Method for calculating belt gradient of tire

Abstract

To provide a method for investigating a gradient to a tire axial direction of a belt included in a tire which prevents measurement errors without requiring much labor.SOLUTION: A method for calculating a belt gradient of a tire includes the steps of: acquiring an image in which a plurality of belt wire rods included in a belt of a tire appear in dot shapes, as a tire axial direction cross-sectional image; subjecting the tire axial direction cross-sectional image to image processing including binarization, thereby acquiring a binarized image where the belt wire rods and other part are divided by white and black; determining an inclination angle to the tire axial direction of a straight line connecting the belt wire rods adjacent to each other in the tire axial direction, in the binarized image; and setting the inclination angle as the belt gradient at one end position of the straight line.SELECTED DRAWING: Figure 12

Description

[Technical field]

[0001] The present invention relates to a method for calculating a belt gradient of a tire. [Background technology]

[0002] A pneumatic tire (hereafter referred to as "tire") is provided with multiple belts. Each belt is made of many parallel belt wires covered with rubber. The belt wires are inclined with respect to the circumferential direction of the tire.

[0003] In addition, each belt is significantly bent radially inward in the tire direction, particularly on both sides in the tire axial direction, so that when viewed in an axial cross section, each belt is curved rather than parallel to the tire axial direction and has a gradient at each position in the tire axial direction.

[0004] This type of belt gradient has been investigated in the past because it affects tire characteristics. The commonly used investigation method was for an operator to cut a tire in the circumferential direction to prepare a cut sample, and then use a protractor to measure the gradient of the belt wire material, which appears on the cross section of the cut sample, relative to the tire axial direction.

[0005] Also, as in Patent Document 1, a method has been proposed in which the belt wire material of the belt itself, rather than the tire, is measured by electrical means. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Publication No. 62-7533 Summary of the Invention [Problem to be solved by the invention]

[0007] However, the conventional investigation methods had problems such as being labor-intensive and prone to measurement errors. Also, with the method of Patent Document 1, it was not possible to measure the gradient of the belt wire used in the tire with respect to the tire axial direction.

[0008] Therefore, an object of the present invention is to provide a method for investigating the gradient of the belt contained in a tire with respect to the tire axial direction, which is relatively less labor-intensive and less prone to measurement errors.

Means for Solving the Problems

[0009] The present invention includes the embodiments shown below.

[0010] [1] In a method for obtaining a belt gradient, which is the inclination angle of the belt with respect to the tire axial direction at each position in the tire axial direction, the steps of: obtaining, as a tire axial direction cross-sectional image, an image in which a plurality of belt wires included in the belt of the tire appear as dots and are arranged side by side; performing image processing including binarization on the tire axial direction cross-sectional image to obtain a binarized image in which the belt wire and other parts are separated into white and black; and obtaining the inclination angle of a straight line connecting the belt wires adjacent to each other in the tire axial direction in the binarized image with respect to the tire axial direction, and setting the inclination angle as the belt gradient at the position of one end of the straight line. A method for calculating the belt gradient of a tire, including the above steps.

[0011] [2] The method for calculating the belt gradient of a tire according to [1], wherein a CT image as the tire axial direction cross-sectional image is obtained based on data acquired by photographing an inflated tire with a CT device.

[0012] [3] As the belt, the tire includes a No. 1 belt that is radially inside the tire and has a wide width, and a No. 2 belt that is radially outside the tire and has a narrow width. In the tire axial cross-sectional image and the binarized image, a plurality of belt wires included in the No. 1 belt and a plurality of belt wires included in the No. 2 belt appear in rows respectively. In the binarized image, the belt wires are sequentially selected from one side to the other side in the tire axial direction, and an association process is performed each time a selection is made. The association process checks for the presence or absence of other belt wires within a search range that is a predetermined range in the tire axial direction including the selected belt wire. If there are no other belt wires within the search range, the selected belt wire is associated with the No. 1 belt. If there are other belt wires within the search range, the belt wire that is the most radially inside within that search range is associated with the No. 1 belt. The belt wires associated with the No. 1 belt by the association process are regarded as the belt wires included in the No. 1 belt, and the belt wires not associated with the No. 1 belt by the association process are regarded as the belt wires included in the No. 2 belt. The belt gradient is calculated for each of the No. 1 belt and the No. 2 belt, and the method for calculating the belt gradient of the tire according to [1] or [2].

[0013] [4] The average value of the belt gradients at a plurality of predetermined positions arranged in the tire axial direction is used as the belt gradient at any one of the plurality of positions, and the method for calculating the belt gradient of the tire according to any one of [1] to [3].

Effect of the Invention

[0014] According to the present embodiment, the gradient of the belt with respect to the tire axial direction can be measured with relatively little effort, and measurement errors are also less likely to occur.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Embodiments for Carrying Out the Invention

[0016] The embodiments will be described with reference to the drawings. Note that the embodiments described below are merely examples, and those appropriately modified without departing from the gist of the present invention are included in the scope of the present invention.

[0017] This embodiment is executed by a CT (Computed Tomography) device that uses X-rays to acquire data on the internal structure of a pneumatic tire (hereinafter referred to as "tire"), and a calculation device that processes the data acquired by the CT device and calculates the inclination angles of the belts and belt wires included in the tire.

[0018] The CT device is a device that uses X-rays to acquire three-dimensional data on the internal structure of the tire and generates a tire internal image, which is an image representing the inside of the tire, based on the three-dimensional data. In the tire internal image, parts that are more likely to absorb X-rays appear brighter, and parts that are more likely to transmit X-rays appear darker. Therefore, metal parts such as belt wires (also called cords), bead cores, and rims appear bright, and rubber parts (hereinafter referred to as "rubber parts") appear dark. According to the CT device, information on the internal structure of the tire can be acquired based on the tire internal image with light and shade without cutting the tire.

