JP2014159965A - Tire visual inspection device and tire visual inspection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a tire visual inspection device and a tire visual inspection method capable of acquiring an accurate inspection image without being affected by decentering even when a tire is rotating while decentered with respect to a camera in visual inspection of the tire.SOLUTION: A tire visual inspection device includes: feature portion identification means 120A for acquiring images of the cross-sectional shape of the surface of a tire in a tire radial direction, the images corresponding to one round of the tire in a tire circumferential direction, and for determining whether or not a common feature portion of the tire T is included in an aligned image group G; correction curve acquisition means 130A for acquiring a correction curve AL by Fourier transforming a line segment connecting the feature portion in each image in the image group G; and image position correction means 140A for correcting the alignment of the images on the basis of the acquired correction curve AL.

Description

  The present invention relates to a tire appearance inspection apparatus and a tire appearance inspection method for performing an appearance inspection by processing an appearance image of a tire surface.

  Conventionally, in the appearance inspection of a tire, while rotating a tire placed sideways on a turntable, a slit light is irradiated in a radial direction from the irradiation means to the tire surface facing upward, and photographing provided corresponding to the irradiation means An image of the tire surface is acquired by photographing a light trace of the light on the tire surface irradiated from the irradiation means by a camera as the means. One frame of images taken by the camera is continuously acquired in a cross-sectional shape on the tire surface along the circumferential direction, and sequentially output to the image processing means. In the image processing means, images for each frame of the tire taken by the camera are combined to form an inspection image, and the inspection image is further subjected to image processing to inspect the presence or absence of a defective portion on the tire surface. The quality of the tire is determined. In the inspection by image processing, changes in the luminance of the pixels constituting the photographed image and the presence or absence of scratches or molding defects on the tire surface are compared by comparing the height information of each pixel with a threshold value.

JP 2011-196741 A

However, in order to inspect the tire surface by image processing as described above, the inspection image needs to be acquired with high accuracy. That is, when the rotation center of the turntable and the rotation center of the tire that is the subject do not coincide with each other and the tire is rotated and photographed in an eccentric state, the image of the tire surface captured by each camera is If the images are arranged on the frame basis in the order in which they were photographed to form an inspection image, the photographed tire surface inspection image is wavy in the tire circumferential direction. It will be distorted. In this case, since the inspection image is distorted, it is not known whether the distortion is caused by the acquisition of the image or the product itself at the time of molding the tire. May not be possible. Therefore, in order to remove the distortion in the inspection image, the method of acquiring the image after adjusting the rotation center of the turntable and the rotation center of the tire to be exactly coincident with great care, or from the camera It is necessary to take a measure for acquiring an inspection image by a method of aligning the cross-sectional shapes of the tire surface acquired in each image with reference to the cross-sectional shape of the image having the closest distance to the tire surface.
However, the former method hinders inspection efficiency, and the latter method forcibly aligns images taken with the closest distance to the tire surface as a reference, and is excluded from the images by alignment. Will occur. For this reason, even if a defect is included in the excluded portion, the defect is removed from the image, so that it is determined that there is no abnormality in the inspection by image processing even though the defect actually occurs. There is a risk that.
Therefore, in order to improve the inspection efficiency, when composing an inspection image from each captured image, the image can be excluded so that the influence of the eccentricity can be excluded even in an eccentric state where the rotation center of the turntable and the tire center do not coincide. There is a need for a method of automatically acquiring an inspection image by processing and performing an appearance inspection.

  In order to solve the above-described problems, the present invention can acquire a high-precision inspection image without being affected by the eccentricity even when the tire rotates in an eccentric state with respect to the camera in the appearance inspection of the tire. A tire appearance inspection apparatus and an appearance inspection method are provided.

