diagram showing another example of a screen displayed on the display unit in an embodiment of this disclosure, showing what the image looks like after the selection range has been binarized according to the NDSI value. Detailed Implementation

[0034] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings.

[0035] Figure 1 An example of the structure of an evaluation apparatus for rough surface treatment of metal surfaces according to an embodiment of the present disclosure is shown. The evaluation apparatus 1 is a device with a simple structure comprising an irradiation unit 2 for irradiating light for inspection, an imaging unit 3 for acquiring image data of the object to be inspected, a display unit 4 for displaying various visual information, an operation unit 5 for inputting operations on each of the irradiation unit 2, imaging unit 3, and display unit 4, and a power supply unit 6 for supplying power to these units.

[0036] Illumination unit 2 is, for example, an LED lighting device that illuminates the object to be inspected with light for inspection. The light illuminated by illumination unit 2 must include light with wavelengths corresponding to the two wavelengths of reflected light described later. Here, "light with wavelengths corresponding to the reflected light of a certain wavelength (λnm)" means "light with wavelengths of reflected light that result in a wavelength of λnm when that light is incident on the object to be inspected." Furthermore, if the object to be inspected is illuminated by other light sources, thereby allowing for the smooth execution of the image acquisition and inspection procedures described later, illumination unit 2, as a component of evaluation device 1, is not necessarily required.

[0037] The imaging unit 3 includes a light-receiving unit 7, an image processing unit 8, and a resolution unit 9. The light-receiving unit 7 receives reflected light from the surface of the object being inspected, and the image processing unit 8 generates image data of the surface of the object being inspected based on the light received by the light-receiving unit 7. The resolution unit 9 performs analysis, described later, based on the image data generated by the image processing unit 8.

[0038] The light-receiving unit 7 needs to be able to detect at least the two wavelengths of reflected light described later. In addition, the light-receiving unit 7 is preferably able to detect visible light, and particularly preferably able to detect the three primary colors of RGB. As an imaging unit 3 equipped with such a light-receiving unit 7, a hyperspectral camera can be used, for example. However, as long as the imaging unit 3 is a device that can detect light including the two wavelengths mentioned above, it is sufficient to perform the analysis and inspection described later. It can also be a device with a narrower wavelength range of detectable light than a typical hyperspectral camera.

[0039] Display unit 4 is a display that shows visual information such as images acquired by imaging unit 3, images processed by analysis unit 9, and text information showing the analysis results of analysis unit 9.

[0040] The operation unit 5 is an input device for the user to input operations on various parts such as the illumination unit 2, the imaging unit 3, and the display unit 4. For example, it can be a button-type device on the main body of the imaging unit 3, or a touch panel display connected to the main body of the imaging unit 3. Furthermore, when the operation unit 5 is configured as a touch panel display, the operation unit 5 can also perform the functions of the display unit 4.

[0041] The power supply unit 6 is, for example, a battery box that houses a rechargeable battery, and supplies power to the illumination unit 2, the imaging unit 3, the display unit 4, and the operation unit 5. Furthermore, if the devices corresponding to the illumination unit 2 or the display unit 4 each have their own power supply devices such as rechargeable batteries, it is not necessary to supply power to them from the power supply unit 6 (for example, if the display unit 4 and the operation unit 5 are configured as touch panel displays, the touch panel displays usually come with a power supply device as standard equipment).

[0042] The structure of the inspection using the evaluation device 1 described above will be explained. In the inspection, a method called NDSI (Normalized Difference Spectral Index) analysis or tilt analysis, which utilizes two wavelengths of reflected light, is used. NDSI analysis refers to a method that detects two specific wavelengths of light from the surface of the object being inspected and determines the properties of the surface of the object based on the difference in their intensities. It is known that the reflectivity of light on a metal surface varies depending on the surface roughness, but the degree of change in reflectivity caused by the surface roughness also varies depending on the wavelength of the reflected light. Therefore, if two specific wavelengths of light are detected from the reflected light from the surface of the object being inspected and their reflectivity is compared, the surface roughness can be determined based on their magnitude. Specifically, light is irradiated onto the surface of the object being inspected, and two wavelengths of reflected light, such as λ1 and λ2, are detected, and their respective reflection intensities are calculated. Then, the difference between the two is used as a relative value to evaluate the magnitude using the following formula. Furthermore, in the following formula (1), R λ1 R is the reflectivity of light with wavelength λ1. λ2 It is the reflectivity of light with wavelength λ2.

