JP2019015615A - Three-dimensional imaging device, three-dimensional imaging method, and three-dimensional imaging program - Google Patents

Abstract

To provide a technique that shortens the time needed for imaging.SOLUTION: A three-dimensional imaging device is constituted comprising: an imaging mechanism capable of imaging an object to be imaged at a plurality of imaging positions differing in the angle of rotation around a rotation axis by a radiant ray outputted in a direction inclined or orthogonal to the rotation axis, as well as capable of changing the position of the rotation axis in a direction perpendicular to the rotation axis; a regular hexagon arrangement unit for arranging a regular hexagon internally contacting one reconfigurable region that is reconfigurable by imaging around the rotation axis virtually in plurality in a row while sharing the sides in an adjoining state on the cross section in which the object to be imaged is virtually cut in a direction orthogonal to the rotation axis; an imaging unit for imaging the object to be imaged at the plurality of imaging positions, with the center of the regular hexagon serving as the axis of rotation; and a reconfiguration computation unit for executing reconfiguration computation on the basis of the imaged image per rotation axis.SELECTED DRAWING: Figure 5

Description

  The present invention relates to a three-dimensional imaging apparatus, a three-dimensional imaging method, and a three-dimensional imaging program.

  Conventionally, a technique such as three-dimensional CT that performs three-dimensional imaging by reconstructing a plurality of transmission images captured by radiation such as X-rays is known. In such three-dimensional imaging, in general, images are captured at a plurality of angles using a mechanism in which a radiation generator and a radiation detector rotate relatively around a rotation axis passing through an imaging target. In such three-dimensional imaging, generally, a region of a rotating body (a shape in which two cones are combined on the bottom surface) around the rotation axis is a reconfigurable region (viewing region). Therefore, a circular or polygonal area centered on the rotation axis is a reconfigurable area on the cut surface perpendicular to the rotation axis.

  In imaging with respect to one rotation axis, three-dimensional information outside the reconfigurable area cannot be obtained. Therefore, if information about the outside is necessary, imaging with respect to a plurality of rotation axes is necessary. In Patent Document 1, in order to obtain a tomographic image of a subject that is larger than the size of one reconfigurable area in the cut plane, the reconfigurable area is regarded as a circle, and squares inscribed in the circle are arranged on the cut plane. A configuration for obtaining a tomographic image of a large subject by reconstructing each square is disclosed.

JP 2006-3122 A

When inspecting an inspection object in a region wider than the reconfigurable region obtained by imaging with respect to one rotational axis, imaging with respect to a plurality of rotational axes is required. Therefore, the imaging time increases as the number of rotational axes increases. It will take.
The present invention solves the above-described problems, and aims to shorten the time required for imaging.

  In order to achieve the above object, the three-dimensional imaging apparatus can image an imaging target with radiation output in a direction inclined or orthogonal to the rotation axis at a plurality of imaging positions having different rotation angles around the rotation axis. In addition, an imaging mechanism capable of changing the position of the rotation axis in a direction perpendicular to the rotation axis, and a cut surface obtained by cutting a reconstruction operation based on an image captured for each rotation axis in a direction perpendicular to the rotation axis , A regular hexagonal arrangement part that virtually arranges a plurality of regular hexagons inscribed in a reconfigurable area by reconstruction operation in a state where they share adjacent sides and a plurality of imaging positions with the center of the regular hexagon as the rotation axis An imaging unit that images an imaging target and a reconstruction calculation unit that executes a reconstruction calculation based on the captured image for each rotation axis.

  That is, the reconfigurable area on the cut surface is a circular shape (or polygon) having rotational symmetry about the rotation axis. On the other hand, when comparing a regular hexagon inscribed in the same circle with a square, the regular hexagon can fill a plane more efficiently (with a smaller number) than a square. Therefore, when regular hexagons are arranged on the cut surface and imaging for three-dimensional image structure analysis is performed with the center of the regular hexagons as the rotation axis, in many cases, the number of imaging targets is smaller than when the squares are arranged. Imaging that covers a wide area can be performed. As a result, the time required for imaging can be shortened.

  Here, the imaging mechanism can image an imaging target with radiation output in a direction inclined or orthogonal to the rotation axis at a plurality of imaging positions having different rotation angles around the rotation axis, and the position of the rotation axis can be determined. What is necessary is just to be able to change to the direction perpendicular | vertical to a rotating shaft. That is, the imaging mechanism can perform imaging multiple times at different rotational positions around the rotation axis, and a three-dimensional image within a predetermined range around the rotation axis can be obtained by reconstruction calculation of the imaging. It only needs to be reconfigured. In addition, the imaging mechanism can select the position of the rotation axis, and by performing imaging at different imaging positions around the rotation axis for each selected rotation axis, it is possible to perform reconstruction calculations for a wide area If it is.

  The imaging mechanism includes a radiation generator and a radiation detector, and an imaging target can be arranged in the radiation irradiation range. In the imaging mechanism, the radiation generator and the radiation detector are configured to rotate with respect to the rotation axis, but these rotations only need to be relative. Accordingly, it is only necessary to perform imaging at a plurality of imaging positions having different rotation angles by being able to move and rotate in at least one of the radiation generator, the radiation detector, and the imaging target.

  The plurality of imaging positions may be set so that the three-dimensional reconstruction of the imaging target is performed by performing the reconstruction calculation based on the captured images at the plurality of imaging positions. Is preferably an imaging position obtained by performing a plurality of rotations at regular intervals. Note that in order to assume a regular hexagon inscribed in the reconfigurable region, the number of imaging is preferably 6 or more, and the higher the number of imaging, the more accurate imaging can be performed. The radiation may be any radiation that passes through the inspection target, and X-rays, gamma rays, and the like can be used.