As a configuration of a tire appearance inspection device for solving the above-mentioned problem, an image of a cross-sectional shape of a tire surface along the tire radial direction is obtained for one round of the tire along the tire circumferential direction, and is common to the aligned image group. Characteristic part specifying means for determining whether or not the tire characteristic part is included, correction curve acquisition means for acquiring a correction curve by Fourier transforming a line segment connecting the characteristic parts in each image of the image group, and Image position correction means for correcting the alignment of the images based on the corrected curve, so that the tire cross-sectional shape acquired in the image can be accurately obtained even if the image is acquired with the rotation of the tire decentered. Can be aligned. In other words, even if the rotation of the tire is acquired as an image in an eccentric state, the position of each image constituting the image group is corrected by using the tire characteristic portion included in each image so that there is no influence of the eccentricity. A group of images can be constructed.
Further, as another configuration of the tire appearance inspection apparatus for solving the above-described problem, the feature portion detected by the feature portion specifying means is set as the position of the rim guard on the tire side surface in the image, so that the image in each frame is A reference position can be set. That is, since the rim guard protrudes largely in the tire width direction on the side surface of the tire for the purpose of protecting the wheel rim, the rim guard is clearly photographed in the image, so that the position in each frame can be accurately identified.
Further, as another configuration of the tire appearance inspection apparatus for solving the above-described problem, the feature portion detected by the feature portion specifying unit is an arc that connects the position of the rim line on the tire side surface and the position of the shoulder line in the image. By setting the radial center position, the reference position can be accurately set in the image of each frame even if the rim guard is not provided on the tire side surface. That is, the shoulder line and the rim line are formed on the side surfaces of all tires, and the reference position can be accurately set for each image. That is, since the cross-sectional shape of the tire is designed to have a continuous curve having a predetermined curvature, an arc that fits the cross-sectional shape is always present in the captured image. Furthermore, the side surface of the tire that is placed sideways that is not assembled with a rim has a large undulation state along the circumferential direction, but the image taken in such a state is in the tire radial direction due to the influence of eccentricity in the frame. And an image that is rotated with respect to the axis along the circumferential direction. Therefore, when images are aligned using one of the detected shoulder line or rim line, the influence of rotation on the axis along the circumferential direction cannot be removed. Therefore, by setting the radius center of the arc connecting the position of the shoulder line and the position of the rim line in the image of each frame as the characteristic portion, it is possible to remove the influence of rotation.

Further, as an aspect of a tire appearance inspection method for solving the above-described problem, an image of a cross-sectional shape of the tire surface in the tire radial direction is obtained for one round of the tire along the tire circumferential direction, and an aligned image group is obtained. A feature determination step that determines whether or not a common tire feature is included, and a feature that is common to the cross-sectional shapes included in each image is detected when the feature determination step determines that there is a feature. A feature detection step, a feature setting step for creating and setting a feature common to the cross-sectional shapes included in each image when it is determined that there is no feature in the feature determination step, and a detected feature A correction curve acquisition step for acquiring a correction curve by Fourier-transforming a line segment that connects a set part or a set feature, and an image position for correcting the position of each image in the image group using the correction curve By providing a positive step, it is acquired as an image in a state where the rotation of the tire is eccentric, acquired images the sectional shape of the tire can be accurately aligned. In other words, even if the rotation of the tire is acquired as an image in an eccentric state, the position of each image constituting the image group is corrected by using the tire characteristic portion included in each image so that there is no influence of the eccentricity. A group of images can be constructed.
Further, as another aspect of the tire appearance inspection method for solving the above-mentioned problem, the feature setting step detects a shoulder line molded in the shoulder portion of the tire and a rim line molded in the bead portion from the image group. In addition, the center position of the arc that includes the detected shoulder line and rim line and fits the cross-sectional shape of each image is calculated, and the center position is set as a feature portion in each image, thereby providing a feature portion on the tire surface. Even if not, the cross-sectional shape of the tire can be accurately aligned. That is, since the shoulder line and rim line are always captured in the image in each image, the shoulder line and rim line are detected from the image, arcs that fit the shoulder line and rim line are calculated, and the center of the arc is calculated. By setting as a feature in the image, it is possible to set a common feature in the image. Since the cross-sectional shape of the tire is designed to have a continuous curve having a predetermined curvature, an arc that fits the cross-sectional shape is always present in the captured image. In addition, the side surface of the tire placed sideways that is not assembled with a rim is greatly wavy along the circumferential direction, but the image taken in such a state is in the tire radial direction due to the effect of eccentricity in the frame. And an image that is rotated with respect to the axis along the circumferential direction. Therefore, when images are aligned using one of the detected shoulder line or rim line, the influence of rotation on the axis along the circumferential direction cannot be removed. Therefore, by setting the radius center of the arc connecting the position of the shoulder line and the position of the rim line in the image of each frame as the characteristic portion, it is possible to remove the influence of rotation.