[0019] The calculation device is realized by a computer including a processing device, a storage device, an input device, and a display device. As the storage device, a RAM (Random Access Memory), a ROM (Read Only Memory), an HDD (Hard Disk Drive), etc. are provided. Programs for executing this embodiment and tire internal images acquired by the CT device are stored in the storage device. The processing device is composed of a CPU (Central Processing Unit), etc. The processing device reads out and executes the program stored in the ROM, etc. on the RAM to execute the method of this embodiment. The input device is, for example, a mouse and a keyboard, and accepts input from a user of the calculation device (hereinafter referred to as "user"). The display device is, for example, a display, and displays an input screen for inputting by the input device, a tire internal image, various data, calculation results, etc.

[0020] In addition, the tire in the present embodiment is a tire having two belts. The two belts are a first belt that is wide and located on the inner side in the tire radial direction, and a second belt that is narrow and located on the outer side in the tire radial direction.

[0021] Each belt is formed by arranging a number of belt wires in parallel and covering them with rubber to form a single sheet. The interval between the belt wires is the same for the first belt and the second belt. Also, each belt wire is inclined with respect to the tire circumferential direction. Further, each belt wire is formed by twisting together a plurality of thinner metal wires to form a single wire. Such a belt wire is also referred to as a cord.

[0022] In the present embodiment, first, the user takes a photograph of the tire with a CT device to obtain three-dimensional data of the internal structure of the tire. Here, the tire is photographed in an inflated state, that is, a state where it is mounted on a rim and internal pressure is applied. Next, the user specifies a portion to be imaged in the three-dimensional data of the tire and obtains an internal image of that portion of the tire.

[0023] In the present embodiment, as the internal image of the tire, a belt image and a tire axial direction cross-sectional image are obtained. The belt image is an image obtained by photographing one of the plurality of belts included in the tire from the outer side in the tire radial direction. The actual belt is a curved surface shape that bends particularly greatly on both sides in the tire axial direction, but the belt image is a planar image. The belt image is a rectangular image that includes both ends of the belt in the tire axial direction and is short in the tire circumferential direction. Also, the tire axial direction cross-sectional image is an image when a plane passing through a perpendicular to the outer peripheral surface of the tire and the tire rotation axis is taken as a cross-section.

[0024] For the belt image and the tire axial cross-sectional image, coordinates are set with an arbitrary reference point being 0 (for example, the axial center position of the tire is set as the image left-right direction coordinate 0). Since the relationship (calibration value) between 1 pixel of the image and the actual length in the tire is known in advance, the coordinates can be expressed in mm. When the calibration values are different between the belt image and the tire axial cross-sectional image, the coordinates (mm) in each of the belt image and the tire axial cross-sectional image are calculated using their respective calibration values.

[0025] Figure 1 is a belt image showing the 1st belt, Figure 2 is a belt image showing the 2nd belt, and Figure 3 is an image obtained by cutting out the part of the tire axial cross-sectional image where the belt is shown. As shown in Figures 1 and 2, in each belt image, a number of belt wires are shown as bright lines extending obliquely with respect to the tire circumferential direction. Also, as shown in Figure 3, in the tire axial cross-sectional image, the cross-section of each belt wire is shown as a bright point, the rubber parts such as the tread rubber and the sidewall rubber are shown relatively dark, and the air part is almost black.

[0026] Next, the user loads the tire internal image into the calculation device. The calculation device performs the processing shown in Figure 4 on the loaded tire internal image. Specifically, based on the belt image, the calculation device calculates the inclination angle of the belt wire when viewed from the outside in the tire radial direction (this inclination angle at this time is referred to as the "temporary inclination angle") (S1), based on the tire axial cross-sectional image, calculates the belt gradient with respect to the tire axial direction (S2), and based on the temporary inclination angle and the belt gradient, calculates the three-dimensional inclination angle with respect to the tire circumferential direction (S3).

[0027] First, the calculation of the inclination angle of the belt wire (temporary inclination angle) when viewed from the outside in the tire radial direction (S1 in Figure 4) will be described. Since the same processing is performed on the two belt images, the 1st belt will be described here.

[0028] As can be seen from FIG. 5, which is an enlarged view of a part of the belt image, the belt wire shown in the belt image does not have a constant thickness or brightness. There are some slightly thicker parts (indicated by reference numeral 1 in FIG. 5) and some slightly thinner parts (indicated by reference numeral 2 in FIG. 5), and the brightness is slightly darker in some of the slightly thinner parts. This is because there are changes in the twist of the belt wire formed by twisting fine wires, resulting in parts where the fine wires are dense and parts where they are sparse. Due to such changes in thickness and brightness in each belt wire, the belt image appears to have wavy unevenness. Also, the belt wire shown in the belt image appears to be straight, but in reality, it is curved.

[0029] First, the belt image read into the calculation device is converted into a grayscale image (S1-1 in FIG. 6). By this conversion, the belt image becomes an image with 256 gradations from 0 to 255, where black is 0 and white is 255. In the following description, the numerical value representing the gradation is referred to as luminance. The larger the luminance, the closer it is to white, and the smaller the luminance, the closer it is to black. If the belt image is substantially a 256-gradation grayscale at the time of being read into the calculation device, the conversion to grayscale may be omitted.

[0030] Next, the part where the belt wire is not shown is cut out from the belt image, and only the part where the belt wire is shown is cut out (S1-2 in FIG. 6). In the following description, when simply referring to the belt image, it indicates the image after the part where the belt wire is shown is cut out. The belt image shows the 1st belt from one end to the other end in the tire axis direction. The left-right direction of the belt image corresponds to the tire axis direction, and the up-down direction of the belt image corresponds to the tire circumferential direction. The belt image is shorter in the up-down direction than in the left-right direction.

[0031] The length of the belt image in the left-right direction corresponds to the length of the 1st belt in the tire axis direction when viewed from the outside in the tire diameter direction. Also, the length of the belt image in the up-down direction is preferably 58 pixels or more. The reason why 58 pixels or more is preferable will be described later.

[0032] As shown in FIG. 7, there is unevenness in the brightness of the belt image. Specifically, unevenness occurs such that it is dark at the center in the left-right direction of the image and bright on both sides in the left-right direction of the image.

[0033] Therefore, brightness correction of the belt image is performed so that the average brightness at each location in the left-right direction of the belt image becomes the same (S1-3 in FIG. 6). Specifically, first, the average value (average brightness) of the brightness of all the pixels arranged in the up-down direction of the image at each position in the left-right direction of the image is calculated. The average brightness is distributed such that it is small at the center in the left-right direction of the image and large on both sides in the left-right direction of the image, as indicated by reference numeral 3 in FIG. 8.