  As a configuration for changing the position of the rotation axis in a direction perpendicular to the rotation axis, various configurations can be adopted. For example, a movable plane (XY plane) of an XY stage on which an imaging target is set. The structure etc. which orient | assign to the plane perpendicular | vertical to a rotating shaft are mentioned. Of course, since the movement is relative, at least one of the radiation generator and the radiation detector may be movable two-dimensionally.

  The regular hexagonal arrangement part is a regular hexagon inscribed in a reconfigurable region that can be reconstructed by imaging around one rotation axis on a cut surface obtained by virtually cutting the imaging target in a direction perpendicular to the rotation axis. It suffices if a plurality of lines can be virtually arranged in a state where the sides are shared and adjacent to each other. That is, the regular hexagonal arrangement unit assumes a reconfigurable area on the cut surface obtained by virtually cutting the imaging target in the direction perpendicular to the rotation axis, and arranges the regular hexagons on the cut surface.

  The cutting plane may be a plane that passes through the imaging target, and is preferably a plane that cuts the imaging target at a position where the three-dimensional structure should be analyzed. The reconfigurable area may be an area where an image to be inspected can be obtained from a three-dimensional reconstructed image formed by the reconstruction calculation. In the three-dimensional imaging apparatus, analysis such as inspection of an inspection target should be performed based on the reconstructed images in the regular hexagon arranged on the cut surface, and the reconfigurable area is defined to define the regular hexagon. Used.

  Therefore, it is not essential that the reconfigurable area can be reconstructed with high accuracy, and the shape thereof may be a virtual shape. For example, since the number of imaging positions around the rotation axis is a finite number, the reconfigurable area on the cut surface has a shape close to a polygon, but may be regarded as a circle. In addition, since it can be assumed that the image becomes unclear at the end of the reconfigurable area (the part near the circumference if it is circular) or affected by artifacts, the boundary of the reconfigurable area is the reconstructed image. Therefore, the reconfigurable area may be virtually determined based on the geometric arrangement of the imaging mechanism, or a predetermined fixed area may be assumed.

  A regular hexagon inscribed in the reconfigurable area may be defined by actually defining a reconfigurable area and calculating a regular hexagon inscribed in the area, or a regular hexagon inscribed in the reconfigurable area. It may be defined by specifying the size of. As an example of the latter, for example, the imaging mechanism includes a radiation detector in which one side of the rectangular effective imaging range is parallel to the cut surface, and one side of the regular hexagon is the length of one side of the effective imaging range parallel to the cut surface. The structure which is 1/2 of the length which remove | divided by the imaging magnification is mentioned.

  That is, if a reconstruction operation is performed on an image within the rectangular effective imaging range of the radiation detector and one side of the image is parallel to the cut surface, the length of the one side is divided by the imaging magnification. The range imaged on the surface can be estimated. Since the range imaged on the cut surface is a reconfigurable region, the value obtained by dividing the effective imaging range of the radiation detector by the imaging magnification is regarded as the total length (or diameter if circular) of the reconfigurable region. Can do. Therefore, if the value is halved, one side of a regular hexagon inscribed in the reconfigurable area can be defined.

  Once a regular hexagon is determined, it is possible to fill the plane efficiently by arranging the regular hexagons in a state where the sides are shared and adjacent to each other on the cut surface. Thus, it becomes possible to perform imaging of the imaging target.

  The imaging unit only needs to be able to image the imaging target at a plurality of imaging positions with the regular hexagonal center as the rotation axis. That is, the image capturing unit only needs to be able to capture an image that can be three-dimensionally reconstructed by controlling the image capturing mechanism and capturing images at a plurality of image capturing positions for each rotation axis. Moreover, the reconstruction calculation part should just be able to perform a reconstruction calculation based on the captured image for every rotating shaft. The result of the reconstruction calculation can be used for various purposes, and can be used for inspection of an inspection target, for example, quality determination based on the presence or absence of a void.

It is a block diagram of a three-dimensional imaging device. FIG. 2A is a diagram illustrating imaging by the imaging mechanism, and FIG. 2B is a diagram illustrating a reconfigurable area. FIG. 3A is a diagram showing regular hexagons and squares inscribed in the reconfigurable region, and FIG. 3B is a diagram showing the size of the reconfigurable region. It is a flowchart of a X-ray inspection process. FIG. 5A is a diagram showing a regular hexagonal arrangement, and FIG. 5B is a diagram showing a square arrangement. 6A and 6B are graphs comparing the arrangement efficiency of regular hexagons and squares. FIG. 7A is a diagram illustrating an arrangement example of the imaging target, and FIG. 7B is a diagram illustrating an example of a mask superimposed on the imaging target.

Here, embodiments of the present invention will be described in the following order.
(1) Configuration of 3D imaging device:
(2) X-ray inspection process:
(3) Other embodiments:

(1) Configuration of 3D imaging device:
FIG. 1 is a schematic block diagram of a three-dimensional imaging apparatus according to an embodiment of the present invention. The three-dimensional imaging apparatus includes an imaging mechanism 10 and a control unit 20. The imaging mechanism 10 includes an X-ray generator 11, an X-ray detector 12, and an XY stage 13. The imaging mechanism 10 can make the X-ray generator 11, the X-ray detector 12, and the XY stage 13 have a predetermined relative positional relationship. A substrate P or the like including an imaging target W (BGA solder bumps or the like on the substrate) is set on the XY stage 13 and can be moved in a two-dimensional plane (XY plane) direction. Therefore, the imaging mechanism 10 can irradiate the imaging target W with X-rays in a state where the X-ray generator 11, the X-ray detector 12, and the imaging target W are in a predetermined relative positional relationship.

