Method for quantifying evaluation index of engraved mark formed on object by laser marking, method for evaluating engraved mark formed on object by laser marking, and method for predicting optimum laser irradiation condition for forming engraved mark on object by laser marking

Abstract

Provided is a method with which an evaluation index of an engraved mark formed on an object by laser marking can be quantified. According to one embodiment of the present invention, provided is a method for quantifying an evaluation index of an engraved mark formed on an object by laser marking, the evaluation index including at least contrast, color unevenness, hue, and engraved mark density. The method comprises: a step of calculating, as the contrast, the absolute value of the difference between the most frequent value of a laser irradiation region and the most frequent value of a laser non-irradiation region; a step of calculating, as the color unevenness, a standard deviation of the laser irradiation region; a step of calculating, as the hue, a ratio of the most frequent value of the laser irradiation region for each the colors of red, green and blue to the sum of the most frequent values of the laser irradiation region for the respective colors; and a step of calculating, as the engraved mark density, the ratio of the area of a laser marking good region to the total area of the laser marking good region and a laser marking defective region in the laser irradiation region.

Description

Method for quantifying evaluation index of inscription formed by laser marking on object, method for evaluating inscription formed by laser marking on object, and method for predicting optimal laser irradiation conditions for forming inscription by laser marking on object REFERENCE TO RELATED APPLICATIONS

[0001] This application benefits from the priority of an earlier Japanese application, Patent Application No. 2024-148384 (filing date: August 30, 2024), the entire disclosure of which is incorporated herein by reference.

[0002] The present invention relates to a method for quantifying an evaluation index of an inscription formed on an object by laser marking, a method for evaluating an inscription formed on an object by laser marking, and a method for predicting optimal laser irradiation conditions for forming an inscription on an object by laser marking.

[0003] In recent years, various codes and symbols such as barcodes and two-dimensional codes have been used in fields such as product management in markets. With the spread of traceability, many industries have adopted systems in which optical information reading devices called barcode readers or code readers are installed in factories, logistics centers, etc., and identification codes, symbols, etc. are printed or engraved on products or goods (code printing), and the information on this code printing is read by the optical information reading device.

[0004] Laser marking is a processing technology that uses laser light to irradiate metal, resin, and other target objects (products) and change their surface condition to create markings (lettering). Compared to printing and other methods, laser marking directly changes the target object, meaning engraved letters, logos, and codes are less likely to fade. Other advantages include good visibility, ease of use for QR codes (registered trademarks) and serial numbers from the perspective of traceability management, flexibility in the use of materials since it can be used on metal, resin, and glass, and consistent quality due to automated processing.

[0005] However, even if a user visually determines that there is no problem with the characters or designs (code printing) engraved by laser marking, it may be difficult to read them with a reading device (QR code reader, barcode reader, etc.). Also, even if there are multiple code printings and the users visually determine that there is no problem in selecting any of them, depending on the barcode reader, only one specific code printing may be reliably read.

[0006] To address these issues, a technology has been developed that sets printing conditions that improve the reading stability of the reading device that reads the laser-processed symbol by determining and adjusting contrast and other factors using a computer (see, for example, Patent Document 1). Also, a technology has been developed that calculates the difference in chromaticity between a printed area printed by laser marking and an unprinted area (unprocessed area) nearby as a color difference ΔE, and determines the print clarity based on the size of the color difference ΔE (see, for example, Patent Document 2).

[0007] Patent No. 5635917 Japanese Patent Application Publication No. 7290000

[0008] However, the evaluation index of the markings formed on the object by laser marking has not yet been quantified.

[0009] The present invention has been made to solve the above problems. That is, an object of the present invention is to provide a method capable of quantifying an evaluation index of an inscription formed by laser marking on an object. The present invention also provides a method for evaluating an inscription formed by laser marking on an object, and a method for predicting optimal laser irradiation conditions for forming an inscription by laser marking on an object.

[0010] [1] A method for quantifying an evaluation index of an inscription formed on an object by laser marking, the evaluation index including at least contrast, color unevenness, color tone, and inscription density, the method comprising the steps of: photographing an area of ​​the object including the inscription with an imaging device to obtain an original image which is a color image; converting the original image into a grayscale image to obtain pixel values ​​of the grayscale image; obtaining a first pixel value distribution graph from the pixel values ​​of the grayscale image; and calculating, as the contrast, the absolute value of the difference between the mode of the pixel values ​​in a laser-irradiated area, which is an area of ​​the object irradiated with a laser, and the mode of the pixel values ​​in a non-laser-irradiated area, which is an area of ​​the object not irradiated with a laser, based on the first pixel value distribution graph; and calculating, as the color unevenness, the standard deviation of the pixel values ​​in the laser-irradiated area, which is an area of ​​the object not irradiated with a laser, based on the first pixel value distribution graph. a color calculation step including a calculation step, a step of acquiring pixel values ​​of each of the colors red, green, and blue from the original image, a step of acquiring a second pixel value distribution graph for each color from the pixel values ​​of each color acquired from the original image, and a step of calculating, as the color hue, the ratio of the most frequent pixel value of the pixel values ​​of the laser irradiation area of ​​each color to the sum of the most frequent pixel values ​​of the pixel values ​​of the laser irradiation area of ​​each color based on each of the second pixel value distribution graphs; and an engraving density calculation step including a step of extracting an image of at least a portion of the laser irradiation area from the grayscale image and acquiring pixel values ​​of the extracted image, a step of acquiring a third pixel value distribution graph from the pixel values ​​of the laser irradiation area, and a step of calculating, as the engraving density, the ratio of the area of ​​the good laser marking area to the total area of ​​the good laser marking area and the poor laser marking area in the laser irradiation area based on the third pixel value distribution graph.

[0011] [2] The method according to [1] above, wherein the marking is an information code, a letter, a number, a symbol, a picture, a figure, or a combination thereof.

[0012] [3] The method according to [2] above, wherein the information code is a one-dimensional code or a two-dimensional code.

