JP4846295B2 - Three-dimensional coordinate measuring apparatus and method - Google Patents

Description

  The present invention relates to a three-dimensional coordinate measuring apparatus and method, and more particularly to a three-dimensional coordinate measuring apparatus and method suitable for measuring a measured object having a large number of protrusions on its surface, such as a BGA. About.

  As one of semiconductor devices, BGA (Ball Grid Array) is used. In this BGA, a large number of fine solder balls are arranged on the surface of a substrate, and the height and position of each solder ball are within a specified range. necessary.

  That is, with the spread of this BGA, a technique for measuring the height and position of a ball having a three-dimensional curved surface is required. In recent years, since the number of balls on the BGA device is also increasing, a technique for measuring at high speed is particularly required.

  By the way, the following five types of techniques are known as typical techniques for measuring the three-dimensional position of a curved surface (Non-Patent Documents 1 to 5).

Non-Patent Document 1 relates to a three-dimensional measurement method using a confocal optical system. Non-Patent Document 2 relates to a method using optical triangulation. Non-Patent Document 3 relates to a method using laser light scanning (light cutting method). Non-Patent Document 4 relates to a method using moire interferometry. Non-Patent Document 5 relates to a three-dimensional inspection apparatus (applying stereo optics) in which an optical system is installed right above and in the direction of 45 degrees to the left and right.
Carl Zeiss Co., Ltd. Catalog (3D measuring device) Optical 3D Measurement, Chapter 2 Toru Yoshizawa, New Technology Communications March 8, 1993 Optical 3D Measurement, Chapter 3, Toru Yoshizawa, New Technology Communications, March 8, 1993 Optical 3D Measurement, Chapter 4 Toru Yoshizawa, New Technology Communications March 8, 1993 Algol Co., Ltd. Catalog (3D inspection equipment)

  However, in the method of Non-Patent Document 1, the optical system becomes complicated and the cost is very high. In addition, it is difficult to increase the speed because a minute movement in the height direction is required. According to the method of Non-Patent Document 2, light must be scanned as in the method of Non-Patent Document 3 for measuring the three-dimensional coordinates of a solder ball of a BGA device having a two-dimensional spread. And there exists the same problem as the method of the nonpatent literature 3 mentioned later.

  According to the method of Non-Patent Document 3, the laser beam must be scanned so as to pass through the apex of the solder ball as the object to be measured. There is a problem that the plane position of the ball cannot be measured, that is, the accuracy cannot be ensured due to a deviation in the direction.

  According to the method of Non-Patent Document 4, since a contour line of a solder ball as an object to be measured is measured as a black and white image pattern, in order to obtain the three-dimensional coordinates of each ball, a complicated algorithm (ball Because there is a problem that some of the contour lines overlap at the radius, etc.). On the other hand, when the phase shift method is adopted in order to increase accuracy, there are problems that processing time is required and costs are increased.

  The three-dimensional inspection apparatus of Non-Patent Document 5 is composed of a total of three optical systems, that is, an optical system directly above and two optical systems arranged symmetrically in the 45-degree direction. Forming optics. As described above, when the left and right optical systems are arranged in the 45-degree direction, when there is a change in the height direction, it can be detected as the largest displacement in the left and right image data as compared with the case where it is installed at another angle.

  However, in an optical system tilted at 45 degrees, unless a sufficient depth of focus is ensured, there will be a portion in the image data that is out of focus due to the difference in height. Such an out-of-focus portion cannot be measured. In this inspection apparatus, an illuminating device having a large aperture is disposed directly under the optical system directly above for illumination.

  Such an inspection apparatus is very effective when measuring the height of an object having a sharp edge. However, for an object to be measured having a three-dimensional curved surface, when the illumination light beam is specularly reflected on the surface of the object to be measured, in the image data obtained by the left and right optical systems, the same point on the object is one of the objects. A phenomenon occurs in which the optical system looks bright but the other optical system looks dark. This state will be described with reference to FIG.

  In the three-dimensional inspection apparatus shown in FIG. 14, a hemispherical measurement object 1 is irradiated with an illumination device 2 having a large area installed directly above and is imaged by left and right optical systems 3 and 4. In the figure, 2A and 2B are illumination rays. On the surface (curved surface) of the hemispherical object to be measured 1, a light ray 3 </ b> A that is specularly reflected and is connected to the left optical system 3 is indicated by a solid line, and a light ray 4 </ b> A that is connected to the right optical system 4 is indicated by a broken line.

  On the surface of the hemispherical object 1 to be measured, the angles of incident light and reflected light are equal to the normal of the object 1 to be measured. The brightened area on the captured image data is due to light rays incident on the optical systems 3 and 4 in parallel with the direction of the optical axis of the optical systems 3 and 4.

  As can be seen from FIG. 14, the region 5 (displayed with a solid line) that becomes bright on the left image data does not match the region 6 (displayed with a broken line) that becomes bright on the right image data. As described above, in the inspection apparatus shown in FIG. 14, the left optical system 3 and the right optical system 4 have different appearances with respect to the measurement object 1, that is, the same point on the measurement object 1 does not have the same brightness ( There is a problem that it is difficult to measure accurately (because bright areas on the image data are different from each other). In particular, in a three-dimensional curved surface, it is difficult to associate the same point on the left and right image data.

  In summary, as in Non-Patent Document 5, when a stereo optical system is applied, the left and right optical systems have different imaging points and imaging directions, so that the object to be measured has a three-dimensional curved surface. However, there is a problem that the same point cannot be identified.

  The present invention has been made in consideration of such a situation, and can solve the above-mentioned problems, and can measure precise three-dimensional coordinates of an object to be measured having a large number of protrusions on the surface, such as BGA. It is an object of the present invention to provide a three-dimensional coordinate measuring apparatus and method suitable for the purpose.