[0043] (R-R2)(R+R2)(1)

[0044] The principle of such NDSI analysis is already well known, but the inventors of this application have specifically developed a method for determining the optimal wavelength of reflected light when analyzing the surface roughness of metals, and thus invented a technique that can also evaluate indicators other than surface roughness in the evaluation of rough surface treatment.

[0045] First, we will explain how to determine the appropriate wavelength of reflected light for analyzing surface roughness. Figure 2This is a graph showing the wavelength of reflected light in steel and the reflectivity of that wavelength. The five curves shown in the graph each represent the measured reflectivity of steels with different surface roughness. Furthermore, the surface roughness of the steels corresponding to the curves shown in the graph is greatest in steel A, decreasing in the order of steel B, steel C, steel D, and steel E.

[0046] As shown here, even for the same object being inspected, the intensity of reflected light varies with each wavelength. Furthermore, the rate of change in reflected light intensity caused by wavelength is not the same as that independent of surface roughness; for example, the intensity of reflected light at wavelength p varies considerably with surface roughness compared to reflected light at wavelength q.

[0047] Based on this, two wavelengths (λ1, λ2) of reflected light were selected for the inspection. The following two conditions are important in the selection of wavelengths.

[0048] Condition 1) The magnitude of reflectivity is positively correlated with the magnitude of surface roughness.

[0049] Condition 2) On the basis of satisfying condition 1, in the two wavelengths of reflected light, the amplitude of the reflectivity caused by the surface roughness of the object being inspected should be as small as possible, and the amplitude of the reflectivity should be as large as possible.

[0050] Condition 1 is a condition used for verifying the basic correctness of the guarantee. For example, in Figure 2 In the diagram shown, the intensity of reflected light at wavelengths p and q is related to surface roughness (i.e., the smaller the surface roughness, the lower the reflectivity, and the larger the surface roughness, the higher the reflectivity). However, for reflected light at wavelength r, the reflectivity is reversed in a portion of the surface roughness region (if considering steels D and E, the reflectivity of reflected light in steel E, which has lower surface roughness, is higher than that in steel D, which has higher surface roughness). The wavelengths at which this phenomenon is observed do not meet condition 1 and are unsuitable for inspection.

[0051] Condition 2 is a condition used to improve inspection accuracy. In NDSI analysis using the above formula (1), the greater the difference in reflectivity between the two wavelengths of light, the higher the detection sensitivity, which is suitable for evaluating surface roughness. Furthermore, the phrases "as small as possible" and "as large as possible" do not mean that the reflectivity caused by surface roughness is "maximum" or "minimum" among the wavelengths that meet condition 1. Of course, wavelengths with "maximum" or "minimum" reflectivity caused by surface roughness can also be selected, but they are not necessarily limited to these wavelengths. "Selecting a wavelength with the smallest amplitude of reflectivity caused by the surface roughness of the object being inspected and the largest amplitude of the other wavelength among the two wavelengths of reflected light" means selecting two wavelengths such that, without obstruction, the difference in amplitude caused by surface roughness is large when using the light of these two wavelengths for NDSI analysis of rough surface treatment. As a benchmark, for example, when arranging the wavelengths of light in descending order of the amplitude of reflectivity caused by surface roughness (the difference between the reflectivity of the surface roughness with the highest reflectivity and the surface roughness with the lowest reflectivity), the wavelength selected from approximately the upper third is set as λ1, and the wavelength selected from approximately the lower third is set as λ2. Alternatively, the wavelengths can be selected as follows: first, the wavelengths whose reflectivity amplitude is approximately the lower third are selected as wavelength λ2, and for wavelength λ1, wavelengths with reflectivity amplitude greater than or equal to λ2 are selected.