[0013] [4] The method according to any one of [1] to [3] above, wherein the contrast is 90 or more, the color unevenness is 10 or less, the color tint is 0.30 or more and 0.40 or less, and the marking density is 0.60 or more.

[0014] [5] The method according to any one of [1] to [4] above, wherein the step of calculating the absolute value of the difference between the mode of the pixel values ​​in the laser irradiation area and the mode of the pixel values ​​in the non-laser irradiation area as the contrast includes a distribution curve acquisition step of acquiring a first distribution curve of the laser irradiation area and a second distribution curve of the non-laser irradiation area by performing normal distribution fitting on the laser irradiation area and the non-laser irradiation area of ​​the first pixel value distribution graph, respectively; a mode extraction step of extracting the mode of the pixel values ​​in the laser irradiation area based on the first distribution curve and the mode of the pixel values ​​in the non-laser irradiation area based on the second distribution curve, respectively; and a mode difference calculation step of calculating the absolute value of the difference between the respective modes.

[0015] [6] The method according to [5] above, wherein at least one of the first distribution curve obtaining step, the mode extracting step, and the mode difference calculating step is performed by a program executable by a computer.

[0016] [7] The method according to any one of [1] to [6] above, wherein the step of calculating the standard deviation of the laser irradiation area as the color unevenness is carried out by a program executable by a computer.

[0017] [8] The method according to any one of [1] to [7] above, wherein the step of calculating as the color hue the proportion of the mode of the pixel values ​​of the laser irradiation area of ​​each color to the sum of the mode of the pixel values ​​of the laser irradiation area of ​​each color includes a mode extraction step of extracting the mode of the pixel values ​​of the laser irradiation area of ​​each color based on the second pixel value distribution graph of each color, and a mode proportion calculation step of calculating the proportion of the mode of the pixel values ​​of the laser irradiation area of ​​each color to the sum of the mode of the color.

[0018] [9] The method according to [8] above, wherein at least one of the second distribution curve acquisition step, the mode extraction step, and the mode ratio calculation step is performed by a program executable by a computer.

[0019]

[10] The method according to any one of [1] to [9] above, wherein the step of calculating as the marking density the ratio of the area of ​​the good laser-marked areas to the total area of ​​the good laser-marked areas and the poor laser-marked areas in the laser irradiation area includes a third distribution curve acquisition step of performing normal distribution fitting on the third pixel value distribution graph to obtain a third distribution curve, and an area ratio calculation step of calculating the areas of the good laser-marked areas and the poor laser-marked areas based on the third distribution curve, and determining the ratio of the total area of ​​the good laser-marked areas to the total area of ​​the good laser-marked areas and the poor laser-marked areas.

[0020]

[11] The method according to

[10] above, wherein at least one of the third distribution curve acquisition step and the area ratio calculation step is performed by a program executable by a computer.

[0021]

[12] A method for evaluating an inscription formed by laser marking on an object using the contrast, color unevenness, color tone, and inscription density quantified by the method described in any one of [1] to

[11] above.

[0022]

[13] A method for predicting optimal laser irradiation conditions for forming an inscription on an object by laser marking, the method using as input data at least first information regarding the laser irradiation conditions and second information regarding the contrast, color unevenness, color tone, and inscription density defined in claim 1 associated with the first information, and output data of the optimal laser irradiation conditions, by inputting at least the second information into a trained model generated by machine learning, the trained model predicting the optimal laser irradiation conditions that satisfy the input second information.

[0023]

[14] The method described in

[13] above, wherein the input data further includes third information regarding the material of the object associated with the first information and the second information, and by inputting the second information and the third information into the trained model, the optimal laser irradiation conditions that satisfy the input second information and the third information are predicted.

[0024] According to one aspect of the present invention, it is possible to quantify evaluation indices for an inscription formed by laser marking on an object. According to another aspect of the present invention, it is possible to evaluate an inscription formed by laser marking on an object. According to yet another aspect of the present invention, it is possible to predict optimal laser irradiation conditions for forming an inscription by laser marking on an object.

[0025] FIG. 1A is a system for quantifying an evaluation index of an inscription formed on an object by laser marking according to an embodiment, and FIG. 1B is a plan view of an object having an inscription used in the system shown in FIG. 1A. FIG. 2 is a diagram showing the configuration of an image processing device according to an embodiment. FIG. 3 is a flowchart of a contrast calculation step according to an embodiment. FIG. 4 is an image diagram of a first pixel value distribution graph. FIG. 5 is an image diagram of a first distribution curve obtained by performing normal distribution fitting on a laser-irradiated region of the first pixel value distribution graph. FIG. 6 is an image diagram of a second distribution curve obtained by performing normal distribution fitting on a laser-unirradiated region of the first pixel value distribution graph. FIG. 7 is a flowchart of a color calculation step according to an embodiment. FIG. 8 is an image diagram of a second pixel value distribution graph. FIG. 9 is a flowchart of an inscription density calculation step according to an embodiment. FIGS. 10A and 10B are image diagrams of a third distribution curve obtained by performing normal distribution fitting on a third pixel value distribution graph. FIG. 11 is a diagram showing the configuration of an optimal laser irradiation condition prediction device according to an embodiment. FIG. 12A is a photograph showing the marking state of Sample 1, FIG. 12B is a photograph showing the marking state of Sample 2, FIG. 12C is a photograph showing the marking state of Sample 3, and FIG. 12D is a photograph showing the marking state of Sample 4. FIG. 13 is a first pixel value distribution graph for Sample 1. FIG. 14 is a graph of a first distribution curve obtained by performing normal distribution fitting on the laser-irradiated region of the first pixel value distribution graph shown in FIG. 13. FIG. 15 is a graph of a second distribution curve obtained by performing normal distribution fitting on the laser-non-irradiated region of the first pixel value distribution graph shown in FIG. 13. FIG. 16 is a first pixel value distribution graph for Sample 2. FIG. 17 is a graph of a first distribution curve obtained by performing normal distribution fitting on the laser-irradiated region of the first pixel value distribution graph shown in FIG. 16. FIG. 18 is a graph of a second distribution curve obtained by performing normal distribution fitting on the laser-non-irradiated region of the first pixel value distribution graph shown in FIG. 16. FIG. 19 is a graph showing the second pixel value distribution of red, green, and blue for Sample 1.Fig. 20 is a second pixel value distribution graph for red, green, and blue for Sample 2. Fig. 21 is an image of a portion of the laser irradiation area extracted from the image of Sample 1. Fig. 22 is an image of a portion of the laser irradiation area extracted from the image of Sample 2. Fig. 23 is a graph showing a third distribution curve obtained by performing normal distribution fitting on the third pixel value distribution graph for Sample 1. Fig. 24 is a graph showing a third distribution curve obtained by performing normal distribution fitting on the third pixel value distribution graph for Sample 2.