1 is an overall schematic view of a tire appearance inspection device. FIG. 4 is a layout diagram of tires placed on a rotary table and image acquisition means. It is a block diagram of a photography control image processing means. It is a figure which shows the contour line without a rim protector and with a rim protector. It is a flowchart of an image process. It is a figure which shows the image group of the arranged image frame. It is a figure which shows the outline when it aligns on the image frame reference | standard. It is a figure which shows the rim line and shoulder line in an image group. It is a figure which shows the deflection width | variety of the circular arc center along a tire circumferential direction. It is a figure which shows the image of the aligned tire side surface.

  Hereinafter, the present invention will be described in detail through embodiments of the invention. However, the following embodiments do not limit the invention according to the claims, and all combinations of features described in the embodiments are included in the invention. It is not necessarily essential to the solution, but includes a configuration that is selectively adopted.

FIG. 1 shows a schematic view of a tire appearance inspection apparatus 1 as an example for executing the appearance inspection method of the present invention, and FIG. 2 shows a layout of tires placed on a rotary table and image acquisition means. . FIG. 3 shows a block diagram of the shooting control image processing means 5. Hereinafter, the tire appearance inspection apparatus 1 will be described with reference to FIGS. 1 to 3.
The tire appearance inspection apparatus 1 generally takes an image of the appearance of an inspection table 2 having a table 12 (described later) on which a tire T as a subject is placed sideways and the surface of the tire T placed sideways facing upward. Thus, the image acquisition means 3 is acquired as an image, and the imaging control image processing means 5 is used to inspect the appearance of the tire surface by processing the image acquired by the image acquisition means 3.

  The examination table 2 includes a pedestal 11, a table 12 on which a tire T as a subject is placed sideways, a drive mechanism that rotates the table 12, and an encoder 14 that detects the rotation angle of the table 12. The pedestal 11 is a rectangular box, for example, installed on the floor surface, and includes a table 12 on the upper surface and a drive mechanism for driving the table 12 inside. The drive mechanism includes, for example, a motor 13 and a power transmission mechanism (not shown) that transmits the rotational force of the motor 13. The motor 13 is fixed in the pedestal 11 by a fixing means (not shown) with the output shaft directed vertically in the pedestal 11. Further, the motor 13 is driven to rotate by a signal output from the shooting control image processing means 5 described later.

  The pedestal 11 includes a rotating shaft (not shown) protruding in the vertical direction. The rotating shaft is rotatably supported by a bearing or the like provided on the upper wall, and the rotational force of the motor 13 is transmitted by a power transmission mechanism (not shown). Examples of the power transmission mechanism include a gear mechanism, a pulley-belt mechanism, a chain-sprocket mechanism, and the like. Further, the output shaft of the motor 13 and the rotation shaft may be directly connected without using a power transmission mechanism.

  In addition, an encoder 14 for detecting the rotation angle of the shaft is provided at the lower end of the rotation shaft, and the rotation angle of the shaft is detected and output to the photographing control image processing means 5 described later.

  At the upper end of the shaft, a table 12 for placing a tire T as a subject is provided. The table 12 is, for example, a flat disk, the axis of the shaft and the center of the disk are concentric, and the upper surface 12a of the table 12 serving as a mounting surface for mounting the tire T as the subject. Is fixed to the upper end of the shaft so that is horizontal.