[0052] The following is an example of a specific selected order of tasks. First, in Figure 2 In this process, the reflected light with wavelength q that satisfies condition 1 and has the smallest possible amplitude of reflectivity caused by surface roughness is selected as a reflected light (reflected light with wavelength λ2) for inspection.

[0053] Next, the reflectance of light at wavelength λ2 in each inspected object is used as a benchmark to normalize the reflectance of light at other wavelengths. Then, the reflectance is re-drawn based on this normalized reflectance. Figure 2 When dealing with graphics, such as Figure 3 As shown. In this Figure 3 In the graph, another wavelength (wavelength λ1) that meets the above conditions 1 and 2 is selected. Under the condition that the magnitude of surface roughness is positively correlated with the magnitude of reflectivity, and the amplitude of reflectivity caused by surface roughness is as large as possible, for example, the reflected light with wavelength p meets the condition, so the reflected light with wavelength p is selected as another reflected light (reflected light with wavelength λ1) for inspection.

[0054] Furthermore, the inventors of this application have discovered that when the object of inspection is steel, it is particularly suitable for inspection when light with a wavelength of approximately 620 nm to 700 nm is selected as wavelength λ1, and light with a wavelength of approximately 450 nm to 520 nm is selected as wavelength λ2, respectively (i.e., these wavelengths well meet conditions 1 and 2 above). However, these values ​​can, of course, vary depending on the type of metal constituting the object of inspection. When applying the evaluation method of this invention to metals other than iron, the two wavelengths (wavelengths λ1 and λ2) of light used for inspection can be determined using the same method as described above. Moreover, even when the object of inspection is steel, the wavelengths suitable as wavelengths λ1 and λ2 can differ from the values ​​described above, depending on the experimental method, etc. It should be noted that the values ​​illustrated above are merely examples.

[0055] Based on this principle, when using, such as Figure 1 When the evaluation device 1 shown is used to inspect the surface roughness of an object, light is first irradiated onto the object from the irradiation unit 2, and image data of the surface of the object is acquired by the imaging unit 3. That is, image data is generated by the image generation unit 8 based on the light received by the light receiving unit 7. The image data acquired here is, for example, an image taken within a range of a few cm × a few cm to about 1 m × 1 m of the surface of the object. The analysis unit 9 calculates the NDSI value using the above formula (1) based on the intensity of light with wavelengths λ1 and λ2, for each pixel included in an appropriate range in the image data (either selecting a region in the obtained image that is part of the object to be inspected, or taking the entire image data as the object). If the average value of the NDSI value obtained for each pixel is calculated, it can be evaluated as a value indicating the surface roughness.

[0056] Figure 4 This diagram illustrates an example of the actual relationship between the calculated NDSI value and the surface roughness of steel. The horizontal axis represents the surface roughness of various steels treated with different degrees of roughening using a roughness gauge, and the vertical axis represents the NDSI value calculated for each steel using the method described above. As shown here, the two values ​​exhibit a strong correlation (furthermore, with a sample size of n = 24, the correlation coefficient r = 0.958), demonstrating the usefulness of the NDSI value as an indicator of surface roughness.

[0057] Furthermore, this NDSI value can be used not only for surface roughness but also for evaluating rust removal. The inventors of this application have shown that the NDSI value based on reflected light from a metal surface varies according to surface roughness, as described above, but is also affected by rust removal. In areas where rust exists, it shows a large value regardless of surface roughness. For example, in steel, when the NDSI value is calculated using the above formula (1) based on light of wavelengths determined by the above method (620nm≤λ1≤700nm, 450nm≤λ2≤520nm), the NDSI value in areas where rust exists (rusted areas) is approximately a certain value (the specific value varies depending on the measurement environment, for example, around 60). Therefore, if the NDSI value is 60 or higher, it can be determined that rust exists in that area. That is, for example, if there is an area in the image data obtained by the evaluation device 1 where the NDSI value is 60 or higher, it can be determined that this area is a rusted area. Then, the proportion of pixels with an NDSI value less than 60 can be used to determine the degree of rust removal.