  In order to achieve the above object, the present invention provides an imaging means for a telecentric optical system having a sample stage on which an object to be measured is placed and a coaxial epi-illumination, wherein the optical axis is relative to the vertical axis of the sample stage. Is a first imaging means provided so as to form a predetermined inclination angle, and an imaging means of a telecentric optical system provided with coaxial epi-illumination, wherein the optical axis is the first in the previous period with respect to the vertical axis of the sample stage Second imaging means provided so as to be symmetrical with the optical axis of the imaging means, and calculation means for obtaining three-dimensional coordinates of the surface of the object to be measured from the images taken by the first and second imaging means, Provided is a three-dimensional coordinate measuring apparatus characterized by comprising the above.

  For this purpose, the present invention is an imaging means of a telecentric optical system provided with coaxial epi-illumination, and the optical axis makes a predetermined inclination angle with respect to the vertical axis of the sample stage on which the object to be measured is placed. The first imaging means provided as described above and the imaging means of a telecentric optical system provided with coaxial epi-illumination, the optical axis being in line with the optical axis of the first imaging means in the previous period with respect to the vertical axis of the sample stage Using the second imaging means provided so as to be symmetric, imaging the surface of the object to be measured illuminated by the first imaging means with the second imaging means, and the first imaging means; Using the second imaging means, the step of imaging the surface of the object to be measured illuminated by the second imaging means with the first imaging means, and the images captured by the first and second imaging means Obtaining three-dimensional coordinates of the surface of the object to be measured. Providing three dimensional coordinate measuring method comprising and.

  According to the present invention, an imaging unit of a telecentric optical system having a coaxial epi-illumination, wherein first and second imaging units provided symmetrically on the left and right sides are used, and the measurement target illuminated by the first imaging unit The object surface is imaged by the second imaging means, and the surface of the object to be measured illuminated by the second imaging means is imaged by the first imaging means, and the three-dimensional coordinates of the object surface to be measured are obtained from these images. .

  That is, the present invention applies stereo optics as in Non-Patent Document 5, but the same image can be obtained with the left and right optical systems even if the object to be measured has a three-dimensional curved surface. This is a three-dimensional coordinate measuring apparatus and method using the reciprocity of light (the principle that incident light and reflected light can be reversed) by a telecentric optical system provided with coaxial epi-illumination arranged symmetrically.

  Therefore, according to the present invention, the above-mentioned problems are solved, and the three-dimensional coordinates suitable for measuring the precise three-dimensional coordinates of the object to be measured having a large number of protrusions on the surface, such as BGA. A measuring device and method are obtained.

  The telecentric optical system refers to a telescope optical system in which a diaphragm is placed at one of the focal points of the objective lens.

  In the present invention, the distance between the mark provided on the surface of the sample stage for detecting the distance to the object to be measured and the center of gravity of the region where the object to be measured is detected with high luminance is the first imaging means. And the second imaging means, and the coordinates in the height direction of the object to be measured are preferably calculated from the difference between the detected distances.

  In this way, the distance between the mark on the sample table and the center of gravity of the detection area of the object to be measured is detected by the first and second imaging means, respectively, and the coordinate in the height direction of the object to be measured is determined from the difference between the distances. Is calculated, suitable three-dimensional coordinate measurement can be performed.

  In the present invention, the region where the object to be measured is detected with high luminance by the first imaging unit and the region where the object to be measured is detected with high luminance by the second imaging unit are respectively recognized by pattern matching. And it is preferable to calculate the coordinate of the to-be-measured object in the height direction from the recognized distance difference between the regions. In this way, highly accurate three-dimensional coordinate measurement can be performed by recognizing the areas detected by the first and second imaging means using the pattern matching technique.

  In the present invention, an image captured by the first imaging unit and an image captured by the second imaging unit using the imaging position parameters of the first imaging unit and the second imaging unit. Thus, it is preferable to take correspondence on the image of the object to be measured and thereby obtain the three-dimensional coordinates of the object to be measured. It is preferable to calibrate the imaging position parameters of the first imaging means and the second imaging means. As described above, if the imaging position parameters of the first imaging unit and the second imaging unit are calibrated, the image captured by the first imaging unit and the image captured by the second imaging unit are used. By taking correspondence on the image of the object to be measured, highly accurate three-dimensional coordinate measurement of the object to be measured can be performed.

Further, as the imaging position parameters of the imaging means, the imaging position of the camera _(X C, _{Y C,} and _Z C), rotation around the direction (X-axis of the camera imaging angle _R X, the rotational angle _R Y around the Y axis, And the rotation angle R _Z around the Z axis), the principal point position (difference X _O and Y _O between the center point of the image data and the camera optical axis), and the focal length f of the camera.

  According to the present invention, the conventional problems are solved, and the three-dimensional coordinate measurement suitable for measuring the precise three-dimensional coordinates of an object to be measured having a large number of protrusions on the surface, such as BGA. An apparatus and method are obtained.

  Hereinafter, a preferred embodiment (first embodiment) of a three-dimensional coordinate measuring apparatus and method according to the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a configuration diagram of a three-dimensional coordinate measuring apparatus 10 according to the present invention. In this apparatus, a sample stage 12 on which an object to be measured (BGA in the figure) W is placed is provided. A left optical system 14 as a first imaging unit is provided so that the optical axis 14A forms a predetermined inclination angle θ with respect to the vertical axis CL of the sample stage 12. The left optical system 14 is provided with a telecentric optical system barrel 14B, and a half mirror 14C is disposed in the barrel 14B at an angle of 45 degrees with respect to the optical axis 14A.