[0026] The following describes a method for quantifying an evaluation index of a mark formed by laser marking on an object, a method for evaluating a mark formed by laser marking on an object, and a method for predicting optimal laser irradiation conditions for forming a mark by laser marking on an object, according to embodiments of the present invention. FIG. 1A illustrates a system for quantifying an evaluation index of a mark formed by laser marking on an object according to this embodiment. FIG. 1B illustrates a plan view of an object bearing a mark used in the system illustrated in FIG. 1A. FIG. 2 illustrates the configuration of an image processing device according to this embodiment. FIG. 3 is a flowchart of a contrast calculation step according to this embodiment. FIG. 4 illustrates a first pixel value distribution graph. FIG. 5 illustrates a first distribution curve obtained by performing normal distribution fitting on the laser-irradiated region of the first pixel value distribution graph. FIG. 6 illustrates a second distribution curve obtained by performing normal distribution fitting on the laser-unirradiated region of the first pixel value distribution graph. FIG. 7 illustrates a flowchart of a color calculation step according to this embodiment. FIG. 8 illustrates a second pixel value distribution graph. Figure 9 is a flowchart of the marking density calculation step according to this embodiment, Figures 10A and 10B are image diagrams of the third distribution curve obtained by performing normal distribution fitting on the third pixel value distribution graph, and Figure 11 is a diagram showing the configuration of an optimal laser irradiation condition prediction device according to this embodiment.

[0027] <<Method for Quantifying Evaluation Indicators of Markings Formed by Laser Marking>> First, an object 10 having a marking 11 formed by laser marking, as shown in Figures 1A and 1B, is prepared. In this specification, the term "marking" is a concept that includes information codes, letters, numbers, symbols, pictures, figures, or combinations of these. The information code may be one that can be read by an information reading device such as a barcode reader, for example, a one-dimensional code such as a one-dimensional barcode, or a two-dimensional code such as a QR code (registered trademark) or a data matrix. The marking 11 shown in Figure 1B is a two-dimensional code.

[0028] The material of the object 10 is not particularly limited, but examples thereof include resin materials, metal materials, glass, etc. When the material of the object 10 is a resin material, the resin is not particularly limited. The above-mentioned resin also includes a resin mixture obtained by blending multiple resins. Furthermore, the resin also includes resin materials to which desired properties have been imparted by adding additives such as inorganic fillers such as glass fiber and carbon fiber, nucleating agents, pigments such as carbon black and inorganic calcined pigments, antioxidants, stabilizers, plasticizers, lubricants, mold release agents, and flame retardants.

[0029] The resin material may be a thermoplastic resin, such as a polyolefin resin, a polyester resin, a polyacetal resin, a polyphenylene sulfide resin, or a polyamide resin.

[0030] The evaluation indexes for such marking 11 include at least contrast, color unevenness, color tone, and marking density. The evaluation indexes may include other indexes in addition to contrast, color unevenness, color tone, and marking density.

[0031] <Contrast> Contrast is calculated in a contrast calculation step (steps S1-1 to S1-4 shown in FIG. 3). In the contrast calculation step, first, an original image including the laser-marked marking of the object is acquired by photographing it with the imaging device 20 (see FIG. 2) (step S1-1). The original image is a color image. The acquired original image is then processed by the image processing device 30, and converted into a grayscale image with, for example, 256 levels, and pixel values ​​of the grayscale image are acquired (step S1-2). When the grayscale image has 256 levels, the pixel values ​​are in the range of 0 to 255. In this case, black is 0 and white is 255.

[0032] The image processing device 30 shown in FIG. 2 includes, for example, a processing unit 31, a storage unit 32, an input unit 33, an output unit 34, and a communication unit 35. The processing unit 31 includes a means for performing arithmetic processing, such as a CPU. The processing unit 31 performs image processing on the original image obtained by the imaging device 20. The storage unit 32 includes storage means, such as a semiconductor memory or a memory card. The storage unit 32 stores computer-executable programs and is electrically connected to the processing unit 31 so as to be accessible. The input unit 33 includes a means for inputting information. The input unit 33 includes, for example, a keyboard, a touch panel, buttons, etc., for receiving input from an administrator. The output unit 34 includes a means for outputting processing results. The output unit 34 includes, for example, a display. The communication unit 35 includes an interface for communicating with external devices via an information and communication network, such as the Internet or a LAN. Communication via the communication unit 35 may be wired or wireless.

[0033] Next, pixel values ​​are obtained from the grayscale image, and then a first pixel value distribution graph is obtained from the pixel values ​​(step S1-3, see FIG. 4). In this specification, a "pixel value distribution graph" is a graph that shows the relationship between pixel values ​​on the horizontal axis and the number of pixels (frequency) on the vertical axis. The pixel value distribution graph may be a histogram or a scatter plot.

[0034] Then, based on the first pixel value distribution graph, the difference between the most frequent pixel values ​​in the laser-irradiated area, which is the area of ​​the object 10 that has been irradiated with the laser, and the most frequent pixel values ​​in the non-laser-irradiated area, which is the area that has not been irradiated with the laser, is calculated, and the absolute value of this difference in the most frequent pixel values ​​is defined as the contrast (step S1-4). Here, since the frequency of pixel values ​​based on the color of the object increases in the non-laser-irradiated area, and the frequency of pixel values ​​based on the engraving increases in the laser-irradiated area, different peaks appear in the first pixel value distribution graph. The tops of these peaks become the most frequent values.