  The X-ray generator 11 includes an X-ray output unit 11a that outputs X-rays, and can irradiate the imaging target W with X-rays with a predetermined intensity. The X-ray detector 12 includes an X-ray detection surface 12 a that detects the intensity of X-rays, and can capture an X-ray image that reflects the amount of X-ray transmitted through the imaging target W. That is, the X-ray detector 12 generates X-ray image data 26b indicating an image of the X-ray transmission amount at each position on the X-ray detection surface 12a. In the present embodiment, the X-ray detection surface 12a is rectangular, and one side thereof is parallel to the XY plane.

  In the present embodiment, the substrate P including the imaging target W is transported along a predetermined plane by a transport mechanism (not shown). That is, the substrate P including the uninspected imaging target W is loaded along a predetermined plane and set on the XY stage 13, and after the imaging and inspection by X-rays are performed, the substrate P is unloaded by the transport mechanism again. Is done. In the present embodiment, the imaging mechanism 10 includes an XY stage 13 and a rotation mechanism (not shown), and can change the relative positional relationship among the X-ray generator 11, the X-ray detector 12, and the imaging target W. . That is, the imaging mechanism 10 can move the imaging target W two-dimensionally along the XY plane within the X-ray irradiation range by the XY stage 13, and a rotation mechanism (not shown) And a moving mechanism for moving at least one position between the X-ray output unit 11a and the X-ray detection surface 12a, and the imaging targets W and X so as to acquire an X-ray image for performing the reconstruction calculation. The relative positional relationship between the line output unit 11a and the X-ray detection surface 12a can be adjusted.

  FIG. 2A is a schematic diagram illustrating a state in which X-rays are irradiated to the imaging target W. In FIG. 2A, the horizontal direction is a direction parallel to the XY plane, and the vertical direction is perpendicular to the XY plane. Z direction. 2A shows an example in which the imaging target W is a solder bump. FIG. 2A shows a chip C on a substrate P (not shown) including the imaging target W and an X-ray generator existing therearound. 11 schematically shows an X-ray output unit 11 a and an X-ray detection surface 12 a of the X-ray detector 12.

  In the present embodiment, the imaging mechanism 10 includes the imaging target W and the X-ray output so that the relative positional relationship between the imaging target W, the X-ray output unit 11a, and the X-ray detection surface 12a rotates with respect to the rotation axis A. At least one of the part 11a and the X-ray detection surface 12a is moved. That is, the imaging mechanism 10 is positioned so that the X-ray output unit 11a and the X-ray detection surface 12a rotate substantially in the rotation directions R and R ′ with respect to the rotation axis A, as shown in FIG. 2A. Can be changed. For example, both the X-ray output unit 11a and the X-ray detection surface 12a may be rotated, or the X-ray output unit 11a is fixed, and the X-ray detection surface 12a and the imaging target W are rotated in the output range. Alternatively, the X-ray output unit 11a and the X-ray detection surface 12a may be fixed, and the imaging target W may be rotated within the output range of the X-ray output unit 11a. When the imaging target W is rotated, the XY stage 13 has a rotation axis A parallel to the Z axis in a state where the X-ray output unit 11a and the X-ray detection surface 12a are fixed at a position indicated by a solid line in FIG. 2A. The imaging target W is configured to rotate.

  The X-ray output unit 11a can output X-rays within a predetermined solid angle range regardless of how the relative positional relationship between the respective parts varies, and the optical axis of the X-rays passing through the imaging target W is the rotation axis A. Is inclined by a predetermined angle. Further, in the imaging mechanism 10, imaging is performed at a plurality of imaging positions with different rotation angles around the rotation axis A with the X-ray output direction inclined to the rotation axis A. The solid line and the broken line shown in FIG. 2A schematically show the imaging positions at two places where the rotation angles differ by 180 °.

  Next, the control unit 20 will be described. The control unit 20 includes a generator control unit 21, a detector control unit 22, an imaging mechanism control unit 23, an input unit 24, an output unit 25, a memory 26, and a CPU 27. The memory 26 is a storage medium capable of storing data, and stores program data 26a and X-ray image data 26b. The CPU 27 reads out and executes the program data 26a, thereby executing calculations for various processes described later. The memory 26 only needs to be able to store data, and various storage media such as RAM, EEPROM, and HDD can be employed.

  The imaging mechanism control unit 23 adjusts at least one position among the imaging target W, the X-ray output unit 11a, and the X-ray detection surface 12a so as to be an imaging position where an X-ray image of the imaging target W is captured. The generator control unit 21 controls the X-ray generator 11 to irradiate the imaging target W with X-rays from the X-ray generator 11. The detector control unit 22 acquires X-ray image data 26b indicating an image of the intensity of X-rays detected by the X-ray detector 12, that is, a transmission amount. The X-ray image data 26b is image data composed of gradation values of a plurality of pixels, and the gradation value of each pixel indicates the intensity of the X-ray detected by the X-ray detector 12. The detector control unit 22 acquires the X-ray image data 26 b from the X-ray detector 12 and stores it in the memory 26. The output unit 25 is a display that displays an X-ray image of the imaging target W, inspection results, and the like, and the input unit 24 is an operation input device that accepts user input.

  The CPU 27 executes the functions of the regular hexagonal arrangement unit 27a, the imaging unit 27b, and the reconstruction calculation unit 27c based on the program data 26a in order to determine whether the imaging target W is acceptable. The imaging unit 27b is a program module that causes the CPU 27 to execute a function of imaging the imaging target W at a plurality of imaging positions around the rotation axis A. That is, the CPU 27 outputs predetermined instructions to the generator control unit 21, the detector control unit 22, and the imaging mechanism control unit 23 by the processing of the imaging unit 27b, and an X-ray image for executing reconstruction calculation. The CPU 27 is caused to execute processing for acquiring the data 26b.