A coaxial epi-illumination system is configured by irradiating illumination light from an external light source in the direction of the arrow onto the illumination barrel 14D provided on the side surface of the barrel 14B. An imaging camera 14E is connected to the base end of the lens barrel 14B, and an output cable 14F of the imaging camera 14E is connected to the image processing device 18. The image processing device 18 corresponds to a calculation unit that obtains three-dimensional coordinates of the surface of the workpiece W from images captured by the left optical system 14 and a right optical system 16 described later.

  Similarly, a right optical system 16 that is a second imaging unit is provided so that the optical axis 16 </ b> A forms a predetermined inclination angle θ with respect to the vertical axis CL of the sample stage 12. That is, in the right optical system 16, the optical axis 16 </ b> A is provided so as to be symmetrical with the optical axis 14 </ b> A of the left optical system 14 with respect to the vertical axis CL of the sample stage 12. The right optical system 16 is provided with a telecentric optical system barrel 16B, and a half mirror 16C is disposed at an angle of 45 degrees with respect to the optical axis 16A inside the barrel 16B.

  A coaxial epi-illumination system is configured by illuminating illumination light from an external light source in the direction of the arrow onto an illumination barrel 16D provided on the side surface of the barrel 16B. An imaging camera 16E is connected to the base end of the lens barrel 16B, and an output cable 16F of the imaging camera 16E is connected to the image processing device 18.

  Although there is no restriction | limiting in particular in inclination-angle (theta) with respect to the vertical axis CL of the left optical system 14 and the right optical system 16, 10-20 degree is preferable and 12.5-17.5 degree is more preferable.

  The sample stage 12 is movable in the X, Y, and Z directions in FIG. 1, and the movement amounts in the X, Y, and Z directions are input to the image processing apparatus 18 by a detection unit (for example, a linear scale) not shown. It has come to be.

  There are no particular restrictions on the type, size, etc. of the object W to be measured on the sample stage 12, but for example, it can be BGA (having a ball diameter of 0.25 mm and a device size of 10 mm square) as shown in the figure. An identification mark (see FIG. 4) is provided in the vicinity of the placement position of the workpiece W on the sample stage 12. This will be described later.

  The lens barrels 14B and 16B are not particularly limited in magnification, aperture size, etc. as long as they are telecentric optical lens barrels. For example, the lens barrels 14B and 16B are 0.1 to 10 times in diameter, and the diameter of the measurement region is 40 to 50 mm. Can be.

  There are no particular restrictions on the types, magnifications, and the like of the imaging cameras 14E and 16E, but, for example, a 1/2 inch CCD (12.7 mm) industrial camera can be used.

  Next, the principle of measurement using the three-dimensional coordinate measuring apparatus 10 of FIG. 1 will be described.

  First, the reciprocity of light used in the three-dimensional coordinate measuring apparatus 10 of the present invention (the principle that incident light and reflected light can be reversed) will be described. FIG. 2A is a schematic diagram in the case where the image data is acquired by imaging the ball (sphere) B placed on the plane by the imaging means 15 of the coaxial incident illumination optical system. The lens barrel 15B is a lens barrel of a telecentric optical system, and an imaging camera 15E is connected to this base end portion. Illustration of the illumination optical system is omitted.

  In FIG. 2A, on the surface of the sample table 12 (which is planar) and the top portion of the ball B, the incident light RI and the reflected light RO are opposite in direction but are on the same line due to specular reflection. In the top part of the ball B, only one point strictly meets this condition. However, when the telecentric optical system is used, the top part of the ball B is shown in the captured image shown in FIG. A predetermined size portion (outlined portion in the figure) becomes bright due to specular reflection. Further, it is also specularly reflected from the plane of the sample table 12 on which the ball B is placed. In FIG. 2A, nothing is depicted in the image processing system, but it is assumed that the image processing system is monitored by an optical camera.

  As described above, in the telecentric optical system, an image of a range slightly tilted with respect to the optical axis is also formed, so that the predetermined range of the top portion of the ball B becomes bright. This is a slightly looser condition than the strict specular condition. The periphery of this bright area becomes dark because the mirror surface condition is not satisfied up to the outer periphery of the ball B. Further, the plane of the sample table 12 on which the ball B is placed becomes brighter because the mirror surface condition is satisfied again.

  The size (diameter) of the brightly visible region near the top of the ball B depends on the size of the ball B, the surface state of the ball B, the characteristics of the telecentric optical system, the characteristics of the illumination device, and the like.

  In FIG. 2A, the number of the imaging means 15 of the telecentric optical system provided with the coaxial incident illumination is one, but two may be used. When there are two imaging means of the telecentric optical system, they are arranged so as to be symmetrical as shown in FIG. 3 (FIG. 1).

  When arranged symmetrically as shown in FIG. 3, image data is acquired by the right imaging optical system 16 using the coaxial epi-illumination of the left optical system 14, and left using the right optical system 16 coaxial epi-illumination. The image pickup optical system 14 can acquire image data.

  Since the light has reciprocity (that is, the incident light and the reflected light can be reversed), in the two telecentric optical systems shown in FIG. 3, the specular condition for the horizontal plane is almost the same as in FIG. I can say that.

  In FIG. 3, as in FIG. 2, the plane of the top portion of the ball B satisfies the specular condition and thus becomes a bright area, and the surrounding area does not satisfy the specular condition until the outer periphery of the ball. So it becomes a dark area.

  FIG. 4 shows image data when imaged by the optical system of FIG. Among these, (A) is image data acquired by the left optical system 14, and (B) is image data acquired by the right optical system 16.

