JP2000018919A - Imaging device, optical measuring apparatus, and optical system inspecting apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an image sensing device, capable of obtaining an accurate image on a projection drawing method. SOLUTION: An image-pickup device 10 is equipped with a parallel light generating part 52 and an image-pickup part 50. The parallel light generating part 52 converts an illuminating light radially outputted from a pin hole 174 of an orifice 152 into parallel light with a concave mirror 148, makes a beam splitter 150 reflect the parallel light at a right angle, and makes it pass vertically the periphery of an object 180 to be sensed from a lower part to the upper part. In the image sensing part 50, the parallel light which has passed the periphery of the object 180 is reflected by a beam splitter 72, and enters a concave mirror 70 from the direction parallel to the optical axis. The concave mirror 70 causes the parallel light to be condensed at a focal point. Only the light condensed in the focal point passes a pin hole 80 of an orifice 74 which exists at the focal point, and spreads radially. This light is converted into a parallel light by a lens system 76, and an image which is formed by the parallel light is pickup by a CCD 82. Thereby an accurate image on a projection drawing method can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image pickup apparatus for picking up an image of an object and obtaining an image represented by image data, and processing the data of the image picked up by the image pickup apparatus to obtain the position (reference) of the image pickup object. An optical measuring device that measures a position on a coordinate plane), a rotation angle (a rotation angle from a reference rotation position), a dimension, and the like, and an optical device that performs a defect inspection of an imaging target by processing data of a captured image. The present invention relates to a type inspection apparatus.

[0002]

2. Description of the Related Art The present applicant has developed an image pickup apparatus, an optical measuring apparatus and an optical inspection apparatus of this kind and applied for a patent. Japanese Patent Application No. 10-9085 and Japanese Patent Application No. 10-162449, which have not been disclosed yet, are such applications. Japanese Patent Application No. 1
The imaging device described in No. 0-9085 includes a reflecting surface that forms a plane, a first lens system that is disposed with the reflecting surface orthogonal to the optical axis, and that condenses parallel light from the reflecting surface to a focal point. An imaging object support device that supports an imaging object at a position between the reflection surfaces and the first lens system, and the reflection surface is provided at a position opposite to the reflection surface with respect to the first lens system, A beam splitter that is inclined with respect to the optical axis of one lens system, has a reflecting surface that transmits part of light, and reflects the rest, and the focal point of the first lens system with respect to the reflecting surface of the beam splitter is An orifice provided at a plane-symmetric position, a light source provided on the side opposite to the beam splitter with respect to the orifice, and a focal point at the focal point of the first lens system, and condensed by the first lens system And passes through the beam splitter A second lens system for converting the light into parallel light, configured to include an imaging device that captures an image formed by light converted into parallel beams by the second lens system.

Light emitted radially from a point light source constituted by a light source and an orifice is reflected by a reflection surface of a beam splitter and travels toward a first lens system. Therefore, since it is located at a plane symmetrical position with respect to the focal point of the first lens system, it is the same as the light radially projected from its own focal point for the first lens system,
An object converted into parallel light by the first lens system and supported by the imaging object support device (there is a case where the object itself is the imaging object, and a case where a part of the object surface is the imaging object. The object supported by the imaging object supporting device is simply referred to as an object unless otherwise required) and is applied to the reflecting surface. The light reflected by the reflecting surface and the light reflected by the object surface enter the beam splitter, and part of the light is returned to the point light source by the reflecting surface, while the rest passes through the beam splitter to reach the second lens system. Of the light reflected by the reflecting surface and the light reflected by the object surface, the light parallel to the optical axis of the first lens system (perpendicular to the reflecting surface) is converted into parallel light and incident on the image sensor. If the point light source constituted by the light source and the orifice is actually installed at the focal point of the first lens system, the reflected light from the reflecting surface or the object surface is blocked by the orifice and the light source and is incident on the image sensor. As a result, an image of the object and its surroundings cannot be obtained, but this problem can be solved by using a beam splitter as in the present embodiment.

[0004] Each of the first and second lens systems may be constituted by one lens or a plurality of lenses. Also, for example, a reflecting surface is provided between the second lens system and the image sensor, so that light converted into parallel light by the second lens system is reflected by the reflecting surface and is incident on the image sensor, A reflecting surface is provided between the focal point of one lens system and the second lens system so that light from the first lens system is reflected by the reflecting surface and then converted into parallel light by the second lens system. can do. The same is possible for the first lens system. in this way,
The direction of light can be arbitrarily changed by the reflection surface, and in this case, the optical axis of the lens system is considered to be bent together with the light. In the optical measuring device of this embodiment, it is not essential that the reflecting surface, the imaging object support device, the first lens system, the second lens system, and the imaging sensor are arranged on a straight line.

[0005] As described above, parallel light orthogonal to the reflective surface is applied to the reflective surface and the object placed in front of the reflective surface, and an image formed by the parallel light reflected by the reflective surface and passing around the object is captured. By doing so, an accurate projected image of the object can be obtained, and by processing the projected image, it is possible to obtain accurate dimensions in the perspective projection or the projection. Since the parallel light is applied to both the object and the reflection surface, not only the light reflected on the reflection surface, but also the light reflected on the surface of the object reaches the imaging surface and forms a surface image of the object, Even if the surface of the object is a ground surface, fine unevenness is present, so that only a part of the parallel light perpendicular to the reflecting surface is reflected in a direction perpendicular to the reflecting surface. On the other hand, if the reflecting surface is a mirror surface, almost all of the parallel light perpendicular to the reflecting surface is reflected in a direction perpendicular to the reflecting surface. Therefore, it is easy to make the image of the reflection surface significantly brighter than the image of the surface of the object, and the contour of the object can be easily specified. A silhouette image of the object formed by the parallel light reflected by the reflecting surface and passing around the object is obtained. By processing the silhouette image, the outer dimension of the object (if there is a penetrating aperture, The inner dimension of the opening can also be determined. However, if there are many components of light reflected from the surface of the object in the direction perpendicular to the reflection surface, the surface of the object is also imaged, so characters such as characters and symbols written on the surface of the object and formed on the surface Accurate projection images of the formed concave portions and scratches can also be obtained, and if necessary, their dimensions and positions can be accurately measured. Also, since the optical system that irradiates the object and the reflecting surface with parallel light, and the optical system that forms an image formed by the parallel reflected light from the object and the reflecting surface on the imaging surface can also be provided on the same side as the object, It is possible to make the entire imaging device compact.

The optical measuring device described in Japanese Patent Application No. 10-9085 is a device for measuring the position of the object in a direction parallel to the reflection surface of the object based on an image taken by the image pickup device. An image processing device that calculates at least one of a rotation angle about an axis perpendicular to the surface and a dimension in a direction parallel to the reflection surface. This optical measuring device can measure the position, the rotation angle, and the dimension without contacting the measurement object, that is, the imaging object, so that the measurement object made of a soft material such as rubber can be measured without any trouble. Further, the image processing apparatus is capable of obtaining dimensions regardless of the position of an image of a measurement target (imaging target) in an image obtained by imaging and a rotation angle (a rotation angle from a reference rotation position). Then, the object to be measured (object) can be simply placed on the flat support surface, and the object to be measured can be easily set on the measuring device.
The efficiency of measurement can be improved. An example of the image processing device in the optical measuring device uses a search template and a measurement template. That is, the image processing apparatus is configured to include a master search template data storage unit, a master measurement template data storage unit, a search template automatic setting unit, a search object determination unit, a measurement template automatic setting unit, and an edge point coordinate calculation unit. It is. The master search template has a plurality of point pairs each including two points separated by a predetermined distance, and the master search template data storage means stores the data of the master search template. The master measurement template is
It has a plurality of measurement seek lines, and the master measurement template data storage means stores the data of the master measurement template. The search template automatic setting means stores the master search template data stored in the master search template data storage means,
A plurality of search templates are automatically set by moving to various positions and rotation angles. When the plurality of search templates set by the search template automatic setting unit are sequentially superimposed on a screen on which an image represented by the image data in the image data storage unit exists, the search object determination unit includes a plurality of point pairs of the search template. When the difference state of the optical characteristic values of the points of each pair constituting the pair is equal to or greater than the set state, the two points of the pair are divided on both sides of the edge where the optical characteristic of the image changes suddenly. It is determined that the imaging target is a search target that matches the search template if more than a set amount of the plurality of point pairs are in the match. The measurement template automatic setting unit is configured to perform a search based on the search template data when the imaging target is determined to be a search target that conforms to the search template and the master measurement template data stored in the master measurement template data storage unit. Then, the measurement template is automatically set. The edge point coordinate calculation means superimposes the measurement template set by the measurement template automatic setting means on a screen on which an image represented by the image data of the image data storage means exists, and measures the measurement target on each of the plurality of measurement seek lines. The coordinates of the edge point of the (imaging target, search target) are calculated. The edge point is the intersection of the edge and the measurement seek line. The edge is a portion where the optical measurement changes suddenly in the image represented by the image data, and corresponds to the contour of the measurement target. Edges are often closed curves, but not always.
Examples are a case where the measurement target is large and does not fit within the field of view of the imaging device, a case where the measurement target is a part of the object, and the like.

An optical inspection apparatus is disclosed in Japanese Patent Application No. Hei 10-1624.
No. 59. The image processing apparatus of the optical inspection apparatus includes an image data storage unit that stores image data that is data of an image captured by an imaging device.
It is configured to include negative seek line setting means and determination means. The negative seek line is a seek line that is not expected to intersect with the edge, and the negative seek line setting means sets the negative seek line. The determining means determines that the imaging target has a defect when the set negative seek line intersects the edge. The seek line of the measurement template in the optical measuring device is expected to intersect the edge, and in that sense, should be called a positive seek line,
By using a negative seek line that is not expected to intersect with the edge, it is possible to easily detect a defect such as a scratch on the surface of the object supported by the imaging object support device, a cutout on the outer surface, or a flash. For example, when inspecting a horizontal part of a square part for flaws, set a number of negative seek lines inside the outline of the surface, and any of those negative seek lines Check if they intersect. If there is a flaw or dirt that becomes a dark image on the surface that becomes a bright image, the outline of the flaw or dirt becomes an edge. Since the negative seek line is set inside the contour of the surface as described above, it does not normally intersect the edge,
If there are scratches and stains on the surface and the negative seek line intersects with the scratches and stains, the negative seek line will intersect the edge. Therefore, if it is determined whether or not there is an edge point on the negative seek line, that is, a point where the change in luminance is large, it is possible to determine whether or not a scratch or a stain exists on the surface of the inspection target.
Further, whether or not a notch is present on the outer surface of the object of the rectangular component can be detected by setting a negative seek line along the outline immediately inside the outline of the object. Similarly, a flash can be detected by setting a negative seek line immediately outside the contour of the object. In the present optical inspection apparatus, defects such as scratches, dirt, notches, and burrs can be considered as imaging targets, and objects having defects can be considered as imaging targets.

[0008]