[0058] Using the evaluation device 1 as described above (refer to...) Figure 1 The order of operations for inspection can be, for example, in the order of operations for inspection. Figure 5 As shown in the flowchart.

[0059] Before the inspection, data from the whiteboard used for calibration is obtained in the imaging unit 3, and the light intensity used in the calculation of the NDSI value is set (step S1). The values ​​(R) used in the calculation of the NDSI value as expressed in the above formula (1) are as follows: λ1 and R λ2 The reflectance is a relative value, but the brightness of the whiteboard is used as the denominator when calculating this relative value. That is, in a pixel of the image obtained in a later step, the reflectance of light of a certain wavelength is obtained by dividing the intensity of light of the same wavelength in the whiteboard data obtained in step S1 by the intensity of light of that wavelength in that pixel. Furthermore, as long as the optical conditions are not significantly different, step S1 can be performed only once for each scene in each scene.

[0060] Light is irradiated onto the surface of the object to be inspected from the irradiation unit 2, and image data of the surface of the object to be inspected is obtained (step S2). Here, image data is generated for light of at least the two wavelengths mentioned above, but in addition, image data generated by light of other wavelengths, such as wavelengths equivalent to RGB, can also be generated.

[0061] Correct the brightness and other properties of the acquired image data (step S3), and smooth the light intensity data of each wavelength acquired for each pixel using a Gaussian filter or the like (step S4).

[0062] Image data is displayed on display unit 4 (step S5). If image data generated by light of wavelengths equivalent to RGB was created in step S2, then in step S5, for example... Figure 6 As shown, the image of the inspection object based on RGB can be displayed on the display unit 4.

[0063] The user of the evaluation device 1 selects the area to be evaluated as the rough surface treatment object from the image displayed on the display unit 4 (step S6). An example of the selection range at this time is... Figure 6 The area is shown as a rectangle. In the example shown here, the central region is selected because it captures the appearance of the surface that can be inspected in the displayed image. Subsequent evaluation steps related to rough surface treatment are performed within this selected area.

[0064] Furthermore, step S6 may be performed when there are foreign objects or other unwanted parts in the image obtained in step S2, and can be omitted if not necessary. Additionally, when selecting the region in step S6, the evaluation device 1 can perform this automatically without user intervention. In this case, for example, a certain region to be selected from the images displayed on the display unit 4 can be pre-stored, and the device can be set to select the stored region as the evaluation object for each image. Alternatively, the device can be set to automatically select the entire region of the displayed images as the evaluation object.

[0065] For each pixel included in the selected region, calculate the reflectance of light at the two wavelengths mentioned above, and calculate the NDSI value (refer to the above formula (1)) (step S7).

[0066] Next, based on the NDSI values ​​calculated for each pixel, evaluations related to the roughness treatment of the inspected object are performed, including surface roughness, shot peening rate, area of ​​rust, and degree of rust removal. Based on these evaluations, a final evaluation related to the quality of the roughness treatment is then conducted. The steps for calculating surface roughness and shot peening rate are steps S8 and S9, and the steps for calculating area of ​​rust and degree of rust removal are steps S10 and S11.

[0067] In step S8, the average NDSI value of each pixel obtained in step S7 is calculated. Based on this average value, the relationship between NDSI value and surface roughness, which was previously determined experimentally (see reference...). Figure 4 ), calculate the surface roughness and shot peening rate (step S9). Here, the shot peening rate refers to the ratio of the area of ​​the object surface that forms an appropriate anchoring pattern through roughening treatment, and is related to the surface roughness. Surface roughness can be calculated based on the NDSI value from Figure 4 The result can be obtained, but since surface roughness is related to the shot peening rate, it can also be based on... Figure 4The shot peening rate is determined by calculating the surface roughness.

