JP2005351760A - Distortion measuring method and system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a distortion measuring method and a system which makes possible, automatization of distortion tests, elimination of skilled testers, stable and fair tests because of quantification of criterion, and distortion tests of gradually curved mirrors (complex curvature mirror) which was conventionally difficult. <P>SOLUTION: By using a predetermined software, a reflection image measured data which is obtained by reflecting a pattern of predetermined dots 12 provided on a screen 4 to the reflection surface (curved surface 5a) of an test object mirror 5, and reflection image theory data of the predetermined dot pattern which is obtained by calculating based on a design value used for forming the reflection surface 5a of the test object mirror 5 are compared so as to measure the distortion of the test object. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a distortion inspection method and apparatus, and more particularly to a method and apparatus for inspecting distortion of a reflecting surface such as a mirror.

  Conventionally, in mirror distortion inspection according to the JIS standard, a concentric circle with a 10 mm graduation and an 8-segment line that passes through the center of the mirror are directly opposed to the mirror. The ratio of the average radius of the mirror image (concentric circle) to the maximum (or minimum) radius (distortion rate [%]) is obtained by using an apparatus capable of projecting a scale image, and the degree of distortion is quantified. However, in an actual mass production factory, it is less efficient to perform the inspection as described above, so a plurality of straight lines are copied on the mirror surface, and the distortion inspection is performed by checking the degree of bending of the mirror image (straight line).

Further, a technique related to a method for measuring distortion of a stamper for forming a CD (compact disc), which is not a mirror, is known. This is a reflection image of a reference pattern formed by reflection on the surface of a stamper to be measured. And a reflection image of a stamper prepared in advance without distortion, and as a measuring method thereof, image data of a video camera is processed by a personal computer and compared (see Patent Document 1).
JP-A-7-83624

  However, the above prior art has the following problems.

First, about the distortion inspection of the mirror by JIS standard,
(1) Since the shape of the mirror is irregular, when a concentric scale is copied, there is a range where a mirror image cannot be obtained.
(2) Generally, since the distortion tends to occur in the outer peripheral portion of the mirror, particularly in an extremely protruding range (a range where a mirror image cannot be obtained), an appropriate inspection may not be performed.
(3) Inspection takes time and efficiency is poor.
(4) This JIS standard can be applied to single plane mirrors and spherical mirrors with a single radius, and is not applicable to gradually changing curved mirrors that have been used in recent years to expand their field of view. Is possible.

Next, about the distortion inspection of the mirror in the current mass production factory,
(5) Since it is a visual sensory test, skill of the inspector is required.
(6) Since the criterion is ambiguous, the test result depends on the subjectivity of the inspector.
(7) In the case of a mirror other than a flat surface and a single curved surface, such as a gradually changing curved surface mirror, the determination becomes more difficult.
(8) For an inspection object close to the strain limit (limit sample), it may take time to determine.

Furthermore, with respect to Patent Document 1 (Japanese Patent Laid-Open No. 7-83624),
(9) It is difficult to form a stamper that is completely free of distortion, and it is also difficult to confirm that there is no distortion. Therefore, the reliability of the reference image data becomes a problem.

  For this reason, the present invention uses predetermined software based on the reflection image actual measurement data obtained by reflecting a predetermined pattern on the reflection surface to be inspected and the design value used for forming the reflection surface to be inspected. The first feature is that the distortion of the reflection surface to be inspected is measured by comparing the reflection image theoretical data of the predetermined pattern obtained by calculation in the above.

  Next, in the distortion measurement method according to claim 1, the predetermined pattern is a pattern composed of uniformly arranged dots, and the positions of the dots in the reflected image actual measurement data are compared with the positions of the dots in the reflected image theoretical data. A second feature is that the distortion of the reflection surface to be inspected is measured.

  Further, in the strain measuring method according to claim 1 or 2, the software determines the shape of the theoretical curved surface of the reflecting surface of the inspection object using a predetermined cubic curve and the cubic curve. A step of approximating a theoretical curved surface of the reflecting surface by a point group using a predetermined number of reference points, a step of obtaining a unit normal vector of the reference point in the point group, and position data of the predetermined pattern; A vector connecting the reference point and the predetermined pattern from the position data of the reference point in the point group on the theoretical curved surface and the position data of the camera to which the predetermined pattern reflected by the reflecting surface is mapped, A vector that connects a reference point and the camera is calculated, and a unit vector that bisects an angle formed by a vector that connects the reference point and the predetermined pattern and a vector that connects the reference point and the camera is calculated. And-up, and the unit vector, a third feature that the difference between the unit normal vector of the reference point in the point group and a step of determining by searching the point where the minimum.

