JPH063371B2 - Evaluation method of circularity of inspected object - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial field of application) The present invention provides a method for evaluating the circularity of an object to be inspected, and in particular, the circularity of a circular hole provided in a spinneret can be properly evaluated by image processing. The present invention relates to a method of evaluating the circularity that can be achieved.

(Prior art and its problems) In the synthetic fiber industry, it is necessary to measure and determine the circularity of circular holes provided in a spinneret used for spinning fibers. That is, if the degree of circularity of the circular holes decreases due to the attachment of deposits to some of the circular holes (hereinafter referred to as "tsumuri"), the quality of the synthetic fibers deteriorates. Therefore, it is necessary to appropriately measure the circularity of the circular hole and appropriately determine whether or not there is a groove.

Conventionally, the above-mentioned determination has been mainly performed by visual determination using a microscope, but in recent years, there has been a growing demand for automating the above determination, and automatic determination using an image processing technique has been proposed.
For example, the image data of a circular hole as an inspection object is stored by a conventionally known image processing technique, the area A and the perimeter L of the circular hole are obtained based on the data, and further substituted into the following equation to obtain a circular shape. A method of evaluating degree is proposed.

(Circularity) = 4πA / L ² (1) In this method, the circularity is 1 if the stored image of the circular hole is a perfect circle, as can be seen from the equation (1).

In addition to the above method, two methods can be cited as methods for evaluating the circularity based on the stored image data. One of them is a method of using the range of radius data, and evaluates the circularity of the circular hole by the difference between the maximum value and the minimum value of the radius data obtained from the image data. The other is the coefficient of variation CV (= coefficient of variation) of the radius data obtained from the image data.
Is a method using. That is, as shown in equation (2),
The circularity is evaluated by a value obtained by dividing the standard deviation s of the radius data by the average value x of the radius data.

CV = s / x (2) In this method, if the stored image of the circular hole is a perfect circle, the standard deviation s becomes 0, so the variation coefficient CV
Also becomes 0.

In other words, the above-mentioned proposed technique stores the image data of the circular hole by the image processing technique and determines the circularity obtained from the image data, the difference between the maximum value and the minimum value of the radius data, or the variation coefficient CV of the radius data from the circular hole Evaluate the circularity of
It is intended to determine whether there is a summary.

By the way, an image stored by using a conventionally known image processing technique is distorted due to the influence of illumination, an optical system, a TV camera, etc. Even if an image of a perfect circle (circularity = 1) is stored, The stored image becomes an ellipse (circularity ≠ 1). However, the above method for evaluating the circularity is based on the assumption that the stored image of the circular hole is a perfect circle, and the method is effective when the stored image is a perfect circle. However, if the stored image is an ellipse, it is difficult to determine whether or not there is a delicate outline.

Here, this will be described with reference to FIG. Fig. 8
(a) is a figure which shows a circular hole in a thumb nail. As shown in FIG. 8 (a), the deposit 81 is attached to a part of the circular hole 5 provided in the spinneret 2. FIG. 8B shows an image 83 of the circular hole 5 stored by a conventionally known image processing technique. As shown in the figure, the circular hole 5
The image 83 is distorted in an elliptical shape as a whole, and a concave portion 82 corresponding to the attached matter 81 is formed in a part thereof. _{Incidentally, r 1, ..., r m} , ..., respectively _{r n} the image 83
Is a radius at a point on each circumference of.

As shown in FIG. 8 (b), the stored image 83 is X, Y.
When the direction has a long axis and a short axis, the maximum value of the radius data is the radius r _m . On the other hand, the minimum value of the radius data differs depending on the radius r ₁ of the recess 82. That is, when the radius r ₁ is longer than the radius r ₄ , the minimum value of the radius data is the radius r ₄ , and when the radius r ₁ is shorter than the radius r ₄ , the minimum value of the radius data is the radius r 4. Becomes ₁ .

Therefore, in the case where the radius r ₁ is longer than the radius r ₄ , the difference in the radius data (= radius r
_{The m} −radius r ₄ ) has the same value as that when the concave portion 82 is not formed. As a result, it is difficult to determine the presence or absence of a summary by a method that uses the range of radius data.