Problems to be Solved by the Invention, Means for Solving Problems, Actions
The present invention relates to an imaging device, an optical device,
Improve scientific measuring and optical inspection equipment, especially
Improve the accuracy of images acquired by imaging devices
And the measurement accuracy of the optical measuring device
To improve the reliability of optical inspection equipment.
This was done as an issue. According to the present invention,
Imaging device, optical measuring device, optical inspection device
Is obtained. Each aspect is divided into terms as in the claims,
Number each section and quote other sections as needed
Describe in the following format. This is a combination of the features described in each section.
In order to facilitate understanding of the possibility of
The technical features described and their combinations are:
Should not be construed as limited to (1) Concave mirror and on the optical axis of the concave mirror
Beam splitter (half mirror)
And the beam splitter as viewed from the concave mirror side.
It is arranged in one of the transmission direction and the reflection direction,
An imaging pair for supporting an object substantially at the focal point of the concave mirror
An object support device and the other of the transmission direction and the reflection direction.
The beam splitter with respect to the focal point of the concave mirror
Are provided on the opposite side and extend radially through the focal point.
Lens system that converts light into substantially parallel light, and its lens
Image formed by parallel light converted by the
An image sensor for imaging and light incident on the image sensor are
Incident on the concave mirror substantially parallel to the optical axis of the concave mirror
An imaging device comprising:
(Claim 1). The imaging target is attached to the imaging target support device.
If the object is a supported object itself, or if it is a part of the object
Sometimes. In the latter case, the imaging object support device
Supports the object to be imaged by supporting the object.
You will have. Supported by the imaging object support device
The illuminated object can be illuminated from the front
May be disclosed. When illuminated from the front side,
Of light reflected by the front of the device passes through the beam splitter
Or a concave mirror reflected by the beam splitter
Incident on. The concave mirror enters from a direction parallel to its optical axis.
Because it has the property of condensing the emitted light at the focal point,
Light reflected radially in all directions by the front of the body
Only the component light parallel to the optical axis of the concave mirror is focused at the focal point.
Can be done. Therefore, at the position of the focal point,
Orifice with pinhole as described in
And only the component light focused at the focal point is allowed to pass.
So that it is incident on the concave mirror parallel to its optical axis
Only the component light enters the lens system. Orifice
This constitutes the limiting means.
Light that spreads radially through the fiss is flattened by the lens system.
The light is converted into row light and is incident on the image sensor to form an image.
However, the pinhole is formed as small as possible
The size of the light cannot be reduced to 0, but is focused near the focal point
Component light, i.e., slightly with respect to the optical axis of the concave mirror.
The component light reflected in the direction inclined to
Incident on the image sensor. As a result, on the imaging surface of the imaging device
Image is formed by component light substantially parallel to the optical axis of the concave mirror.
This image is taken by the imaging device
You. Therefore, the image sensor is located at the focal point of the lens system.
It is desirable to be installed. Supported by the imaging object support device
Even when the held object is illuminated from the back side, the concave mirror
An orifice with a pinhole at the focal point of
The light passing through the periphery of the object and entering the concave mirror
The component light incident parallel to the optical axis of the concave mirror
Imaging device by component light parallel to the center line of the object support device)
An image is formed on the imaging surface of the object and a silhouette image of the object is acquired
Is done. In this case, the light passing around the object is
In addition to the component light parallel to the center line of the elephant support device,
Even if it contains component light in the direction tilted,
The incidence of component light in the direction on the imaging surface of the image sensor
Blocked by Fiss. Therefore, as described in (5),
Use a light background forming device or a uniform illumination device described in (6).
The silhouette image of the object
The cut image is projected onto the concave mirror from a direction substantially parallel to its optical axis.
It is formed only by the incident component light. So
In contrast, the light passing around the object itself is
So that it consists only of component light parallel to the center line of the support device.
If the concave mirror is incident on the concave mirror, the component light parallel to its optical axis
Orifice at the focal point of the concave mirror.
Is not always necessary. Around the object
The light passing through is a component parallel to the center line of the imaging object support device.
Means consisting solely of light constitute limiting means
It will be. Move the object inside the imaging object support device.
To illuminate with light consisting only of component light parallel to the core wire
Is, for example, the parallel light generator described in (7) or (8)
Just use it. When using a parallel light generator
However, it is effective to provide an orifice. Collimated light
Light that passes around the object using the raw device
So that it consists only of component light parallel to the center line of the object support device
Even by the diffraction of light, the concave mirror is
Component light that is incident from an inclined direction
Inevitably impinges on the imaging surface of the image sensor.
is there. The restricting means controls the light incident on the image sensor as a result.
Limited to light incident on a concave mirror substantially parallel to its optical axis
Anything that can do it,
Considering folding, it should be installed as close as possible to the image sensor.
Would be desirable. Collimated light generator and
Refinements are both functions that act as restricting means.
If both are provided, a particularly accurate image can be obtained. Theory
As described above, in the imaging apparatus according to the aspect described in this section,
Is imaged by component light substantially parallel to the optical axis of the concave mirror.
Is formed, the obtained image of the object to be imaged is substantially
It is an accurate image in perspective or projection. Develop first
In the imaging device described in Japanese Patent Application No. 10-9085,
Reflects around the object supported by the imaging object support device
Object passing through the surface and reflected by the reflective surface again
Which is substantially perpendicular to the reflective surface
An image is formed on the imaging surface of the image sensor by the component light
Therefore, the outer contour of the silhouette image of the object is a regular contour
Larger inner contours (through holes in the object)
It is unavoidable that the contour line) is imaged smaller. Soshi
This reduces the accuracy of the image of the imaged
Measurement accuracy of the measuring instrument
Cause a decrease in reliability.
Of the object supported by the imaging object support device
The component light forming the jet image passes once around the object
Only the accuracy of the image of the object to be imaged can be improved.
And easy. In addition, the imaging device of this aspect is manufactured at low cost.
It has the following advantages. Recorded in Japanese Patent Application No. Hei 10-9085
In the above-mentioned imaging device, the reflection including the information of the imaging target is
A lens system for converging component light perpendicular to the surface to one point
Was used, so if the object to be imaged was large,
While a large-diameter lens system was required, the lens system
This is because a concave mirror is used instead of Parallel
Focus the component light at one point and convert it to parallel component light again
In order for the image sensor to capture accurate images,
It is necessary to reduce aberrations of the lens system,
To produce a lens system with small aberration,
It is necessary to combine a number of lenses, and
It is difficult to reduce aberrations, but a concave mirror
Can easily produce one stigmatic lens. (2) The restricting means is located at the focal position of the concave mirror.
Includes orifices with pinholes
Imaging device. (3) The orifice, lens system and image sensor are
Item (1) or provided in the reflection direction of the beam splitter
The imaging device according to item (2). (4) The imaging object support device is the beam splitter.
The shooting described in (1) or (2) provided in the reflection direction of
Imaging device. (5) Further, the camera is supported on the imaging object support device.
The object to be imaged, which is provided on the side opposite to the beam splitter.
Background formation to form a bright and substantially uniform background
The equipment described in any one of (1) to (4)
Imaging device (Claim 3). Light radiates from the light background forming device
Is emitted. Of the radial light, as described above,
The component light substantially parallel to the optical axis of the concave mirror causes
A silhouette image of the elephant is formed on the imaging surface of the image sensor.
You. (6) The light background forming device is the imaging object supporting device.
Lighting device that emits substantially uniform illumination light toward
An imaging device according to (5). Light background forming equipment
The lighting device itself can be configured by the launch surface.
Therefore, it is more energy efficient to construct a light background forming device.
No. (7) Further, with the imaging object supporting device interposed therebetween,
Provided on the opposite side of the beam splitter,
Beam passing through the holding device and around the object to be imaged.
A parallel light generator that projects parallel light that enters the splitter
The parallel light generator functions as the limiting means.
The imaging device according to any one of (1) to (4).
(Claim 4). Flat light generated by a collimated light generator
When illuminating the object to be imaged from the back side with line light,
Useless component light that is not substantially parallel to the optical axis of the concave mirror is taken.
Can be almost non-existent around the image object
Therefore, a high-quality image can be easily obtained. An example
For example, if the imaging target is a sphere,
From the back side with light emitted in all directions
If this is done, some of the illumination light will be in the hemisphere on the front side of the sphere
(Semi-spherical surface opposite to the light-background forming device)
And is reflected there toward the concave mirror. In this light
Includes component light parallel to the optical axis of the concave mirror, which is
And enters the image sensor through the
The Rouette image is smaller than the theoretical silhouette image
Will be able to. In contrast, the sphere is a parallel light generator
Illuminated from behind by the parallel light generated by the
, The parallel light hits the front hemisphere of the sphere
Never. Therefore, an accurate silhouette image of the sphere
You get it. The object to be imaged is, for example, a rectangular parallelepiped
Even if the corners of the rectangular parallelepiped are rounded,
A similar situation occurs for the rounded portion. In addition,
The light emitted from the scene forming device is only a small part (the light of the concave mirror).
Only the component light substantially parallel to the axis) is effective for image acquisition
Used. On the other hand, it is imaged by a parallel light generator
If the object is illuminated from behind, all of the illumination
Since it is effectively used for image formation, images can be efficiently formed.
Can be achieved. (8) The parallel light generating device is an imaging object supporting device.
The beam splitter as the first beam splitter
A second beam splitter provided on the side opposite to the splitter
And the second beams as viewed from the imaging object support device.
The concave mirror as a first concave mirror in the reflection direction of the splitter
From the second concave mirror provided separately from the second concave mirror
The second in the transmission direction of the second beam splitter
Provided at the focal position of the concave mirror, and provided to the second beam splitter
A point light source that emits illumination light radiating toward
(7) The imaging device according to the mode (7) (claim 5). According to this aspect
Thus, parallel light can be easily generated. (9) The point light source is provided at a focal position of the second concave mirror.
Front as first orifice with pinholes
A second orifice separate from the orifice and a second orifice
On the opposite side to the second beam splitter
And an almost parallel illumination light toward the orifice.
A collimated light source that emits light and light from the collimated light source
The second beam splitter through a pinhole in the orifice
Converting the light into illumination light that radiates toward
Including the second lens system separate from the lens system
An imaging device (claim 6). (10) The first concave mirror, the first beam splitter, the first
Orifice, first lens system, second concave mirror, second lens
The beam splitter, the second orifice, and the second lens system.
In each case, the imaging object support by the imaging object support device
The imaging device described in (9) is arranged symmetrically with respect to the position.
(Claim 7). According to this aspect, the imaging device and the parallel light emitting device
The raw device can be configured symmetrically, and the entire device can be
Can be configured to be compact and share parts
As a result, the manufacturing cost can be reduced. (11) Further, the concave mirror of the beam splitter
Includes a tilt angle adjustment device that adjusts the tilt angle with respect to the optical axis.
(M) The imaging device according to any one of (1) to (10).
If an inclination angle adjustment device is provided, manufacturing errors of the imaging device can be reduced.
Counter by adjusting the tilt angle of the beam splitter.
Can be reduced, and inexpensive
High images can be obtained. (12) The tilt angle adjusting device is provided with the beam splitter.
To the device body, which is the body of the imaging device, with the concave surface
Pivotable about a pivot axis perpendicular to the optical axis of the mirror
Mounting shaft and the beam splitter,
An elongated hole formed long in a direction orthogonal to the axis of the shaft,
When the supported shaft portion is parallel to the axis of the mounting shaft in the apparatus body,
The shaft is supported rotatably about the axis of rotation, and its supported shaft is
At the eccentric shaft portion eccentric with respect to
The beam sprocket is rotated and operated.
Eccentric operation for rotating the litter about the axis of the mounting shaft
The imaging device according to (11), including a working member. According to this aspect
If this is the case, the inclination angle adjusting device can be configured at low cost. (13) The orifice, lens system and imaging
Position to adjust the position of the elements in a direction parallel to their optical axis
Including one of the position adjustment devices, described in any one of (1) to (12).
Mounted imaging device. If a position adjustment device is provided,
Error in the axial direction of the orifice, lens system and image sensor
Can be canceled or reduced by adjusting the position of
Obtain high-precision images using inexpensive imaging devices
Can be. (14) The position adjusting device includes the orifice and the lens.
A holding member such as an image sensor for holding the system and the image sensor;
An optical axis of the concave mirror is provided on an apparatus main body which is a main body of the imaging apparatus.
To guide the holding member such as the image sensor.
The guide, the main body and the holding member such as the imaging device
Engage with both of them so that they can rotate relative to each other.
At least one of the engagement with the holding member such as an image element
It is screwed with a female screw, and is rotated.
Moving the holding member such as the image sensor along the guide
The imaging device according to item (13), including an adjustment screw member. This aspect
According to the above, the position adjusting device can be configured at a low cost.
You. (15) Imaging according to any one of (1) to (14)
Device and an image representing an image captured by the imaging device
By processing the data, the position and rotation of the
Image processing device for acquiring at least one of roll angle and dimension
An optical measuring device comprising: According to this aspect
For example, the position, rotation angle, dimensions, etc. of the imaging target can be accurately measured.
A determinable optical measuring device is obtained. (16) The image processing device captures an image using the image sensor.
Image data storage means for storing data of the obtained image,
Multiple seek lines consisting of two points connected in a straight line
Measurement template that stores the data of the measurement template
Data storage means and its measurement template data
The measurement template of the storage means is stored in the image data storage means.
Superimpose on the screen where the image represented by the image data exists, and
Sudden changes in optical properties on each of the book seek lines
Point coordinate operation to calculate the coordinates of the edge point
The optical measuring device according to the above mode (15) including
9). The “optical characteristic values” are, for example, luminance, hue,
And so on. The term “edge point” refers to the seek line and the edge.
It is the point of intersection with Ji. Edges are the optical characteristics of the image
Is the part where the imaging object changes suddenly.
If the whole object is supported by the
The imaging object is supported by the imaging object support device corresponding to the contour line.
Characters, recesses, scratches, etc. on the surface of the object
In this case, these characters correspond to contour lines. True nature
In an optical measuring device like the one described above, imaging is performed by an imaging device.
When an image is taken, the imaging surface is usually broken down into a number of pixels,
Image data is created for each pixel and
Is stored in the data storage means. "Image represented by image data"
Is an image obtained based on the image data,
In the image where the optical characteristic value is obtained for each pixel,
Good, or not in units of pixels, with arbitrary optical characteristics
An image obtained at the point (1) may be used. One example of the former is
Number of photoelectric conversion elements
Each obtained by the image sensor that generates an electric signal according to
An image thought as a set of electrical signals for each photoelectric conversion element
The image is an object on the imaging surface on which the photoelectric conversion elements are arranged.
It can be said that the image is a logically formed image. That
In the sense, this image is called a physical image, and a physical image exists
Screen (corresponding to the imaging surface) is called a physical screen
And The physical image data is the electrical data for each photoelectric conversion element.
Signal data is stored in association with the position of each photoelectric conversion element
It can be stored and reproduced by storing it in the means.
Also, image data representing a physical image actually exists.
In that sense, a physical image is compared with a virtual image described later.
Can also be called a real image, and a real image exists
The screen can be referred to as a real screen. Represents a real image
The image data is the size of the electric signal of each photoelectric conversion element.
It can be analog or digital data representing the body
For example, multi-stage digital values represented by 256 discrete values
Data (referred to as gradation data)
Binary depending on whether the electric signal of each element exceeds the threshold
Binarized data may be used. Light without using pixels
An example of an image in which biological characteristic values are obtained at any point is as follows:
The electric signal data of each photoelectric conversion element on the physical screen is
It is considered to represent the optical characteristic value of the center point of the photoelectric conversion element.
And assuming a curved surface that satisfies the optical characteristic values of these many points
, The continuous optical characteristic value that defines the curved surface
Is an image considered as a set of. Image data of this image
Data is stored as, for example, data of an expression representing the above-mentioned curved surface
This image data is also a kind of real image
There will be. On the other hand, the data of the above
Instead of asking for it, an arbitrary point on the screen is specified.
When each point is specified the optical property value of that point only
It is also possible to calculate by calculation every time. In this case
Is an image that does not actually exist but is assumed to exist
Is called a virtual image, and a virtual image exists.
The screen is referred to as a virtual screen. "Measurement template
On the physical screen, for example,
Corresponding to the pixel specified by the measurement template data
Image data from the image data storage means.
Yes, or specified by seek line on virtual screen
Image data (optical characteristic values)
It is to calculate based on the data. The object to be imaged
In the case of the entire object supported by the image object support device,
The inside of the outline of the object is inside the edge, and the periphery of the object is
The object to be imaged is on the surface of the object
If it is a formed character, recess, scratch, etc.,
The object surface surrounding them is outside the edge or the background. I
Even if the image data is shifted, the image data and background
Is different from the image data corresponding to. Conversely
If there is no difference from the background,
Cannot obtain information about the imaging target
No. The optical characteristics between the object to be imaged and its surroundings are as described above.
Optical property value at the edge point
Change suddenly. Therefore, it is obtained for each pixel on the physical screen
By determining changes in optical characteristic values,
By determining the change gradient of the optical characteristic value on the surface,
The di point can be determined. On the physical screen
Edge point is determined in pixel units, but on the virtual screen
Finds the edge point at any point on the seek line
The resolution of image processing can be increased as necessary.
Wear. According to this aspect, it is possible to obtain the edge points of the imaging object.
Can be. Thereby, for example, the shape of the imaging object
If you know in advance, relatively few seek lines
Calculate the size, position, rotation angle, etc. of the imaging target by setting
be able to. Also, the shape of the imaging target is known in advance.
Increase the number of seek lines and increase the number of edge points
Can be obtained from the set of edge points.
It is possible to obtain dimensions, positions, rotation angles, etc.
You. The imaging object support device moves the imaging object to the correct position and
With a positioning device for positioning at a rotation angle
And when the dimensions of the imaging target are almost constant,
The measurement template is uniquely determined. But the imaging target
When the object support device does not have a positioning device
Means that the position and rotation angle of the
Absent. In this case, the optical measuring device must be
Use the search template described in the next section in addition to the
It is necessary to make things. (17) The two image processing devices are separated by a certain distance.
Search point having a plurality of point pairs
Search template data record that stores plate data
Search of storage means and its search template data storage means
The template is stored in the image data storage
When superimposed on the screen where the image to represent exists,
Difference in optical characteristic value of each pair of points that make up a point pair
If the state is greater than or equal to the set state,
Are edges where the optical characteristics of the image change suddenly
Of the plurality of poi
If more than the set amount of
The imaging target is a search target that matches the search template.
(15) or (16) including a determination means for determining that there is
The optical measuring device according to claim (claim 10). Optics of this embodiment
Expression of image data
The screen on which an image exists '' is also the optical measurement device described in the previous section.
Physical screen or real screen as in
It may be a screen or a virtual screen. Many edges
In most cases, it is a closed curve, and one of the two points of each point pair is
Will be located inside the edge and the other will be outside the edge.
You. So if you don't need strictness,
The expression “inside, outside edge” is used, but the edge is
When strictness is required because it may not be a closed curve
States that the two points of each point pair are located on either side of the edge.
The expression "ru" is used. Search template de
Data is displayed on the screen of the two points that make up the point pair.
Data that defines the location to search
On the screen where the image represented by the image data exists. ''
If the screen is a physical screen, the search template data
The image data of the pixel at each position specified by
Data from the data storage means to obtain the optical characteristic values.
is there. If the screen is a virtual screen, the search template
Assuming that the rate is on the virtual screen,
Optical characteristics of each point on the virtual screen specified by the
That is, the gender value is calculated from the physical image data.
The physical image data in this case is the electrical
Analog data or digital data representing the magnitude of the signal itself
Data or grayscale data.
There is no meaning in coded data. As mentioned earlier, the smell at the edge
One point of a point pair
Is located within the edge of the object to be imaged, and the other point is outside the edge.
(In the background), in other words, one of the point pairs
The image data corresponding to the point
Yes, and the data corresponding to the other point is the data representing the background
If present, the optical properties of the two points that make up the point pair
The difference state of the value should be higher than the setting state,
The number of pairs of point pairs whose status is not
Is less than a preset number (or set rate)
In some cases, it is determined that the imaging target is a search target.
Can be specified. Also, for such point pairs
By searching for a set, the image of the object
You can look for it. `` The optical characteristic value is over the set state ''
Indicates that the difference in luminance, hue, etc. is
And the ratio of color and hue are equal to or greater than a set value. search
Obtain optical characteristic values by overlaying the template on the physical screen
Image data is analog data or gradation data (hereinafter
Below, referred to as gradation data).
Is also good. In the case of gradation data, etc., two point pairs
The difference between one point value and the other point value is greater than or equal to the set value
That the ratio of both values is greater than or equal to the set ratio
The difference state of the optical characteristic values is equal to or greater than the set state.
You. If it is binary data, the two points of the point pair
If one value is 0 and the other value is 1, optical characteristics
The value difference state is equal to or greater than the set state. Imaging pair
If the position of the elephant (search object) is almost constant,
The search template is superimposed on the screen at a certain position.
It is sufficient that at least the position and the rotation angle of the imaging target are
If one of them is indeterminate, the determination means determines whether the search template
At least one of the position and rotation angle of the
Until affirmative or a predefined change limit is reached
The determination is repeated while changing. Search template
When the position is changed, multiple positions and rotation angles differ.
Even if the search template is prepared in advance, the standard position and
Only the standard search template with the standard rotation angle is prepared,
General position and rotation angle search templates are needed
Alternatively, it may be created by coordinate transformation. "Change limit" and
For example, a search area for searching the image processing target part is set.
Change the position of the search template if it is
Then the search template protrudes from the search area
Is reached, or the position of the search template
That is, the number of times has been changed. Of this embodiment
One of the preferred fields of application of optical measuring devices is the
This is a dimensional inspection of the object to be measured. In dimension inspection,
The size of the imaging target as a fixed target is almost constant
Therefore, it is suitable for using the search template. Shooting
Positions the object to be measured at a fixed position with the image object support device
If it is supported, the search
Easy to stack plates, especially efficiency
Inspection can be performed well. Optical measurement of this embodiment
Another preferred application of the device is in electrical circuit assembly lines.
You. For example, in an electrical component mounting device,
Electrical components and printed circuit boards mounted on materials to be mounted such as boards
The reference mark, etc. provided in the
Screen printing press, the reference mark provided on the screen
A mark, a printing through hole, or the like is an imaging target. Also, imaging
The target object is not limited to the outline of the entire object,
Sometimes it is. For example, for a square chip
The object itself is the object to be imaged, but leads, solder bumps, etc.
For electrical components with
In addition to the main body having the
And one solder bump may be the target of imaging.
You. Leads located outside the contour of the main body, as well as inside
Of other members such as leads, solder bumps, etc.
The object located on the side may be the object to be imaged.
You. In the latter case, the search template is the outline of the entire object
Is created by searching for the outline of the object located inside
Along with the physical or virtual screen
Inside the outline image of the whole body,
Overlap the position where the physical image or virtual image is to be formed.
The user is searched for the object to be imaged. It is clear from the above description
As described above, according to this aspect, the image data and the search
The presence or absence of the search object and the
It is possible to know whether or not the part is a search target part. Point
Calculation and comparison of optical characteristic values of two points constituting a
Can be determined by performing the number of point pairs.
And the processing can be performed easily and quickly. Ma
Both search and measurement templates are used.
If the object to be imaged is
If it is determined by one of the
Search template data and master measurement template
The imaging object determined to be the search object based on the data
A measurement template for measuring
Can be set to accurately measure the measurement target (imaging target)
An optical measuring device that does not need to be positioned at a predetermined position is obtained. (18) The image processing apparatus further includes the search template.
Is a search object definition data defining the search object.
Search template automatically set based on data
(17) The optical measuring device according to the above (17), comprising a determining means. Search target
If the position and rotation angle of the object are almost constant, search
One search template based on search target data
What is necessary is just to set it. However, the position of the search object
When at least one of the setting and the rotation angle is not constant
Indicates that the non-constant is in the reference state.
The master search template is set
The cable template has a fixed position and rotation angle
Not corresponding changes have been made to at least one
Need to make sure that the search template is set
Becomes In this embodiment, the search template is automatically
Since it is set, there is no need for the operator to set it,
An easy and highly efficient optical measuring device can be obtained. (19) The image processing apparatus further includes the search template.
Automatically set the measurement template based on the report
Item (17) or (18) including measurement template automatic setting means
An optical measuring device according to the item. Search target position and time
If the turning angle is almost constant, the search template
Because it is uniquely determined, the measurement template is also uniquely determined.
You. However, at least the position and rotation angle of the search object
If one of them is not constant, multiple search templates
And any one of them will
It is determined that the object is an elephant. Therefore, the measurement template
Is a search that has been determined that the imaging target is a search target.
It will be automatically set based on the template. Yes
Effective measurement templates can be easily obtained,
The position, rotation angle, and dimensions of the elephant (imaging target, search target)
It becomes possible to quickly perform measurement such as a method. (20) The image processing apparatus further comprises one measuring template.
Position, rotation angle and dimensions of the measuring object using the rate
For re-measurement based on at least one measurement result of
Re-measurement template
Includes automatic re-measurement template automatic setting means
The optical measurement device according to (19). One measuring template
The position, rotation angle, and dimensions of the
Remeasurement template based on at least one measurement result
Is set, the remeasurement template
Less deviation from measurement object than one measurement template
Measurement template with small deviation
Higher accuracy can be obtained by performing the measurement by using (21) The imaging device includes a large number of photoelectric conversion devices,
Generates electric signals according to the light receiving state of each photoelectric conversion element
The image data storage means stores
The data of the electric signal of the conversion element is transferred to the position of each photoelectric conversion element.
And the image processing
The apparatus is further provided with the image data in the image data storage means.
On the virtual screen assumed corresponding to the physical screen to be formed
A point designating means for designating an arbitrary point;
Each time a point is specified, the data of the optical characteristic value on the physical screen
A virtual point that calculates the optical characteristic value of a specified point based on
Any one of the above items (16) to (20) including data calculation means
The optical measuring device according to any one of the above. Any of (16) to (20)
In the image processing apparatus according to
When an image is captured, the imaging surface is usually
Image data is created for each element
Stored in the data storage means. Image represented by image data
The image is a pixel unit, and the optical characteristic values are obtained for each pixel.
Image may be used, or optical characteristics may not be used for each pixel.
An image whose gender value is obtained at an arbitrary point may be used. This responsibilities
Instead of obtaining the image data of the desired point in advance,
If any point is specified, the optical characteristics of that point only
The gender value can be obtained by calculation each time a point is specified.
It is possible. The image processing apparatus according to this embodiment uses the virtual image
In this mode, the storage screen is relatively small.
Quick and high accuracy while using image data storage means
Measurement of the object to be measured and defect inspection of the object to be inspected
Can be. (22) The edge point coordinate calculation means is configured to
Acquire the optical characteristic values of multiple division points set on the
Dividing point characteristic value acquiring means, and the dividing point characteristic value acquiring means
Based on the optical characteristic values of the division points obtained by
The sharpest change point of the optical characteristic value on the Klein
Edge point searching means for searching as an edge point (16)
The optical measuring device according to any one of items (21) to (21). (23) The pitch of the plurality of division points is equal to the photoelectric conversion element.
Item 22. The optical measuring device according to item 22, which is smaller than the pitch. (24) The imaging described in any one of the above (1) to (14).
Device and an image representing an image captured by the imaging device
By processing the data, it is possible to detect defects in the imaging object.
And an optical inspection device including an image processing device that outputs the image data.
1). According to this aspect, the defect of the imaging target is highly reliable.
Thus, an optical inspection device that can be inspected is obtained. (25) The image processing device captures an image using the imaging device.
Image data that stores the image data that is the data of the
Data storage means and the image data stored in the image data storage means.
The optical characteristics of the image represented by the image data change suddenly
Negatives that are not expected to intersect the edges that are parts
Negative seek line setting
And the set negative seek line
In the case where the object intersects with the
The optical inspection device according to item (24), comprising:
(Claim 12). According to the optical inspection device of this aspect,
Exists on the surface of the object supported by the imaging object support device
Defects such as cuts, burrs, etc. on the outer surface of the object, such as scratches and dirt
Is quickly detected using a negative seek line
be able to. The negative seek line crosses the edge
It is a seek line that is not expected to be inserted. So
If the negative seek line intersects the edge, the object
Edges on unplanned parts of the surface or circumference
Exists, presumes that a defect exists
You can do it. In addition, the inspection of the defect
Not only for the entire object supported by the object support device,
It can be performed on a part of the object. The object itself is detected
Even when the object is inspected, part of the object is
In some cases. (26) The image processing apparatus further includes a
Measurement to measure at least one of position, rotation angle and dimension
Means for setting the negative seek line
Of the position, rotation angle and dimensions measured by the
The negative seek line based on at least one
The optical inspection device according to the above mode (25). Negative
The Klein should not intersect the edge as described above
Defects such as scratches and dirt due to the planned seek line
When used for inspection, the position and rotation angle of the inspection object
It is necessary that the degree and size are known. Inspection object
Has the correct dimensions and the correct position and rotation angle
When positioning, measurement means can be omitted.
Function, but the dimensions of the inspection object are incorrect or
When the object to be inspected is not positioned at the correct position or rotation angle
Inspection before setting the negative seek line
Object (It is undecided whether the object is an inspection object
In the state, the position, rotation angle and dimensions of the imaging object)
What is inaccurate needs to be measured. Paraphrase
Then, according to the image processing apparatus including the measuring means, the position,
Defects in the inspection object whose rotation angle, dimensions, etc. are not constant also hinder
It can be inspected without. (27) The image processing apparatus further includes the image data storage
Image light represented by image data stored in storage means
May intersect with the edge where the characteristic changes suddenly
Positive to set scheduled positive seek line
(25) or (26) including the
Optical inspection equipment. Negative seek line and positive
If you use it together with the live seek line, only one will be used
Inspection of objects is less reliable and efficient than
At least one can be improved. (28) The imaging according to any one of the above items (1) to (14).
Device and an image representing an image captured by the imaging device
By processing the data, the position and position of the
And rotation angle, and the obtained position and rotation
To the data of the roll angle and the data for defect detection given in advance.
And an image processing device that detects a defect of the imaging target based on the
An optical inspection device including: Item (15)
Combining the characteristics of the measurement of
An optical inspection device having
The features of paragraphs 6) to (23) can be adopted and
Therefore, the features described in the above (25) to (27) should be adopted.
Can be. In addition to the above-described defect detection means,
To obtain the dimensions of the imaging object by processing the image data
Dimension inspection
May be provided. Attached
In other words, the features related to measurement described in paragraph (15) and those described in paragraph (24)
The inspection features described in this section and the features in this section
Adopt independently of the features related to the imaging device described in (1).
Can be