[0068] In steps S10 and S11, the NDSI values ​​of each pixel in the acquired image are compared with a pre-set threshold to perform an evaluation related to rust. First, in step S10, the image data is binarized based on the NDSI values ​​of each pixel obtained in step S7. The threshold used for binarization is an NDSI value suitable for determining rust, determined experimentally beforehand. Pixels with NDSI values ​​above the threshold are considered rust areas, and the area and degree of rust removal can be calculated based on the proportion of pixels with NDSI values ​​above the threshold (step S11).

[0069] In steps S7 to S11, the display unit 4 can appropriately display images of the object, various numerical values, etc. For example, after calculating the NDSI value for each pixel in step S7, it can display, as follows: Figure 7 The image is displayed as shown, with each pixel differentiated by color based on its NDSI value. Furthermore, it is possible to... Figure 8 As shown, the image in which pixels were color-differentiated in step S10 is displayed as an image showing the rusted area. The user can, for example, compare these images with... Figure 6 By comparing the images shown, we can understand the situation. Figure 6 The image shows the surface of the object under inspection, indicating which areas have higher (or lower) surface roughness, shot peening rate, or which areas correspond to rust. Furthermore, if the surface of the object is covered with dirt, etc., it is possible to... Figure 6 The image shown indicates its location. Furthermore, in Figure 7 , Figure 8 In the case of an area with abnormal NDSI values ​​observed in the image shown, it is possible to... Figure 6 The images shown are used to confirm what the area looks like visually (whether there is dirt or rust in the area, etc.).

[0070] Furthermore, in display unit 4, besides... Figures 6-8 In addition to the image shown, various textual information such as values ​​can also be displayed (e.g., the average NDSI value in the selected area, the threshold NDSI value for rust detection, surface roughness, shot peening rate, rust removal degree, etc. calculated based on these values). Furthermore, if the display unit 4 is a touch panel display and also functions as the operation unit 5, operation buttons can also be displayed.

[0071] The quality of the roughening treatment is evaluated based on the surface roughness or shot peening rate obtained in step S9 and the rust removal rate obtained in step S10 (step S12). The evaluation here is determined based on the compatibility with various evaluation criteria, considering the inspection object, area, and site. In the quality evaluation of sweep shot peening in ship steel, for example, a shot peening rate of 30% or higher and a rust removal rate of 90% or higher are considered acceptable (although these values ​​are just examples, specific criteria may vary depending on the object and site). Therefore, the evaluation is based on the shot peening rate corresponding to the surface roughness calculated from the NDSI value and the rust removal rate calculated through binarization. Furthermore, the evaluation in step S12 can be performed by the analysis unit 9 of the imaging unit 3 based on pre-inputted criteria (pre-set criteria values ​​related to shot peening rate or rust removal rate corresponding to the environment of the equipment used by the company or ship owner supervision), or it can be performed manually based on the results of step S9 or step S11.

[0072] Record the results displayed in the display unit 4 (some or all of the surface roughness, shot peening rate, rust area, rust removal degree, quality of rough surface treatment, or other information) or print appropriate outputs such as using a printing press not shown (step S13), and end the inspection.

[0073] Thus, in the evaluation method and apparatus as described in this embodiment, the reflectivity of two wavelengths of light, predetermined to be suitable for inspection, is used to perform an evaluation related to surface roughness and rust degree through NDSI analysis. To date, various techniques for evaluating either surface roughness or rust degree have been proposed, but to the best of the inventors' knowledge, a technique that can easily evaluate both using the same NDSI analysis is unprecedented. When evaluating the surface roughness treatment of metal surfaces, both surface roughness and rust degree need to be evaluated. While visual inspection by an inspector fulfills this requirement, machine evaluation techniques developed to date only provide apparatus or methods for evaluating either surface roughness or rust degree, and cannot evaluate both simultaneously. As in this embodiment, both indicators can be evaluated through simple calculations, and based on them, an evaluation of the surface roughness treatment itself can be output.