Further, even if the circular hole 5 has no outline, the stored image is an ellipse, and the radius data varies between the radius r _m and the radius r ₄ , so that the variation coefficient CV has a constant value of 0 or more. ing. Therefore, even if the coefficient of variation CV fluctuates due to the summary, it is difficult to properly evaluate it.

Further, since the method using the circularity is based on the assumption that the image is a perfect circle, if the image becomes elliptical,
It is not appropriate to adapt the circularity.

(Object of the Invention) The present invention has been made in order to solve the problems of the related art, and an object of the present invention is to suppress the influence of image distortion and improve the ability to determine the presence or absence of a summary.

(Means for Achieving the Object) The present invention is directed to a first step of capturing an image of an object to be inspected and storing it as image data, and an image of the object to be inspected stored in the first step based on the image data. A second step of obtaining a center position, a third step of obtaining a radius represented by a distance between a plurality of points on the circumference of the inspection object image and the center position, and a point on the plurality of circumferences A fourth step of obtaining a difference between respective radii corresponding to two points on an adjacent circumference, a fifth step of obtaining a statistic of the plurality of differences obtained in the fourth step, and the statistic A sixth step of evaluating the circularity of the inspection object based on the quantity.

(Operation) According to the present invention, among the points on the plurality of circumferences, the difference between the radii corresponding to the two adjacent points on the circumference is obtained, and the statistic is obtained based on the value of the difference. There is. Therefore, it is possible to suppress the influence of the distortion of the inspection object image stored in the first step.

(Example) A. Outline of Embodiment FIG. 2 is a block diagram of an image reading apparatus 1 to which an embodiment of the present invention can be applied. This device 1 has an inspection table 3 on which a spinneret 2 is placed. A light projector 4 is provided below the inspection table 3, and the light emitted from the light projector 4 is incident on the lower surface of the spinneret 2 and passes through a circular hole 5 provided in the spinneret 2 to inspect the table. 3 is projected onto the image sensor 6 provided on the upper side. And the image sensor 6
Thus, the spinneret 2 is imaged and input to the image processing circuit 7 as an output signal S _I , and predetermined image processing is performed. Further, the image data subjected to the predetermined image processing in the image processing circuit 7 is supplied from the image processing circuit 7 to the arithmetic circuit 8, various operations are performed based on the image data, and the circular hole 5 has a margin. Whether or not it is determined. The determination result is output to the output device 9 such as a display or a printer.

Next, an outline of one embodiment of the present invention will be described. In the present embodiment, first, the image data of the circular hole 5 as the inspection object is captured and stored in the image processing circuit 7, and the arithmetic circuit 8 calculates the center position of the circular hole 5 and the points on the circumference based on the image data. Ask in. Then, the value of the radius at each point on each circumference is obtained, the difference between the radii at two points on the circumference that are closer to each other is calculated, and various statistical quantities (for example, maximum value,
Average value, standard deviation, etc.) Then, the circularity of the circular hole 5 is evaluated based on the statistics obtained as described above. That is, the point that this embodiment is greatly different from the conventional proposal is
In the proposed example, the radius data was obtained, and the circularity was evaluated by various statistics of the data, whereas in the present embodiment, the difference in radius between two points on the adjacent circumference was obtained, and various statistics of the difference were obtained. The point is that the circularity is evaluated by the amount.

Therefore, in this embodiment, even if the radius data greatly varies due to the distortion of the stored image as a result of the image processing, the difference between the radii at two points on the adjacent circumferences is obtained, and this is used as a basis for evaluating the circularity. Therefore, it is possible to evaluate the circularity in accordance with the substance by suppressing the influence of image distortion.

B. Detailed Operation of Embodiment FIG. 1 is a diagram showing an operation flow when one embodiment of the present invention is carried out by using the apparatus 1 shown in FIG.

First, while the light projector 4 is turned on, the image data of the circular hole 5 as the inspection object is captured by the image sensor 6 and stored in the image processing circuit 7 (step S1).

Then, the approximate center position P _c ′ (x _c ′, y _c ′) of the circular hole 5 is obtained based on the stored image data (step S2). The following is one example of the method.

(i) As shown in FIG. 3, the X coordinate value x at the intersection of the circular hole image and the scanning line that first intersects the circular hole image by scanning in the X direction at predetermined intervals d.
_{Find a} and x _b .