[0034] Next, each brightness of all the pixels is multiplied by a value obtained by dividing the target brightness by the average brightness at the left-right position of the image to which the pixel belongs. Here, the target brightness is the brightness desired as the average value of the brightness of all the pixels in the belt image after brightness correction, and is, for example, about 170.

[0035] Expressed by a mathematical formula, the brightness B'ij after brightness correction of the pixel at the i-th position in the left-right direction and the j-th position in the up-down direction of the image is obtained by the following formula using the brightness Bij before brightness correction of the pixel, the target brightness BG, and the average brightness BAi at the left-right position of the image to which the pixel belongs.

[0036]

Equation

[0037] By this brightness correction, the unevenness in brightness in the left-right direction of the image is eliminated. The distribution of the average brightness after brightness correction is shown by reference numeral 4 in FIG. 8.

[0038] Next, binarization processing of the belt image after brightness correction is performed (S1-4 in FIG. 6). By the binarization processing, pixels with a brightness greater than the threshold value are converted into white pixels. Also, pixels with a brightness less than the threshold value are converted into black pixels. The threshold value for the binarization processing is determined as appropriate, for example, about 190. By the binarization processing, a binarized image is obtained in which the original belt wire material with high brightness is white and the rest is black. The binarized image of the belt image is referred to as a binarized belt image. In the binarized belt image, each belt wire material is a set of a plurality of white pixels.

[0039] As can be seen from FIG. 9, the binarized belt image contains a plurality of noises. As the noises, there are white linear noises extending in the left-right direction of the image (this noise is pointed by an arrow in the figure) and small white dot-like noises. Therefore, removal of these noises is performed from the binarized belt image (S1-5 in FIG. 6).

[0040] As can be seen from FIG. 10, which is an enlarged view of a part of the binarized belt image, in the part of the belt wire material, a large number of white pixels are continuous in the up-down direction. On the other hand, in the part of the white linear noise extending in the left-right direction of the image (the part pointed by an arrow in FIG. 10), the number of continuous white pixels in the up-down direction is smaller than that in the part of the belt wire material. This can also be confirmed by counting the number of continuous white pixels on the straight line L-L in FIG. 10. Therefore, a process is performed to convert the part where the number of continuous white pixels in the up-down direction of the image is smaller than the number of pixels in the up-down direction of the image of one said belt wire material from white pixels to black pixels.

[0041] Specifically, a predetermined number is determined such that a part where the number of continuous white pixels in the up-down direction of the image is more than the predetermined number becomes the part of the belt wire material, and a part where the number of continuous white pixels in the up-down direction of the image is less than or equal to the predetermined number becomes the white linear noise. The predetermined number is set as the "linear noise removal reference number". The linear noise removal reference number is set as appropriate according to the thickness of the belt wire material and the linear noise, for example, 3 pixels.

[0042] Then, the part where the number of consecutive white pixels in the vertical direction of the image is less than or equal to the linear noise removal reference number is converted from white pixels to black pixels. By repeating this conversion process from one end to the other end in the horizontal direction of the image, white linear noise extending in the horizontal direction of the binary belt image is removed from the entire binary belt image.

[0043] However, in the vicinity of the upper and lower sides of the binary belt image, even for the belt wire material, the number of consecutive white pixels is less than or equal to the linear noise removal reference number. Therefore, in the range of the linear noise removal reference number in the vertical direction of the binary belt image on the upper side and the lower side, it is excluded from the range of this noise removal process.

[0044] Also, for small white dot-like noise, compared with the part of the belt wire material, the number of aggregated white pixels is small. Therefore, a process of converting the part where the number of aggregated white pixels is less than the number of pixels representing one belt wire material from white pixels to black pixels is executed. Here, aggregation means being in contact with each other.

[0045] Specifically, a predetermined number is determined such that the part where the number of aggregated white pixels is more than the predetermined number becomes the part of the belt wire material, and the part where the number of aggregated white pixels is less than or equal to the predetermined number becomes small white dot-like noise. This predetermined number is set as the "dot-like noise removal reference number". The dot-like noise removal reference number is appropriately set according to the size of the belt wire material and dot-like noise, and is, for example, 10 pixels.

[0046] Then, the part where the number of aggregated white pixels is less than or equal to the dot-like noise removal reference number is converted from white pixels to black pixels. By executing this conversion process on the entire binary belt image, small white dot-like noise is removed from the entire binary belt image.

[0047] Furthermore, as one of the noise removal processes, dilation and erosion processing is performed. The execution order of the dilation process and the erosion process depends on the type of noise to be removed. For example, when trying to remove chips occurring on the belt wire, one erosion process is performed after one dilation process.

[0048] Here, three types of noise removal were performed, but the execution order is, for example, the removal of white linear noise extending in the left - right direction of the image, the removal of small white dot - like noise, and the dilation - erosion processing, in that order.

[0049] Next, labeling is performed to assign numbers to each belt wire in the binarized belt image after noise removal (S1 - 6 in FIG. 6). Labeling is performed by assigning different numbers to each continuous part of white pixels in the binarized belt image. Such labeling can be performed using the bwlabel function, which is a function of MATLAB (registered trademark), a numerical analysis software by MathWorks. By labeling, each belt wire can be managed by a number.

[0050] By the way, since the belt wire has a thickness corresponding to a plurality of pixels, each belt wire has a thickness of a plurality of pixels in the left - right direction of the image at the upper and lower sides of the binarized belt image. Therefore, next, the center coordinates in the left - right direction of the image of each belt wire (that is, the part of white pixels) at the upper and lower side positions of the binarized belt image are calculated (S1 - 7 in FIG. 6). This calculation can be performed using the regionprops function of MATLAB. The center points in the left - right direction of the image of the belt wire at the upper and lower side positions of the binarized belt image are used as the end points of the belt wire.