  And it is a distortion measuring device, which is used to form a reflection image actual measurement data by reflecting a predetermined pattern on the reflection surface of the inspection object, and to form the reflection image actual measurement data and the reflection surface of the inspection object A fourth feature is provided with a comparison means for comparing the reflection image theoretical data of a predetermined pattern obtained by calculation using predetermined software based on the design value.

  5. The distortion measuring apparatus according to claim 4, wherein the predetermined pattern is a pattern composed of uniformly arranged dots, and measuring means for measuring the positions of the dots of the reflected image actual measurement data and the dots of the reflected image theoretical data is provided. The fifth feature is that it is provided.

  In addition, in the distortion measuring apparatus according to claim 4 or 5, the software uses a means for defining a shape of a theoretical curved surface of a reflection surface of the inspection object with a predetermined cubic curve, and the cubic curve. Means for approximating the theoretical curved surface of the reflecting surface by a point group using the determined predetermined number of reference points; means for determining a unit normal vector of the reference point in the point group; and position data of the predetermined pattern A vector connecting the reference point and the predetermined pattern from the position data of the reference point in the point group on the theoretical curved surface and the position data of the camera to which the predetermined pattern reflected by the reflecting surface is mapped, Means for calculating a vector connecting the reference point and the camera, and calculating a unit vector that bisects an angle formed by a vector connecting the reference point and the predetermined pattern and a vector connecting the reference point and the camera; And simply And a vector, and the sixth aspect of the difference between the unit normal vector of the reference point in the point cloud is equipped with means for determining by searching point to be minimized.

The vehicle door handle device according to the present invention has the following excellent effects.
(1) The distortion inspection can be automated, and a skilled inspector becomes unnecessary.
(2) Since determination criteria are quantified, stable and fair inspection is possible.
(3) Distortion inspection of a gradually changing curved mirror (composite curvature mirror), which has been difficult in the past, can be performed.

  Hereinafter, the best mode for carrying out the present invention will be described based on an embodiment shown in the drawings. FIG. 1 is a schematic view of an embodiment of a strain measuring device according to the present invention as viewed from the top. FIG. 2 is a schematic view of an embodiment of the strain measuring apparatus according to the invention as viewed from the front. FIG. 3 is a plan view of a screen on which a dot pattern is drawn, and FIG. 4 is an explanatory diagram showing a state in which the dot pattern of the screen of FIG. 3 is reflected on a mirror to be inspected. FIG. 5 is an explanatory diagram in the case where the speed of inspection is divided by region. FIG. 6 is a conceptual diagram of pass / fail judgment by comparing the dot distribution of theoretical data and the dot distribution of measured data. FIG. 7 is a conceptual diagram when a theoretical curved surface is displayed as a point group using a cubic curve, and FIG. 8 is an explanatory diagram showing a situation where reference points for searching are displayed. 9 to 13 are explanatory diagrams showing the concept of the search method. FIG. 14 is an explanatory diagram showing the concept of the mapping position calculation method. FIG. 15 is a plan view showing the calculated positions of the dots on the curved surface, and FIG. 16 is a plan view obtained by converting FIG. 15 onto a plane photograph.

  As shown in FIG. 1, the distortion measuring apparatus 1 according to the present invention includes a mirror carry-in unit 2, an inspection unit 3, and a carry-out unit 8. The index 9 is rotated by a direct drive rotary motor (not shown), The mirror 5 as an inspection object is transferred to each stage 10. The inspection unit 3 is a dark box, and as shown in FIGS. 2 and 3, a lens insertion hole 4a for inserting the lens unit 6a of the camera 6 is formed in the central portion on the inner top surface, and the pitch is equal. A screen 4 on which a plurality of dots 12 arranged in a grid is drawn is attached. A mirror 5 is placed in front of the screen 4, and a camera 6 is installed above the center of the mirror 5 to capture the entire mirror 5. In addition, an illuminating device 7 is disposed at a position facing the screen 4 in order to increase the contrast during imaging. FIG. 4 shows a state in which a part of the dots 12 drawn on the screen 4 is reflected on the reflection surface 5a of the mirror 5, and the mirror 6 is attached to the camera 6 at the center of the reflection surface 5a. The lens 6a is shown. At this time, depending on the size and shape of the mirror 5, all the dots 12 on the screen 4 may be reflected on the reflection surface 5 a of the mirror 5. Since the mirror 5 is a gradually changing curved surface mirror, in the range from the left end of FIG. 4 to the right center portion, the dots 12 are shown in a lattice pattern at substantially equal pitches, but in the range from the right center portion to the right end side, It is shown that the pitch of the dots 12 is gradually shortened toward the right end side. The measuring device 1 is connected to a personal computer 11 on which a reference table that records software or theoretical data calculated by software is mounted.