(ii) By substituting the X coordinate values x _a and x _b obtained in (i) into the following equation, the X coordinate value x _c ′ is obtained.

x _c ′ = (x _a + x _b ) / 2 (3) (iii) The X coordinate is set to the X coordinate value x _c ′, scanning is performed in the Y direction, and Y at the intersection of the scanning line and the circular hole image. coordinate value _y a, obtaining a _{y b.}

(iv) Motoma' y-coordinate value y _a in _(iii), finding a y-coordinate value y _{c 'by} substituting y _b in the following equation.

_{y c '= (y a +} y b) / 2 ... (4) the approximate center coordinates of the circular hole 5 to be determined whether a clogging the above _{P c' (x c ',} y c') is obtained.

In the above method, in order to obtain the X coordinate value x _c ′, X
The coordinate values x _a and x _b are obtained one by one, but X
The X coordinate value x _c ′ may be obtained by obtaining two or more coordinate values x _a and x _b and then performing statistical processing (for example, obtaining an average value Ix _c ′). In this case, X
This has an advantage that the coordinate value x _c ′ can be obtained with high accuracy.

Next, based on the center position P _c ′ obtained in step S2 and the hole diameter of the circular hole 5, as shown in FIG. 4, the area for sampling in the horizontal direction and the vertical direction and the interval for sampling are determined, and the horizontal scanning line And the circular hole image and the coordinates (x _i , y _i ) (where i = 1, 2, ..., N) of the vertical scanning line and the circular hole image are obtained (step S3). here,
To obtain the coordinates (x _i , y _i ) of the points on the circumference,
As shown in FIG. 4, the sampling area is set to 45 ° in each of the X direction and the Y direction, and the sampling interval is set to a constant sampling interval d _s . In this case, the intervals between the adjacent points on the circumference are not equal to each other, but they can be sufficiently used in carrying out the present invention. Although in the process of evaluating the degree of circularity is known diameter of the circular hole 5, even if this value is not known, the y-coordinate values obtained in (iii) in step S2 y _a, of y _b It goes without saying that the difference may be the diameter of the circular hole 5.

Then, by substituting the coordinates (x _i , y _i ) of the points on these circumferences into the following equation, the central position P _c (x _c , x _c , which is more accurate than the central position P _c ′ obtained in step S2:
y _c ) is _calculated (step S4).

Further, the center position P obtained in step S4
By substituting _c (x _c , y _c ) and the coordinates (x _i , y _i ) of the points on each circumference into the following equation, the radius r at the points on each circumference
_i (i = 1, 2, ..., N) is obtained (step S5).

Then, the radius r _i obtained in step S5 is substituted into the following equation, and the radius difference Δr (= Δ
r ₁₂ , Δr ₂₃ , ..., Δr _i1 ) is calculated (step S
6).

Finally, various statistics (for example, maximum value,
(Average value, standard deviation, etc.) is calculated (step S7), and the circularity of the circular hole 5 is evaluated based on the statistic.
Whether or not there is a summary is determined (step S8).

As described above, in the present embodiment, two adjacent points on the circumference are selected as two adjacent points on the circumference. Then, since the difference Δr in radius between adjacent points on the circumference is obtained and the circularity is evaluated by various statistics of the difference, it is possible to suppress the influence of the distortion of the image generated by the image processing. That is, even if the image of the circular hole 5 having no groove becomes an ellipse due to the distortion of the image and the radius r _{i at} the points on each circumference greatly varies, the difference in the radius difference Δr at the points on the adjacent circumferences is equal to the radius. It is considerably smaller than the variation of r _i . Therefore, when the circular hole 5 has a groove, the variation in the radius difference Δr between the points on the adjacent circumferences is different from that when there is no groove. However, the variation depends on the shape and size of the deposit.

Therefore, next, general models of thumbs are given, and how the circularity, that is, the presence or absence of the thumbs is determined when the above-described embodiment is applied to the models.