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, the present invention is applied to an object having a three-dimensional shape (an object to be imaged at the time of imaging, an object to be measured at the time of measurement, and an inspection object at the time of inspection. An embodiment in which the present invention is applied to an apparatus that performs imaging, position, rotation angle, and dimension measurement, dimension inspection, and defect inspection (using an appropriate name) will be described with reference to the drawings. This apparatus includes an imaging device and an optical measuring device according to an embodiment of the present invention, and is an embodiment of an optical inspection device according to the present invention. The optical inspection device includes an imaging device 10 shown in FIGS. 1 to 8, an image processing device 12, a monitor television 14, and an input device 16 shown in FIG. The input device 16 includes a keyboard 18, a mouse 20, a foot switch 22, and the like. The image processing device 12 operates in response to information, commands, and the like input from the input device 16,
The data of the image captured by the image capturing device 10 is processed to measure the dimensions of the inspection object, and the measured dimensions are compared with the reference dimensions to determine the pass / fail. In addition, an inspection is performed to determine whether or not the surface of the inspection object has a defect such as a scratch or dirt, and whether or not the outer peripheral portion has a defect such as a chip. The captured image, the measured dimensions, the determination result, the presence or absence of a defect, and the like are displayed on a monitor television 14 as a display device. Further, it is also possible to connect a printing device for printing the display contents of the monitor television 14 on recording paper as a recording medium to the image processing device 12.