  The “theoretical data” calculated by the software is the reflection surface of the mirror 5 under the imaging conditions in the inspection unit 3 dark box (distance h1 between the mirror 5 and the camera 6, distance h2 between the mirror 5 and the screen 4, etc.). Refers to the dot coordinates theoretically calculated by the search method by vector analysis based on the design value such as the curvature value of the reflection surface of the reflection image that will be reflected, the present invention is "theoretical data" in the distortion inspection It is characterized in that pass / fail judgment is performed with reference to the above. According to this method, it is possible to detect not only a curved surface having a simple curvature (that is, a single radius) but also a complex curvature surface having a plurality of curvatures. In the present invention, “theoretical data” calculated using this vector analysis search method is used, but data calculated by other methods may be used.

When the index 9 rotates and the mirror 5 is conveyed to the inspection unit 3 in a predetermined posture, the dots 12 on the screen 4 are reflected as a mirror image on the surface of the mirror 5. These dot distributions are imaged by the camera 6 on the top of the dark box, the edges of all the dots 12 are detected by a general image processing method, and fitting with a circle is performed. The coordinates are output as “dot distribution (x ′ _n , y ′ _n )”. This coordinate value (x ′ _n , y ′ _n ) is compared with the coordinate value (x _n , y _n ) obtained from “theoretical data”. The personal computer 11 calculates the deviation (difference) between the coordinate values of the same coordinates, and quantifies the degree of distortion of the mirror 5 by an appropriate arithmetic expression. Here, the arithmetic expression may be an accumulation of the squares of the deviation amounts of all coordinate values (the following mathematical expression 1), or an inspection may be provided depending on the region (the following mathematical expression 2). And it considers that it is distorting, so that the (DELTA) value shown to following Numerical formula 1 and Numerical formula 2 is large, and it determines with a thing less than a regulation threshold value as a pass.

(Formula 1)
Δ = (x ₁ -x ' ₁ ) ² + (y ₁ -y' ₁ ) ² + (x ₂ -x ' ₂ ) ² + (y ₂ -y' ₂ ) ² + ... + (x _n- x ' _n ) ² + (y _n -y' _n ) ²