FIG. 5 (a) shows the first model of Tsumari, which has a circular hole 5
The needle-shaped deposit 51 is attached to a part of the. Also, the fifth
FIG. 5 (b) shows the difference Δr in radius between adjacent points on the circumference when the above embodiment is applied to the circular hole 5 shown in FIG. 5 (a).
FIG. 3 is a diagram showing the distribution of ## EQU1 ## where the horizontal axis shows the value of the radius difference .DELTA.r between adjacent points on the circumference, and the vertical axis shows the frequency with respect to that value. As can be seen from FIG. 7B, the first model has two large and small peaks p ₁ and p ₂ . The larger peak p ₁ is formed in a region close to the region where the radius difference Δr between adjacent points on the circumference is 0, and the smaller peak p ₂ has a relatively small radius difference Δr. It is formed in a region having a large value. That is, the smaller peak p ₂ corresponds to the difference in radius at points on the circumference existing in the range R ₁ shown in FIG. 5 (a), in other words, due to the presence of the deposit 51. It can be said that the smaller peak p ₂ is formed. Therefore, if the deposit 51 is not present, the smaller peak p ₂ is not formed and only the larger peak p ₁ is formed.

From the above, in the case of the first model, it is possible to determine the presence / absence of a summary based on the maximum value Δr _max of the radius difference Δr at points on the adjacent circumferences. That is, the difference Δr between the adjacent radii is obtained as described above, and the maximum value Δ
The presence / absence of a summary can be determined by checking the magnitude relationship between r _max and a predetermined value (criterion determination criterion for presence / absence).

Next, the second model will be described. FIG. 6 (a) is a diagram showing the second model, and FIG. 6 (b) is a diagram showing points on adjacent circumferences when the above embodiment is applied to the circular hole 5 shown in FIG. 6 (a). 6 is a diagram showing a distribution of radius differences Δr in FIG. The horizontal axis and the vertical axis in FIG. 5B are the same as those in FIG. 5B. As can be seen from FIG. 6 (b), in the case of the second model, there is only one peak, and the center value of the peak p ₃ is the average value of the radius differences Δr at points on the adjacent circumferences. It almost agrees with IΔr, and the average value IΔr or maximum value Δ
It is possible to determine the presence or absence of a summary by r _max . That is, the radius difference Δr between adjacent points on the circumference is obtained as described above, and the size difference between the maximum value Δr _max or the average value IΔr thereof and a predetermined value (criterion presence / absence determination) is checked to find a summary. It is possible to determine the presence or absence of.

Next, the third model will be described. FIG. 7 (a) is a diagram showing a third model, in which the circular hole 5 has an arc shape due to the attachment of the deposit 71. Also, the same figure
(b) is a diagram showing a distribution of radius differences Δr at points on adjacent circumferences when the above-described embodiment is applied to the circular hole 5 shown in FIG. The horizontal axis and the vertical axis in FIG. 5B are the same as those in FIG. 5B. 7 As can be seen from Figure (b), in the case of the third model, the number of peaks is one, the peak p ₁ in the case of the shape of the peak p ₄ are have second model first is , P ₂ and p ₃ are broader. In such a case, it is possible to determine the standard deviation σ of the difference Δr of the radii at the points on the adjacent circumferences, and determine the presence or absence of the summary from the value. That is, as described above, the difference Δr between the radii at the points on the adjacent circumferences is obtained, and the standard deviation σ and a predetermined value (criterion presence / absence criterion)
Whether or not there is a summary can be determined by examining the magnitude relationship with.

By the way, in terms of reliability when determining the characteristics of the event in statistical processing, when the number of processed data exceeds 10, the maximum value obtained from the processed data is the average value obtained from the same processed data or It is generally known that it is inferior to the standard deviation. Also, the average value It is not clear which is the effective means for judging the characteristic of the event by comparing the standard deviation σ with the standard deviation σ.
Therefore, the above average value A method of substituting the standard deviation σ into the following equation and comparing the obtained value with a predetermined value to determine the presence or absence of a summary can be considered.

This method uses the average value as shown in equation (8). And the standard deviation σ, the mean value 1 and the standard deviation σ are weighted by 3, and it becomes possible to evaluate the circularity in practice.

It should be noted that in the above-described embodiment, the points on the adjacent circumferences are selected as the two points on the adjacent circumferences, but in addition to this, every other adjacent points on the circumference is selected, and the difference in radius of all the points is set as a whole. You may ask. In this case, instead of the radius difference Δr (= Δr ₁₂ , Δr ₂₃ , ..., Δr _i1 ) calculated by the equation (7), the radius difference Δr ′ (= Δr 1
Various statistics are obtained as described above by using r ₁₃ , Δr ₂₄ , ..., Δr _i2 ) and the circularity of the circular hole 5 is evaluated.
The presence / absence of a summary may be determined.