  In the example shown in FIG. 2A, the CPU 27 changes the rotation angle around the rotation axis A (rotation angle when a certain imaging position is used as a reference) at regular intervals (for example, 30 ° for imaging 12 times). The imaging mechanism 10 is controlled so that imaging is performed at each rotation angle. In the present embodiment, the rotation axis A can be set at a different position in the XY plane as will be described in detail later.

  The reconstruction calculation unit 27c is a program module that causes the CPU 27 to execute a function of executing the reconstruction calculation based on the captured image for each rotation axis A. That is, in the present embodiment, the CPU 27 refers to the X-ray image data 26b, acquires each image captured about each rotation axis A, performs a reconstruction operation on each rotation axis A, and performs three-dimensional imaging of the imaging target W. Obtain an imaging device.

  The reconstruction calculation only needs to be able to reconstruct the three-dimensional structure of the imaging target W, and various processes can be employed. For example, a filter-corrected back projection method can be employed. In this process, the CPU 27 first performs a Fourier transform on any one of a plurality of X-ray images imaged about the same rotation axis A, and a filter correction function in the frequency space for the result obtained by the Fourier transform. Multiply Furthermore, an image subjected to filter correction is obtained by performing inverse Fourier transform on this result. As the filter correction function, a function for enhancing the edge of the image can be adopted.

  Subsequently, the image after the filter correction is back-projected into a three-dimensional space along a locus on which the image is projected. That is, the locus corresponding to the image at a certain position on the X-ray detection surface 12a of the X-ray detector 12 is a straight line connecting the focal point of the X-ray generator 11 and this position. To do. When the back projection described above is performed for all of a plurality of X-ray images imaged about the same rotation axis A, the X-ray absorption coefficient distribution in the portion where the imaging target W exists in the three-dimensional space is emphasized, and the imaging target Reconstruction information indicating the three-dimensional shape of W is obtained.

  In the reconstruction calculation as described above, a portion where the back projection results in each of a plurality of X-ray images captured with respect to the same rotation axis A overlap is a reconfigurable area by the reconstruction calculation. In FIG. 2A, the reconfigurable area Zc is shown by hatching. FIG. 2B is a diagram schematically showing a reconfigurable area Zc extracted from a perspective view. As shown in FIG. 2B, the reconfigurable region Zc has a shape in which cones having different heights and matching bottom surfaces are joined to face each other on the bottom surface. Note that the shape of the reconfigurable region Zc approaches a cone as the number of times of imaging around the rotation axis A increases, and in reality is close to a polygonal pyramid, but here, the reconfigurable region Zc is assumed to be a cone.

  By imaging a plurality of times with respect to one rotation axis A, a three-dimensional captured image can be obtained for the range of the reconfigurable region Zc as shown in FIG. 2B. Accordingly, assuming a cut surface cut in a direction perpendicular to the rotation axis A (Z axis), the reconfigurable region Zc on the cut surface can be considered as a circle (or a polygon). In FIG. 2B, the reconfigurable region Zc on the cut surface when the cutting is performed at the position Z1 in the Z-axis direction where the diameter is the largest in the reconfigurable region Zc is shown as the reconfigurable region Zc1. . As is apparent from FIG. 2B, the size of the reconfigurable region Zc on the cut surface can vary according to the change in the position Z1 (cutting position) in the Z-axis direction.

  The CPU 27 executes the reconstruction calculation based on X-ray images captured at a plurality of imaging positions around the rotation axis A in the imaging mechanism 10, but when the imaging target W is distributed over a wide range, one In some cases, imaging around the rotation axis A cannot cover all imaging targets W. For example, when the tomographic image of the imaging target W at the position Z1 shown in FIG. 2B is analyzed, if the imaging target W is distributed over a wider range than the reconfigurable region Zc1 shown in FIG. 2B, one rotation is performed. Imaging around the axis A cannot cover all imaging targets W.

  In this case, the CPU 27 covers all the imaging targets W by setting the rotation axes A at a plurality of positions and performing imaging a plurality of times with respect to the plurality of rotation axes A. Therefore, if a small number of rotation axes are set for the same imaging object W, imaging of the imaging object W can be completed in a shorter time compared to the case where the number of rotation axes is large.

  Therefore, in this embodiment, the imaging object W is covered with a small number of rotation axes A by setting the position of the rotation axis A using a regular hexagon. The regular hexagonal arrangement portion 27a is located in a reconfigurable area Zc that can be reconfigured by imaging around one rotation axis on a cut surface obtained by virtually cutting the imaging target W in a direction perpendicular to the rotation axis A. This is a program module that causes the CPU 27 to execute a function of virtually arranging a plurality of tangent regular hexagons in a state where the sides are adjacent to each other in a shared state.

  That is, the CPU 27 virtually sets a cut surface for obtaining a tomographic image of the imaging target W in a direction perpendicular to the rotation axis A. In the present embodiment, the cut surface is a surface that is parallel to the XY plane and passes through the imaging target W. Further, the CPU 27 virtually sets a regular hexagon inscribed in the reconfigurable area Zc1 instead of a square inscribed in the reconfigurable area Zc1 on the cut surface. In the present embodiment, the CPU 27 sets a regular hexagon by specifying one side r of the regular hexagon. Note that one side r of the regular hexagon is ½ of the diameter of the reconfigurable region Zc1, and details of calculation of the side r will be described later.