  Above the inspection table 2, an image acquisition means 3 for acquiring an image of the tire surface, specifically, one of the tire side surfaces Ts is arranged. The image acquisition means 3 photographs the irradiation means 16 that irradiates the surface of the tire T with slit-shaped two-dimensional laser light (hereinafter referred to as slit light) La, and the light trace Lc of the slit light La that is irradiated on the tire side surface Ts. And a photographing means 17. That is, the image acquisition means 3 of the present embodiment acquires an image of the tire side surface Ts by a so-called light cutting method.

  The irradiation means 16 is fixed by a fixing means (not shown) so as to irradiate slit light in the radial direction of the tire side surface Ts placed sideways on the table 12, for example. The slit light La emitted from the irradiation means 16 is from the tire inner diameter to the tire outer diameter of the tire T placed on the table 12, in other words, from the bead portion edge TA to the shoulder portion through the shoulder portion. Even if the rotation center C2 of the tire T is placed in an eccentric state that is shifted by δR relative to the center C1 of the table 12, the tire outer diameter is changed from the tire inner diameter. The irradiation range is set so that it is always within the irradiation of the slit light La.

  The photographing unit 17 is a so-called area camera, and the light trace Lc of the slit light La irradiated from the irradiation unit 16 to the tire side surface Ts is compared with the irradiation angle of the slit light La from the irradiation unit 16 to the tire side surface Ts. Take a picture from an oblique direction with a predetermined angle. Specifically, the photographing unit 17 is configured such that the optical axis of the camera is a predetermined light receiving angle with respect to the irradiation angle of the slit light La emitted from the irradiation unit 16 to the tire side surface Ts. It is fixed by a fixing means (not shown) so as to have an angle for receiving the totally reflected light Lb.

  In the present embodiment, the slit light La is irradiated on the tire side surface Ts and photographed with the area camera. However, the diffused light may be irradiated on the tire side surface Ts and photographed with the line camera. In this case, it arrange | positions so that the sequence direction of the light receiving element of a line camera may follow a tire radial direction.

  As shown in FIGS. 4A and 4B, images of the tire side surface Ts photographed by the irradiation unit 16 and the photographing unit 17 are contours of the cross-sectional shape in the radial direction on the tire side surface Ts in the image frame F. Obtained as line P. Specifically, the cross-sectional shape of the tire surface in the range extending from the inner diameter to the outer diameter of the tire T along the tire radial direction and the range extending in the tire width direction so as to include the tread surface continuous thereto is acquired. The The contour line P includes, for example, height information from the table 12 and luminance information representing the cross-sectional shape. The acquisition of the contour line P of the tire side surface Ts is acquired by dividing the circumference of the tire by 10,000 along the circumferential direction, that is, along the direction in which the tire side surface extends. That is, every time the tire T rotates 360/10000 degrees together with the table 12, photographing is performed, and 10,000 image frames F are acquired. Note that the image frame F is one image having the size of the field of view of the photographing unit 17. Further, the number of divisions is not limited to the above 10,000 divisions, and may be changed as appropriate. The photographing unit 17 and the irradiation unit 16 are connected to the photographing control unit 20 of the photographing control image processing unit 5, and an image photographed by the photographing unit 17 is output to the image processing unit 21.

The photographing control image processing means 5 is a so-called computer, and includes a CPU as a calculation means, a ROM and a RAM as storage means, and an input / output interface as a communication means. The storage means stores a control program as the photographing control means 20 for controlling the photographing operation of the tire side surface Ts, and an image processing program as the image processing means 21 for processing the photographed image. As shown in FIG. 3, the image processing means 21 includes an image alignment means 110A, a feature part specifying means 120A, a correction curve acquisition means 130A, an image position correction means 140A, and the like.
The photographing control image processing means 5 is connected to a display means (not shown), a keyboard, a mouse, and the like for inputting a control program and information for executing the image processing program. Information input from the input means is stored in the storage means. The photographing control image processing means 5 is connected to a monitor 5A, and an image as shown in FIG. 4 is displayed on the monitor 5A.