Input of information necessary for operating the image processing apparatus 12 is performed using the keyboard 18 and the mouse 20.
The input can be made from either the mouse 20 or the foot switch 22. In particular, in the case of continuously inspecting a large number of inspection objects, if an operation start command is input by the foot switch 22, both hands can be used to detect the inspection objects. It can be used for other work such as replacement, which is convenient. In addition, as a result of inspection,
When the inspection object is rejected, a buzzer built in the image processing apparatus is activated, and the operator does not need to watch the monitor television 14 at all times. In this respect, it is easy to use.

Next, the imaging device 10 will be described. The imaging device 10 has a configuration conceptually shown in FIG. 8, and specifically has a structure shown in FIGS. 1 to 3, reference numeral 30 indicates an apparatus main body. The apparatus main body 30 is formed by assembling a plurality of members, and has a rectangular box shape as a whole. A first member of the apparatus main body 30 is a vertically rising support wall 32, and a support wall 34 is fixed to one vertical side surface of the support wall 32. The two support walls 32,
A base plate 36 is fixed to the lower end of 34, and a frame 38 is fixed to the center of the support wall 32. A cover member 40 is fixed to the upper ends of the support walls 32 and 34, a cover member 42 parallel to the support wall 32 is fixed, and a box-shaped structure having only one side opened. In FIG. 1, the cover member 42
Are removed, and only the position is indicated by a two-dot chain line. Also,
As is clear from FIG. 2, the support wall 32 is a shallow box-shaped member itself, and has a plurality of reinforcing ribs (not shown) formed therein. 2, the cover member 40 and the beam splitter 72 are omitted.

An image pickup section 50 and a parallel light generating section 52 are formed in the main body 30 of the apparatus. The imaging unit 50 includes an imaging object support device 58 having a frame 38 and two glass plates 54 and 56. A through-hole 60 penetrating the frame 38 in the up-down direction is formed, and two glass plates 54 and 56 are disposed in openings above and below the through-hole 60 (see FIG. 2). is there. Frame 3
8 also has a through-hole 62 penetrating it in a direction perpendicular to the support wall 32 and the cover member 42, and the support wall 32 and the cover member 42 also have openings 64, 66 corresponding to the through-hole 62. Is formed. Parallel light generator 52
Generates parallel light (see FIG. 8) from the bottom to the top through the through hole 60 of the frame 38 by a configuration described later. Therefore, if an object is inserted from the opening 66 of the cover member 42 and placed on the lower glass plate 54, the object is illuminated by parallel light from the back side (lower side). In the present imaging device 10, a silhouette image of an object is captured, and the entire object is an imaging target.
Note that the opening 64 of the support wall 32 is formed to allow insertion of the imaging target into the frame 38 even when the imaging target is long.

The imaging unit 50 includes the imaging object support device 5.
8, a concave mirror 70, a beam splitter 72, an orifice 74, a lens system 76, and a CCD camera 78 (FIG. 2).
Reference). The concave mirror 70 is a parabolic mirror, and reflects light incident parallel to the optical axis toward the focal point. The concave mirror 70 is fixed to the support wall 34, and the pinhole 80 of the orifice 74 (see FIG. 8) is located at the focal point of the concave mirror 70. These concave mirror 70 and orifice 74
The beam splitter 72 is disposed at a position between the two. The beam splitter 72 is supported by the support wall 32 at a position where the center line of the through hole 60 of the frame 38 intersects with the optical axis of the concave mirror 70 at an angle of 45 degrees with respect to the center line and the optical axis. ing. Beam splitter 7
Numeral 2 denotes a part which transmits a part of the incident light and reflects the remaining part. As shown in FIG. 8, the parallel light is generated in the parallel light generating part 52 and passes through the through hole 60 of the frame 38. Among them, the one reflected by the beam splitter 72 enters the concave mirror 70. This light is incident on the concave mirror 70 from a direction substantially parallel to the optical axis of the concave mirror 70, and thus is collected at the focal point of the concave mirror 70.

The condensed light passes through the pinhole 80 of the orifice 74 and spreads radially.
The light is converted into a parallel light by the light source 6. This is because the focal point of the lens system 76 is matched with the focal point of the concave mirror 70. The orifice 74 has a function of allowing only light parallel to the optical axis of the concave mirror 70 to pass through among light incident on the concave mirror 70 (that is, light passing around the object to be imaged) and not passing light in other directions. The function is pinhole 8
The smaller the diameter of 0, the sharper. If the pinhole 80 is a theoretical point having no area, only strictly parallel light passes through the pinhole 80, but in reality, it must have a small area, The light in the direction inclined by a small angle with respect to the light also passes through the pinhole 80, and is substantially parallel to the optical axis of the concave mirror 70, and the CCD 82 as an image sensor of the CCD camera 78 (see FIGS. 2 and 8). An image is formed on the imaging surface of. Therefore, in order to obtain a clear image, the pinhole 80 of the lens system 76 is required.
The CCD 82 is located at the position of the focus opposite to the focus at the position of
It is necessary to arrange. The CCD 82 has a large number of photoelectric conversion elements (charge-coupled elements) formed in a lattice shape, and converts an electric signal corresponding to the intensity of light incident on each photoelectric conversion element to a position of each photoelectric conversion element. It can be taken out.

The orifice 74 and the lens system 76
Are integrated into a common cylinder to constitute an integrated optical system. In the apparatus of this embodiment, as shown in FIGS.
0, 92, and 94, three types of optical systems 96, 98, and 100 having different fields of view, small, medium, and large, are prepared. It is attached. In that sense, hereinafter, the optical system 9 will be described.
6, 98, 100 are referred to as interchangeable lenses. These interchangeable lenses 96, 98, 100 are
2 and is attached to the device main body 30 through
A CD camera 78 is detachably attached by screwing.

The direction of incidence of the parallel light on the concave mirror 70 must be exactly parallel to the optical axis of the concave mirror 70, and the pinhole 80 of the orifice 74 is exactly the concave mirror 7
Since it is necessary to be at the focal position of 0, the tilt angle adjusting device 1 for adjusting the tilt angle of the beam splitter 72
08 and a position adjusting device 110 for adjusting the axial position of the interchangeable lens 96 and the like.

As shown in FIG. 4, the inclination angle adjusting device 108 includes a mounting shaft 112, an eccentric operation member 114, and a long hole 1
16 is included. The mounting shaft 112 is fixed to the support wall 32 at a right angle, and the beam splitter 72 can be rotated around a rotation axis perpendicular to both the center line of the through hole 60 of the frame 38 and the optical axis of the concave mirror 70. I support it.
The eccentric operation member 114 includes a supported shaft portion 118 held rotatably on the support wall 32 about an axis parallel to the axis of the mounting shaft 112 and an eccentric cam portion 120 eccentric to the shaft portion. 120 is fitted into the elongated hole 116 formed in the beam splitter 72. Slot 116
Is formed long in a direction orthogonal to the rotation axis of the beam splitter 72, and the eccentric cam 120 is
6, the beam splitter 72 is attached to the mounting shaft 11 as the eccentric operation member 114 is rotated.
2 is rotated about a rotation axis which is the axis of
Is adjusted with respect to the optical axis. After the angle adjustment, the beam splitter 72 is fixed to the support wall 32 by the bolt 122.

The position adjusting device 110 includes a key 126, a key groove 128, an adjusting screw member 130, an adjusting screw member holding member 132, and the like, in addition to the bracket 102, as shown most clearly in FIG. I have. The key 126 is fixed to the apparatus main body 30 in parallel with the optical axis of the concave mirror 70 to form a guide, and the key groove 128 of the bracket 102 is fitted to the key 126. Also, the bracket 10
2 is fixed to the apparatus main body 30 by a plurality of bolts 134, but in a state where the bolts 134 are loosened, the bracket 102 is allowed to move relative to the apparatus main body 30 while preventing excessive lifting from the apparatus main body 30. I do.
As described above, the movement of the bracket 102 is restricted in a direction parallel to the optical axis of the concave mirror 70. Adjustment screw member 13
Numeral 0 is fixed to the apparatus main body 30 and held rotatably and axially immovable by an adjustment screw member holding member 132 forming a part thereof, and is screwed to the female screw portion 140 of the bracket 102 at the male screw portion 138. ing. Therefore, as the adjusting screw member 130 is rotated,
The bracket 102 moves along the key 126 so that the interchangeable lens 96 held by the bracket 102 is moved.
And the axial position of the CCD camera 78 are adjusted. The pinhole 80 of the orifice 74 is just the concave mirror 70
When the bolt 13 is positioned at the focal point of
4 is tightened, and the positions of the interchangeable lens 96 and the like are fixed. The adjusting screw member may have two male screw portions having different screw directions or pitches, and each of the male screw portions may be screwed with the bracket 102 and the apparatus main body 30.