[0051] Next, the inclination angle of each belt wire in the binarized belt image with respect to the tire circumferential direction is calculated (S1 - 8 in FIG. 6). The inclination angle is calculated as the inclination angle of the straight line connecting the end points on both sides in the extension direction of the belt wire with respect to the up - down direction of the image (the up - down direction of the image coincides with the tire circumferential direction).

[0052] Specifically, first, the upper and lower two end points belonging to the same belt wire are identified. Here, the fact that the upper and lower two end points belong to the same belt wire is recognized by the fact that those end points belong to the belt wire with the same number assigned in the above labeling. Next, from the coordinates of the two end points, the horizontal interval and the vertical interval between the two end points in the image are calculated, and the inclination angle is calculated by the arctangent function using those intervals. In this way, since the inclination angle of the straight line connecting the end points on both sides in the extension direction of the belt wire is calculated, an inclination angle that is not affected by the influence of the bend or the change in thickness of the belt wire between the upper side and the lower side of the binarized belt image is calculated. By this specific method, the inclination angles of all the belt wires in the binarized belt image are calculated.

[0053] The calculated inclination angle of each belt wire is treated as the inclination angle of each belt wire at the position of the lower side of the binarized belt image.

[0054] By the way, when the binarized belt image is short in the vertical direction of the image, even if the position of the end point of the belt wire changes by 1 pixel in the horizontal direction of the image, the inclination angle of the belt wire with respect to the vertical direction (tire circumferential direction) of the image changes greatly. Therefore, the number of pixels in the vertical direction of the binarized belt image and the belt image on which it is based is determined so that the influence on the inclination angle of the belt wire with respect to the vertical direction of the image is less than 1° even if the position of the end point changes by 1 pixel in the horizontal direction of the image. The number of such pixels P is obtained as a value that satisfies the following formula.

[0055]

Equation

[0056] The minimum value of the number of pixels P that satisfies this formula is 58. Therefore, the length of the belt image in the vertical direction is preferably 58 pixels or more.

[0057] In this way, the inclination angles of the belt wire materials at respective positions in the left-right direction of the image (tire axis direction) are obtained. However, in order to improve the accuracy, at each position in the left-right direction of the image, the average of the inclination angles of a plurality of belt wire materials at the same position in the left-right direction of the image is calculated (S1-9 in FIG. 6). Then, the average value of the inclination angles of the plurality of belt wire materials at the same position in the left-right direction of the image is taken as the inclination angle of the belt wire material at that position in the left-right direction of the image.

[0058] For example, in FIG. 9, since there are 7 belt wire materials at the position of the vertically long frame F, the average of the inclination angles of those 7 belt wire materials is calculated. Then, that average value is taken as the inclination angle of the belt wire material at the position of the vertically long frame F.

[0059] Note that in the regions on both left and right sides in the left-right direction of the binarized belt image (the regions indicated by reference numeral 5 in FIG. 9), the belt wire materials reach only one of the upper side and the lower side of the binarized belt image. For such belt wire materials, the inclination angle may not be calculated, or it may be calculated. When it is calculated, the endpoints of the upper side or the lower side in the binarized belt image and the endpoints of the left side or the right side in the binarized belt image are specified, and it is calculated as the inclination angle with respect to the up-down direction of the image of the straight line connecting those two endpoints belonging to the same belt. Note that the endpoints of the left side or the right side in the binarized belt image are the center points in the up-down direction of the image of the belt wire material at the position of the left side or the right side of the binarized belt image.

[0060] When the inclination angles are not calculated for the belt wire materials that reach only one of the upper side and the lower side of the image on both left and right sides in the left-right direction of the binarized belt image, the average value of the inclination angles of the belt wire materials at the same position in the left-right direction of the image is calculated based on a smaller number of belt wire materials on both left and right sides than at the center in the left-right direction. Also, when the inclination angles are calculated for the belt wire materials that reach only one of the upper side and the lower side of the image, the average value of the inclination angles of the belt wire materials at the same position in the left-right direction of the image is calculated including those belt wire materials. Then, such an average value is taken as the inclination angle of the belt wire material at that position in the left-right direction of the image.

[0061] In this way, the inclination angle of the belt wire with respect to the tire circumferential direction (image vertical direction) at each position in the tire axial direction (image left - right direction) is calculated. The distribution of the calculated inclination angle with respect to the image left - right direction is, for example, as shown in FIG. 11.

[0062] For the second belt as well, in the same way as for the first belt, the inclination angle of the belt wire when viewed from the outside in the tire radial direction is calculated.

[0063] In this way, the inclination angle of the belt wire with respect to the tire circumferential direction, which is calculated based on the belt image when the belt is viewed from the outside in the tire radial direction, is defined as the provisional inclination angle.

[0064] Next, the calculation of the belt gradient with respect to the tire axial direction based on the tire axial cross - sectional image (S2 in FIG. 4) will be described.

[0065] First, the tire axial cross - sectional image read into the calculation device is converted into a grayscale image (S2 - 1 in FIG. 12). By the conversion, the tire axial cross - sectional image becomes a 256 - gradation image with black being 0 and white being 255. When the tire axial cross - sectional image is substantially in 256 - gradation grayscale at the time of being read into the calculation device, the conversion to grayscale may be omitted.

[0066] Note that in the tire axial cross - sectional image, the lower side is the inside in the tire radial direction and the upper side is the outside in the tire radial direction. Therefore, the second belt is shown above the first belt. The first belt extends on both sides in the tire axial direction more than the second belt. Also, the cords of the carcass ply are made of organic fibers and are not shown in the tire axial cross - sectional image.

[0067] Next, from the tire axial cross - sectional image, the part on the outer - end side rather than the center of the height from the inner - end to the outer - end in the tire radial direction is cut out. Thereby, unnecessary parts such as the bead core are cut off, and the part where the whole belt is shown is cut out (S2 - 2 in FIG. 12). The cut - out image is the image in FIG. 3.

[0068] In the following description, an image in which unnecessary parts such as bead cores are cut off and a part where the entire belt is shown is cut out is defined as a tire axial direction cross-sectional image. The tire axial direction cross-sectional image includes the first belt and the second belt from one end to the other end in the tire axial direction. The left-right direction of the tire axial direction cross-sectional image corresponds to the tire axial direction, and the up-down direction of the tire axial direction cross-sectional image corresponds to the tire diameter direction. The tire axial direction cross-sectional image is shorter in the up-down direction than in the left-right direction.