  The imaging control unit 20 controls the operations of the irradiation unit 16 and the imaging unit 17 based on information related to imaging input to the imaging control image processing unit 5 from the input unit. Specifically, when a signal for starting photographing is output, the photographing control unit 20 outputs a signal for starting irradiation of the slit light La to the irradiating unit 16 and rotates the tire T so that the tire T The slit light La is irradiated from the start of rotation to the stop of rotation. The photographing unit 17 outputs a signal for performing photographing based on the rotation angle of the table 12 detected by the encoder 14.

  Hereinafter, the image processing means 21 will be described. As shown in FIGS. 5A and 5B, the image processing means 21 is roughly an image alignment step 110 executed by the image alignment means 110A and a feature portion specifying step 120 executed by the feature portion specifying means 120A. And a correction curve acquisition step 130 executed by the correction curve acquisition means 130A, and an image position correction step 140 executed by the image position correction means 140A.

  The image alignment step 110 aligns the image frames F over the entire circumference of the tire so that the corners of the captured image frames F are aligned in the order in which the image frames F captured by the imaging unit 17 are acquired. For example, the outline P of the image group G aligned in the image alignment step 110 constitutes a stereoscopic image composed of the tire radial direction (tire height direction), the tire width direction, and the tire circumferential direction, as shown in FIG. To do. As shown in FIG. 6, the contour lines P when aligned are in a state of being displaced in the tire radial direction due to the influence of eccentricity.

The feature specifying step 120 specifies the position of the feature in the actual tire on the contour line P of the image frame F as shown in FIGS.
Specifically, the feature part specifying step 120 includes a tire information reading step 121, a feature part determining step 122, a feature part detecting step 123, and a feature part setting step 124.

  In the tire information reading step 121, the tire T data input from the input means to the photographing control image processing means 5 is read. In this embodiment, the information about the presence or absence of the rim guard RG molded on the tire shown in FIG. 4B is read.

  In the present embodiment, the rim guard RG is treated as a characteristic portion of the tire itself, is a portion that is formed on the tire side surface Ts concentrically with the rotation center C2 of the tire T, and protrudes greatly in the tire width direction. This is different from the rim line RL molded smaller than the rim guard RG in the vicinity of the bead portion. The rim guard RG has a large convex shape that prevents damage to the hole rim by protruding in the tire width direction from the rim seat of the wheel rim on the tire side surface Ts when the subject tire T is assembled to a recommended wheel rim. It is molded into In the present embodiment, the rim guard RG formed on the tire side surface Ts is used as a characteristic part, and the determination is made based on the information on the presence / absence of the rim guard RG. Even if it is formed in a concavo-convex shape, it may be a concave portion or a convex portion that continues in the circumferential direction at the same radial position on the tire side surface Ts.

  In the characteristic part determination step 122, the determination is made based on the presence or absence of a characteristic part, in the present embodiment, based on the presence or absence of the rim guard RG as a characteristic part from the information on the tire T of the subject read in the tire information reading step 121.

  When it is determined by the feature determination step 122 that the rim guard RG as the feature is present, the feature detection step 123 detects the position of the rim guard RG from the contour line P of each image frame F. Specifically, the position of the tip of the rim guard RG that protrudes most in the tire width direction on the contour line P is detected. Further, when it is determined that there is no rim guard in the feature determination step 122, a feature is created based on the target on the tire side surface Ts in the feature setting step 124, and the target is set as the feature.

  As shown in FIG. 5B, the feature detection step 123 is executed when the rim guard RG as the feature exists, and has a sharp height in the contour line P acquired in the image frame F. The position of the rim guard RG is detected by detecting the portion where the change occurs. For example, by calculating the gradient between adjacent pixels of the pixel representing the contour line P along the contour line P, and detecting the position of the pixel in the image frame F when a gradient equal to or greater than the threshold value is detected, the feature portion As a result, the position of the rim guard RG can be detected.