[0074] Furthermore, in the evaluation device 1 described above, an area of ​​approximately 1m x 1m can be inspected on the surface of the object being inspected in a single operation. This is the same range as that evaluated visually by an inspector. While conventional devices developed for evaluating the quality of rough surface treatments exist that can only inspect a very small area of ​​the object's surface at a time, the evaluation device 1 described above allows for the inspection of a sufficiently large area in a single operation. Moreover, for the evaluation of surface roughness and rust removal, only very simple calculations are required. Therefore, even for images with a relatively large pixel count, the time required to obtain the calculation and evaluation results is very short (at most a few seconds). Thus, by acquiring an image, the evaluation related to the rough surface treatment of the object being inspected can be confirmed on the spot. In this embodiment, aside from the accuracy of the evaluation, the evaluation items, the area to be evaluated, and the time required for the evaluation can all fully replace the conventional visual inspection by an inspector.

[0075] Furthermore, since only two specific wavelengths of light are used in the evaluation, it is not necessarily necessary to utilize an expensive hyperspectral camera, and it is possible to manufacture such cameras inexpensively. Figure 1 The evaluation device 1 shown is further characterized by its battery-powered operation (6) and the elimination of power cables, allowing it to be easily brought into inspection sites such as the interior of ship hull structures for convenient inspection. Furthermore, the imaging unit 3 can be configured as a device of the same size as a standard camera, and the display unit 4 and operation unit 5 can be configured as a touch panel display, enabling even a single person to carry the evaluation device 1 to the site and perform evaluations through simple operation.

[0076] The evaluation method and apparatus for roughness treatment of metal surfaces according to this embodiment can be effectively used as a quantitative evaluation technique, replacing the current visual evaluation, for example, in inspections after shot peening before painting in shipbuilding. For instance, it can be used as a method to provide objective benchmarks when the evaluation of roughness treatment varies from person to person, where the assessment is either acceptable or unacceptable. Alternatively, it can be envisioned as a tool for training and educating beginners in painting operations. In training and education settings, it is sometimes difficult to always have an inspector or personnel with equivalent knowledge and experience; however, in such cases, it is possible to provide evaluation standards equivalent to those of an inspector, thus aiding in training and education.

[0077] Furthermore, the structure and operation sequence of the evaluation device 1 described above are merely one example. The structure and operation sequence can be appropriately modified as long as the evaluation can be performed appropriately using the same principle. For example, regarding the image generation unit 8 and the analysis unit 9 constituting the evaluation device 1, the above description illustrates the case where an image is generated and analyzed based on light received by the light-receiving unit 2 provided in the imaging unit 3. However, for example, the image generation unit 8 can obtain image data for inspection from an external device, and the analysis unit 9 can analyze it. In this case, it is sufficient for the evaluation device 1 to have at least the image generation unit 8 and the analysis unit 9. Moreover, regarding the evaluation process, for ease of explanation, the above description assumes that the evaluation operation sequence related to surface roughness and the evaluation sequence related to rust are performed in parallel. However, these can also be performed individually and sequentially.

[0078] As described above, in the evaluation method for rough surface treatment of metal surfaces in this embodiment, the following steps are performed: step S2, obtaining image data of the surface of the object to be inspected after rough surface treatment; step S7, calculating the NDSI value of each pixel in the image data based on light with respect to two pre-selected wavelengths; and steps S8 to S12, performing an evaluation related to the rough surface treatment of the object to be inspected based on the NDSI values. This allows for an objective evaluation, replacing the conventional visual inspection by an inspector.

[0079] In the above embodiments, the wavelength of light used in the calculation of the NDSI value can be selected based on the following conditions. This allows for a correct and accurate evaluation of the surface roughness treatment of the metal surface.

[0080] Condition 1) The magnitude of reflectivity is positively correlated with the magnitude of surface roughness.