[0069] By the way, the luminance of the tire axial direction cross-sectional image is not constant in the left-right direction of the image, and there is unevenness such that it is dark in a part of the left-right direction of the image and bright in other parts. Therefore, luminance correction for eliminating such unevenness is performed (S2-3 in FIG. 12). However, in the case of the tire axial direction cross-sectional image, the average luminance calculated from all the pixels arranged in the up-down direction of the image is small at the location where the main groove is present (therefore, at the location where there are many pixels with low luminance), and the average luminance calculated from all the pixels arranged in the up-down direction of the image is large at the location where there is no main groove. Therefore, the luminance correction performed in S1-3 for the belt image is not optimal as a method of luminance correction for the tire axial direction cross-sectional image.

[0070] Therefore, the following luminance correction is performed on the tire axial direction cross-sectional image. First, for pixels with a luminance of a predetermined value (for example, 50) or less, the luminance is converted to 0. This is a process for making the luminance of the air portion around the tire 0. By this process, the tire axial direction cross-sectional image becomes an image in which only the tire portion made of belt wire and rubber has a luminance greater than 0.

[0071] Next, a calculation range is set that has a length greater than the interval between the two belt wire materials (the length of the portion without the belt wire materials between the two belt wire materials) in the left-right direction of the image, and has a range that includes at least from the outer surface to the inner surface of the tire (for example, the range from the upper side to the lower side of the tire axial cross-sectional image) in the up-down direction of the image. Here, the reason why the length of the calculation range in the left-right direction of the image is greater than the interval between the two belt wire materials is to ensure that pixels of the belt wire material with high luminance always fall within the calculation range.

[0072] Next, the calculation range moves in the left-right direction of the image, and the maximum luminance and the minimum luminance in each calculation range are detected. Here, the amount of movement of the calculation range in one step in the left-right direction of the image is the same as the length of the calculation range in the left-right direction of the image, so that all pixels are included in the calculation range once. Also, the maximum luminance is the luminance of the pixels of the belt wire material, and the minimum luminance is the luminance of the pixels of the rubber portion. Pixels with a luminance of 0 are not treated as pixels with the minimum luminance.

[0073] For each calculation range, a luminance correction value C is obtained from the maximum luminance Bmax and the minimum luminance Bmin in that calculation range according to the following formula.

[0074]

Equation

[0075] Next, the luminance B'ij after luminance correction of the pixel at the i-th position in the left-right direction of the image and the j-th position in the up-down direction of the image is obtained from the luminance Bij before luminance correction of that pixel, the minimum luminance Bmin of the calculation range to which that pixel belongs, and the luminance correction value C of the calculation range to which that pixel belongs according to the following formula.

[0076]

Equation

[0077] As can be seen from this formula, the luminance of each pixel after luminance correction is calculated based on the difference between the maximum luminance and the minimum luminance in the calculation range to which the pixel belongs, and as the ratio of the difference between the luminance of the pixel with respect to this reference and the minimum luminance of the calculation range to which the pixel belongs. Therefore, in the tire axial cross-sectional image after luminance correction based on this formula, not only is the unevenness of luminance in the left-right direction of the image eliminated, but also the contrast increases and the luminance of the belt wire becomes very large. It can also be seen from FIG. 13, which is the tire axial cross-sectional image after luminance correction, that the unevenness of luminance has been eliminated and the contrast has increased compared to before luminance correction.

[0078] By the way, the difference between the maximum luminance and the minimum luminance becomes smaller in the left-right direction of the image, that is, in a place where there is no belt wire, namely, a place where there are no pixels with high luminance. Therefore, as can be seen from the formula [Equation 3], and as shown in FIG. 14, which is a distribution diagram of the luminance correction values in the left-right direction of the image, in a place where there is no belt wire (the place surrounded by a dashed circle in the figure), the luminance correction value becomes large.

[0079] As a result, as can be seen from FIG. 13, in a place where there is no belt wire in the left-right direction of the image (the place surrounded by a dashed ellipse in the figure), pixels with high luminance are generated even though it is not a belt wire. Such pixels with high luminance although not a belt wire are scattered on both the left and right sides of the tire axial cross-sectional image and are later removed as noise.

[0080] Next, binarization processing of the belt image after luminance correction is executed (S2-4 in FIG. 12). By the binarization processing, pixels with luminance greater than the threshold value are converted into white pixels. Also, pixels with luminance less than the threshold value are converted into black pixels. The threshold value for the binarization processing is determined as appropriate and is, for example, about 120. By the binarization processing, a binarized image is obtained in which the original belt wire with high luminance is white and the rest is black. The binarized image of the tire axial cross-sectional image is referred to as a binarized cross-sectional image. In the binarized cross-sectional image, each belt wire is a set of a plurality of white pixels.

[0081] Next, noise removal from the binarized cross-sectional image is performed (S2-5 in FIG. 12).

[0082] First, on both the left and right sides of the tire axial direction cross-sectional image before binarization, as described above, due to the formula of [Equation 3], there are pixels with high luminance even though they are not belt wires. These pixels remain as white pixels in the binarized cross-sectional image. Therefore, from the binarized cross-sectional image, the belt portion and a slightly upper and lower portion thereof are cut out. Thereby, the left and right side portions having pixels with high luminance even though they are not belt wires are removed. Such cutting is performed manually by the user or by automatic detection.

[0083] Also, the binarized cross-sectional image contains small white dot-like noises. The small white dot-like noises have fewer aggregated white pixels compared to the belt wire portion. Therefore, a process of converting portions where the number of aggregated white pixels is less than the number of pixels representing one belt wire from white pixels to black pixels is performed. Here, aggregation means being in contact with each other.

[0084] Specifically, a predetermined number is determined such that a portion where the number of aggregated white pixels is more than the predetermined number becomes the belt wire portion, and a portion where the number of aggregated white pixels is less than or equal to the predetermined number becomes small white dot-like noises. The predetermined number is set as the "dot noise removal reference number". The dot noise removal reference number is appropriately set according to the size of the belt wire and dot-like noises, and is, for example, 3 pixels. Then, portions where the number of aggregated white pixels is less than or equal to the dot noise removal reference number are converted from white pixels to black pixels. By performing this conversion process on the entire binarized cross-sectional image, small white dot-like noises are removed from the entire binarized cross-sectional image.