  When the image data in FIG. 4 is compared with the image data in FIG. 2B, the position of the top portion of the ball B (the position of the white portion in the figure) is shifted in the image data in FIG. (Note that the black region on the outer periphery of the ball B is also misaligned between the left image and the right image, but is not used for measurement). The size (length) of this deviation can be easily detected by placing a mark on the plane of the sample stage 12 and measuring it as the length from the mark. Therefore, in FIG. 4, “+” marks are drawn on the upper left and lower right of the sample stage 12 as identification marks. The identification marks in FIG. 4 are emphasized for the sake of explanation. If they are fixed to the sample stage 12, the same measurement as shown in FIG. 4 can be performed.

  When the optical systems 14 and 16 are tilted in the horizontal axis (X-axis) direction of the image data as shown in FIG. 3 (FIG. 1), this deviation D occurs only in the X direction. As shown in FIG. 4, when a mark placed on a plane is used as a position reference, in the left image data and the right image data, the area of the top portion of the ball B is shifted in the X-axis direction by DL and DR ( Position difference).

  The deviation in the X direction of the top partial area of the ball B (difference between the left image data and the right image data) changes substantially in proportion to the height thereof. Therefore, if the deviation (positional difference) on the left and right image data due to the height is calibrated in advance, the height of the part to be measured can be measured. In the present invention, the height of the measurement site is detected from the size of the deviation on the left and right image data.

  As shown in FIG. 3 (FIG. 1), when the left optical system 14 and the right optical system 16 of the telecentric optical system with coaxial epi-illumination are arranged so as to be symmetric, the left optical system 14 uses the right illumination device. Thus, the object to be measured can be imaged, and the right optical system 16 can image the object to be measured using the left illumination device. Such optical systems 14 and 16 can satisfy the mirror condition with respect to the horizontal plane.

  Then, the periphery of the top of the object to be measured placed on the plane of the sample table 12 is a horizontal plane. That is, the top periphery of the object to be measured satisfies the mirror condition. Further, since the incident light and the reflected light can be reversed (reciprocity of light), the area around the top of the object to be measured is imaged under the same conditions with respect to the left and right optical systems. That is, the image data has the same shape. This is true even if the object to be measured is a curved surface such as a bump electrode C (see FIG. 9 or the like), which will be described later, and provides means for identifying the same point on the curved surface with the left and right image data. In addition, since the optical systems 14 and 16 are tilted in the X direction, when the height of the measurement object is changed, the measurement site changes in the X direction on the image data.

  In this way, the feature of the present invention is to use a telecentric optical system arranged symmetrically. With such an optical system, the area around the top of the object to be measured is imaged in the same shape on the left image data and on the right image data. Then, using this characteristic, the object to be measured is made to correspond on the left and right image data, and the height can be measured.

  Next, a measuring method using the three-dimensional coordinate measuring apparatus 10 of FIG. 1 will be described. The left optical system 14 and the right optical system 16 of the telecentric optical system provided with coaxial epi-illumination can illuminate the object W to be measured with parallel rays and form an image with parallel rays. Therefore, the left optical system 14 images the object W to be measured using the right illumination device, and the right optical system 16 images the object W to be measured using the left illumination device.

  As described above, when the BGA device is imaged by the three-dimensional coordinate measuring apparatus 10 of FIG. 1, the left and right images as shown in FIG. 4 are obtained from the balls B, B. As described above, when the height of the ball B changes, the amount of displacement of the bright area at the top of the ball B changes on the image data (the amount of deviation changes). In order to explain the relationship between the amount of displacement on the image data and the height difference, a description will be given with reference to FIG.

  When the object to be measured W is illuminated with light beams RI, RI,... That are substantially parallel by the illumination device provided in the left optical system 14 of FIG. 1, the light beams RO, RO,. The image is formed at 16 and becomes image data. FIG. 5 shows how the same point in the plane coordinates (X, Y coordinates) changes with the height (Z-axis direction). The displacement dx on the image data with respect to the height change h can be easily expressed by the following formula 1.

[Equation 1]
dx = (imaging magnification) × h × sin θ (Formula 1)
Further, the displacement on the image data imaged by the left optical system 14 is in the opposite direction.

  In order to calculate the displacement dx, a lattice plate shown in FIG. 6 is used. This lattice plate is a reference lattice plate in which a lattice is formed by etching so that the distance and thickness are constant on a glass plate, and this is used for adjusting the left and right optical systems 14 and 16. .

  The lattice plate shown in FIG. 6 has a lattice interval P of 1 mm and a lattice line thickness t of 0.05 mm. In addition, an extra mark (a so-called register mark) is added to identify the center lattice point. FIG. 6 shows image data captured in a state where the lattice plate is installed at a reference height (Z-axis direction) position.

  Specifically, the left and right optical systems 14 and 16 are adjusted such that the grid plate is placed horizontally on the reference height (Z-axis direction) position and the grid positions overlap on the left and right image data. In FIG. 6 described above, the distance between the grating plate and the object W to be measured is 110 mm, the crossing angle of the left and right optical systems 14 and 16 is 30 degrees (θ is 15 degrees), and the optical system has a magnification of 1.0 ( This is an example of imaging with the left optical system 14).

  When the reference grid plate shown in FIG. 6 is accurately moved in the height direction (Z-axis direction) and the displacement of the grid point (one point) on the image data is accurately measured, as shown in FIG. A graph showing the relationship between the height (vertical axis) and the difference (horizontal axis) on the left and right image data is obtained. That is, in FIG. 7, the height is in the Y-axis direction, and the difference on the measured image data (measured value on the right image data−measured value on the left image data) is in the X-axis direction.

  As described above, the simplified expression of the graph of FIG. Note that the difference in image data is twice dx.