  The feature setting step 124 includes a shoulder line detection step 125, a rim line detection step 126, and an arc setting step 127, which are detected as a common target included in the contour line P of each image frame F. Set the features using. In the present embodiment, a shoulder line SL and a rim line RL formed on the tire side surface Ts shown in FIG. 4A are used as targets.

  As shown in FIG. 8, the shoulder line detection step 125 detects a shoulder line SL molded near the shoulder portion of the tire side surface Ts from the image group G aligned in the image alignment step 110. The shoulder line SL is a convex portion formed in an annular shape around the rotation center C2 of the tire T along the tire circumferential direction in the shoulder portion of the tire T, and a plurality of concentric circles are provided. In the present embodiment, the shoulder line SL that is located closest to the bead portion is detected. The detection of the shoulder line SL is executed by adopting a curve on the bead portion side among curves continuous in the circumferential direction with a predetermined luminance value in the tread side region of the image group G in the radial direction. The

  The rim line detection step 126 detects a rim line RL molded near the bead portion of the tire T from the image group G. The rim line RL is for visually confirming whether the rim flange is assembled at a predetermined position when the tire T is assembled to the wheel, and can be visually recognized outside the edge of the rim flange of the wheel. Thus, it is a convex part continuously formed concentrically along the tire inner diameter on the tire side surface. The detection of the rim line RL is performed by adopting a tread-side curve among curves continuous in the circumferential direction at a predetermined luminance value in a region closer to the bead portion than a half in the radial direction in the image group G. .

  In the arc setting step 127, as shown in FIG. 4A, an arc A passing through the shoulder line SL and the rim line RL is added to each contour line P detected in the contour line P of each image frame F constituting the image group G. Fit each one. Specifically, the arc setting step 127 calculates a radial center position C3 of one arc A passing through a plurality of points in the section of the shoulder line SL and the rim line RL on the contour line P, and sets the center position C3 to each of the center positions C3. It is set as a feature portion of the contour line P of the image frame F. The center position C3 of the arc A set in the arc setting step 127 is stored in the storage means.

The method of setting the arc A includes, for example, a position corresponding to the rim line RL and shoulder line SL of each contour line P in each image frame F, and a maximum width portion Wmax of the tire T that is a top portion in each contour line P, for example. A method of calculating the arc A from three points is conceivable. Then, the radius center position C3 of the arc A in the image frame F of the calculated arc A is calculated, and the center position C3 is set as a feature portion of the contour line P in each image frame F. The method for calculating the radius center position C3 includes, for example, a normal h1 extending from the midpoint of a line segment connecting a position corresponding to the rim line RL and a position corresponding to the maximum width portion Wmax in each image frame F, and a shoulder line. The intersection of the normal h2 extending from the midpoint of the line segment connecting the position corresponding to SL and the position corresponding to the maximum width portion Wmax is the radial center position C3 of the arc A, and can be easily calculated in this way. Can do.
As a reason for forming the circular arc A and using the radial center position C3 as a characteristic portion, when trying to specify the position of the shoulder line SL or the position of the rim line RL from the individual contour line P, the shoulder line SL does not protrude. Since there are many, it is difficult to detect and the rim line RL is also erroneously detected. Therefore, the shoulder line SL and the rim line RL are detected from the image group G, and both are employed to provide accuracy. In calculating the arc A, the position of the maximum width portion Wmax is also added to improve the reliability.

  As another method for setting the arc A, the arc A may be set using all points from the rim line RL to the shoulder line SL of the contour line P. By setting the arc A using all points between the rim line RL and the shoulder line SL of the contour line P, the contour line P on the tire side surface Ts can be approximately represented by the arc A. The arc A can be accurately set for each frame F. That is, since the tire T is continuous in the circumferential direction, the shape does not change greatly between the adjacent image frames F. Therefore, the rim line RL of the contour line P included in each image frame F to the shoulder line SL. The arc A along the cross-sectional shape can be set by using all the points up to. Therefore, the arc A set between the adjacent image frames F is set as the arc A having substantially the same size, and the correction curve AL with high accuracy is acquired in the correction curve acquisition step 130 described later. to enable.