[0085] Next, labeling is performed to assign numbers to each belt wire in the binarized cross-sectional image after noise removal (S2-6 in FIG. 12). The labeling is performed by assigning different numbers to each continuous part of white pixels in the binarized cross-sectional image. Such labeling can be performed using the bwlabel function in MATLAB. By labeling, each belt wire can be managed by a number.

[0086] Next, separation of the first belt and the second belt is performed (S2-7 in FIG. 12). For this purpose, first, the centroids of all belt wires appearing in the binarized cross-sectional image are calculated. The regionprops function in MATLAB can be used to calculate the centroids.

[0087] Also, a predetermined range in the left-right direction of the image is set as the search range. This search range is approximately the same as the interval in the left-right direction of the images of two adjacent belt wires within one belt. As a specific method for setting this search range, first, the coordinate data of the centroids of all belt wires appearing in the binarized cross-sectional image are arranged in the order from the left to the right of the image. Next, the difference between the coordinates of two adjacent centroids in the left-right direction of the image (the difference in the left-right direction coordinates of the image) is calculated. Then, the largest of the calculated differences is set as the search range. The search range set by this method is usually the interval between two adjacent belt wires in the part of the first belt that does not overlap vertically with the second belt on both sides in the left-right direction of the image.

[0088] Next, belt wires are sequentially selected from one side to the other in the left-right direction in the binarized cross-sectional image, and an association process is executed each time a selection is made. In the association process, the presence or absence of the centroids of other belt wires within the search range including the centroid of the selected belt wire (itself) is searched. If there are no other belt wires, itself is associated with the first belt. On the other hand, if there are centroids of a plurality of belt wires including itself within the search range, the belt wire with the centroid at the bottommost side (i.e., the inner side in the tire diameter direction) is associated with the first belt.

[0089] Such linking processing is executed for all the belt wire materials from one belt wire material in the left - right direction of the image to the other belt wire material. Then, all the belt wire materials that were not linked to Belt 1 are linked to Belt 2. The belt wire materials linked to Belt 1 are treated as constituting Belt 1, and the belt wire materials linked to Belt 2 are treated as constituting Belt 2.

[0090] Next, the belt gradient at the position of each belt wire material is calculated as an angle (S2 - 8 in FIG. 12). The method of calculating the belt gradient will be described taking Belt 1 as an example.

[0091] First, as shown in FIG. 15, from the coordinates of the centroids of two adjacent belt wire materials, the length l in the left - right direction (tire axis direction) of the image and the length h in the up - down direction (tire circumferential direction) of the image of these two centroids are calculated. The lengths l and h are calculated, for example, in terms of the number of pixels.

[0092] Here, the length h in the up - down direction of the image is a value with positive and negative signs. Specifically, with respect to two adjacent belt wire materials, taking one belt wire material in the previously determined left - right direction as a reference, if the other belt wire material in the left - right direction is in one of the up - down directions, the length h is positive, and if the other belt wire material in the left - right direction is in the other up - down direction, the length h is negative.

[0093] Next, by the following formula, the belt gradient β (°) at the position of the reference belt wire material (strictly speaking, the centroid position of the belt wire material) is calculated as an angle.

[0094]

Equation

[0095] By this formula, the belt gradients at the positions of all the belt wire materials belonging to Belt 1 are calculated.

[0096] However, due to the influence of the resolution of the binarized cross-sectional image, the calculated center-of-gravity position of the belt wire may not be accurate. For this reason, among others, the change in the image left-right direction position (tire axis direction position) of the calculated belt gradient may not be smooth. In that case, smoothing processing is executed (S2-9 in FIG. 12). The smoothing processing is a process of setting the average value of the belt gradients at a plurality of positions arranged in the image left-right direction as the belt gradient at any one of the plurality of positions.

[0097] FIG. 16 shows the distribution of the belt gradient before and after smoothing superimposed. The zigzag line depicts the distribution of the belt gradient before smoothing, and the smooth line depicts the distribution of the belt gradient after smoothing.

[0098] The belt gradient at each position in the image left-right direction (tire axis direction) of the second belt is also calculated in the same way as for the first belt.

[0099] Next, the calculation of the three-dimensional inclination angle of the belt wire with respect to the tire circumferential direction based on the provisional inclination angle calculated based on the belt image and the belt gradient calculated based on the tire axis direction cross-sectional image (S3 in FIG. 4) will be described.

[0100] The belt image and the tire axis direction cross-sectional image are obtained in such a form that the tire circumferential direction position at which the provisional inclination angle of the belt wire is calculated (i.e., the position of the lower side of the belt image) and the tire circumferential direction position at which the belt gradient is calculated (i.e., the position of the tire axis direction cross-sectional image) coincide, and the tire axis direction coordinates (tire axis direction position) of each belt wire at the tire circumferential direction position at which the provisional inclination angle is calculated (i.e., the position of the lower side of the belt image) and the tire axis direction coordinates (tire axis direction position) of each belt wire at the tire circumferential direction position at which the belt gradient is calculated (i.e., the position of the tire axis direction cross-sectional image) coincide. However, there may be cases where these do not coincide.

[0101] Therefore, the tire axial coordinates of each belt wire appearing in the belt image are aligned with the tire axial coordinates of each belt wire appearing in the tire axial cross-sectional image, and a process of calculating the belt gradient at the aligned tire axial coordinates is executed. In this paragraph and the next paragraph, the coordinates are strictly the coordinates of the center of gravity of the belt wire. Also, the tire axial coordinates of the same scale with a common reference point (for example, the axial center position of the tire) set to 0 are set in the belt image and the tire axial cross-sectional image.

[0102] For this process, the tire axial coordinates of two adjacent belt wires in the tire axial cross-sectional image and the data of the belt gradients at these two coordinates are used. Based on this data, the belt gradient at any tire axial coordinate between two adjacent belt wires appearing in the tire axial cross-sectional image can be calculated by interpolation (for example, linear interpolation). Therefore, the belt gradient can be calculated at the tire axial coordinates of the belt wires appearing in the belt image.