[0035] Step S1-4 may be performed by the following steps. First, normal distribution fitting is performed on the laser-irradiated region and non-laser-irradiated region of the first pixel value distribution graph to obtain a first distribution curve (see FIG. 5 ) that is the distribution curve of the laser-irradiated region and a second distribution curve (see FIG. 6 ) that is the distribution curve of the non-laser-irradiated region (distribution curve acquisition step). Next, the most frequent value of the peak of the laser-irradiated region is extracted from the first distribution curve, and the most frequent value of the peak of the non-laser-irradiated region is extracted from the second distribution curve (mode extraction step). Then, the absolute value of the difference between the respective modes is calculated (mode difference calculation step).

[0036] The contrast calculation step may be performed manually, but it is preferable to automate at least some of the steps (e.g., steps S1-4) by executing them using a computer-executable program (e.g., a program written in a programming language such as Python (registered trademark)) or image processing software (e.g., Image J) stored in the storage unit 32. By automating at least some of the steps in the contrast calculation step, it is possible to reduce the time required for analyzing a large number of conditions.

[0037] When step S1-4 is performed by a computer-executable program or image processing software, it is preferable to perform at least one of the steps of, for example, the first distribution curve acquisition step, the mode extraction step, and the mode difference calculation step by a program, and it is more preferable to perform all of the steps by a program or image processing software.

[0038] For example, when a program written in Python is used, execution of this program can automatically perform the acquisition of a grayscale image, the acquisition of a first pixel value distribution graph, the acquisition of the first and second distribution curves (performing normal distribution fitting), the extraction of the mode, and the calculation of the mode difference, thereby automatically performing steps S1-2 to S1-4. Furthermore, when Image J is used, this software can automatically perform the acquisition of a grayscale image, the acquisition of a first pixel value distribution graph, and the acquisition of the first and second distribution curves (performing normal distribution fitting), but the extraction of the mode and the calculation of the mode difference are performed using other software (e.g., Excel (registered trademark) manufactured by Microsoft Corporation).

[0039] From the viewpoint of obtaining good visibility when viewed by the human eye and good readability when read by a reading device such as a barcode reader, the contrast is preferably 90 or more. The lower limit of the contrast is more preferably 95 or more, or 97 or more.

[0040] <Color Unevenness> Color unevenness is calculated in a color unevenness calculation step. In the color unevenness calculation step, the standard deviation of the pixel values ​​in the laser irradiation area is calculated based on the first distribution curve, and the standard deviation is taken as the color unevenness (step S2-1). Note that the steps up to the step of acquiring the first distribution curve are the same as steps S1-1 to S1-3.

[0041] The color unevenness calculation step may be performed manually, but from the standpoint of time saving, it is preferable to automate step S2-1 by executing it using a computer-executable program stored in memory unit 32 (for example, a program written in Python, etc.).

[0042] From the viewpoint of obtaining good visibility and good readability, the color unevenness is preferably 10 or less. The upper limit of the color unevenness is more preferably 8 or less or 6 or less.

[0043] <Color> Color is calculated by a color calculation step (steps S3-1 to S3-3 shown in FIG. 7). First, pixel values ​​of each of the colors red, green, and blue, for example, at 256 gradations, are obtained from the original image, which is a color image (step S3-1). Note that the step of obtaining the original image is the same as step S1-1. Next, a second pixel value distribution graph is obtained for each color from the pixel values ​​of each color obtained from the original image (step S3-2, see FIG. 8). That is, three types of second pixel value distribution graphs are obtained: specifically, a second pixel value distribution graph for red, a second pixel value distribution graph for green, and a second pixel value distribution graph for blue.

[0044] Then, the ratio of the mode of the pixel values ​​of the laser irradiation area in each second pixel value distribution graph (the mode of the pixel values ​​of the laser irradiation area in the red second pixel value distribution graph, the mode of the pixel values ​​of the laser irradiation area in the green second pixel value distribution graph, and the mode of the pixel values ​​of the laser irradiation area in the blue second pixel value distribution graph) to the sum of the mode of the pixel values ​​of the laser irradiation area in each second pixel value distribution graph (the sum of the mode of the pixel values ​​of the laser irradiation area in the red second pixel value distribution graph, the mode of the pixel values ​​of the laser irradiation area in the green second pixel value distribution graph, and the mode of the pixel values ​​of the laser irradiation area in the blue second pixel value distribution graph) is calculated, and this ratio is used as the color (step S3-3). There are three types of color: red, green, and blue.

[0045] Step S3-3 may be performed by the following steps: First, extract the most frequent value of the peak of the laser irradiation area in the second pixel value distribution graph for each color (mode extraction step), and then calculate the ratio of the most frequent value of the pixel values ​​in the laser irradiation area for each color to the total of the most frequent values ​​for each color (mode ratio calculation step).

[0046] The color calculation step may be performed manually, but from the viewpoint of time saving, it is preferable to automate at least some of the steps, for example, step S3-3, by executing them using a computer-executable program (for example, a program written in Python, etc.) stored in the memory unit 32 or image processing software (for example, Image J, etc.).

[0047] When step S3-3 is performed by a computer-executable program, it is preferable to perform at least one of the steps of, for example, the mode extraction step and the mode ratio calculation step by the program, and it is more preferable to perform all steps by the program or image processing software.

[0048] For example, when a program written in Python is used, the acquisition of pixel values, the acquisition of the second pixel value distribution graph, the extraction of the mode, and the calculation of the mode ratio can be automatically performed by executing this program, so steps S3-1 to S3-3 can be performed automatically. Also, when Image J is used, this software can automatically acquire pixel values ​​and the acquisition of the second pixel value distribution graph, but the extraction of the mode and the calculation of the mode ratio are performed using other software (for example, Excel (registered trademark) manufactured by Microsoft Corporation).