[0081] Condition 2) On the basis of satisfying condition 1, in the two wavelengths of reflected light, the amplitude of the reflectivity caused by the surface roughness of the object being inspected should be as small as possible, and the amplitude of the reflectivity should be as large as possible.

[0082] In the above embodiments, it is possible to calculate the average NDSI value of the pixels that are objects in the acquired image data, and to perform an evaluation related to surface roughness based on the average value.

[0083] In the above embodiments, the NDSI value of each pixel as the object can be compared with a preset threshold in the acquired image data to perform an evaluation related to rust.

[0084] In the above embodiments, it is possible to select a range of evaluation objects from the acquired image data, and to perform an evaluation related to rough surface processing within the selected range.

[0085] Furthermore, the metal surface roughness evaluation apparatus 1 of the above embodiment includes: an image generation unit 8, which generates image data based on at least two pre-selected wavelengths of light; and a resolution unit 9, which performs NDSI resolution based on the image data, and is configured to perform the metal surface roughness evaluation method. Thus, the above-mentioned effects can be achieved with a simple device.

[0086] Therefore, according to the above embodiment, the roughness treatment of metal surfaces can be evaluated easily and appropriately.

[0087] Furthermore, the evaluation method and apparatus for rough surface treatment of metal surfaces described in this disclosure are not limited to the embodiments described above, and various modifications can be made without departing from the spirit of the present invention.

[0088] Explanation of reference numerals in the attached figures

[0089] 1: Evaluation device

[0090] 2: Irradiation part

[0091] 3: Filming Department

[0092] 4: Display Section

[0093] 5: Operations Department

[0094] 6: Power Supply Section

[0095] 7: Light-receiving part

[0096] 8: Image Production Department

[0097] 9: Analysis section.

Claims

1. A method for evaluating the roughness treatment of a metal surface, wherein, implement: The steps for obtaining image data of the surface of an object to be inspected after roughening the surface; Based on the image data, the step of calculating the NDSI value of each pixel in the image data for light of two pre-selected wavelengths; as well as The steps for evaluating the surface roughness of the inspected object based on the NDSI value.

2. The evaluation method for rough surface treatment of metal surfaces according to claim 1, wherein, The wavelength of light used in the calculation of NDSI values ​​is selected based on the following conditions: Condition 1) The magnitude of reflectivity is positively correlated with the magnitude of surface roughness; Condition 2) On the basis of satisfying condition 1, in the two wavelengths of reflected light, the amplitude of the reflectivity caused by the surface roughness of the object being inspected should be as small as possible, and the amplitude of the reflectivity should be as large as possible.

3. The evaluation method for rough surface treatment of metal surfaces according to claim 1, wherein, In the acquired image data, the average NDSI value of the pixels representing the object is calculated, and an evaluation related to surface roughness is performed based on this average value.

4. The evaluation method for rough surface treatment of metal surfaces according to claim 2, wherein, In the acquired image data, the average NDSI value of the pixels representing the object is calculated, and an evaluation related to surface roughness is performed based on this average value.

5. The evaluation method for rough surface treatment of metal surfaces according to any one of claims 1 to 4, wherein, In the acquired image data, the NDSI value of each pixel representing the object is compared with a pre-set threshold to evaluate its rust-related properties.

6. The evaluation method for rough surface treatment of a metal surface according to any one of claims 1 to 4, wherein, Within the acquired image data, a range is selected as the evaluation object, and an evaluation related to rough surface processing is performed within the selected range.

7. The evaluation method for rough surface treatment of metal surfaces according to claim 5, wherein, Within the acquired image data, a range is selected as the evaluation object, and an evaluation related to rough surface processing is performed within the selected range.

8. An evaluation device for rough surface treatment of a metal surface, wherein, The evaluation method is configured to perform the rough surface treatment of the metal surface as described in claim 1 or 2. The evaluation device includes: The image processing unit generates image data based on at least two pre-selected wavelengths of light; and The analysis unit performs NDSI analysis based on the image data.

Images