(Formula 2)
Δ = a {(x ₀ -x ' ₀ ) ² + (y ₀ -y' ₀ ) ² + (x ₀ -x ' ₀ ) ² + (y ₁ -y' ₁ ) ² + (x _-1 -x ' _-1 ) ² + (y ₁ -y' ₁ ) ² + (x _-1 -x ' _-1 ) ²
+ (y ₀ -y ' ₀ ) ² + (x _-1 -x' _-1 ) ² + (y _-1 -y ' _-1 ) ² + (x ₀ -x' ₀ ) ² + (y _-1- y ' _-1 ) ² + (x ₁ -x' ₁ ) ² + (y _-1 -y ' _-1 ) ²
+ (x ₁ -x ' ₁ ) ² + (y ₀ -y' ₀ ) ² + (x ₁ -x ' ₁ ) ² + (y ₁ -y' ₁ ) ² }
+ b {(x ₀ -x ' ₀ ) ² + (y ₂ -y' ₂ ) ² + (x _-1 -x ' _-1 ) ² + (y ₂ -y' ₂ ) ² + (x _-2- x ' _-2 ) ² + (y ₂ -y' ₂ ) ² + (x _-2 -x ' _-2 ) ²
+ (y ₁ -y ' ₁ ) ² + (x _-2 -x' _-2 ) ² + (y ₀ -y ' ₀ ) ² + (x _-2 -x' _-2 ) ² + (y _-1- y ' _-1 ) ² + (x _-2 -x' _-2 ) ²
_{+ (y -2 -y '-2)} 2 + (x -1 -x' -1) 2 + (y -2 -y '-2) 2 + (x 0 -x' 0) 2 + (y - ₂ -y ' _-2 ) ² + (x ₁ -x' ₁ ) ²
+ (y _-2 -y ' _-2 ) ² + (x ₂ -x' ₂ ) ² + (y _-2 -y ' _-2 ) ² + (x ₂ -x' ₂ ) ² + (y _-1- y ' _-1 ) ² + (x ₂ -x' ₂ ) ² + (y ₀ -y ' ₀ ) ²
+ (x ₂ -x ' ₂ ) ² + (y ₁ -y' ₁ ) ² + (x ₂ -x ' ₂ ) ² + (y ₂ -y' ₂ ) ² + (x ₁ -x ' ₁ ) ² + (y ₂ -y ' ₂ ) ² }
+ c {(x ₀ -x ' ₀ ) ² + (y ₃ -y' ₃ ) ² + (x _-1 -x ' _-1 ) ² + (y ₃ -y' ₃ ) ² + (x _-2- x ' _-2 ) ² + (y ₃ -y' ₃ ) ² + (x _-3 -x ' _-3 ) ²
+ (y ₃ -y ' ₃ ) ² + (x _-3 -x' _-3 ) ² + (y ₂ -y ' ₂ ) ² + (x _-3 -x' _-3 ) ² + (y ₁ -y ' ₁ ) ² + (x _-3 -x' _-3 ) ² + (y ₀ -y ' ₀ ) ²
+ (x _-3 -x ' _-3 ) ² + (y _-1 -y' _-1 ) ² + (x _-3 -x ' _-3 ) ² + (y _-2 -y' _-2 ) ² + ( x _-3 -x ' _-3 ) ² + (y _-3 -y' _-3 ) ²
+ (x _-2 -x ' _-2 ) ² + (y _-3 -y' _-3 ) ² + ((x _-1 -x ' _-1 ) ² + (y _-3 -y' _-3 ) ² + (x ₀ -x ' ₀ ) ² + (y _-3 -y' _-3 ) ²
+ (x ₁ -x ' ₁ ) ² + (y _-3 -y' _-3 ) ² + (x ₂ -x ' ₂ ) ² + (y _-3 -y' _-3 ) ² + (x ₃ -x ' ₃ ) ² + (y _-3 -y' _-3 ) ² + (x ₃ -x ' ₃ ) ²
+ (y _-2 -y ' _-2 ) ² + (x ₃ -x' ₃ ) ² + (y _-1 -y ' _-1 ) ² + (x ₃ -x' ₃ ) ² + (y ₀ -y ' ₀ ) ² + (x ₃ -x' ₃ ) ² + (y ₁ -y ' ₁ ) ²
+ (x ₃ -x ' ₃ ) ² + (y ₂ -y' ₂ ) ² + (x ₃ -x ' ₃ ) ² + (y ₃ -y' ₃ ) ² + (x ₂ -x ' ₂ ) ² + (y ₃ -y ' ₃ ) ² + (x ₁ -x' ₁ ) ²
+ (y ₃ -y ' ₃ ) ² }

  In the case of Equation 2, the mirror 5 is divided into three areas A, B, and C as shown in FIG. 5, and a coefficient a is assigned to the calculation of the area A, and a coefficient b is assigned to the calculation of the area B. In addition, a coefficient c is assigned to the calculation of the region C. A large numerical value is assigned to each of the coefficients a, b, and c as the inspection is severer. If the area A is the center part of the mirror 5, the area C is the outer peripheral part of the mirror 5, and the area B is the middle part thereof, the distortion inspection needs to be stricter as it is closer to the center part. The value decreases in the order of b and c.

  FIG. 6 shows an image of pass / fail judgment by comparing “dot distribution of theoretical data of reflected image” and “dot distribution of measured data of reflected image”. In this embodiment, as shown in FIG. 6, the dot size of the actual measurement data is enlarged and displayed as the difference from the dot position of the theoretical data is larger. A method may be used in which the dot color is changed depending on the size of the difference from the dot distribution.