[0049] From the viewpoint of obtaining white laser marking and good visibility, each hue is preferably 0.30 or more and 0.40 or less. The lower limit of each hue is preferably 0.31 or more or 0.32 or more, and the upper limit is preferably 0.39 or less or 0.38 or less.

[0050] <Marking density> The marking density is calculated by the marking density calculation step (steps S4-1 to S4-3 shown in FIG. 9). First, an image of at least a portion of the laser irradiation area is extracted from the grayscale image, and pixel values ​​are obtained (step S4-1). Note that the steps up to obtaining the grayscale image are the same as steps S1-1 and S1-2.

[0051] Thereafter, a third pixel value distribution graph is obtained from the pixel values ​​of the laser irradiation area (step S4-2).

[0052] After obtaining the third pixel value distribution graph, the ratio of the good laser-marked areas to the total area of ​​the good laser-marked areas and the poor laser-marked areas in the laser irradiation area is calculated based on the third pixel value distribution graph, and this ratio is defined as the marking density (step S4-3). Specifically, the marking density is calculated using the following formula (1): Marking density = area of ​​good laser-marked areas / (area of ​​good laser-marked areas + area of ​​poor laser-marked areas) ... formula (1)

[0053] Even in a laser irradiated area, in addition to areas where laser marking is sufficient (good laser marking areas), there is a risk of areas where laser marking is not performed and / or is insufficient (poor laser marking areas). For this reason, the marking density is calculated to evaluate the extent to which good laser marking areas exist in the laser irradiated area.

[0054] Step S4-3 may be performed by the following steps: First, a third distribution curve such as that shown in Figure 10A or 10B is obtained by performing normal distribution fitting on the third pixel value distribution graph (third distribution curve obtaining step), Next, the area of ​​the good laser-marked regions and the area of ​​the poor laser-marked regions are calculated, and the ratio of the total area of ​​the good laser-marked regions to the total area of ​​the good laser-marked regions and the poor laser-marked regions is calculated (area ratio calculating step).

[0055] The determination of whether a laser marking area is good or bad is made as follows: if the base of the peak in the third distribution curve is not broad as shown in FIG. 10A, then it is determined that no laser marking area is good; and if the base of the peak in the third distribution curve is broad as shown in FIG. 10B, then a predetermined lower probability (for example, the lower 1% point) is used as the threshold, and the threshold and the area including the peak top below the threshold (the area with pixel values ​​equal to or greater than the threshold in FIG. 10B) are determined to be good laser marking areas, and the area below the threshold that does not include the peak top (the area with pixel values ​​less than the threshold in FIG. 10B) are determined to be bad laser marking areas.

[0056] The marking density calculation step may be performed manually, but from the standpoint of saving time, it is preferable to automate at least some of the steps (e.g., step S4-3) by performing them using a computer-executable program (e.g., a program written in Python, etc.) or image processing software (e.g., Image J, etc.).

[0057] When step S4-3 is performed by a computer-executable program or image processing software, it is preferable to perform at least one of the steps of, for example, the third distribution curve acquisition step and the area ratio calculation step by the program, and it is more preferable to perform all steps by the program.

[0058] For example, when a program written in Python is used, execution of this program can automatically perform the acquisition of a grayscale image, the acquisition of a third pixel value distribution graph, the acquisition of a third distribution curve (performing normal distribution fitting), the calculation of the area, and the calculation of the area ratio, thereby automatically performing steps S4-1 to S4-3. Also, when Image J is used, this software can automatically perform the acquisition of a grayscale image, the acquisition of a third pixel value distribution graph, and the acquisition of a third distribution curve (performing normal distribution fitting), but the calculation of the area and the calculation of the area ratio are performed using other software (for example, Excel (registered trademark) manufactured by Microsoft Corporation).

[0059] From the viewpoint of obtaining good visibility, the marking density is preferably 0.60 or more. The lower limit of the marking density is preferably 0.70 or more, 0.80 or more, or 0.90 or more.

[0060] According to this embodiment, contrast, color unevenness, color tone, and marking density are used as evaluation indices for the marking formed on the object by laser marking, and these are expressed as numerical values, so that the evaluation indices for the marking formed on the object by laser marking can be quantified.

[0061] An engraving with a contrast of 90 or more, a color unevenness of 10 or less, a color tint of 0.30 to 0.40, and an engraving density of 0.60 or more can achieve good readability and visibility with a barcode reader. To achieve such a uniform engraving using laser marking, the laser irradiation conditions are important. For example, with regard to the printing speed (scanning speed), which is the speed at which the marking head engraves characters, symbols, etc., the laser light spreads evenly over the area that constitutes the line, so if the printing speed is doubled, the laser irradiation area also doubles. In other words, as the printing speed increases, the power density received by the surface to be printed is halved, so it is preferable not to increase the printing speed too much. For this reason, the printing speed is preferably, for example, between 1000 mm / sec and 2000 mm / sec. Furthermore, with regard to the frequency, which is the number of waves repeated per second (number of irradiated points), it is preferable to set it higher because increasing the frequency increases the density of the laser dots and makes the marked spots appear smoother. Therefore, the frequency of the laser is preferably, for example, 20 kHz or more and 40 kHz or less.

[0062] <<Method for Evaluating Markings Formed by Laser Marking>> The contrast, color unevenness, color tint, and marking density quantified by the above-mentioned method for quantifying evaluation indices are used to evaluate a marking formed by laser marking on an object. For example, if the contrast of the marking is 90 or more, the color unevenness is 10 or less, the color tint is 0.30 to 0.40, and the marking density is 0.60 or more, it can be considered that the marking is formed evenly on the object, and the marking can be evaluated as "good." On the other hand, if at least one of the contrast, color unevenness, color tint, and marking density of the marking does not satisfy the above ranges, it can be considered that the marking is uneven on the object, and can be judged as "poor." This makes it easy to determine whether the marking is good or bad.

[0063] <<Method for predicting laser irradiation conditions>> Contrast, color unevenness, color tone, and marking density change depending on the laser irradiation conditions when irradiating a laser onto an object, so the optimal laser irradiation conditions are predicted using an optimal laser irradiation condition prediction device 40 shown in Figure 11.