[0103] By executing such calculations for at least all belt wires where the tire axial coordinates of the belt wires do not match between the belt image and the tire axial cross-sectional image, the belt gradient can be obtained for the tire axial coordinates of all belt wires appearing in the belt image. This process can be performed using the resample function in MATLAB. By this process, both the provisional inclination angle and the belt gradient can be obtained for the tire axial coordinates of all belt wires appearing in the belt image.

[0104] Next, at each tire axial position, the three-dimensional inclination angle of the belt wire with respect to the tire circumferential direction is calculated from the provisional inclination angle and the belt gradient. The calculation method will be described using FIGS. 17 to 20.

[0105] FIG. 17 is a perspective view of the belt 10, FIG. 18 is a view of the same belt 10 as seen from the outside in the tire radial direction, FIG. 19 is a view of the same belt 10 on the tire axial cross-section, and FIG. 20 is a view for explaining the length and angle for obtaining the inclination angle. Although only a part of the belt 10 is depicted in these figures, actually the belt 10 extends around the tire in the circumferential direction. As can be seen from these figures, the belt wire 20 included in the belt 10 is inclined not only with respect to the tire circumferential direction when viewed from the outside in the tire radial direction, but also along the belt gradient, so it is three-dimensionally inclined with respect to the tire circumferential direction.

[0106] First, looking at FIG. 18, between the inclination angle α of the belt wire 20 with respect to the tire circumferential direction (this inclination angle α is a provisional inclination angle), the length a in the tire circumferential direction of the belt wire 20 (the length between P and Q), and the length b1 in the tire axial direction of the belt wire 20 (the length between Q and R), when each is viewed from the outside in the tire radial direction, the following equation holds. Note that the lengths a and b1 are the lengths within a certain range in the tire circumferential direction and within the range shown in FIGS. 17 to 20.

[0107]

Equation

[0108] Therefore, the following equation holds.

[0109]

Equation

[0110] Next, looking at FIG. 19, between the inclination angle β of the belt wire 20 with respect to the tire axial direction (this inclination angle β is the belt gradient), the length b1 in the tire axial direction of the belt wire 20, and the length b2 in the direction of the inclination angle β of the belt wire 20 (the length between Q and S), when each is viewed on the tire axial cross-section, the following equation holds. Note that the length b2 is the length within a certain range in the tire circumferential direction and within the range shown in FIGS. 17 to 20.

[0111]

Equation

[0112] Next, referring to FIG. 20, it can be seen that the following equation holds for the three-dimensional inclination angle θ of the belt material 20 with respect to the tire axis direction.

[0113]

Equation

[0114] Therefore, the three-dimensional inclination angle θ of the belt material 20 with respect to the tire axis direction is obtained by the following equation.

[0115]

Equation

[0116] For each belt material, the three-dimensional inclination angle with respect to the tire circumferential direction is obtained by this equation. Also, for each of the first belt and the second belt, the three-dimensional inclination angles of the respective belt materials with respect to the tire circumferential direction are obtained.

[0117] The obtained three-dimensional inclination angles are displayed on the display device as a graph as shown in FIG. 21. Although only the inclination angle of the first belt is shown as a representative in FIG. 21, the inclination angle of the second belt is actually also displayed.

[0118] In FIG. 21, the solid lines are the three-dimensional inclination angles of the belt materials with respect to the tire circumferential direction at the respective tire axis direction coordinates. Also, the dashed lines are the provisional inclination angles of the belt materials at the respective tire axis direction coordinates. The center in the left-right direction of FIG. 21 coincides with the center in the tire axis direction, and both sides in the left-right direction of FIG. 21 correspond to both sides in the tire axis direction.

[0119] From the comparison between the solid line and the broken line in Fig. 21, it can be seen that there is a difference between the three-dimensional inclination angle of the belt wire and the provisional inclination angle which is the inclination angle of the belt wire when viewed from the outside in the tire radial direction on both sides in the tire axial direction. This is the influence of the large belt gradient on both sides in the tire axial direction.

[0120] When the calculation device captures the belt image and the tire axial cross-sectional image, it automatically executes up to the display of the graph as shown in Fig. 21 on the display device. However, between the capture of the belt image etc. and the display of the graph, the user may participate in the form of checking the results up to the intermediate stage of the process and instructing the continuation of the process, or inputting necessary information with the input device.

[0121] The three-dimensional inclination angle of the belt wire obtained in this way is used for various purposes. For example, in a tire completed through molding, it is possible to check whether the belt wire is arranged as designed, or to check the influence on the belt wire due to the formation of grooves in the tread during molding.

[0122] Next, the effects of the present embodiment will be described. This embodiment includes a step of acquiring, as a tire axial cross-sectional image, an image in which a plurality of belt wires included in the belt of the tire appear as dots and are arranged side by side, and a step of performing image processing including binarization on the tire axial cross-sectional image to obtain a binarized image (binarized cross-sectional image) in which the belt wire and other parts are separated into white and black, and a step of obtaining the inclination angle of a straight line connecting adjacent belt wires in the tire axial direction in the binarized image with respect to the tire axial direction, and setting the inclination angle as the belt gradient at the position of one end of the straight line (that is, one of the adjacent belt wires).

[0123] According to this method, compared with the conventional method of measuring the gradient of the belt wire appearing in the cross-section of the cut sample with a protractor in the tire axial direction, it is possible to measure the gradient of the belt in the tire axial direction with relatively little effort and it is also difficult to generate measurement errors.

[0124] In addition, since a CT device capable of obtaining data on the internal structure nondestructively is used to obtain a tire axial cross-sectional image, it is not necessary to prepare a cut sample, and it does not take much time to obtain the tire axial cross-sectional image.

[0125] In addition, since a CT device is used, it is also possible to obtain a tire axial cross-sectional image of an inflated tire. Since the tire bulges outward in the tire radial direction when inflated, the belt gradient in the tire changes compared to before inflation. Although it has been impossible to obtain the belt gradient in such an inflated tire conventionally, it can be obtained according to the present embodiment.