  Before measuring the height of the ball B of the BGA device that is the workpiece W, such calibration is performed and calibration data is stored. If the orthogonality of the X, Y, and Z axes of the three-dimensional coordinate measuring apparatus 10 is accurate, this calibration may be performed at one point in the center of the image. However, for adjustment of the X, Y, and Z axes, Since there is an allowable value, it is practical to have calibration data for the central 25 points of the lattice plate in FIG. 6 so as to cover the measurement region (the range indicated by the white broken line in FIG. 6).

  Since many balls B, B... Are imaged on the image data of the BGA device, it is preferable to use calibration data at a position close to the measured ball B when determining the height of the ball B. (Interpolation calculation may be performed at an intermediate position between grid points).

  In a normal BGA device, the diameter of the ball B is 250 μm, and the distance (pitch) between the balls B and B is 500 μm. As described above, when the crossing angle is 30 degrees (θ is 15 degrees) and the optical system is 1 × (one pixel is 10 μm), the height of the ball B is 6.47 pixels on the image data. (In the left and right images, this is twice as much), while the distance between the balls B and B is 50 pixels.

  Therefore, if a bright area is found in one image data, a bright area is easily found in the other image data. For example, in the case of a 250 μm ball, it can be found at a position shifted by about 13 pixels in the X direction.

  That is, it is not difficult to correspond the left and right images to the same ball B. Further, since the bright area around the top of the ball B is the same arrangement as the arrangement of the balls B and has a substantially constant interval, it can be easily found even if the size of the area changes.

  In order to measure the three-dimensional coordinates of the ball B of the BGA device, the position of the ball B (the arrangement of the balls B, B... At the same time) is taught in advance on the image data. The position of the ball B on the image data varies to the same extent as the repeated positioning accuracy of the BGA device loaded on the sample stage 12, but this size is the same for the left and right image data.

  Next, the flow of measuring the three-dimensional coordinates of the ball B will be described. FIG. 8 is a flowchart for explaining the process of measuring the three-dimensional coordinates of the ball B.

  First, a bright area at the top of the ball B is extracted from the left and right image data (step S-1). The bright area has the same arrangement as the arrangement of the taught balls B. Then, the left and right image data are measured as the same array with a substantially constant shift.

  Next, the barycentric coordinates of the bright area at the top of the ball B are obtained (step S-2). In order to obtain the three-dimensional coordinates of the top of the ball B, it is necessary to extract one point as a coordinate from a bright area. This one point is calculated as the shape centroid or luminance centroid. The shape centroid is binarized in a range slightly wider than the bright area, the bright area is determined, and the centroid coordinates are obtained. The luminance center of gravity is a little wider than the bright area, and the coordinates are obtained by weighted averaging with the brightness as a weight.

  Next, the height of the ball B is obtained from the X-coordinate difference of the barycentric coordinates obtained from the left and right image data (step S-3). At this time, as described above, since there is calibration data for 25 points, data closest to the ball B is used.

  Next, the X coordinate difference corresponding to the height (half the size obtained in the graph of FIG. 7) is corrected for the image data (step S-4).

  As in step S-4, the height difference from the coordinates on the image coordinates and the reference height is obtained, and the three-dimensional coordinates are obtained from this (step S-5). At this time, the coordinates on the image data and the three-dimensional coordinates can be converted by measuring the reference lattice plate at the reference height.

  If the above steps (step S-1 to step S-5) are repeated for the number of balls B arranged, all the three-dimensional coordinates of the registered balls B can be obtained.

  The shape and size of the bright image area at the top of the ball B depends on the state of the ball B. However, even if the shape and size of the bright image area at the top of the ball B change greatly, according to the present invention, the left and right image data have substantially the same shape due to the reciprocity of light (the second described later). 9 and 10 of the embodiment). If the shape of the ball B is the same, the center of gravity position in the left and right images is also calculated equally, so that the height accuracy is ensured. The present invention has been made in consideration of this point.

  If it is possible to identify a deviation of 0.2 pixel on the image data, the deviation of the optical system on one side is 0.1 pixel. When calculated by substituting into the above-described equation 1, the deviation of this pixel is 2.5 μm as the height difference. As described above, when the 1 × telecentric optical system is arranged at an intersecting angle of 30 degrees, the accuracy in the height direction is 2 to 3 μm. If a difference smaller than 0.2 pixels can be identified, the measurement accuracy is further improved.

  Next, a second embodiment of the present invention will be described. In this embodiment, the imaging position parameters of the left and right optical systems 14 and 16 are three-dimensionally calibrated with respect to the first embodiment described above shown in FIG. The two points of finding corresponding points by pattern matching are improved. Therefore, the configuration of the three-dimensional coordinate measuring apparatus 10 is substantially the same as that in FIG.

  The telecentric optical system employed in the present invention is designed to illuminate and image with light rays parallel to the optical axis. However, since it is an imaging system using a lens, it is the same as when imaged with a normal camera. In addition, the imaging position parameter can be calibrated three-dimensionally.

  Although the imaging position parameters of the imaging optical system (imaging camera) will be described later, it consists of nine parameters. When the imaging position parameters of the left and right optical systems 14 and 16 are obtained, the coordinates on the left and right image data and the imaging position parameters can be obtained by measuring on the left and right image data points on the object to be measured whose three-dimensional coordinates are to be measured Can be used to easily calculate the three-dimensional coordinates of the point. This point is known as a measurement method using a stereo optical system.

  In the present embodiment, the three-dimensional coordinates of the object to be measured are measured by measuring the position on the left image data and the position on the right image data as a stereo optical system. That is, in the first embodiment described above, the corresponding points on the left and right image data are obtained from the barycentric coordinates, but in this embodiment, the corresponding positions of the left image and the right image are extracted by pattern matching.