Further, instead of selecting points on the adjacent circumferences or two points on the adjacent circumferences as two points on the adjacent circumferences, n (n
Needless to say, it is also possible to select adjacent points on the circumference at intervals of ≧ 2) and obtain the difference in radius for all points as a whole. That is, it may be appropriately selected according to the size of the object to be inspected, the number of coordinates obtained in step S3, and the like. In addition, as this selection method, the device 1 shown in FIG.
It may be configured to automatically set appropriately according to the size of the object to be inspected, the number of the coordinates, or the like, or to give an appropriate signal or the like from the outside.

Further, in the above-described embodiment, the example in which the difference between the radii of two points on the adjacent circumference is obtained for all the points in step S6 is shown. , For example, Δr ₁₂ , Δr ₂₃ , Δr ₃₄ , Δ in the equation (7).
r ₄₅ . Δr ₅₆ , Δr ₆₇ , Δr ₇₈ ... Is Δr ₁₂ , Δr ₄₅ , Δ
It is also possible to obtain the difference in radius as r ₇₈ , and further use the difference in radius to obtain various statistics as described above to evaluate the circularity of the circular hole 5 and determine the presence or absence of the outline.

Also, determine the approximate center position p _{c 'in} step S2 in the above embodiment, although further re seeking again the central position p _c in step S4, if Motomare the center position with a sufficient accuracy in step S2 step S
The operation of 4 is unnecessary.

Further, generally, a spinneret is provided with a plurality of circular holes, and it is necessary to determine the presence or absence of a tsunami in all of these circular holes, but the presence or absence of the tsunami is determined by the above method. If the number of circular holes having a trim at a certain stage becomes a certain number or more and the spinneret is washed by stopping the determination of the presence or absence of the trim, the time required for the above determination can be shortened.

Further, by outputting the number of circular holes for each statistic (for example, maximum value, average value, standard deviation, etc.) obtained by the above method for each spinneret, it becomes effective information for quality control in the manufacturing process.

(Effects of the Invention) As described above, according to the present invention, the difference between the radii corresponding to two points on the adjacent circles among the points on the plurality of circles is further statistically calculated based on the value of the difference. By determining the amount and evaluating the circularity from the total metric, it is possible to suppress the influence of the distortion of the reproduced image stored in the first step.

[Brief description of drawings]

FIG. 1 is a diagram showing an operation flow of one embodiment of the present invention, FIG. 2 is a configuration diagram of an image reading apparatus to which one embodiment of the present invention is applicable, and FIG. FIG. 4 is a diagram showing an example of a method of obtaining an approximate center position of the circular hole from image data of the circular hole, and FIG. 4 is a diagram showing an example of a method of obtaining points on each circumference in the present embodiment. FIG. 5 (a) is a view showing a first model of the thumb, and FIG. 5 (b) is adjacent to the circular hole shown in FIG. 5 (a) when this embodiment is applied. It is a figure which shows the distribution of the difference of the radius in the point on the circumference, FIG.6 (a) is a figure which shows the 2nd model of a summary, FIG.6 (b) is shown in FIG.6 (a). It is a figure which shows distribution of the difference of the radius in the point on the circumference which adjoins when a present Example is applied with respect to a circular hole, FIG.7 (a) is a figure which shows the 3rd model of a summary. FIG. 7 (b) is a diagram showing the distribution of radius differences at points on adjacent circumferences when this embodiment is applied to the circular hole shown in FIG. 7 (a). The figure is a diagram for explaining the problems of the conventional method. 5 ... Circular hole, S1-S8 ... Step

Claims (1)

[Claims] 1. A first step of capturing an image of an object to be inspected and storing it as image data, and a second step of obtaining a center position of an image of the object to be inspected stored in the first step based on the image data. And a third step of obtaining a radius represented by a distance between a plurality of points on the circumference of the image of the object to be inspected and the center position, and two points on the adjacent circumference of the plurality of points on the circumference. A fourth step of obtaining a difference between respective corresponding radii, a fifth step of obtaining a statistic amount of the plurality of differences obtained in the fourth step, and the inspection object based on the statistic amount. And a sixth step of evaluating the circularity of the object.