Next, the parallel light generator 52 will be described. The parallel light generating unit 52 is configured to be plane-symmetric with the imaging unit 50 except that a light source device 146 is provided instead of the lens system 76 and the CCD camera 78. The concave mirror 148, the beam splitter 150, and the orifice 152 are respectively symmetrical with the concave mirror 70, the beam splitter 72, and the orifice 74 of the imaging unit 50 and the horizontal plane passing through the center in the vertical direction of the imaging object support device 58 as a plane of symmetry. It is arranged at the position. further,
The parallel light generating unit 52 also includes the tilt angle adjusting device 1 having the same configuration as the tilt angle adjusting device 108 and the position adjusting device 110.
58 and a position adjusting device 160 are provided.

The light source device 146 includes a halogen lamp 170 as a light source and a glass fiber 17 for guiding the light.
2 and a ball lens 176 as a lens system for condensing light emitted almost in parallel from the tip of the glass fiber 172 to the pinhole 174 of the orifice 152. The light collected by the ball lens 176 passes through the pinhole 174 and spreads radially. The light source device 146 functions as a point light source. Pinhole 1
Some of the light emitted from 74 is split by beam splitter 150
The rest is transmitted through the beam splitter 150 and enters the concave mirror 148. Since the pinhole 174 is located at the focal point of the concave mirror 148, the light emitted from the pinhole 174 is converted by the concave mirror 148 into parallel light parallel to the optical axis of the concave mirror 148. The parallel light is reflected by the beam splitter 150 toward the imaging object support device 58.

Since the present imaging apparatus is constructed as described above, the parallel light generated in the parallel light generating section 52 moves the imaging object supporting device 58 from bottom to top in parallel with the center line of the through hole 60. Pass through. Therefore, if the imaging object 180 (see FIG. 8) is placed on the glass plate 54 of the imaging object supporting device 58, a part of the parallel light will be part of the imaging object 180.
And the rest passes around the imaging object 180.
The light passing through the CCD camera 7 as described above
An image is formed on the eight CCDs 82, and this image includes an image of the dark object 180 and an image of a light background. Since a silhouette image of the imaging target 180 is obtained, and a bright image of the background is formed by parallel light passing around the imaging target 180, the silhouette image of the imaging target 180 is drawn by the perspective projection method or the projection projection method. Dimensions exactly the same as the figures in the drawing. As shown in FIG.
When the imaging target 180 is imaged by the CCD 184 through the normal optical system 182, an image of a part of the imaging target 180 far from the optical system 182 is imaged smaller than an image of a near part. If the portion far from the optical system 182 is located at the same position as the other portions as indicated by a two-dot chain line, a larger image is obtained but a smaller image is taken. It is. This problem is caused by the optical system 18 of the imaging object 180.
2 is generated unless the plane opposite to the optical axis of the optical system 182 is a plane, and even if it is a plane, the plane is generated unless the position of the plane is constant. Dimension measurement becomes impossible. On the other hand, according to the imaging apparatus 10, this problem is solved, and the same image as the drawing drawn by the perspective projection method or the projection projection method is obtained even with the three-dimensional imaging object 180. If the dimensions are obtained based on the dimensions, the dimensions in the perspective projection or the projection can be obtained.

An imaging device 190 shown in FIG. 9 can be used in place of the imaging device 10 described above. The imaging device 190 includes a parallel light generation unit 52 in the imaging device 10 and a uniform illumination device 192 that is a type of a bright background forming device.
It is changed to. The uniform illumination device 192 is, for example,
A light source in which a large number of light emitting diodes are arranged on a grid may be covered with a milky white diffusion plate. Light emitted from a large number of light emitting diodes is diffused in the diffusion plate, and is uniformly emitted in all directions from the surface of the diffusion plate. Therefore, the light passing around the imaging object 180 includes component lights in various directions. However, as described above, the orifice 74 having the pinhole 80 moves the concave mirror 70 parallel to its optical axis. Since the CCD 82 as an image sensor has a function of selectively passing component light incident from the direction, the CCD 82 as the image sensor is exactly the same as the figure of the drawing drawn by the perspective projection method or the projection method. Dimensions.

It is also possible to use the imaging device 200 shown in FIG. The imaging device 200 is the same as the imaging device 1
The imaging object 180 (that is, the imaging object support device), the orifice 74, the lens system 76, and the CCD at 90
The position of the camera 78 is changed. In the imaging device 190, the imaging object support device is disposed in the reflection direction of the beam splitter 72 when viewed from the concave mirror 70 side, and the orifice 74 and the like are disposed in the transmission direction.
In the imaging device 200 of FIG. 10, the imaging object support device is disposed in the transmission direction of the beam splitter 72, and the orifice 74 and the like are disposed in the reflection direction. With the imaging device 200 having this configuration, it is possible to obtain an image in the perspective projection method or the projection projection method.

It is also possible to use the image pickup device 204 shown in FIG. This imaging device 204 is used to mount an electric circuit on a printed wiring board to assemble a printed circuit board. In an electric component mounting system, the position of an electric component 208 as an imaging target held by a suction nozzle 206 serving as an electric component holding member. And for measuring the rotation angle. The suction nozzle 206 includes a fluorescent plate 212 that emits visible light upon receiving ultraviolet rays from the ultraviolet radiator 210 (a diffuse reflection plate that diffusely reflects visible light may be used). The imaging device 204 uses a fluorescent plate 212 (or a diffuse reflection plate) as a light background forming device, and acquires an accurate silhouette image of an imaging target (electric component 208).

Although all of the imaging devices described above with reference to the drawings obtain a silhouette image of an object to be imaged, the parallel light generating section 52, the uniform illumination device 192, and the fluorescent screen 2
Instead of 12 or the like, it is also possible to illuminate the imaging target object from the front side with a lighting device to obtain a front image. In one example, in the imaging device 204 of FIG. 11, the ultraviolet radiator 210 is changed to a ring light that emits visible light. By providing both the ultraviolet radiator 210 and the ring light and selectively using them, it is also possible to obtain both a silhouette image and a front image of the electric component 208 as an object to be imaged. If a front image of the object to be imaged is obtained, it is possible to obtain an image of a defect such as a concave portion, a protrusion, a character printed on the surface, and a flaw or dirt generated on the surface.

Next, the image processing apparatus 12 will be described. The image processing device 12 processes the data of the image acquired by the CCD camera 78 as described above, and based on the outlines inside and outside the image pickup object 180, the dimensions, the position, and the reference of the image pickup object 180. Obtain the rotation angle from the rotation position. Further, based on the contours of the characters, figures, recesses, scratches, dirt and the like existing on the surface of the imaging target 180, the characters, figures, recesses and the like are identified, and defects such as scratches and dirt are detected. Since this image processing is complicated, it will be collectively described later in detail and omitted here. The image processing device 12 further calculates the size, position,
The data such as the rotation angle is compared with the reference data to make a pass / fail judgment, and the result of the judgment is displayed on the monitor television 14 together with the obtained values of the dimensions, position, rotation angle and the like. In addition, the monitor television 14 also displays the identification result of a character, a figure, a concave portion, or the like existing on the surface of the imaging target 180, and the detection result of a defect such as a scratch or dirt.

The image processing apparatus 12 is designed to hold an electric component supplied from an electric component supply device by a component holding head such as a suction head, transport the electric component, and mount the electric component on a circuit substrate such as a printed circuit board. It was developed as an image processing device for a component mounting device. In the electric component mounting apparatus, the holding position and holding direction of the electric component by the component holding head (represented by the rotation angle from the reference rotation position.
Detecting the rotation angle), or detecting the position and rotation angle of the circuit substrate supported by the substrate supporting device, correcting the errors, and then mounting the electric component on the circuit substrate. For this purpose, an image pickup device for picking up an image of a reference mark on an electric component or a circuit substrate is required, and the image pickup device has been developed as an image processing device. However, the position, rotation angle,
It can also be used for measuring and inspecting dimensions and inspecting for defects.

The image processing apparatus 12 is mainly composed of a computer, and as shown in FIG.
It has a DRAM (dynamic ram) 256, an SRAM (static ram) 258, a PROM (programmable rom) 260, a kanji ROM 262, a frame grabber memory 264, and an overlay display memory 266 for four planes, which are not shown on the substrate 267. They are connected to each other by an internal bus.

A two-channel serial interface 270 is connected to the internal bus.
Is connected. The input device 16 mainly includes the keyboard 18 having numeric keys, alphabet keys, and the like. The input device 16 includes information and commands necessary for the operation of the entire optical inspection apparatus, and types of the image processing target (the entire object or a part of the object). This is a device for inputting information necessary for image processing, such as the number, number, and the like. The bus also has an Ethernet interface 27
4 and a memory card interface 276 are connected.

The Ethernet interface 274 is an interface for performing communication with a computer that controls parts other than the image processing apparatus, such as an electric component mounting apparatus. An optional control device can be connected to the image processing apparatus 12.
Performs data exchange via the P1 connector 268. Further, a control device for controlling various driving devices of the electric component mounting apparatus is also a computer-based image processing apparatus 1.
2 and P1 via an external bus (not shown).
Connected to connector 268. Since this other control device has little relation to the present invention, its illustration and description are omitted. The memory card stores a program created in advance for performing image processing. When the memory card is set in the image processing apparatus 12, the CPU 254 uses the PROM 260 to execute programs and programs in the memory card. The necessary data is read out via the memory card interface 276 and stored in the DRAM 256.

Further, a CCD camera interface 280 is connected to the bus.
The D camera 78 (or the CCD camera 78 of the electric component mounting device) is connected. Image data which is data of an image obtained by the CCD camera 78 is stored in the frame grabber memory 264 via the CCD camera interface 280. As described above, the frame grabber memory 26
4 are provided, and, for example, image data of four imaging targets 180 that are continuously imaged are sequentially stored in each frame grabber memory 264.

The bus further includes a television interface 28.
6 is connected, and the monitor television 14 is connected. The monitor television 14 is capable of both color display and monochrome display. As described above, the image data of the four monochrome images obtained by imaging the imaging object 180 is stored in parallel in the frame grabber memory 264, while the overlay display memory 266 stores the There are four memories each capable of storing color image data to be displayed in 16 colors. The monitor TV 14 displays one of the four monochrome images with the four
Among the color images for the planes, those corresponding to the monochrome images are displayed in a superimposed manner, and the progress and results of the image processing are displayed.
Data input using the input device 16 is also displayed in color on the same monitor television 14. At the time of this display, the kanji ROM 262 is used.

Hereinafter, the processing of image data obtained by the imaging by the CCD camera 78 will be described. The programs and data for image processing are stored in the memory card as described above, and are read out and stored in the DRAM 256 when the memory card is set. 14 to 16 show an image processing program read from the memory card.
Are shown below. By executing these programs, measurement of dimensions, positions, rotation angles, and the like of the image processing target (imaging target), inspection of the dimensions, and inspection for the presence or absence of a defect are performed.

The program shown in FIG. 14 is a pre-processing program. The pre-processing program is executed at the start of one inspection program, that is, the DRAM of the pre-processing program
It is executed after storing in 256. First, it is determined whether or not to perform pattern matching on one of all image processing objects necessary for executing one inspection program, and if so, a search template is generated based on the master search template. Stored in the DRAM 256. Similar processing is sequentially performed on all image processing objects.

The master search template has a plurality of point pairs each including two points. Is the same as That is, the origin and the direction of the coordinate axis of the master search template coordinate plane match the origin and the coordinate axis of the reference coordinate plane in which the origin is set at the center of the field of view of the CCD camera 78. The master search template is created in advance based on the shape and dimensions of the image processing target, is stored in the memory card, and is read into the DRAM 256 together with the preprocessing program.

FIG. 17 shows an example of setting data of the master search template when the image processing object is a square portion, and FIG. 18 shows a master search template 300 set by the data. In the data of FIG.
Data on lines 7, 8, 10, 11 and hs on line 5
(Half span) = 5.5 is the setting data of the master search template. Pair means that another point pair is set symmetrically with respect to the center line of the image processing object on an extension of two points forming the point pair. For example, for each point pair 302 of (7), (8), (10), and (11) shown in FIG. 18, (7) ', (8)', (10) ',
(11) Each point pair 302 is set. The numbers in parentheses given to these point pairs 302 match the line numbers in FIG. Also, 30
Reference numeral 4 denotes an image processing target.