  FIG. 3A shows a state in which the reconfigurable area Zc1 is viewed from a direction parallel to the Z axis, and the circumference of the reconfigurable area Zc1 is indicated by a circle. When the diameter of the reconfigurable region Zc1 is 2r, one side of the regular hexagon inscribed in the reconfigurable region Zc1 is r as shown by a broken line. On the other hand, in FIG. 3A, a square inscribed in the reconfigurable region Zc1 is indicated by a one-dot chain line. As apparent from FIG. 3A, when comparing a regular hexagon inscribed in the same reconfigurable region Zc1 with a square, the area of the regular hexagon is larger. Therefore, when the regular hexagon and the square are arranged on the cut surface, in many cases, the regular hexagon can efficiently fill the cut surface.

  Therefore, the CPU 27 assumes a regular hexagon having a side r of ½ of the diameter of the reconfigurable region Zc1, and virtually arranges the regular hexagon in a state where the sides are adjacent to each other, thereby cutting the cut surface. A virtual arrangement of regular hexagons covering the imaging target W extending along the line is specified.

  In the present embodiment, the CPU 27 identifies the center of each regular hexagon based on the regular hexagonal arrangement identified by the process of the regular hexagonal arrangement unit 27a by the process of the imaging unit 27b, and sets it as the position of the rotation axis A. Then, the CPU 27 performs imaging at a plurality of imaging positions around the rotation axis A for each of the rotation axes A by the processing of the imaging unit 27b. Therefore, in this embodiment, there is a high possibility that the imaging target W can be covered even if the number of positions of the rotation axis A is small as compared with the case where the cut surface is filled with squares. For this reason, in many cases, the time required for imaging can be shortened.

  The quality determination unit 27d is a program module that causes the CPU 27 to execute a function of determining an inspection range based on a tomographic image of the imaging target W and determining the quality of the imaging target W. That is, the CPU 27 refers to the reconstruction information acquired by the processing of the reconstruction calculation unit 27c by the processing of the pass / fail determination unit 27d, and sets the imaging target W in the direction perpendicular to the Z axis based on the reconstruction information. A tomographic image of the cut surface is obtained. Then, the CPU 27 compares the tomographic image with a predetermined determination condition (for example, a condition indicating a pass / fail criterion based on the presence / absence of a void, the size, shape, position, etc.) of the void, and determines pass / fail based on the determination condition. .

(2) X-ray inspection process:
FIG. 4 is a flowchart showing the X-ray inspection process. The CPU 27 moves the XY stage 13 to a position where the imaging target W of the substrate can be imaged by the processing of the imaging unit 27b (step S100). That is, the CPU 27 specifies the imaging target W on the board based on the design information of the board (not shown) and moves the imaging target W to the initial position at the time of inspection.

  Next, the CPU 27 acquires an inspection region by the function of the regular hexagonal arrangement unit 27a (step S105). The inspection area is an area that includes the imaging object W to be inspected, and is an area that should be covered by imaging with respect to one or more rotation axes A. The imaging target W may be acquired by various methods. For example, an area including the imaging target W may be acquired as an inspection area based on the above-described design information or the like, or may be acquired by the X-ray detector 12 or the like. In a state where an image obtained by imaging the imaging target W from the axial direction is displayed on the output unit 25, the user may designate an examination area from the image by the input unit 24, and various configurations can be employed. In the present embodiment, the inspection area is rectangular.

  Next, the CPU 27 acquires imaging conditions by the function of the regular hexagonal arrangement unit 27a (step S110). As the imaging condition, control information specifying the output intensity of the X-ray from the X-ray generator 11, specification information indicating an effective imaging range of the X-ray detector 12, and the position of the movable part in the imaging mechanism 10 are specified. Location information and the like are included. Among these imaging conditions, predetermined information is stored in advance in the memory 26, and the CPU 27 refers to the memory 26 and acquires each piece of information.

  On the other hand, among the imaging conditions, default values are set in advance for variable information, but the user may input the values via the input unit 24 to obtain desired values. For example, the imaging magnification in the X-ray image is a variable value and may be input by the user. The imaging magnification can be realized by a mechanism that can change the XY stage 13 in the Z-axis direction. In this case, even if the imaging magnification is input indirectly by inputting a position in the Z-axis direction, etc. The imaging magnification itself may be directly input.

  The imaging magnification is (the size of the X-ray image of the imaging target W / the size of the imaging target W). For example, a straight line (light) passing through the focal point of the X-ray output unit 11a and the center of the X-ray detection surface 12a. The distance from the X-ray output unit 11a to the X-ray detection surface 12a (a + b shown in FIG. 2A) on the axis) is the distance from the X-ray output unit 11a to the rotation axis A (shown in FIG. 2A) on the same straight line. Calculated by the value divided by a). The effective imaging range is the size of an effective imaging area in which an X-ray image can be captured on the rectangular X-ray detection surface 12a, and is defined by the size of two sides.

  In the present embodiment, an effective imaging range is defined by the size of a horizontal side that is a side parallel to the XY plane and a vertical side that is perpendicular to the horizontal direction. The unit of the effective imaging range may be specified by various units, and may be a unit of length (mm or the like) or the number of dots. When the effective imaging range is expressed by the number of dots, it can be converted into a unit of length by multiplying the number of dots and the dot pitch (μm / dot, etc.).

  Next, the CPU 27 obtains the length of one side of the regular hexagon based on the imaging magnification and the effective imaging range by the function of the regular hexagonal arrangement unit 27a (step S115). In the present embodiment, one side of the regular hexagon is ½ of the size obtained by dividing the effective imaging range in the direction parallel to the cut surface in the X-ray detection surface 12a of the imaging mechanism 10 by the imaging magnification. That is, the effective imaging range of the X-ray detection surface 12a in the present embodiment is a rectangle, and the length in the horizontal direction (the direction parallel to the cut surface) is longer than the length in the vertical direction. In this effective imaging range, it is the length of the effective imaging range in the lateral direction that substantially defines the size of the reconfigurable area Zc.