  The reason why the arc A is fitted to the contour line P by the arc setting step 127 is that the actual cross-sectional shape of the tire T is designed to have a continuous curve having a predetermined curvature. In the contour line P of the frame F, there is always an arc A that fits the cross-sectional shape. Furthermore, the tire side surface Ts placed sideways that is not assembled with a rim is in a large wavy state along the circumferential direction. The contour line P photographed in such a state is, for each image frame F, It moves in the tire radial direction due to the influence of the eccentricity, and is rotated with respect to the axis along the circumferential direction. Therefore, when the contour line P is aligned using one of the detected shoulder line SL or rim line RL, the influence of rotation on the axis along the circumferential direction cannot be removed. Therefore, it is possible to eliminate the influence of rotation by setting the radial center position C3 of the arc A connecting the position of the shoulder line SL of the contour line P of each image frame F and the position of the rim line RL as a characteristic portion. . In the present embodiment, the description will be made assuming that the contour line P moves in the radial direction for each image frame F due to the influence of eccentricity.

  In the correction curve acquisition step 130, a common feature is connected to the contour line P of each image frame F aligned in the image alignment step 110, a polygonal line is acquired, and the polygonal line is Fourier transformed, as shown in FIG. A correct correction curve AL is acquired.

  The correction curve AL is calculated by calculating a broken line connecting the center position of the arc A in each image frame F along the tire circumferential direction by a line segment, and Fourier-transforming the broken line to calculate the correction curve AL. The Specifically, the correction curve AL is a curve obtained by performing an inverse Fourier transform after validating up to a fourth-order component obtained by Fourier transforming a polygonal line and discarding the fifth-order component or less. By obtaining the correction curve AL in this way, it is possible to effectively remove the influence of the eccentricity in the image group G and align the image group G with high accuracy. That is, the primary component by Fourier transform shows the influence by the eccentricity at the time of acquisition of the image group G, and the secondary component to the quaternary component show the thing accompanying the deformation | transformation which occurred at the time of tire conveyance after shaping | molding. Therefore, it is possible to obtain a highly accurate correction curve AL by performing Fourier transform, leaving the quaternary component from the first order component and excluding the rest, and performing inverse Fourier transform. If the tire T is not affected by deformation, the correction curve AL may be acquired by leaving the primary component and excluding other components. Further, when an influence due to deformation is observed, in accordance with the influence due to deformation, other components are excluded except for the primary component and the secondary component, and other components are excluded from the primary component, leaving the tertiary component. Further, the correction curve AL may be obtained by excluding the quaternary component from the primary component and excluding other components. By obtaining the correction curve AL in this way, as shown in FIG. 9, a curve obtained by Fourier transforming a polygonal line connecting the center position of the arc A in each image frame F along the tire circumferential direction. A correction curve AL with higher accuracy can be obtained.

  In the image position correction step 140, the position of the feature portion detected in the feature portion detection step 123 or the feature portion detected in the feature portion setting step 124 in each image frame F is made to coincide with the correction curve AL, and each image is corrected. The position of the contour line P in the frame F is corrected.

That is, the image position correction step 140 corrects the position of the contour line P in each image frame F using the correction curve AL calculated in the correction curve calculation step. Specifically, the position of the contour line P of each image frame F is corrected by arranging the position of the feature in each image frame F on the correction curve AL. In the correction, the position of the contour line P may be corrected so that the image frame F is arranged on the correction curve AL.
The image corrected in this way can be an image that is not displaced in the radial direction, as shown in FIG.

  As shown in FIG. 10 described above, the corrected image is obtained by executing a test program to compare the change in luminance of pixels constituting the image and the height information of each pixel with a threshold value to compare the tire surface. The presence or absence of scratches or molding defects is performed.

  As described above, according to the appearance inspection method of the present invention, the contour line P, which is a cross-sectional image of the tire side surface Ts of the tire T placed in an eccentric state with respect to the table 12, is automatically prevented from being eccentric. Therefore, a more accurate inspection can be performed in the inspection by the subsequent inspection means.