The master search template has dimensions, positions,
For a master image processing object having no error in any of the rotation angles, one of the two points forming each point pair is located inside the image processing object, the other is located outside, and the points It is created such that the midpoint of the two points of the pair is located on the edge (contour) of the master image processing object. If it is shown in the figure, it will be as shown in FIG. In general, it is not essential that the midpoint between the two points of the point pair be located on the edge of the master image processing object, and the two points are located inside the edge of the image processing object. What is necessary is just to specify outside. Also,
In the example shown in FIG. 18, one of the two points of the point pair 302 is shared with one of the points of another point pair 302, but this is not essential.
Further, in FIG. 18, two points constituting the point pair 302 are connected by a straight line in order to easily indicate which point constitutes the point pair 302. , And the data of the straight line is not actually set.

FIG. 21 shows the data of the master search template when the image processing object is the disk 306 of FIG. 22 which is partially cut away. In this master search template 308, (15)-(17)
Is paired with the point pair 310 of (15) ′ to (17) ′, but the other point pair 3
10 is not paired with another point pair.

In the optical inspection apparatus, the image processing object is, for example, the entire image of the imaging object 180, and master search template data for a plurality of types of imaging objects 180 to be inspected is created in advance. It is stored in a memory card, and is stored in a DRAM 25 when image processing is executed.
6 is stored. Therefore, at the time of generating the search template, the imaging object 180 for which the search template is to be generated.
Master search template data is DRA according to the type of
It is read from M256. As described above, the master search template is a search template having a rotation angle of 0 degrees. By rotating the master search template at a set pitch within a set angle range as shown by a two-dot chain line in FIG. A search template is generated and the data is stored in DRAM 256.

The data specifying the generation angle range and the set pitch of the search template are stored together with the data of the master search template as shown in FIG. PitchA = 4.5 on the 15th line is the data of the set pitch, startA = −45 on the 16th line, and en on the 17th line.
dA = 45 is data that defines the search template generation angle range. The angle range and the pitch for generating the search template are set according to the imaging target 180. For example, when it is expected that the rotation angle when the imaging target 180 is placed on the glass plate 54 of the imaging target supporting device 58 is largely shifted, the generation angle range is widened. Incidentally, in the example of the pre-processing program shown in FIG. 14, the generation angle range is from -45 degrees to +45 degrees, and the set pitch is 5 degrees.

When the generation of the search template for one imaging target 180 is completed, the execution of the program returns to the beginning, and it is determined whether or not pattern matching is to be performed for the next imaging target 180, and if so, the search is performed. A template is generated. If the pattern matching is not performed, the execution of the program returns to the beginning, and it is determined whether or not pattern matching is to be performed on the next imaging target 180. If it is determined whether or not pattern matching is to be performed for all of the imaging targets 180, and if a search template has been generated for the imaging targets 180 to be subjected to pattern matching, the execution of the program in FIG. 14 ends.

Next, the execution processing program shown in FIG. 15 will be described. This program is executed after the imaging object 180 is imaged by the CCD camera 78 and the image data is stored in the frame grabber memory 264. First,
If the imaging object 180 can be processed only by pattern matching, such as a square object, image processing is performed according to the pattern matching program shown in FIG.

Next, the object to be image-processed is an electric component having a complicated shape, such as QFP (Quad Flat Package), PLCC (Plastic Leaded Chip Carrier), SOP (Small Outline Package), etc. Ball grid
Array) solder bumps with a large number of image processing objects such as solder bumps, image processing objects with complicated shapes such as solar cell grid patterns, mechanical parts that are not simple circles or squares, and image processing It is determined whether it is necessary to operate a pattern matching manager that combines pattern matching. The combination of pattern matching will be described later.

If the operation of the pattern matching manager is not necessary, it is determined whether or not to perform image processing on the virtual screen. If the result of the determination is NO, whether or not to perform image processing on the physical screen is determined. Is determined. The physical screen is
An optical characteristic value is obtained for each pixel and a screen of an image in which image data actually exists, and a virtual screen is a screen where an optical characteristic value of an arbitrary point which is not constrained by pixels is obtained by calculation as needed. . As described later, both the pattern matching and the pattern matching manager are processes performed on the virtual screen. However, in the present embodiment, in addition to these, on the virtual screen and the physical Image processing can be performed on the screen and on the screen. The determination of “whether or not to perform image processing on the virtual screen” and “whether or not to perform image processing on the physical screen” are instructed to perform pattern matching or image processing of the pattern matching manager. It is a judgment of whether or not

The pattern matching program shown in FIG. 16 will be described. First, a search window is set, and a search area for searching for an image processing object is set. The search window is set by designating a part or the whole of the imaging surface of the CCD camera 78 by coordinate values. The identity of the image processing object is known from the data during the work procedure, and the position of the image of the image processing object formed on the imaging surface is roughly known. Is also set to a size sufficient to include the image processing target. In this way, the search area can be narrow, and the search can be performed in a short time.

The full set of pattern matching processing is as follows:
A search step for searching a search target portion, a re-search step for searching for an approximate edge point of the search target portion, a measurement step for calculating an edge point of the search target portion, a re-measurement step for repeating the measurement step, and an inspection step. Including steps. Normally, pattern matching ends when all five steps end. If there is at least one abnormal step, the next step is not executed, and the pattern matching is immediately terminated.

First, the search step will be described. In the search step, search templates are sequentially read one by one from the DRAM 256, and as shown in FIG. 24, the image 320 of the image processing object and the background If the object is a concave portion or the like on the surface of the component, an image including the image of the surface of the component becomes a background).
Optical characteristic values (luminance in the present embodiment) of two points (hereinafter, referred to as point pair configuration points) constituting the two plurality of point pairs 324 are calculated. In the example illustrated in FIG. 17, the search templates are sequentially read from the search template whose rotation angle is −45 degrees.

The point pair composing points are points on the virtual screen, and the luminance of the point pair composing points is obtained by interpolation from the luminance of a plurality of pixels on the physical screen as image data. The optical characteristic value of a point on the virtual screen designated by the data of the search template is obtained based on the image data of the physical screen, and FIG. 24 visually shows this. This means that "the search template is superimposed on the screen on which the image represented by the image data exists". The screen 321 in FIG. 24 is a virtual screen, and the image 320 of the image object in the screen 321 is only imagined to exist at this position, and there is no actual image data representing the image 320. The same applies to other figures.

Interpolation calculation of the brightness of the point pair composing points is as follows.
For example, it can be performed using a curved surface such as a bicubic spline curved surface defined by image data of 4 × 4 control points on the X and Y coordinate planes. The simplest linear interpolation is performed based on the image data of four pixels adjacent to the constituent point. In FIG. 25, (u ₀ , v ₀ ) is the point pair composing point, f (u ₀ , v ₀ ) is the luminance of the point pair composing point,
(U ', v'), (u '+ 1, v'), (u ', v' +
1) and (u '+ 1, v' + 1) are the center positions of four pixels used for linear interpolation, f (u ', v'), f
(U '+ 1, v'), f (u ', v' + 1), f (u '
(+1, v ′ + 1) is the luminance of each of the four pixels, and the luminance of the point pair composing point is calculated by equation (1). f = (u ₀ , v ₀ ) = f (u ′, v ′) (1−α) (1−β) + f (u ′ + 1, v ′) α (1−β) + f (u ′, v ′) +1) (1-β) β + f (u ′ + 1, v ′ + 1) αβ (1)

The above operation is performed by the physical screen / virtual screen conversion driver 380 shown in FIG. As shown in the figure, the physical screen / virtual screen conversion driver 380 is configured separately from general image processing application software 382, and is based on image data on the physical screen 384 in the image processing application software 382. Every time it is necessary to calculate image data on the virtual screen 386, the physical screen / virtual screen conversion driver 380
Is called to calculate the image data on the virtual screen 386.

Each time the luminance is calculated for the two constituent points of each pair of point pairs 324, the luminance of these two point pair constituent points is compared. At the time of capturing a silhouette image by the CCD camera 78, a difference occurs in the charge amount of the solid-state imaging device between a portion corresponding to the image processing target (the imaging target 180) and a portion corresponding to the background. The image of the image processing object is dark and the background is light. Therefore, 2
If one of the point pair constituent points is located within the edge of the image processing target object and the other point is located outside the edge, the luminance value of the two point pair constituent points is equal to or greater than a preset value. A difference occurs when the set value is positive or less than the set value (when the set value is negative).

The set value of the luminance difference is stored together with the master search template data. For example, in FIG. 17, the setting value diff is set to 20, as described in the fifth row. In this case, if the luminance of the point pair composing point inside the edge of the image processing target object is smaller than the luminance of the point pair composing point outside the edge by 20 gradations or more among the two point pair composing points, the setting value is exceeded. Are determined to be different. Conversely, if the setting value diff is set to -20, the brightness of the point pair composing point outside the edge of the image processing object is smaller than the brightness of the point pair composing point within the edge by 20 gradations or more. It is determined that there is the above difference. In any case, these two point pair constituent points straddle the edge of the image processing target, and are in a suitable state. In this case, the expression “two point pair constituent points satisfy the set luminance difference condition” will be used.

(1) The image processing target is not the search target portion, and the two point pair constituent points do not straddle the edge. (2) The imaging target 180 is placed on the inspection target support plate 72 due to a supply error. The two point-pair configuration points do not satisfy the set luminance difference condition because they are not mounted or (3) image data cannot be obtained due to dust or the like adhering to the solid-state image sensor There are things you can't say. This state is called a point pair failure. The number of failures for determining that there is no search target portion that matches the search template is set in advance. For example,
In FIG. 17, the number of failures is 0 as shown in the third row.
If two point pair constituent points do not satisfy the set luminance difference condition for all the point pairs, it is not determined that there is a search target portion that matches the search template.

If the number of failures is set to one or more, if the number of point pairs that do not satisfy the set luminance difference condition exceeds the set number of failures among a plurality of point pairs, the search target matching the search template It is determined that the part does not exist. If it is determined that the search target portion exists at a rotation angle of −45 degrees, the search step ends,
A re-search step is performed, but if it is determined that the search template does not exist, search templates having different rotation angles are read and a search target portion is searched.

Until it is determined that the search target portion exists,
A plurality of types of search templates are sequentially read, and a search target portion is searched. Even if the search is performed using all types of search templates, if it is not determined that there is a search target portion that matches the search template, then the search is performed by shifting the position of the search template. The search target portion is searched using a plurality of types of search templates having different rotation angles at each position while being shifted by a constant pitch in the X-axis direction and the Y-axis direction.

The moving pitch is set in advance and stored in the memory card together with data defining the master search template. In FIG. 17, pitchX = 2.2 and pitchY = 2.2 shown in the 13th and 14th rows are movement pitches. First, the set pitch is moved in the Y-axis direction. Specifically, the coordinate conversion is performed such that the coordinates of each point pair of the plurality of types of search templates are shifted from the Y-axis direction in the positive direction by the set pitch. Using this search template, a search of the search target portion is performed. At this position, if it is not determined that the search target portion exists even if all types of search templates having different rotation angles are used, then the search template is shifted in the X-axis direction in the forward direction by the set pitch. Further, if it is not determined that the search target portion exists, the search template is then shifted in the Y-axis direction in the negative direction by the set pitch. If it is not determined here that the search target portion exists, the search template is further shifted in the Y-axis direction in the negative direction by the set pitch, and it must be determined that the search target portion also exists here. Then, the search template is shifted by the set pitch in the X-axis direction in the negative direction. The search template is moved in the search window so as to draw a square spiral.

Even if the search template is moved, it cannot be determined that the search target portion exists. If the coordinate transformation is performed and a point pair protrudes from the search window, the search template cannot be moved. It is determined that there is no search target portion that is possible and matches the search template, and the search step is abnormally terminated. The occurrence of an abnormality is displayed on the monitor TV 14, and
The occurrence of the abnormality is stored. When the image processing object is the inspection object 180, the operator checks the mounting position of the inspection object 180 determined to be abnormal in the image processing result on the inspection object support plate 72, and If the position deviates significantly, the position is corrected. If the position is not largely deviated, the inspection object 180 is determined to be unscheduled and is removed.

As shown in the image 320 of the image processing object shown in FIG. 24, two point pair constituting points of all the point pairs 324 are inside and outside the edge of the image 320, and the set luminance difference condition is satisfied. For example, the position and rotation angle of the search template at that time are stored in the DRAM 256,
A re-search step is performed. In the re-searching step, as shown in FIG. 27, the edge point of the image 320 of the image processing object is changed to the re-searching template coordinate plane (searching template coordinate plane) which is the coordinate plane of the re-searching template 328 by using the re-searching template 328. Is the same as above). The search template 328 includes a plurality of seek lines 330. The seek line 330 is set based on a search template that has found the image 320 of the image processing object in the search step. The two point pair constituting points of the point pair are defined as points defining both ends of the seek line 330, respectively.