  FIG. 3B is a diagram showing a set of the X-ray output unit 11a and the X-ray detection surface 12a rotating around the rotation axis at a plurality of rotation positions, viewed from a direction parallel to the Z axis. FIG. 3B shows three rotational positions, and the range of X-rays reaching the X-ray detection surface 12a from the X-ray output unit 11a at each rotational position is indicated by a one-dot chain line. Among these X-ray ranges, the overlapping range is a range where projection trajectories overlap in the back projection described above, and corresponds to the reconfigurable region Zc1. If there are many pairs of the X-ray output unit 11a and the X-ray detection surfaces 12a, the overlapping range approaches a circle indicated by a broken line in FIG. 3B, and the range of the circle can be regarded as the reconfigurable region Zc1. .

  In FIG. 3B, the length of the effective imaging range in the horizontal direction is Hd. As described above, the imaging magnification is the distance from the X-ray output unit 11a to the X-ray detection surface 12a (a + b shown in FIG. 2A), and the distance from the X-ray output unit 11a to the rotation axis A on the same straight line. It is the value divided by (a shown in FIG. 2A). Accordingly, the CPU 27 obtains the diameter D1 of the reconfigurable region Zc1 on the cut surface of the imaging target W by dividing the length Hd of the effective imaging range in the horizontal direction by the imaging magnification (a + b) / a. Can do. Therefore, if the CPU 27 acquires ½ of the diameter D1, the radius r of the reconfigurable region Zc1 can be acquired. Since the radius r is one side of a regular hexagon inscribed in the reconfigurable area Zc1, the length r of one side of the regular hexagon is acquired by the above processing.

  Next, the CPU 27 uses the functions of the regular hexagon arrangement unit 27a and the imaging unit 27b to match another regular hexagon with respect to a regular hexagon in which one corner and vertex of the inspection region Zr coincide with each other and one side of the inspection region Zr coincides with one side. The regular hexagonal arrangement including the inspection area Zr is obtained by packing the squares in a line (step S120).

  In FIG. 5A, the regular hexagon Hex is aligned so that the vertex of the regular hexagon Hex matches the corner of the rectangular inspection region Zr, the side of the inspection hexagon Hex matches one side of the regular hexagon Hex, and the inspection region Zr is included. It shows the state of lining up. In FIG. 5A, the inspection region Zr is indicated by a broken line, and the regular hexagon Hex is indicated by a solid line. Such an arrangement may be acquired by virtually arranging regular hexagons Hex with respect to the inspection region Zr, or may be calculated by calculation based on the sizes of the inspection region Zr and the regular hexagon Hex.

  Various methods can be adopted as a method for acquiring the arrangement of the regular hexagon Hex by calculation. For example, the position of the center Hr1 of the regular hexagon Hex existing at the corner is calculated based on the corner Tr of the inspection region Zr. A process of calculating a position at a distance of 2r in the direction perpendicular to each side of the regular hexagon from the center Hr1 (for example, Hr2 in FIG. 5A if the position is at a distance of 2r in the j direction) This is repeated until a new position cannot be calculated inside the region Zr, and can be realized by a configuration in which the position existing inside the inspection region Zr is the center of a regular hexagon. That is, paying attention to the outermost periphery of the regular hexagon arranged in this way, as shown in FIG. 5A, it can be said that the inspection region Zr is included inside the outermost periphery of the regular hexagon and is included.

In FIG. 5B, for a square Scu (one side 2 ^−1/2 r) inscribed in a circle inscribed by a regular hexagon of one side r, the corners and vertices of the inspection region Zr coincide, and the side and one side of the inspection region Zr are This is an example in which an arrangement of squares including the inspection region Zr is acquired with reference to a matching square. Comparing FIG. 5A and FIG. 5B, the total number of regular hexagons arranged to contain the inspection area Zr of the same size is 17, and the total number of squares is 20. Accordingly, in the example shown in FIGS. 5A and 5B, the regular hexagonal shape includes the inspection region Zr with a smaller number than the square.

  FIG. 6A and FIG. 6B show regular hexagons necessary for enclosing a rectangular inspection region having the same vertical and horizontal lengths by the regular hexagon and the square inscribed in the reconfigurable region Zc1 of the same size. It is a figure which shows the number of squares (total number of visual fields). FIG. 6A is a graph showing the inspection area size (size) on the horizontal axis and the total number of fields of view on the vertical axis. The size is defined with the radius (equal to one side of the regular hexagon) of the reconfigurable area Zc1 as 1. ing. Therefore, the inspection area size 10 indicates that the vertical and horizontal lengths of the inspection area are 10 times the radius of the reconfigurable area Zc1. In FIG. 6A, the total number of fields when a square is arranged is indicated by a broken line, and the total number of fields when a regular hexagon is arranged is indicated by a solid line.

  FIG. 6B is a diagram showing (total number of fields of square / total number of fields of regular hexagon) in% notation, and the horizontal axis is the same as FIG. 6A. As shown in FIGS. 6A and 6B, over almost the entire inspection area size, the total number of fields obtained by the regular hexagon is less than the total number of fields obtained by the square. As described above, when the regular hexagon inscribed in the reconfigurable region Zc1 having the same size is compared with the square, the regular hexagon can generate an arrangement that can include the inspection region with a smaller number than the square. it can.

  When the arrangement of regular hexagons is acquired as described above, the CPU 27 displays the regular hexagons superimposed on the inspection area Zr (step S125). That is, the CPU 27 outputs a control signal to the output unit 25, and displays the regular hexagons side by side as in the arrangement acquired in step S120 with respect to the inspection region Zr acquired in step S105. For example, the CPU 27 displays an image as shown in FIG. 5A on the output unit 25. Of course, here too, an image obtained by imaging the imaging target W from the Z-axis direction may be displayed in an overlapping manner. At this time, the CPU 27 displays an option for selecting any one of the re-input of the imaging condition, the re-input of the inspection area, and the start of imaging on the output unit 25.