  Next, a software configuration for calculating “theoretical data of the reflected image” will be described. This software is a simulation software for mapping the curved reflection image defined by the cubic curve. Based on the basic cubic curve, several reference points are calculated, and these reference points are rotated and moved. After the curved surface shape is expressed as a point cloud, a predetermined pattern (on the background screen) is drawn from the normal vector of each reference point in the point cloud, the camera position (lens position) to capture the mapping, etc. In the present embodiment, the position of the dot in the mapping is simulated. The present invention is characterized by a method for searching for a reflection position on a curved surface.

  The use of a door mirror attached to a car has a wider field of view and can be made more compact by using a curved surface composed of a plurality of radii than a curved surface having a single radius. However, it is not easy to confirm whether such a curved surface is accurately formed. Therefore, it is possible to inspect whether the product is manufactured with high accuracy by obtaining an ideal curved surface, simulating how the background appears on the ideal curved surface, and comparing it with the product.

  In the present invention, in the state where the screen 4 depicting the pattern of dots arranged at an equal pitch is placed on the background, the position where the dots appear on the cubic curved mirror (mirror 5) is searched, and the mapping on the ideal curved surface is performed. Can be sought. The method will be described below.

(Approximate expression of curved surface by point cloud)
First, the approximate expression by the point group of the curved surface 5a of the mirror will be described with reference to FIG. As shown in FIGS. 7A and 7B, the center of the sphere GL is the intersection of the axes X, Y, and Z that are orthogonal to each other, and this intersection is defined as the origin (0, 0, 0). To do. Further, at the intersection of the outer circumferential surface of the sphere GL and the axis Y, the intersection on the left side of the drawing is the north pole NP, the intersection on the right side of the drawing is the south pole SP, and the great circle on the outer circumference of the sphere GL connecting the north pole NP and the south pole SP is the meridian ME To do. The meridian ME is a single radius curve. First, as shown in FIG. 7A, the curved surface of the mirror 5 is partially defined by a cubic curve 21 from the middle of a single radius curve 20 near the north pole NP on the meridian ME. Several reference points 22a are obtained on the single radius curve 20, and several reference points 22b are obtained on the cubic curve 21, and these reference points (the reference points 22a and 22b are combined). The point 5) is rotationally moved along the outer peripheral surface of the sphere GL with the axis Y as the rotation axis, and as shown in FIG. 7B, the curved surface 5a is approximated by a point group 23 including reference points 22a and 22b. . In FIG. 7A and FIG. 7B, a circle C1 indicates a rotation trajectory at a predetermined point in the reference point 23b. Further, FIG. 8 shows a situation in which the reference points 22a and 22b for search created by this method are displayed. The point group 23 includes a reference point 22a on a curved surface having a single radius and a reference point 22b on a gradually changing curved surface. The boundary between the single radius curved surface and the gradually changing curved surface is indicated by an arc. In the present embodiment, the reference points 22a and 22b are arranged near the north pole NP and rotated. However, the reference points 22a and 22b can be arranged and rotated at other positions.