[0126] In the tire axial cross-sectional image and the binarized image (binarized cross-sectional image) obtained by binarizing it, a plurality of belt wires included in the first belt on the inner side in the tire radial direction and a plurality of belt wires included in the second belt on the outer side in the tire radial direction appear in columns, respectively. In such a binarized image, belt wires are sequentially selected from one side to the other side in the tire axial direction, and an associating process is executed each time a belt wire is selected. Here, the associating process checks for the presence or absence of other belt wires within a search range including the selected belt wire. If there are no other belt wires within the search range, the selected belt wire itself is associated with the first belt. If there are other belt wires within the search range, the lowermost (inner side in the tire radial direction) belt wire within the search range is associated with the first belt.

[0127] By this process, the belt wires associated with the first belt can be treated as the belt wires included in the first belt, and the belt wires not associated with the first belt can be treated as the belt wires included in the second belt. Thereby, each belt wire appearing in the binarized image can be separated into the belt wires of the first belt and the belt wires of the second belt, and it becomes easy to calculate the belt gradient in each of the first belt and the second belt.

[0128] After the belt gradient at each belt wire position is calculated, the average value of the belt gradients at a plurality of predetermined belt wire positions arranged in the tire axis direction is set as the belt gradient at any one of the plurality of belt wire positions by smoothing, so the accuracy of the belt gradient at each belt wire position is improved.

[0129] Various changes can be made to the above embodiments. Any one of the modification examples described below may be applied to the above embodiments, or any two or more of them may be combined and applied to the above embodiments. The combination can be freely made.

[0130] <Modification Example 1> The tire may be photographed by a CT device in a deflated state where it is not cut and no internal pressure is applied. Also, a cut sample in which the tire is cut and cut out from one end to the other end in the tire axis direction at two positions in its circumferential direction may be photographed by a CT device.

[0131] In any method, 3D data of the whole tire or the cut sample can be acquired by a CT device, and a belt image and a tire axis direction cross-sectional image can be acquired based on the 3D data.

[0132] <Modification Example 2> One or both of the belt image and the tire axis direction cross-sectional image may be acquired by an imaging device other than a CT device.

[0133] For example, the belt image may be an image obtained by peeling off the tread-side rubber of the cut sample to expose the belt wire and scanning and capturing the exposed surface where the belt wire is exposed with a scanner. Also, the tire axis direction cross-sectional image may be an image obtained by scanning and capturing the cut surface of the cut sample with a scanner.

[0134] <Modification Example 3> As a method for obtaining both the provisional inclination angle and the belt gradient for each tire axial coordinate (tire axial position), the tire axial coordinates of each belt wire appearing in the tire axial cross-sectional image may be matched with the tire axial coordinates of each belt wire appearing in the belt image, and a process of calculating the provisional inclination angle at the matched tire axial coordinates may be executed.

[0135] For this process, the tire axial coordinates of two adjacent belt wires in the belt image and the data of the provisional inclination angles at these two coordinates are used. Based on this data, the provisional inclination angle at any tire axial coordinate between two adjacent belt wires appearing in the belt image can be calculated by interpolation (for example, linear interpolation). Therefore, the provisional inclination angle can be calculated at the tire axial coordinates of the belt wires appearing in the tire axial cross-sectional image.

[0136] By executing such calculations for at least all the belt wires where the tire axial coordinates of the belt wires do not match between the belt image and the tire axial cross-sectional image, the provisional inclination angle can be obtained for the tire axial coordinates of all the belt wires appearing in the tire axial cross-sectional image. By this process, both the provisional inclination angle and the belt gradient can be obtained for the tire axial coordinates of all the belt wires appearing in the tire axial cross-sectional image.

[0137] <Modified Example 4> The method of the above embodiment is applicable to tires with a number of belts other than two.

Explanation of Signs

[0138] 10... Belt, 20... Belt Wire

Claims

1. A method for obtaining a belt gradient which is the inclination angle of a belt with respect to the tire axis direction at each position in the tire axis direction, comprising: obtaining, as a cross-sectional image in the tire axis direction, an image in which a plurality of belt wires included in the belt of the tire appear as dots and are arranged side by side; performing image processing including binarization on the cross-sectional image in the tire axis direction to obtain a binarized image in which the belt wires and other portions are separated into white and black; in the binarized image, obtaining the inclination angle of a straight line connecting the belt wires adjacent to each other in the tire axis direction with respect to the tire axis direction, and setting the inclination angle as the belt gradient at the position of one end of the straight line; A method for calculating the belt gradient of a tire, including the above steps.

2. The method for calculating the belt gradient of a tire according to claim 1, wherein a CT image is obtained as the cross-sectional image in the tire axis direction based on data obtained by photographing the tire with a CT device.

3. As the belt, the tire includes a No. 1 belt which is wide and located inside the tire in the radial direction and a No. 2 belt which is narrow and located outside the tire in the radial direction. In the cross-sectional image in the tire axis direction and the binarized image, a plurality of belt wires included in the No. 1 belt and a plurality of belt wires included in the No. 2 belt appear in columns respectively. In the binarized image, the belt wires are sequentially selected from one side to the other side in the tire axis direction, and an association process is performed each time a selection is made. The association process is to check for the presence or absence of other belt wires within a search range which is a predetermined range in the tire axis direction including the selected belt wire. If there are no other belt wires within the search range, the selected belt wire is associated with the No. 1 belt. If there are other belt wires within the search range, the belt wire which is the innermost in the tire radial direction within the search range is associated with the No. 1 belt. The belt wires associated with the No. 1 belt by the association process are regarded as the belt wires included in the No. 1 belt, and the belt wires not associated with the No. 1 belt by the association process are regarded as the belt wires included in the No. 2 belt. Calculating the belt gradient for each of the No. 1 belt and the No. 2 belt. The method for calculating the belt gradient of a tire according to claim 1 or 2.

4. The method for calculating the belt gradient of a tire according to claim 1 or 2, wherein the average value of the belt gradients at a plurality of predetermined positions arranged in the tire axial direction is taken as the belt gradient at any one of the plurality of positions.

Images