  Next, the CPU 27 determines whether or not re-input of the imaging conditions has been accepted (step S130). That is, when the user operates the input unit 24 and selects the re-input of the imaging condition displayed on the output unit 25, the CPU 27 determines that the re-input of the imaging condition has been accepted. If it is determined in step S130 that re-input of the imaging conditions has been accepted, the CPU 27 repeats the processing from step S110.

  If it is not determined in step S130 that re-input of imaging conditions has been received, the CPU 27 determines whether or not re-input of the inspection area has been received (step S135). That is, when the user operates the input unit 24 to select re-input of the inspection area displayed on the output unit 25, the CPU 27 determines that re-input of the inspection area has been accepted. If it is determined in step S135 that re-input of the inspection area has been accepted, the CPU 27 repeats the processing from step S105.

  If it is not determined in step S135 that re-input of the inspection area has been received, the CPU 27 sets the center of the regular hexagon to the position of the rotation axis (step S140). That is, the CPU 27 identifies the center of the regular hexagon based on the regular hexagonal arrangement on the cut surface acquired in step S120, and sets it to the position of the rotation axis A. In the example shown in FIG. 5A, in addition to the positions Hr1 and Hr2, the centers of all the regular hexagons shown in FIG.

  Next, the CPU 27 executes CT imaging processing by the function of the imaging unit 27b (step S145). That is, the CPU 27 outputs predetermined instructions to the generator control unit 21, the detector control unit 22, and the imaging mechanism control unit 23 by the processing of the imaging unit 27b, and the center of the rotation axis A set in step S140. Select one of the following. Then, the CPU 27 adjusts the arrangement of the X-ray generator 11, the X-ray detector 12, and the like so that the imaging target W is irradiated with the X-rays at an angle inclined with respect to the selected rotation axis A. Further, the CPU 27 outputs a predetermined instruction to the generator control unit 21 by the processing of the imaging unit 27b, causes the X-ray generator 11 to output X-rays with a predetermined output, and causes the detector control unit 22 to The X-ray detector 12 obtains X-ray image data 26b by outputting a predetermined instruction.

  Further, the CPU 27 executes the process of imaging the X-ray image data 26b as described above at a plurality of imaging positions after rotation about the rotation axis A, and X-rays imaged at the plurality of imaging positions. The image data 26 b is recorded in the memory 26. When imaging for one rotation axis A is completed as described above, the CPU 27 selects an unphotographed rotation axis A and performs imaging for the newly selected rotation axis A. The above processing is repeated until imaging is performed for all the rotation axes A set in step S140.

  Next, the CPU 27 executes a reconstruction calculation process by the process of the reconstruction calculation unit 27c (step S150). The reconstruction calculation only needs to reconstruct the three-dimensional structure of the imaging target W, and various processes can be employed. The CPU 27 acquires reconstruction information by, for example, a filter-corrected back projection method. The CPU 27 performs processing for acquiring the reconstruction information for each rotation axis A set in step S140. As a result, the reconstruction information in the inspection area Zr on the cut surface as shown in FIG.

When the reconstruction information in the inspection area Zr is acquired as described above, the CPU 27 performs a quality determination process by the process of the quality determination unit 27d (step S155). That is, the CPU 27 extracts the reconstruction information of the imaging target W from the reconstruction information in the inspection area Zr by the processing of the quality determination unit 27d, and determines quality based on a predetermined determination condition.
The above embodiment is an inclined CT in which a straight line (optical axis) passing through the focus of the X-ray output unit 11a and the center of the X-ray detection surface 12a obliquely intersects the rotation axis A. Various values can be adopted as the crossing angle between the axis A and the optical axis. For example, as disclosed in Patent Document 1, an orthogonal CT in which the optical axis of the X-ray is orthogonal to the rotation axis may be used.

(3) Other embodiments:
The above embodiment is an example for carrying out the present invention, and various other embodiments can be adopted as long as a regular hexagon is arranged on the cut surface and the center of the regular hexagon is used as a rotation axis. . For example, various inspection objects other than the bumps in the BGA are envisaged, and may be through holes filled by plating, bumps used in the C4 method, or solder for mounting lead components. Etc. In addition, various methods may be employed for the quality determination process. For example, the quality determination may be performed based on the shape of the bump, the distance between the void and the center of gravity of the inspection target, or the like.

  Furthermore, the processing when the CPU 27 arranges the regular hexagon by the regular hexagon arranging portion 27a can employ various processes other than the above-described processes. For example, the CPU 27 overlaps the regular hexagons arranged on the cut surface with the regular hexagonal arrangement unit 27a and the imaging target W arranged on the cut surface, and the imaging unit 27b performs the imaging target. Alternatively, a regular hexagon that includes a center of the selected hexagon may be selected, and the center of the selected regular hexagon may be the rotation axis.

  That is, the CPU 27 virtually generates a mask in which regular hexagons inscribed in the reconfigurable area Zc1 are two-dimensionally arranged without a gap. Then, if a regular hexagonal mask is arranged on the cut surface and the imaging target W arranged on the cut surface is overlapped, it can be seen that the center of the regular hexagon including the imaging target W may be the rotation axis A. FIG. 7A is a diagram schematically illustrating bumps that are imaging targets W included in the chip C. FIG. In FIG. 7A, it is assumed that the widest surface of the rectangular parallelepiped chip C is parallel to the cut surface, and the imaging target W can be regarded as being arranged on the cut surface as shown in FIG. 7A.