[0064] The optimal laser irradiation condition prediction device 40 may include a control unit 41, a memory unit 42 in which a trained model 42A is stored, an input unit 43, and an output unit 44. The control unit 41 includes a means for performing arithmetic processing, such as a CPU. The control unit 41 may have a function for storing at least first information and second information (described later) input to the memory unit 42, and a function for generating a trained model 42A using at least the first information and second information and storing the trained model 42A in the memory unit 42. The memory unit 42 includes a storage means, such as a semiconductor memory or a memory card. The storage unit 42 is accessible and connected to the processing unit 41 and stores trained parameters and other information necessary for processing in the optimal laser irradiation condition prediction device 40. The input unit 43 includes a means for inputting information including input data. The input unit 43 may include, for example, a keyboard, a touch panel, buttons, etc. for receiving input from an administrator. The output unit 44 includes a means for outputting processing results. The output unit 36 ​​may include, for example, a display for outputting processing results.

[0065] In the optimal laser irradiation condition prediction device 40, at least first information on laser irradiation conditions and second information on the contrast, color unevenness, color tone, and marking density associated with the first information are input as input data, and optimal laser irradiation conditions are output as output data. By inputting the second information into a trained model 42A generated by machine learning, optimal laser irradiation conditions that satisfy the input second information are predicted. Alternatively, the input data may further include third information on the material of the object associated with the first information and the second information, and the second information and third information may be input into the trained model to predict optimal laser irradiation conditions that satisfy the input second information and third information.

[0066] The first information may include at least one of the scanning speed, laser frequency, focus, number of scans, line spacing, laser irradiation energy, etc., but it is preferable to use all of these.

[0067] The third information includes the type of material that constitutes the object (for example, if the object is made of resin, the type of resin), the type and amount of additives added to the material, etc. If the object is made of resin, examples of additives include inorganic fillers such as glass fiber and carbon fiber, nucleating agents, pigments such as carbon black and inorganic calcined pigments, antioxidants, stabilizers, plasticizers, lubricants, release agents, and flame retardants.

[0068] The machine learning method used to generate the trained model 42A is not particularly limited as long as it performs supervised learning, but for example, neural networks, deep learning, etc. can be used.

[0069] According to this embodiment, by inputting the second information into a trained model generated by machine learning, which uses as input data at least first information regarding laser irradiation conditions and second information regarding contrast, color unevenness, color tone, and engraving density associated with the first information, and outputs optimal laser irradiation conditions, the optimal laser irradiation conditions that satisfy the input second information are predicted, making it possible to easily predict optimal laser irradiation conditions.

[0070] In order to explain the present invention in detail, the following examples are given, but the present invention is not limited to these. Fig. 12A is a photograph showing the engraved state of Sample 1, Fig. 12B is a photograph showing the engraved state of Sample 2, Fig. 12C is a photograph showing the engraved state of Sample 3, and Fig. 12D is a photograph showing the engraved state of Sample 4. Fig. 13 is a first pixel value distribution graph for Sample 1, Fig. 14 is a graph of a first distribution curve obtained by performing normal distribution fitting on the laser-irradiated region of the first pixel value distribution graph shown in Fig. 13, and Fig. 15 is a graph of a second distribution curve obtained by performing normal distribution fitting on the laser-unirradiated region of the first pixel value distribution graph shown in Fig. 13. Fig. 16 is a first pixel value distribution graph for Sample 2, Fig. 17 is a graph of a first distribution curve obtained by performing normal distribution fitting on the laser irradiation region of the first pixel value distribution graph shown in Fig. 16, and Fig. 18 is a graph showing a second distribution curve obtained by performing normal distribution fitting on the laser non-irradiation region of the pixel value distribution graph for Sample 2. Fig. 19 is a second pixel value distribution graph for red, green, and blue for Sample 1, and Fig. 20 is a second pixel value distribution graph for red, green, and blue for Sample 2. Fig. 21 is an image obtained by extracting a portion of the laser irradiation region from the image of Sample 1, and Fig. 22 is an image obtained by extracting a portion of the laser irradiation region from the image of Sample 2. Fig. 23 is a graph showing a third distribution curve obtained by performing normal distribution fitting on the third pixel value distribution graph for Sample 1, and Fig. 24 is a graph showing the third distribution curve obtained by performing normal distribution fitting on the third pixel value distribution graph for Sample 2.

[0071] <Sample Preparation> First, four black test pieces made of flat polyacetal resin measuring 100 mm long x 100 mm wide x 3 mm thick were prepared as the objects to be laser marked. Next, the surface of each test piece was irradiated with a laser under the different processing conditions shown in Table 1 to mark the test piece with a data matrix, thereby obtaining Samples 1 to 4 with the markings shown in Figures 12A to 12D. Sample 1 was marked under processing condition A, Sample 2 was marked under processing condition B, Sample 3 was marked under processing condition C, and Sample 4 was marked under processing condition D.

[0072]

[0073] <Contrast Calculation> The contrast was calculated for Samples 1 to 4. Specifically, first, an image including the two-dimensional code of Samples 1 to 4 was photographed using an imaging device (Canon LiDE210 manufactured by Canon Inc.) to obtain an original image. The original image was a color image. The obtained original image was then converted into a grayscale image with 256 gradations, and pixel values ​​were obtained from the grayscale image. The pixel values ​​ranged from 0 to 255, with black being 0 and white being 255.

[0074] After obtaining pixel values ​​from the grayscale image, a first pixel value distribution graph was obtained from the pixel values ​​( FIGS. 13 and 16 ). After obtaining the first pixel value distribution graph, normal distribution fitting was performed on the laser-irradiated region and non-laser-irradiated region of the first pixel value distribution graph to obtain a first distribution curve and a second distribution curve ( FIGS. 14 , 15 , 17 , and 18 ). Next, the mode of pixel values ​​in the laser-irradiated region was extracted based on the first distribution curve, and the mode of pixel values ​​in the non-laser-irradiated region was extracted based on the second distribution curve. The absolute value of the difference between the respective modes was then calculated, and this absolute value of the difference between the modes was used as the contrast.