  In addition, since the contour line P is aligned on the basis of the portion formed on the actual tire side surface Ts for alignment of the contour line P in the image frame F and acquired in the contour line P of each image frame F, accuracy is improved. It is possible to perform a good alignment. That is, when an image is taken in an eccentric state in which the rotation center C2 of the tire T is displaced by δR relative to the center C1 of the table 12, the cross-sectional shape of the tire T photographed by the photographing means 17 is the tire radius in the image frame F. Will move in the direction.

In consideration of the fact that the tire T is rotating with respect to the photographing means 17 for photographing from a specific position, this movement is a tire having a contour line P having a cross-sectional shape that moves within the image frame F for each image frame F. Since the increase / decrease in the amount of movement accompanying the rotation of T is substantially equal to the increase / decrease in the sine wave, it can be approximately represented by a sine wave. Therefore, by obtaining a correction curve AL by performing a Fourier transform on the broken line connecting the features calculated in the correction curve acquisition step 130, it is possible to approximate the movement of the contour line P to be approximately equal to the sine wave. . In particular, the eccentric component of the tire T with respect to the table 12 can be expressed with high accuracy by adopting up to the fourth-order component of the broken line subjected to Fourier transform and discarding the fifth-order and subsequent components.
Therefore, the image group G free from the influence of eccentricity can be obtained by arranging the characteristic portion in each image frame F forming the image group G on the correction curve AL obtained in the correction curve acquisition step 130.

1 tire appearance inspection device, 2 inspection table, 3 image acquisition means,
5 shooting control image processing means, 20 shooting control means, 21 image processing means,
110 image alignment step, 110A image alignment means,
120 feature specifying step, 120A feature specifying means,
121 tire information reading step, 122 characteristic portion determination step,
123 feature detection step, 124 feature setting step,
125 shoulder line detection step, 126 rim line detection step,
127 arc setting step,
130 correction curve acquisition step, 130A correction curve acquisition means,
140 image position correcting step, 140A image position correcting means,
A arc, G image group, F image frame, P contour line,
RG rim guard, RL rim line, SL shoulder line.

Claims (5)

An image of the cross-sectional shape of the tire surface along the tire radial direction is acquired for one round of the tire along the tire circumferential direction, and a characteristic portion for determining whether or not a common tire characteristic portion is included in the aligned image group Specific means,
Correction curve acquisition means for acquiring a correction curve by Fourier-transforming a line segment connecting the feature portions in each image of the image group;
A tire appearance inspection apparatus comprising: an image position correction unit that corrects the alignment of the images based on the acquired correction curve.   The tire appearance inspection apparatus according to claim 1, wherein the feature portion detected by the feature portion specifying means is a position of a rim guard on a tire side surface in the image.   The tire appearance inspection according to claim 1 or 2, wherein the feature portion detected by the feature portion specifying means is a radial center position of an arc connecting the position of the rim line on the tire side surface and the position of the shoulder line in the image. apparatus. Feature section determination for acquiring an image of the cross-sectional shape of the tire surface in the tire radial direction for one round of the tire along the circumferential direction of the tire and determining whether or not a common tire feature is included in the aligned image group Steps,
A feature detection step for detecting a feature common to the cross-sectional shapes included in each image when it is determined that there is a feature in the feature determination step;
A feature setting step for creating and setting a feature common to the cross-sectional shapes included in each image when it is determined that there is no feature in the feature determination step;
A correction curve acquisition step of acquiring a correction curve by Fourier transforming the detected feature or a line segment connecting the set feature;
A tire appearance inspection method comprising: an image position correction step of correcting the position of each image in the image group using the correction curve. The feature portion setting step detects a shoulder line molded on a shoulder portion of a tire and a rim line molded on a bead portion from the image group, includes the detected shoulder line and the rim line, The tire appearance inspection method according to claim 4, further comprising: calculating a radial center position of an arc that fits the cross-sectional shape, and setting the center position as a feature in each image.

Images