  FIG. 7B is a diagram in which the imaging target W arranged on the cut surface and a regular hexagonal mask Mk inscribed in the reconfigurable region Zc1 are overlapped. As described above, when the imaging target W arranged on the cut surface and the mask Mk are overlapped, the imaging with the rotation axis A as the center only in the regular hexagon in which the imaging target W exists in the mask Mk may be performed. It turns out.

  Therefore, the CPU 27 extracts a regular hexagon whose center is the rotation axis A from the mask Mk, automatically or manually. When the regular hexagon is automatically extracted, the CPU 27 specifies the position of the imaging target W on the cut surface based on the design information of the chip C, the X-ray image including the imaging target W, and the like. Further, the CPU 27 specifies the position of the circumference of each regular hexagon constituting the mask Mk, and determines whether or not the imaging target W is included inside the circumference. Then, a regular hexagon in which the imaging target W is included inside the circumference is specified, and the center of the regular hexagon is acquired as the position of the rotation axis A.

  On the other hand, when the regular hexagon is manually extracted, the CPU 27 causes the output unit 25 to display an image in which the mask Mk and the imaging target W as illustrated in FIG. Then, the CPU 27 receives a regular hexagonal selection performed by the user operating the input unit 24. The selection of a regular hexagon may be a selection of a regular hexagon to be left, or a selection of a regular hexagon to be removed from the mask Mk. In any case, the CPU 27 acquires the center of the regular hexagon specified by the selection as the position of the rotation axis A. In addition, according to the manual selection, when there is an object that is similar to the imaging target W but is not an imaging target or an imaging target W that does not need to be inspected, it is easy to exclude them.

  In FIG. 7B, the regular hexagon excluded from the mask Mk is illustrated by hatching. According to the processing as described above, it is possible to determine the regular hexagonal arrangement without defining the inspection region. Further, even when the imaging target W is arranged in an arbitrary shape on the cut surface, it is possible to easily obtain a minimum number of regular hexagonal arrangements.

  DESCRIPTION OF SYMBOLS 10 ... Imaging mechanism, 11 ... X-ray generator, 11a ... X-ray output part, 12 ... X-ray detector, 12a ... X-ray detection surface, 13 ... XY stage, 20 ... Control part, 21 ... Generator control , 22 ... Detector control unit, 23 ... Imaging mechanism control unit, 24 ... Input unit, 25 ... Output unit, 26 ... Memory, 26a, 27b ... Imaging unit, 27c ... Reconstruction calculation unit, 27d ... Pass / fail judgment unit

Claims (6)

The imaging object can be imaged with radiation output in a direction inclined or orthogonal to the rotation axis at a plurality of imaging positions with different rotation angles around the rotation axis, and the position of the rotation axis is perpendicular to the rotation axis An imaging mechanism that can be changed in any direction,
Shares a side of a regular hexagon inscribed in a reconfigurable area that can be reconfigured by imaging around one rotation axis on a cut surface obtained by virtually cutting the imaging target in a direction perpendicular to the rotation axis And a regular hexagonal arrangement section that virtually arranges a plurality of them adjacent to each other,
An imaging unit that images the imaging object at the plurality of imaging positions with the center of the regular hexagon as the rotation axis;
A reconstruction calculation unit that executes a reconstruction calculation based on the captured image for each rotation axis;
A three-dimensional imaging device. The imaging mechanism includes a radiation detector in which one side of a rectangular effective imaging range is parallel to the cut surface,
One side of the regular hexagon is ½ of the length obtained by dividing the length of one side of the effective imaging range parallel to the cut surface by the imaging magnification.
The three-dimensional imaging device according to claim 1. The regular hexagonal arrangement part is
With respect to the rectangular inspection area on the cut surface, one corner and vertex of the inspection area coincide with each other, and the regular hexagon whose one side coincides with one side of the rectangle is filled with the other regular hexagon. ,
The imaging unit
The center of the regular hexagon containing the rectangle is the rotation axis,
The three-dimensional imaging device according to claim 1. The regular hexagonal arrangement part is
The regular hexagons arranged side by side on the cut surface and the imaging object arranged on the cut surface are overlapped,
The imaging unit
Select the regular hexagon containing the imaging object, and the center of the selected regular hexagon as the rotation axis,
The three-dimensional imaging device according to claim 1. The imaging object can be imaged with radiation output in a direction inclined or orthogonal to the rotation axis at a plurality of imaging positions with different rotation angles around the rotation axis, and the position of the rotation axis is perpendicular to the rotation axis A three-dimensional imaging method using an imaging mechanism that can be changed in any direction,
Shares a side of a regular hexagon inscribed in a reconfigurable area that can be reconfigured by imaging around one rotation axis on a cut surface obtained by virtually cutting the imaging target in a direction perpendicular to the rotation axis And a regular hexagonal arrangement step in which a plurality of virtual hexagons are virtually arranged in an adjacent state,
An imaging step of imaging the imaging object at the plurality of imaging positions with the center of the regular hexagon as the rotation axis;
A reconstruction calculation step of performing a reconstruction calculation based on the captured image for each rotation axis;
A three-dimensional imaging method. The imaging object can be imaged with radiation output in a direction inclined or orthogonal to the rotation axis at a plurality of imaging positions with different rotation angles around the rotation axis, and the position of the rotation axis is perpendicular to the rotation axis A computer that controls the imaging mechanism that can be changed in any direction,
Shares a side of a regular hexagon inscribed in a reconfigurable area that can be reconfigured by imaging around one rotation axis on a cut surface obtained by virtually cutting the imaging target in a direction perpendicular to the rotation axis And a regular hexagonal arrangement function that virtually arranges them adjacent to each other,
An imaging function for imaging the imaging object at the plurality of imaging positions with the center of the regular hexagon as the rotation axis;
A reconstruction calculation function for executing a reconstruction calculation based on the captured image for each rotation axis;
3D imaging program that realizes