  FIG. 9 is an enlarged view when a bump electrode is employed as the object to be measured W, (B) is a front view (sectional view) of the bump electrode C, and (A) is provided with coaxial epi-illumination. This is image data obtained by imaging the bump electrode C from directly above with a telecentric optical system.

  In a telecentric optical system, an image is picked up with light rays parallel to the optical axis, but light rays in a range that makes a small angle with respect to the optical axis contribute to imaging. The condition of specular reflection on the surface of the bump electrode C in FIG. 9 is only the horizontal plane (XY plane), but in the telecentric optical system, the reflected light returns from a plane inclined at a slight angle from the horizontal plane. Come.

  In the vicinity of the plateau-like top of the bump electrode C as shown in FIG. 9, the condition of a plane within a certain inclination is paraphrased as the condition of a range from the top to a substantially constant height. FIG. 9 shows a state in which the illumination light is reflected from the top to the dh lower region from such an approximate view. This dh is a small amount (described later) that can be ignored in practice.

  FIG. 10 shows image data obtained by imaging the bump electrode C in FIG. 9 with a telecentric optical system (left and right optical systems 14 and 16 in FIG. 1) arranged symmetrically. In FIG. 10, the left image data is image data on the imaging surface of the imaging camera 14E, and the right image data is image data on the imaging surface of the imaging camera 16E. This corresponds to FIG. 4 of the first embodiment.

  In FIG. 10, incident light and reflected light that are specularly reflected on the surface of the bump electrode C are shown. The light beam of the coaxial epi-illumination provided in the left optical system 14 (see FIG. 1) forms a right image data by forming an image, and the light beam of the coaxial epi-illumination provided in the right optical system 16 (see FIG. 1) The left image data. Due to the principle of light reciprocity, incident light and reflected light can be interchanged, so the shape of the area around the top of the bump electrode C and the shape of the bottom surface of the bump electrode C are the same in the left and right image data. .

  In addition, the shape of the region around the top of the bump electrode C and the shape of the bottom surface of the bump electrode C are different in height, and therefore, a deviation occurs between the left image data and the right image data. In FIG. 10, the imaging surfaces of the imaging cameras 14E and 16E and the image data to be captured are drawn so as to be orthogonal to the illumination beam and the imaging beam. Since there is an optical system upstream of the imaging surfaces of the imaging cameras 14E and 16E, the actual image data is switched up and down and left and right (in FIG. 10, the imaging optical system is omitted so as not to complicate the figure). ). That is, when FIG. 4 and FIG. 10 of 1st Embodiment are compared, the area | region of the top part of an electrode has shifted | deviated to the reverse direction. Since an actual image is interchanged with FIG.

  In the technical field of image processing, if the shapes are the same, the best overlapping position can be easily obtained by pattern matching. In order to execute pattern matching, it is necessary to register an image model for pattern matching. In a stereo image system having left and right optical systems, if a position (coordinates) on the left image data and a position (coordinates) on the right image data for a point on the object to be measured are given, the three-dimensional coordinates of that point Can be calculated.

  In FIG. 10, the top peripheral area (bright area on the image data) of the bump electrode C is registered as a model image on the left image data, and the right image data is searched using the registered model image (most in pattern matching). The corresponding points of the left image and the right image are obtained by obtaining the overlapping position. The black + mark on the right image data is the model origin, and is the point considered as the top at the center of gravity of the bright area. A black + mark on the right image data indicates the searched position. If such a correspondence is taken, the three-dimensional coordinates of the point regarded as the top can be calculated from the image coordinates of the left and right image data.

  When the substrate of the device is specularly reflected, the shape of the bottom surface of the electrode is a region surrounded by a black boundary as shown in FIGS. 9 and 10 described above. If the shape of the bottom surface of the electrode is registered as a model image, the position of the bottom surface of the electrode (for example, the position of the center of gravity) can be easily obtained by pattern matching. If the position of the bottom surface of the electrode is obtained, the top peripheral region of the bump electrode C can be easily obtained, and the image model can be registered by setting the top peripheral region of the bump electrode C.

  Further, when the device substrate is not specularly reflected, there is no black boundary as shown in FIGS. 9 and 10 described above, and an image of only the bright area of the top peripheral area of the bump electrode C is obtained. In such a case, a region that appears bright on the image data is extracted (it is necessary to confirm that the size of the region is equal to or smaller than the size of the bump electrode C), and the extracted region is the arrangement of the bump electrodes C. Check the coordinates on the plane.

  Thereafter, the model image is registered so as to surround the bright area of the left image (the same operation as in FIG. 10). If the right image data is searched using the registered image model, the coordinates of the corresponding points on the left image data and the right image data can be obtained. As a result, three-dimensional coordinates can be calculated as described above.

  Next, the flow of measuring the three-dimensional coordinates of the bump electrode C will be described. FIG. 11 is a flowchart for explaining the process of measuring the three-dimensional coordinates of the bump electrode C.

  First, a bright area at the top of the bump electrode C is extracted on the left image (step S-11). Next, it is checked whether or not the extracted area has the same arrangement as that of the bump electrode C (step S-12).

  Next, the area around the top of the bump electrode C is registered as an image model (step S-13). Then, the right image is searched using the registered image model (step S-14).

  Next, three-dimensional coordinates are calculated from the origin of the image model (coordinates on the left image) and the searched position (coordinates on the right image) (step S-15).

  If the above steps (step S-13 to step S-15) are repeated for the number of the bump electrodes C arranged, all the three-dimensional coordinates of the registered bump electrodes C are obtained (step S-16).