An edge point of the image 320 of the image processing object is searched for each of the plurality of set seek lines 330. In this sense, the search template 328 can also be considered a measurement template. This search divides the seek line 330 at a predetermined pitch (for example, 0.05 mm) as shown in FIG.
This is performed by calculating the luminance for each of the plurality of division points P1 to P15. This pitch is CCD
The length is set shorter than the diagonal line of the solid-state imaging device 332 of the camera 78. Therefore, one solid-state imaging device 33
Two or three division points are included in 2.
The division point is also a point on the virtual screen, and linear interpolation is performed as in the search step, and the luminance of the division points P1 to P15 is calculated.

FIG. 29 shows an example of each luminance of the fifteen division points P1 to P15 calculated by the linear interpolation. In addition,
The luminance value is represented by a positive value, and the value increases as the luminance value acquisition target is brighter. The imaging device 10 of the present embodiment irradiates the inspection object 180 as an image processing object with parallel light from the back side, and acquires an image based on the parallel light passing around the inspection object 180. In the following description, it is assumed that the image processing target is dark and has a small luminance value, and the background is bright and has a large luminance value. From the luminance value calculated by the linear interpolation, it is not known where the luminance changes most, as shown in the graph of FIG. Therefore, a differential value of the luminance value is obtained using a difference filter. FIG. 33 is a graph showing the result of obtaining a differential value using the difference filter shown in FIG. This difference filter sets the brightness of a point located on the upstream side of two adjacent points from one point defining the seek line to the other point to a negative value, and sets the point located on the downstream side to a negative value. This is a filter that takes a positive value and calculates the sum of those two values. This differential value is not the value of the division point. In the graph of FIG. 33, the position at which the luminance differential value is obtained is indicated by ".5" which is the center position of the adjacent two division points. As is clear from this graph, the magnitude of the luminance change can be determined, but it is not clear which is the maximum. It should be noted that depending on the calculation direction, that is, whether the calculation is performed from the division point inside the image processing object to the division point outside the image processing object (inside the background) or vice versa, The sign of the value is reversed. In the former case, the luminance differential value is a positive value, and the position where the luminance differential value is the maximum is the position where the absolute value of the luminance change gradient is the maximum (the luminance differential value at the position where the absolute value of the change gradient is the maximum) Maxima). In the latter case, the luminance differential value is a negative value, and the position where the luminance differential value is the minimum is the position where the absolute value of the luminance change gradient is the maximum (the luminance differential value at the position where the absolute value of the change gradient is the maximum) Called the local minimum). FIG. 33 and FIG. 34 described next.
Is a value obtained by the former calculation.

On the other hand, if differentiation is performed using the difference filter shown in FIG. 31, f
A maximum value 177 of the luminance differential value is obtained at the position 8.5, and it is understood that this position is the position where the absolute value of the luminance change gradient is the maximum. The difference filter shown in FIG. 31 sets one of the four division points on the upstream side in the calculation direction including one of the division points set on the seek line to a negative value. , A filter for obtaining the sum of the luminance values of the four continuous division points on the downstream side as positive values.

If dirt or the like adheres to the portion corresponding to the edge point of the solid-state image pickup device and the charge amount changes, a maximum value or a minimum value of the luminance differential value may be obtained at a position other than the edge point. However, the absolute value of the luminance change gradient at such a position is small. On the other hand, in the vicinity of the edge point, there is a remarkable difference in brightness between the image processing target and the background, and the absolute value of the luminance change gradient is large. Therefore, a set value is provided, and it is determined whether or not the luminance differential value at the position where the absolute value of the change gradient is the largest is a value obtained in the vicinity of the edge point, and the former case is excluded. This setting value is set as a positive value when the luminance derivative value is obtained as a positive value, and it is determined whether or not the maximum value of the luminance derivative value is equal to or greater than the set value. If, the local maximum value is a value obtained at a position near the edge point, and it is determined that the maximum value can be used for the calculation of the edge point, and the calculation of the edge point is performed. If the luminance derivative is obtained as a negative value, the set value is set as a negative value, and it is determined whether or not the minimum value of the luminance derivative is equal to or less than the set value. If it is below, it is determined that the minimum value can be used for the calculation of the edge point. In other words, the failure of the seek line in the re-searching step is that the maximum value of the luminance differential value is smaller than the set value or the minimum value is larger than the set value and the edge point is not calculated.

In the example shown in FIG. 17, similarly to the examples shown in FIGS. 32 and 33, the calculation of the luminance differential value is performed from the division point inside the image processing object to the division point outside. Is set at the position where the absolute value of the change gradient of the brightness value is the largest, and the setting value for determining whether or not the maximum value is a value obtained near the edge point is , A positive value, that is, 11 = 200 as shown in the fifth row. In this example, since the image processing object is darker than the background, the luminance differential value is obtained as a positive value, and if the value is not 200 or more, the calculation of the position of the edge point is not performed. I have.

Also, as shown in FIG. 17, it is predetermined in advance what kind of difference filter is to be used for the calculation. This difference filter coefficient N is calculated according to the equation (2). N = gUnit / pitch between division points (2) where gUnit is the length of a diagonal line of the solid-state imaging device.

If the maximum value (or the minimum value) of the luminance differential value is obtained by performing differentiation using a difference filter, the position where the absolute value of the luminance change gradient is the maximum, that is, the edge point is obtained according to the following equation. Can be Expressions (3) and (4) are expressions taking the case of N = 4 as an example, and f
_max , f _{(max-4) to} f _(max-1) , f _{(max + 1) to} f
_{(max + 4)} is the luminance differential value (f _max is local maximum (small))
Value). “F” is, as shown in FIG. 34, numbered to indicate the acquisition position of the luminance differential value on the seek line, and is designated by the number added to f in the equations (3) and (4). Represents the luminance differential value of the position to be set. f _max is a position where the luminance differential value is maximum (small) (in the example shown in FIG. 33, f 8.
5) is the luminance differential value of f _(max-1) , f _(max-2) , f
_(max-3) and f _(max-4) are respectively f
The four locations upstream of _max (f7.5 in the example shown in FIG. 33,
f6.5, f5.5, f4.5).
_{(max + 1)} , f _{(max + 2)} , f _{(max + 3)} , f _{(max + 4)} are respectively four locations downstream of f _{max in} the calculation direction (FIG. 3
In the example shown in FIG. 3, the luminance differential values are f9.5, f10.5, f11.5, and f12.5). dl = _fmax × 4− (f _(max−1) + f _(max−2) + f _(max−3) + f _(max−4) ) (3) dr = _fmax × 4− (F _{(max + 1)} + f _{(max + 2)} + f _{(max + 3)} + f _{(max + 4)} ) (4) edgePitch = (dl × N) / (dl + dr) −N / 2. ... (5) Edge point = (Luminance value local maximum (small) value point pitch number + edgePitch) × division point pitch... (6) N = 4 in equations (3) and (4) Is the expression for
In general, when dl is obtained, the luminance differential value of N points upstream of the local maximum (small) value point in the calculation direction is calculated from the value obtained by multiplying the local differential value of the local brightness value by N. Is subtracted and dr is obtained, the value obtained by multiplying the luminance differential value of the local maximum (small) value point by N is N from the value downstream of the local maximum (small) value point.
The sum of the luminance differential values of the points is subtracted. When the edge point is obtained when the calculation result shown in FIG. 29 is differentiated by using the difference filter shown in FIG. 31, the pitch number of the maximum (small) value of the luminance differential value in equation (6) is 8.8. 5

When calculating the edge point, first, the luminance of the division point is calculated by linear interpolation, the differentiation is performed according to the difference filter coefficient N, and then the calculation is performed according to the equations (3) to (6). The maximum change position, that is, the edge point is obtained. In the case of the seek line 330 shown in FIG.
The calculation result of the equation (6) is 0.403 mm, and it is understood that the edge point is located 0.403 mm from the division point P1 of the seek line 330.

In this manner, a plurality of seek lines 33
An edge point is computed for each of the zeros. The number of failures of the seek line 330 (if the number of failures of the point pair 302 is less than or equal to the set number, if it is determined that there is a search target portion that matches the search template, the number of failures of the point pair 302 If the sum of the number and the number of failures of the seek line 330) is equal to or smaller than the set number, it is determined that the operation is normal, and the measurement step is executed. The processing when an abnormality occurs is the same as the search step. In FIG. 17, the fail Count is set to 0 as shown in the third row, and if there is at least one failure, the re-search step is abnormally terminated.

If the number of failures is equal to or less than the set number and the re-search step is completed normally, the measurement step is executed next. The re-search template is set based on the search template determined to have the search target portion in the search step, and an edge point can be found on the seek line 330, but the edge point and the seek line 3
Usually, there is a deviation from the midpoint of 30 (shown with an X mark in FIG. 27 and hereinafter referred to as an ideal point). The two points forming the point pair are set such that the middle point of the two points is located on the edge of the search target portion if there is no deviation in the size, position, and rotation angle of the search target portion. Although the ideal point and the edge point should match, in reality, the object to be image-processed has a deviation, and a deviation occurs between the ideal point and the edge point obtained by the calculation.

For this reason, if the re-search step is executed without any abnormality, the measurement step is executed next, and the position of the edge point is calculated. In the measurement step, first, FIG.
The measurement template 336 as shown in FIG. 5 is automatically set. The measurement template 336 includes a plurality of seek lines 3
38, data of a preset master measurement template, data of a relative position of the re-search template coordinate plane with respect to the reference coordinate plane in the re-search step, and calculation of an edge point on the re-search template coordinate plane. It is set based on the result.

As shown in FIG. 17, the master measurement template data is stored together with the master search template data and the like. The data in the 20th to 33rd rows in FIG.
The data of one line hs = 3.5, lines 23 to 27 and lines 29 to 33 are master measurement template data. FIG. 19 shows a master measurement template 340 obtained from this data. 342 is a seek line.
The master measurement template 340 has more seek lines than the master search template for the same electrical component. It should be noted that the image processing object is a disk 3 partially cut away.
In the case of 06, a master measurement template 344 having a plurality of seek lines 346 is set as shown in FIG.

Some or all of the seek lines of the measurement template are paired. Another seek line is set on the extension of the seek line symmetrically with respect to the center line of the image processing object. These paired seek lines are referred to as pair seek lines. The measurement template is set by coordinate conversion of the master measurement template data. The coordinate plane of the re-search template (the coordinate plane of the re-search template is the same as the coordinate plane of the search template of the search template when the image processing object is determined to be the search target portion that matches the search template in the search step. ) Relative position and relative rotation angle with respect to the reference coordinate plane, and the relative position and relative rotation angle of the image processing object with respect to the re-search template coordinate plane (these are based on the coordinate values of the edge points calculated in the re-search step). The calculation is performed, and this calculation will be described later.) This is set by performing a coordinate conversion corresponding to the master measurement template data (set on the master measurement template coordinate plane coincident with the reference coordinate plane). is there.

When the automatic setting of the measurement template is completed, the calculation of the edge point on each seek line of the measurement template is performed in the same manner as in the re-search step. Division points are set at a constant pitch on the seek line, luminance is calculated by linear interpolation for each division point, a luminance differential value is calculated using a difference filter, and an edge point is calculated. Also in the measurement step, the number of allowable failures is set. A failure here means that no edge point is calculated for the seek line, as in the re-search step. If the number of failures is equal to or less than the set number, it is determined to be normal, and then a re-measurement step is executed. If the number of failures exceeds the set number, the measurement step is abnormally terminated. The processing when an abnormality occurs is the same as in the search step.

In the re-measurement step, a re-measurement template is set, and an edge point is calculated. The re-measurement template is automatically set based on the measurement template and the edge points calculated in the measurement step. Based on the edge points obtained in the measurement step, the measurement template is rotated by coordinate transformation to a position where the ideal point is expected to be located on the edge point. The calculation of the edge point in the re-measurement step is performed in the same manner as in the re-search step.

The determination of abnormality in the re-measurement step uses the allowable number of failures set for the measurement step. If the number of seek lines for which edge points cannot be obtained is larger than the set number, it is abnormal and the image processing is terminated. . The processing when an abnormality occurs is the same as in the search step. If the number is equal to or smaller than the set number, the process ends normally, and then the object vector, that is, the size, position, and rotation angle of the image processing object are calculated. The greater the number of times the re-measurement step is performed, the smaller the difference between the ideal point and the edge point, and the detection accuracy of the edge point is improved. The set number of re-measurement steps is set and stored in advance. The re-measurement template used for executing the second and subsequent re-measurement steps is automatically set based on the re-measurement template at the time of the immediately preceding re-measurement step and the calculation result of the edge point in the re-measurement step.

If the re-measurement step is completed normally, the difference between the measured dimension of the image processing object and the reference dimension is calculated, and the calculation result is compared with the allowable error range. If the value is outside the allowable error range, it is determined to be unacceptable. The reference size and the allowable error range are predetermined for each image processing object, and are stored in the memory card. Therefore, when the memory card is read into the image processing device 12, it is stored in the DRAM 256, and these reference dimensions and the allowable error range are used in the dimension inspection.