(Search method)
Next, the search method will be described with reference to FIGS. First, as shown in FIG. 9, the position Pt at which the dot Pm (dot 12) of the background (on the screen 4) is projected onto the theoretical curved surface is the normal vector n (hereinafter, the vector is The angle α formed by the vector PtPm connecting the projection position Pt and the dot Pm and the vector PtPc connecting the projection position Pt and the center position Pc of the lens 6a is bisected. Is a known fact. In other words, the β angle between vector PtPm and the normal vector n, is equal to β the angle between the vector PtPm and the normal vector n. Under such conditions, the unit vector that bisects the angle formed by the vectors PtPm and PtPc is equal to the unit normal vector at the projection position Pt, so the difference between the unit vector and the unit normal vector is “ 0 ". Therefore, the search was performed according to the following procedure.
(1) As shown in FIG. 10, with respect to each reference point Pt _i (i is an integer) in the point group that approximates the curved surface 5a of the mirror 5, the unit normal vector n _{i of} all points is obtained from the basic equation of the curved surface. (I is an integer) data is theoretically calculated in advance.
(2) Next, the dots Pm _i (i is an integer) drawn on the background screen 4, the camera lens center position Pc, and a predetermined position Pt _i (i is an integer) on the curved surface are connected. A unit vector N _i (i is an integer) that bisects an angle formed by two vectors ( PtPm _i (i is an integer) and Pt _i Pc (i is an integer)), and the unit method calculated in advance in (1) above The point where the difference from the line vector n _i (i is an integer) is minimized is theoretically calculated. In this embodiment, as shown in FIG. 11, first, the lens insertion hole 4a reference point Pm _1, which is located closer to the on screen 4, perform the above calculation for all the points Pt _i on the curved surface, The difference between the unit vector and the unit normal vector is compared, and a reference point that minimizes the difference between the unit vector and the unit normal vector is obtained. The point theoretically obtained in this way is defined as Pa. In this embodiment, the point Pa is a point at the same position as Pt _{11 as} shown in FIG.
(3) Next, as shown in FIG. 12, each intermediate point (16 white points in FIG. 12) is obtained based on the surrounding eight points adjacent to the obtained point Pa.
(4) Then, from the total of 25 reference points of the obtained 16 intermediate points and the original point (9 black points in FIG. 12), unit normal vectors of these 25 reference points And a reference point having a minimum difference between the above-described background dot Pm ₁ , the position Pc of the camera lens center, and the unit vector that bisects the angle formed by the two vectors connecting the 25 reference points. This reference point is defined as a point Pb. Here, since the unit normal vector data of the original point (9 black points in FIG. 12) is calculated in advance in the above (1), there are 16 intermediate points (Ptw _i (i = 1 to 1)). It is necessary to newly obtain the unit normal vector nw _i (i = 1 to 16) of 16)). In this embodiment, as shown in FIG. 12, the PTW ₅ that it has a unit normal vector nw _5, a point Pb.
(5) Further, the obtained point Pb is set as Pa, and the process returns to (3) to repeat the process. Specifically, as shown in FIG. 13, each intermediate point (16 in the two-dot chain line frame in FIG. 13) is based on the surrounding eight points adjacent to the point Pb (replaced by Pa) obtained in (4) above. White points), and from the total of 25 points including the obtained 16 intermediate points and the original points (9 black dots in the two-dot chain line frame in FIG. 13), the unit method A unit vector that bisects the angle formed by the two vectors connecting the line vector, the background dot Pm ₁ , the camera lens center position Pc, and the 25 reference point positions on the curved surface; A reference point having a minimum difference between the obtained point and the unit normal vector is obtained. Here, since the unit normal vector data of the original point (9 black dots in the two-dot chain line in FIG. 13) has already been calculated in the above (4), 16 intermediate points (see FIG. It is necessary to newly obtain a unit normal vector (16 white circles in 13 two-dot chain lines).
(6) the search for the (3) to (6), repeated while narrowing the search range, the reference points Pm ₁ is a position to be mapped to the theoretically mirror curved surface 5a can be determined approximately. In the above operations (3) to (6), if the difference between the unit vector and the unit normal vector is within a predetermined range, it is considered that the unit vector and the unit normal vector match. End the search.
(7) In this way, by repeatedly searching for all the dots Pm _i (i is an integer) on the screen 4 while narrowing the search range, the positions where the dots Pm _i are mapped onto the theoretical curved surface 5a are determined. , All can be obtained approximately.

(Request map)
Then, after the projection position on the curved surface is obtained, the position of the dot when this situation is made a single photograph can be obtained. At this time, as shown in FIG. 14, the vertex of the curved surface shape is placed at the origin of the XY plane, and the position of the projected position of the dot on the curved surface on the XY plane may be obtained. At this time, the intersection position Pf between the line segment connecting the dot position Pt projected on the curved surface and the camera lens center position Pc and the XY plane is the position to be obtained. FIG. 15 is a plan view showing the calculated dot projection position Pt on the curved surface, and FIG. 16 is a plane showing the position Pf when the point Pt on the curved surface in FIG. 15 is mapped to the camera film. The position data indicating the dot Pf is reflection image theory data.

  The reflected image theoretical data is calculated using the software configured as described above, stored in advance as a reference table in a personal computer, and the reflected image theoretical data in the reference table and the reflected image actual measurement data are compared at the time of inspection. The method of searching for the position where the predetermined pattern is reflected by comparing the vector direction consisting of each point normal vector on the curved surface and the predetermined pattern of the background and the position of the camera as described above is very unique, High-speed calculation processing can be realized.