[0075] The acquisition of the grayscale image, the acquisition of the first pixel value distribution graph, the acquisition of the first distribution curve and the second distribution curve (performing normal distribution fitting), the extraction of the mode, and the calculation of the mode difference were performed using a program written in Python.

[0076] <Calculation of Color Unevenness> Color unevenness was calculated for Samples 1 to 4. Specifically, the standard deviation of pixel values ​​was calculated from the first distribution curve of the laser irradiation area obtained in the contrast column, and this was taken as color unevenness. The calculation of the standard deviation was performed using a program written in Python.

[0077] <Color Calculation> Color was calculated for Samples 1 to 4. Specifically, first, the pixel values ​​of 256 gradations of red, green, and blue were obtained from the original image, which is the color image obtained in the contrast column, and a second pixel value distribution graph for each color was obtained from the pixel values ​​of each color (FIGS. 19 and 20).

[0078] After obtaining the second pixel value distribution graph, the most frequent peak value of the laser irradiation area for each color was extracted from the second pixel value distribution graph for each color.Then, the ratio of the most frequent pixel value of the laser irradiation area for each color to the sum of the most frequent pixel values ​​of each color was calculated, and this ratio was defined as the color hue.

[0079] The acquisition of the second pixel value distribution graph, the extraction of the most frequent value, and the calculation of the most frequent value ratio were performed using a program written in Python.

[0080] <Marking Density> The marking density was determined for Samples 1 to 4. Specifically, as shown in Figures 21 and 22, an image of at least a portion of the laser irradiation area was extracted from the grayscale image acquired in the contrast column, and pixel values ​​were acquired. Then, a third pixel value distribution graph was acquired from the pixel values ​​of the laser irradiation area.

[0081] After obtaining the third pixel value distribution graph, a third distribution curve was obtained by performing normal distribution fitting on the third pixel value distribution graph. Next, the areas of the laser-marked regions and the laser-marked regions were calculated based on the third distribution curve. Here, for sample 1, as shown in FIG. 23 , the base of the peak in the third distribution curve was not broad, so it was determined that no laser-marked regions were present. Similarly, for sample 4, no laser-marked regions were present. Furthermore, for sample 2, as shown in FIG. 24 , the base of the peak in the third distribution curve was broad, so the threshold value was set to 1% below the third distribution curve, pixel values ​​below this threshold were determined as laser-marked regions, and pixel values ​​above this threshold were determined as laser-marked regions. For sample 3, as with sample 2, the base of the peak in the third distribution curve was broad, so the threshold value was set to 1% below the third distribution curve, pixel values ​​below this threshold were determined as laser-marked regions, and pixel values ​​above this threshold were determined as laser-marked regions.

[0082] The area of ​​the separated good laser marking areas and the area of ​​the poor laser marking areas were then calculated, and the ratio of the total area of ​​the good laser marking areas to the total area of ​​the good laser marking areas and the poor laser marking areas was calculated using the above formula (1), and this ratio was used as the marking density.

[0083] The acquisition of the above grayscale image, acquisition of the third pixel value distribution graph, acquisition of the third distribution curve (performing normal distribution fitting), calculation of the areas of the good laser marking areas and the poor laser marking areas, and calculation of the area ratio were performed using a program written in Python.

[0084] <Evaluation Results of Samples 1 to 4> Table 2 shows the contrast, color unevenness, color tone, and marking density of Samples 1 to 4.

[0085] <Readability Evaluation> The two-dimensional codes engraved on Samples 1, 2, and 4 were read using a barcode reader with a grading function (Cognex Corporation, "DM282-9D97EE") in verification mode, Standard-Base Grading (SBG), to assess readability. The following test items were in accordance with ISO 15416, and the imaging method was based on ISO 15416, with a fixed focal distance of approximately 30 cm. Note that optimal conditions, such as illumination angle, fixed focal distance, and readability, were set on the barcode reader with a grading function through calibration. (Test Items) Unused Error Correction (UEC): The percentage of error correction capacity available for inaccurate modules. Module (MOD) and Reflection Margin (RM): Grades based on the magnitude of variation in the module's reflectivity. Axial Nonuniformity (ANU): The degree to which the module is "out of square," i.e., a measure of the overall aspect ratio of the symbol. Grid Non-Uniformity (GNU): The worst case distance between the calculated module center and the ideal location of the module center based on perfectly equally spaced modules. Fixed Pattern Damage (FPD): The combined score of all fixed pattern components. Left 'L' Side (LLS): The grade based on defects on the left 'L' side of the finder pattern. Bottom 'L' Side (BLS): The grade based on defects on the bottom 'L' side of the finder pattern. Left Quiet Zone (LQZ): The grade based on defects in the quiet zone of a one module area to the left of the left 'L' side. Bottom Quiet Zone (BQZ): The grade based on defects in the quiet zone of a one module area below the bottom 'L' side. Top Quiet Zone (TQZ): The grade based on defects in the quiet zone of a one module area above the top clock track. Right Quiet Zone (RQZ): A rating based on the imperfection of the quiet zone in the area of ​​one module to the right of the right clock track. Top Transition Ratio (TTR): A rating based on the imperfection of the top clock track in relation to the adjacent quiet zone.Right Conversion Ratio (RTR): The conversion ratio of the right clock track to the right quiet zone. Top Clock Track (TCT): A grade based on defects in the top clock track. Right Clock Track (RCT): A grade based on defects in the right clock track. Average Grade (AG): A grade that takes into account the cumulative effect of damage to several parts of the finder pattern. Decode: Reports whether the 2D symbol was decoded according to the reference decoding algorithm at the specified aperture.

[0086] The evaluation results are shown in Table 3. In the evaluation results, A is the best, and the closer to F, the worse the results.

[0087] As shown in Table 3, in Sample 2, the decoding result was "F", and in Sample 4, there were no "F" results, but there were some "C" results. In contrast, in Sample 1, all items were either "A" or "B".