  As described above, according to the present invention, the height measurement is performed using the principle of the reciprocity of light and paying attention to the characteristics of the light beam reflected by the plane near the vertex of the object to be measured. Thus, it is suitable for measuring the three-dimensional coordinates of the object to be measured formed by a three-dimensional curved surface. And since it can measure simultaneously in the to-be-measured object in a visual field, processing can be sped up.

  As described above, in the telecentric optical system, a stop is placed at the focal position of the objective lens (object-side telecentric optical system). Since there is this diaphragm, on the object side from the objective lens, light rays other than those parallel to the optical axis are shielded by the diaphragm, so that they do not contribute to image formation. The imaging position of the optical system is the same regardless of the stop.

  Since this telecentric optical system is also an imaging system using a lens, it can be calibrated by the same method as the camera positioning. A typical camera positioning method is described below (see, for example, “Three-dimensional measurement with photographs (1983 Kyoritsu Shuppan)”). However, the distance to the object to be measured is large in the camera, but the distance to the object to be measured is not so large in the telecentric optical system. For this reason, the focal length of the camera as the camera positioning parameter is the imaging distance (longer than the focal length) based on the design of the optical system.

The imaging parameter of a camera, the imaging position of the camera _(X C, _{Y C,} and _Z C), the rotational angle _R X around direction of the camera imaging (X-axis, rotation about the Y axis angle _{R Y,} and around the Z axis 9 of the rotation angle R _Z ), the principal point position (difference X _O and Y _O between the center point of the image data and the camera optical axis), and the focal length f of the camera.

  These nine parameters can be obtained from published documents (for example, as described above) by accurately measuring coordinates on image data obtained by imaging from a point where a three-dimensional position is known in a reference coordinate system. "3D measurement with photographs (1983 Kyoritsu Shuppan)"). At this time, it is only necessary to measure five or more points whose positions on the image data (coordinates on the image data) and three-dimensional coordinates are known. In this case, it is necessary to arrange three-dimensionally.

  FIG. 12 is a diagram illustrating the imaging position parameter of the camera. In FIG. 12, the plane that determines the coordinate system is used as a reference plane, a second reference plane and a third reference plane are provided in parallel with the plane, and nine points with known coordinates are arranged on these three planes. (The reference point shall not be on a straight line (but different from the figure) at each reference plane).

  These nine points are imaged on the image data as indicated by straight lines in FIG. Since the telecentric optical system captures an image with light rays parallel to the optical axis, the light rays indicated by the straight lines in the figure are shielded by a diaphragm. However, since the figure position does not depend on the diaphragm, this straight line is easier to obtain the figure position. If these nine image coordinates are accurately measured, camera imaging parameters can be obtained.

  Next, a region where the ball B appears bright in the telecentric optical system will be described. FIG. 13 shows a state where a vertical line passing through the center O of the ball B is BL, and a straight line BH that intersects the vertical line BL at an angle φ at the center O of the ball B is rotated around the vertical line BL. The locus of the straight line BH on the spherical surface of the ball B is a circle as shown in the figure. The height from the center O to the circle (the distance in the height direction) is r × cos φ where the radius of the ball B is r. The height difference dh between the apex T of the ball B and its circle can be expressed by the following equation 2.

[Equation 2]
dh = r × (1-cos φ) = r × φ × φ / 2 (Formula 2)
When cos φ is expanded with φ being small, the right equation of Equation 2 is obtained. If φ is a minute amount, it becomes a minute amount of the second order, and is negligible. FIG. 13 approximates the situation of the telecentric optical system. φ is related to the design element of the telecentric optical system.

  As described above, according to the present invention, by arranging the telecentric optical system symmetrically, the region reflected on the plane near the vertex of the object to be measured is focused on the left image data by paying attention to the reciprocity of the light. And the right image data can be used to measure the height of the object to be measured.

  Therefore, the present invention is suitable for measuring the three-dimensional coordinates of an object formed by a three-dimensional curved surface, such as a BGA or a bump electrode having a shape as shown in FIG. In the case of a BGA or the like having a large number of protrusions within the range imaged as image data (that is, the visual field range of the optical system), there is an advantage that these can be measured simultaneously and at high speed.

  For example, in a 0.5 to 1 × optical system, the visual field is 6 × 5 mm to 12 × 10 mm, and a BGA device having 100 to 400 solder balls can be simultaneously measured with an accuracy of 3 to 5 μm. In addition, there is a measuring range with a height of about ± 0.5 to ± 1 mm. Therefore, there are significant advantages over conventional devices in that the accuracy is high, the cost is low, the speed is high, and the height measurement range is wide.

  As mentioned above, although embodiment of the three-dimensional coordinate measuring apparatus and method which concerns on this invention was described, this invention is not limited to the said embodiment, Various aspects can be taken.

  For example, the BGA has been mainly described as an object to be measured, but it can be similarly performed even when measuring three-dimensional coordinates of other objects.