It is the schematic diagram which looked at one Example of the distortion measuring apparatus which concerns on this invention from the plane. It is the schematic diagram which looked at one Example of the distortion measuring apparatus which concerns on this invention from the front. It is a top view of a screen. It is explanatory drawing which shows the state in which the screen of FIG. 3 is reflected in the mirror to be examined. It is explanatory drawing in the case of classifying the speed of inspection by area. It is an image figure of the quality determination by comparison of the dot distribution of theoretical data and the dot distribution of measured data. It is a conceptual diagram at the time of displaying a theoretical curved surface with a point cloud using a cubic curve. It is explanatory drawing which shows the condition which displayed the reference point for a search. It is explanatory drawing which shows the concept of the search method. It is explanatory drawing which shows the concept of the search method. It is explanatory drawing which shows the concept of the search method. It is explanatory drawing which shows the concept of the search method. It is explanatory drawing which shows the concept of the search method. It is explanatory drawing which shows the concept of the calculation method of a mapping position. It is a top view which shows the projection position of the dot on the calculated curved surface. FIG. 16 is a plan view obtained by converting FIG. 15 onto a plane photograph and projecting it.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Distortion measuring apparatus 2 Mirror carrying-in part 3 Inspection part 4 Screen 4a Lens insertion hole 5 Mirror 5a Curved surface 6 Camera 6a Lens 7 Illuminating device 8 Mirror carrying-out part 9 Index 10 Stage 11 Personal computer 12 Dot 20 Single radius curve 21 Tertiary curve 22a Reference point 22b obtained from a single radius curve 23b Reference point 23b obtained from a cubic curve

Claims (6)

Reflected image actual measurement data obtained by reflecting a predetermined pattern on the reflection surface to be inspected,
By comparing the reflection image theoretical data of the predetermined pattern calculated by using predetermined software based on the design value used to form the reflection surface of the inspection target, Distortion measuring method characterized by measuring distortion.   The predetermined pattern is a pattern composed of uniformly arranged dots, and the distortion of the reflection surface to be inspected is measured by comparing the positions of the dots of the reflection image actual measurement data and the dots of the reflection image theoretical data. The distortion measuring method according to claim 1, wherein: The software is
Defining the shape of the theoretical curved surface of the reflecting surface of the inspection object with a predetermined cubic curve;
Approximating the theoretical curved surface of the reflecting surface with a point group using a predetermined number of reference points determined using the cubic curve;
Obtaining a unit normal vector of a reference point in the point cloud;
From the position data of the predetermined pattern, the position data of the reference point in the point group on the theoretical curved surface, and the position data of the camera to which the predetermined pattern reflected by the reflecting surface is mapped, the reference point and the predetermined pattern Calculate a vector connecting a pattern, a vector connecting the reference point and the camera, and bisect an angle formed by a vector connecting the reference point and the predetermined pattern and a vector connecting the reference point and the camera Calculating a unit vector;
3. The distortion according to claim 1, further comprising a step of searching for and finding a point at which a difference between the unit vector and a unit normal vector of a reference point in the point group is minimized. Measuring method. Means for reflecting a predetermined pattern on the reflection surface to be inspected to obtain reflected image measurement data;
Comparison means for comparing the reflected image actual measurement data with the reflected image theoretical data of a predetermined pattern obtained by calculation using predetermined software based on the design value used for forming the reflection surface to be inspected. A distortion measuring device comprising:   5. The predetermined pattern according to claim 4, wherein the predetermined pattern is a pattern composed of uniformly arranged dots, and comprises a measuring means for measuring the positions of the dots of the reflected image actual measurement data and the dots of the reflected image theoretical data. Strain measuring device. The software is
Means for defining the shape of the theoretical curved surface of the reflecting surface of the inspection object by a predetermined cubic curve;
Means for approximating the theoretical curved surface of the reflecting surface with a point group using a predetermined number of reference points determined using the cubic curve;
Means for determining a unit normal vector of a reference point in the point cloud;
From the position data of the predetermined pattern, the position data of the reference point in the point group on the theoretical curved surface, and the position data of the camera to which the predetermined pattern reflected by the reflecting surface is mapped, the reference point and the predetermined pattern Calculate a vector connecting a pattern, a vector connecting the reference point and the camera, and bisect an angle formed by a vector connecting the reference point and the predetermined pattern and a vector connecting the reference point and the camera Means for calculating a unit vector;
6. The distortion according to claim 4, further comprising means for searching for and finding a point at which a difference between the unit vector and a unit normal vector of a reference point in the point group is minimized. measuring device.

Images