[0088] From the results in Table 2 and the results of the readability evaluation in Table 3, it was confirmed that Sample 1, which satisfied all of the following conditions: contrast of 90 or more, color unevenness of 10 or less, color hue of 0.30 to 0.40, and marking density of 0.60 or more, had better readability than Samples 2 and 4, which did not satisfy any of the conditions. Therefore, it was confirmed that it is appropriate to use contrast, color unevenness, color hue, and marking density as evaluation indexes.

[0089] 10... Object 11... Marking 20... Imaging device 30... Image processing device 42A... Trained model

Claims

1. A method for quantifying evaluation indexes of an inscription formed on an object by laser marking, wherein the evaluation indexes include at least contrast, color unevenness, color tone, and inscription density, comprising: a contrast calculation step including the steps of: photographing an area of ​​the object including the inscription with an imaging device to obtain an original color image; converting the original color image into a grayscale image to obtain pixel values ​​of the grayscale image; obtaining a first pixel value distribution graph from the pixel values ​​of the grayscale image; and calculating, based on the first pixel value distribution graph, the absolute value of the difference between the most frequent pixel value in a laser-irradiated area, which is an area of ​​the object irradiated with a laser, and the most frequent pixel value in a non-laser-irradiated area, which is an area not irradiated with a laser, as the contrast; and a color unevenness calculation step including the step of calculating, based on the first pixel value distribution graph, the standard deviation of the pixel values ​​in the laser-irradiated area, which is an area of ​​the object not irradiated with a laser, as the color unevenness. a color calculation step including the steps of acquiring pixel values ​​for each of the colors red, green, and blue from the original image, acquiring a second pixel value distribution graph for each color from the pixel values ​​of each color acquired from the original image, and calculating, as the color, the ratio of the most frequent pixel value of the pixel values ​​of the laser irradiation area for each color to the sum of the most frequent pixel values ​​of the pixel values ​​of the laser irradiation area for each color based on each of the second pixel value distribution graphs; and an engraving density calculation step including the steps of extracting an image of at least a portion of the laser irradiation area from the grayscale image and acquiring pixel values ​​of the extracted image, acquiring a third pixel value distribution graph from the pixel values ​​of the laser irradiation area, and calculating, as the engraving density, the ratio of the area of ​​the good laser marking area to the total area of ​​good laser marking areas and poor laser marking areas in the laser irradiation area based on the third pixel value distribution graph.

2. The method of claim 1, wherein the indicia is an information code, a letter, a number, a symbol, a picture, a graphic, or a combination thereof.

3. The method according to claim 2, wherein the information code is a one-dimensional code or a two-dimensional code.

4. The method according to claim 1, wherein the contrast is 90 or more, the color unevenness is 10 or less, the color tint is 0.30 or more and 0.40 or less, and the marking density is 0.60 or more.

5. The method of claim 1, wherein the step of calculating as the contrast the absolute value of the difference between the mode of the pixel values ​​in the laser irradiated area and the mode of the pixel values ​​in the non-laser irradiated area comprises: a distribution curve acquisition step of acquiring a first distribution curve for the laser irradiated area and a second distribution curve for the non-laser irradiated area by performing normal distribution fitting on the laser irradiated area and the non-laser irradiated area of ​​the first pixel value distribution graph, respectively; a mode extraction step of extracting the mode of the pixel values ​​in the laser irradiated area based on the first distribution curve and the mode of the pixel values ​​in the non-laser irradiated area based on the second distribution curve, respectively; and a mode difference calculation step of calculating the absolute value of the difference between the respective modes.

6. The method according to claim 5, wherein at least one of the distribution curve obtaining step, the mode extracting step, and the mode difference calculating step is performed by a program executable by a computer.

7. The method according to claim 1, wherein the step of calculating the standard deviation of the pixel values ​​in the laser irradiated area as the color unevenness is performed by a program executable by a computer.

8. The method according to claim 1, wherein the step of calculating as the color hue the proportion of the mode of the pixel values ​​of the laser irradiation area for each color to the sum of the modes of the pixel values ​​of the laser irradiation area for each color includes a mode extraction step of extracting the mode of the pixel values ​​of the laser irradiation area for each color based on the second pixel value distribution graph for each color, and a mode proportion calculation step of calculating the proportion of the mode of the pixel values ​​of the laser irradiation area for each color to the sum of the modes of the colors.

9. The method according to claim 8, wherein at least one of the mode extracting step and the mode ratio calculating step is performed by a program executable by a computer.

10. The method of claim 1, wherein the step of calculating as the marking density the ratio of the area of ​​the good laser-marked area to the total area of ​​the good laser-marked area and the poor laser-marked area in the laser irradiation area includes a third distribution curve acquisition step of performing normal distribution fitting on the third pixel value distribution graph to obtain a third distribution curve, and an area ratio calculation step of calculating the areas of the good laser-marked area and the poor laser-marked area based on the third distribution curve, and determining the ratio of the total area of ​​the good laser-marked area to the total area of ​​the good laser-marked area and the poor laser-marked area.

11. The method according to claim 10, wherein at least one of the third distribution curve acquisition step and the area ratio calculation step is performed by a program executable by a computer.

12. A method for evaluating an inscription formed by laser marking on an object using the contrast, color unevenness, color hue, and inscription density quantified by the method of claim 1.

13. A method for predicting optimal laser irradiation conditions for forming an inscription on an object by laser marking, comprising: inputting at least first information on the laser irradiation conditions and second information on the contrast, color unevenness, color tone, and inscription density defined in claim 1 associated with the first information as input data; and inputting at least the second information into a trained model generated by machine learning, the trained model outputting the optimal laser irradiation conditions as output data, thereby predicting the optimal laser irradiation conditions that satisfy the input second information.

14. The method of claim 13, wherein the input data further includes third information regarding the material of the object associated with the first information and the second information, and by inputting the second information and the third information into the trained model, the optimal laser irradiation conditions that satisfy the input second information and the third information are predicted.

Images