Configuration diagram of a three-dimensional coordinate measuring apparatus according to the present invention Schematic diagram when acquiring image data by imaging a ball with the imaging means of the coaxial incident illumination optical system Schematic diagram when acquiring image data by imaging the ball by arranging the imaging means of the coaxial incident illumination optical system symmetrically The figure which shows the image data at the time of imaging with the optical system of FIG. Schematic explaining a state in which a BGA device is imaged by the three-dimensional coordinate measuring apparatus of FIG. Plan view of lattice plate Graph showing the relationship between the height of the grid plate and the difference in the left and right image data Flow diagram explaining the process of measuring the three-dimensional coordinates of a ball Image data captured when bump electrodes are used as the object to be measured Image data captured by a telecentric optical system with bump electrodes arranged symmetrically Flow chart explaining the process of measuring the three-dimensional coordinates of the bump electrode The figure explaining the imaging position parameter of a camera Diagram explaining the area where the ball looks bright in the telecentric optical system Configuration diagram of a conventional three-dimensional inspection apparatus

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Three-dimensional coordinate measuring device, 12 ... Sample stand, 14 ... Left optical system, 16 ... Right optical system, 14B, 16B ... Lens barrel, 14C, 16C ... Half mirror, 14D, 16D ... Illumination lens barrel, 14E, 16E ... Imaging camera, 18 ... Image processing device

Claims (6)

A sample stage on which an object to be measured having a mirror-like surface is placed;
An imaging means of a telecentric optical system provided with a coaxial epi-illumination for illuminating the object to be measured with parallel rays, the first being provided such that the optical axis forms a predetermined tilt angle with respect to the vertical axis of the sample stage. Imaging means,
Wherein a telecentric optical system imaging unit having a coaxial incident illumination for illuminating the object to be measured in parallel rays, the optical axis and the line symmetry of the optical axis with respect to the sample stage of the vertical axis is the first image pickup means A second imaging means provided to be,
A first image obtained by imaging the surface of the object to be measured illuminated with parallel rays by the first imaging means with the second imaging means, and an object to be measured illuminated with parallel rays by the second imaging means Computation means for obtaining a three-dimensional coordinate of the surface of the object to be measured from a second image obtained by imaging the surface with the first imaging means , wherein each of the objects to be measured is obtained from the first and second images, respectively. A high brightness area is extracted, and the height of the surface of the object to be measured is determined based on a difference between a centroid position of the high brightness area in the first image and a centroid position of the high brightness area in the second image. Calculating means for calculating the coordinates of the direction;
A three-dimensional coordinate measuring apparatus comprising: A sample stage on which an object to be measured having a mirror-like surface is placed;
An imaging means of a telecentric optical system provided with a coaxial epi-illumination for illuminating the object to be measured with parallel rays, the first being provided such that the optical axis forms a predetermined tilt angle with respect to the vertical axis of the sample stage. Imaging means,
An imaging means of a telecentric optical system having a coaxial epi-illumination for illuminating the object to be measured with parallel rays, the optical axis being symmetrical with respect to the optical axis of the first imaging means with respect to the vertical axis of the sample stage A second imaging means provided to be,
A first image obtained by imaging the surface of the object to be measured illuminated with parallel rays by the first imaging means with the second imaging means, and an object to be measured illuminated with parallel rays by the second imaging means A calculation means for obtaining a three-dimensional coordinate of the surface of the object to be measured from a second image obtained by imaging the surface with the first image pickup means, wherein the high-luminance region of the object to be measured on the first image And corresponding points in the high-intensity region on the first image and the high-intensity region on the second image are detected by pattern matching between the high-intensity region of the object to be measured on the second image Calculating means for calculating the coordinate in the height direction of the surface of the object to be measured from the difference between the position on the first image and the position on the second image of the points corresponding to each other;
A three-dimensional coordinate measuring apparatus comprising: An imaging means of a telecentric optical system having a coaxial epi-illumination for illuminating an object to be measured having a mirror-like surface with parallel rays , wherein the optical axis is relative to a vertical axis of a sample stage on which the object to be measured is placed Is a telecentric optical system imaging unit including a first imaging unit provided so as to form a predetermined inclination angle and a coaxial epi-illumination for illuminating the object to be measured with a parallel light beam, the vertical axis of the sample table using the second imaging means in which the optical axis is provided such that the optical axis line symmetry of the first image pickup means with respect to said first object surface that illuminates the parallel light by the imaging means Capturing the first image with the second imaging means; and
Using the first imaging unit and the second imaging unit, the second imaging unit captures a surface of the object to be measured, which is illuminated with parallel rays by the second imaging unit, and acquires a second image. And steps to
Extracting a high-intensity region of the object to be measured from each of the first and second images;
Calculating a coordinate in a height direction of the surface of the object to be measured based on a difference between a centroid position of the high-luminance region in the first image and a centroid position of the high-luminance region in the second image. When,
A three-dimensional coordinate measurement method comprising: 4. The three-dimensional coordinate measuring method according to claim 3, wherein the coordinate in the height direction is calculated with reference to a mark provided on the surface of the sample table. An imaging means of a telecentric optical system having a coaxial epi-illumination for illuminating an object to be measured having a mirror-like surface with parallel rays, wherein the optical axis is relative to a vertical axis of a sample stage on which the object to be measured is placed Is a telecentric optical system imaging unit including a first imaging unit provided so as to form a predetermined inclination angle and a coaxial epi-illumination for illuminating the object to be measured with a parallel light beam, the vertical axis of the sample table The surface of the object to be measured is illuminated with a parallel light beam by the first imaging means using a second imaging means provided so that the optical axis is symmetrical with the optical axis of the first imaging means. Capturing the first image with the second imaging means; and
Using the first imaging unit and the second imaging unit, the second imaging unit captures a surface of the object to be measured, which is illuminated with parallel rays by the second imaging unit, and acquires a second image. And steps to
The pattern matching between the high-intensity region of the object to be measured on the first image and the high-intensity region of the object to be measured on the second image results in pattern matching between the high-intensity region on the first image and the first image. Corresponding points in the high-intensity area on the second image, and the difference between the position on the first image and the position on the second image of the points corresponding to each other is detected. Calculating coordinates in the height direction ;
A three-dimensional coordinate measurement method comprising: 6. The three-dimensional coordinate measurement method according to claim 5 , wherein the three-dimensional coordinates of the object to be measured are obtained using the imaging position parameters of the first imaging means and the second imaging means.

Images