If the dimensional inspection is completed normally, an inspection step for inspecting the presence or absence of a defect is subsequently performed. As illustrated at the bottom of FIG. 17, master inspection template data is stored in the DRAM 256 together with the master search template data and the master measurement template data. The master inspection template represented by the master inspection template data is shown in FIG. 57. The master inspection template data and the data of the position and rotation angle of the image processing object measured by executing the re-measurement step are shown in FIG. Inspection template data is created based on this, and the inspection step is executed. The master inspection template shown in FIG. 57 includes two negative seek lines 350 and 352 for inspecting whether or not a defect such as a notch and a burr exists on the right side of the square image processing object 304. And a positive seek line 354 of FIG. The positive seek line 354 should be originally represented by a rightward arrow, but is represented by a simple line segment for convenience of illustration. Details of the inspection step will be described later.

The progress of the image processing is displayed on the monitor television 14. For example, at the time of executing the search step, an image processing object (inspection target) is based on image data stored in the frame grabber memory 264 (for example, one set of four sets of image data obtained by imaging four inspection targets 304). The image including the image of the object 304) and the background is displayed in monochrome, on which the angle of the search template is changed by the set pitch and the position of the search template is changed in a spiral shape in color are displayed in color. Progress is shown.

The monitor television 14 is capable of displaying in two modes: an automatic selection display mode and a manual selection display mode. Both the display and the input-related data related to the input from the input device 16 are displayed while giving priority to the input-related data. Therefore, if data is input by the input device 16 in the automatic selection display mode, the display of the progress of the image processing is automatically switched to the display of the input-related data. In the manual display selection mode, of the display of the progress of the image processing and the display of the input related data, only the one selected by the manual operation of the operator is displayed.

Next, the calculation of the object vector of the image processing object will be described. An object vector calculation program for performing the calculation described below is stored in a memory card, and is transferred to the DRAM 256 when the memory card is set in the image processing device 12. Size,
The position and rotation angle are calculated when specified. For example, in the example of FIG.
P of f = PA represents a position (Position), A represents an angle (Angle), and it is specified that a position and a rotation angle are calculated.

The dimension calculation is required, for example, in the following cases.
If the dimensions of the image processing object are required, and if you want to identify the image processing object as similar to the search target part but not the search target part, the edge measurement has a failure and the position measurement accuracy is secured. Such as when you want to. This is the case, for example, when it is necessary to distinguish electrical components having the same shape but slightly different dimensions. The reason why the dimension calculation is required in the case of (1) is that if there is a failure, the dimension is required in order to obtain the position of the image processing object in consideration of the failure as described later. Also in this embodiment,
If none of the above-mentioned conditions applies, the dimension calculation is not performed, but if the dimension calculation is necessary, it is performed prior to the calculation of the position and the rotation angle.

Hereinafter, a case will be described as an example where the image processing object is the square inspection object 304, the calculation of the size, position, and angle is specified and there is a failure. First, the dimension calculation will be described. The dimension calculation is performed using a pair seek line instructed to use for the dimension calculation. In other words, if there is no instruction to use the dimension calculation for any of the seek lines, or if the seek line is not a paired seek line even if the instruction is given, the dimension calculation is not performed. The flag is set to 0 when there is no designation to perform the dimension calculation for the seek line in one template, and the flag is set to 1 when there is at least one designation.
Is set, it is determined whether or not to perform the dimension calculation. In this way, when the dimension calculation is unnecessary, it is possible to know that the dimension calculation is unnecessary without determining whether or not it is instructed to use each of the plurality of seek lines for the dimension calculation. Processing time is short.

The first of the dimensional operations is the operation of the size factor. The size factor is an excessive rate of the dimension of the image processing target in the X-axis direction and the Y-axis direction of the measurement template coordinate plane (the measurement template setting coordinate plane). The oversize ratio in the X-axis direction is shown in FIG.
As shown at 58, the distance (referred to as measurement span) between the edge points of each set of seek lines of a plurality of pairs of set seek lines set parallel to the X-axis direction is represented by the ideal point of the pair seek line (× in the figure). It is obtained by averaging the values divided by the distance (referred to as the original span) between the marked points). In addition, the dimensional excess rate in the Y-axis direction is obtained by averaging values obtained by dividing the measurement span of each of the plurality of pairs of seek lines set in the Y-axis direction by the original span. Note that the broken line in FIG. 36 indicates a seek line in which a failure has occurred. For a pair seek line including a seek line in which a failure has occurred, the oversize ratio is not calculated. The dimensional excess rate in the X-axis direction and the dimensional excess rate in the Y-axis direction
The dimension is calculated by multiplying each of the original spans of the rectangular object 358 in the X-axis direction and the Y-axis direction.

If any one of the plurality of seek lines has a failure, a size point, which is a point obtained by correcting an ideal point by a size factor, is calculated for all of the seek lines having no failure. As shown in FIG. 37, for a seek line substantially parallel to the Y-axis direction, the Y-coordinate value of the size point is calculated by multiplying the Y-coordinate value of the ideal point by the dimensional excess rate in the Y-axis direction. Is omitted, a seek line parallel to the X-axis direction is calculated by multiplying the X-coordinate value of the ideal point by the dimensional excess rate in the X-axis direction.

As described above, the calculation of the size factor is performed using only a pair of seek lines parallel to the X-axis and the Y-axis. The calculation of the used size point and the like are also performed for seek lines that are not parallel to the X axis and the Y axis. The calculation of the size point for the inclined seek line and the following calculation are substantially the same as the calculation of the size point of a circular image processing target (inspection target) described later.

Next, the difference between the edge point and the size point
Diff is calculated. The size point corresponds to an ideal point when it is considered that a measurement template is newly set again after approving the dimensional error of an image processing target having a dimensional error. Therefore, the difference Diff between the edge point and the size point indicates the positional deviation amount of the edge point itself. Since the difference Diff is calculated by subtracting the coordinate value of the size point from the coordinate value of the edge point, when the difference Diff is a positive value, the edge point is shifted to the positive side on the measurement template coordinate plane. You will be.

The reason why the size point is calculated when there is a failure is to reduce the effect of the failure on the calculation of the position of the image processing object and reduce the calculation error. For example, if the size of the image processing object is larger than the original size but there is no displacement, if there is a failure in one of the seek lines constituting the calculation error pair seek line and no failure in the other seek line If the calculation is performed with the ideal point without calculating the size point, the sum of the difference between the edge point and the ideal point is larger on the side without the failure than on the side with the failure, and the position is actually shifted. Although the calculation result that the image processing object is shifted to the seek line side where no failure occurs even though there is no
In order to avoid this, the calculation of the size point is performed.

Next, the calculation of the position and the rotation angle will be described. In many cases, only the outer dimensions or inner dimensions of the inspection object 304 need to be acquired by the optical inspection apparatus. However, the image processing object may be a character described on the surface of the inspection object 304 or an inspection object. In the case of a scratch or the like generated on the surface of the object 304, or in the case of the electric component 208 held by the suction head in the electric component mounting apparatus, it is necessary to measure the position and the rotation angle thereof. Often there. The calculation of the rotation angle is 0 degree, 90 degree, 1
It is performed only on the basis of seek lines of 80 degrees and 270 degrees, which are instructed to be used for calculating the rotation angle. 0 and 180 degrees, 9
The processes of 0 ° and 270 ° are performed collectively.

The position and the rotation angle are calculated for the rotation center RC. The rotation center RC is a calculation center of the computer, and may be different from or coincide with the specified center DC specified by the operator as the center of the image processing target. The designated center DC is generally set to the center point when the planar shape of the inspection object 180 has a center point. The point may be set as the designated center DC. When the image processing target is the electric component 208, a point used as a reference of the image processing target is set as the designated center DC in creating a program for mounting the electric component 208 on the circuit substrate. Also, the origins of the master search template coordinate plane, master measurement template coordinate plane, and the like are placed at the designated center DC, and the position correction amount of the electric component 208 in the horizontal plane is calculated for the designated center DC as described later. .

First, taking the straight line L shown in FIG. 38 as an example,
The calculation of the position and the rotation angle will be described. Originally, the straight line L
Four seek lines were set, but one failed
Or three seek lines have been set from the beginning.
3 seek lines SL₁ , SL
_Two , SL_Three If the above-mentioned difference Diff is calculated for
The center of rotation RC is the two seek lines SL on the left.₁ , SL
_Two Set closer. The number of rotation centers RC is plural on one side.
Sum of each distance between the book seek line and the rotation center RC (see
When there is only one kline, the seek line and the rotation center R
Distance to C) and the multiple
The sum of the distance between the in and the center of rotation (for a single seek line)
The distance between the seek line and the rotation center RC)
Set to a position where the value is equal. In other words, rotation
If one side of the center RC is positive and the other side is negative,
Rotating to a position where the sum of the distances to the seek line is zero
The heart RC is set. Rotation angle (radian)
Is calculated by equation (7). Rotation angle = (AO.A '+ BO.B' + CO.C ') / (AO^Two + BO^Two + CO ^Two ) ... (7) where AO: rotation center RC and seek line SL₁ BO: Rotation center RC and seek line SL_Two CO: Rotation center RC and seek line SL_Three A ': Difference Diff on seek line A B': Difference Diff on seek line B C ': Difference Diff on seek line C The position of rotation center RC is calculated by equation (8).
It is. (A '+ B' + C ') / 3 (8)

Using the difference Diff and the distance from the center of rotation of the edge point to calculate the angle for each of the plurality of seek lines and averaging them, the angle can also be obtained. If the angle is greatly deviated even at one location due to the unevenness, the error greatly affects the final angle value. On the other hand, if the angle is calculated according to the above equation (7), the influence of an error at a specific edge point can be reduced, and the calculation accuracy of the rotation angle can be improved.

If any of the plurality of seek lines has a failure and an edge point cannot be obtained, a deviation occurs between the designated center DC and the rotation center RC as shown in FIG. SL ₂ is a Sea Klein was the fail. In general, the designated center DC is set at the center of the image processing object, and the seek line is set symmetrically with respect to the designated center DC. If specified, designated center D
A shift occurs between C and the rotation center RC. As described above, by setting one side of the rotation center RC to be positive and the other side to be negative, the sum of the distance between each edge point of the plurality of seek lines and the rotation center RC becomes zero, as described above. Thus, equation (9) is satisfied, and equation (10) is obtained from equation (9), and the shift v between the rotation center RC and the designated center DC is calculated. t ₀ + t ₁ + t ₃ = 0 (9) (s ₀ −v) + {s ₀ + (s ₁ −s ₀ ) −v} + Δs ₀ + (s ₃ −s ₀ ) −v｝ = 0 (10) where t ₀ is the distance between the rotation center RC and the seek line SL ₀ t ₁ is the distance between the rotation center RC and the seek line SL ₁ t _{3 is the} rotation center Distance between RC and seek line SL ₃ s ₀ : distance between designated center DC and seek line SL ₀ s ₁ : distance between designated center DC and seek line SL ₁ s ₃ : distance between designated center DC and seek line SL ₃ distance

As shown in FIG. 40, when a seek line is set for two sides parallel to each other of a rectangular image processing object, equation (11) is established, and the rotation center RC and the designated center DC are set. Is calculated based on equation (12). t ₀ + t ₁ + t ₂ + t ₃ + t ₄ + t ₅ = 0 (11) s ₀ + ｛s ₀ + (s ₁ -s ₀ ) ｛+ ｛s ₀ + (s ₂ -s ₀ ) _{} + {s 0 + (s} 3 -s 0)} + {s 0 + (s 4 -s 0)} + {s 0 + (s 5 -s 0)} - 6v = 0 ······ (12)

In the case of the image processing object shown in FIG.
Equation 3) holds, and the deviation v between the rotation center RC and the designated center DC is calculated based on equation (14). t ₀ + t ₁ + t ₂ + t ₃ + t ₄ = 0 (13) s ₀ + {s ₀ + (s ₁ −s ₀ )} + {s ₀ + (s ₂ −s ₀ )} + {S ₀ + (s ₃ −s ₀ )} + {s ₀ + (s ₄ −s ₀ )} − 5v = 0 (14)

As shown in FIG. 42, when the seek lines are set in two directions perpendicular to each other, equations (15) and (16) are established in the X-axis direction and the Y-axis direction, respectively, The respective positional deviations v _x and v _y between the center RC and the designated center DC in the X-axis direction and the Y-axis direction are calculated based on the equations (17) and (18), respectively. t ₀ + t ₁ + t ₂ + t ₃ + t ₄ + t ₅ = 0 (15) t ₆ + t ₇ + t ₈ + t ₉ + t ₁₀ + t ₁₁ = 0 (16) s ₀ + ｛ _{s 0 + (s 1 -s 0} )} + {s 0 + (s 2 -s 0)} + {s 0 + (s 3 -s 0)} + {s 0 + (s 4 -s 0)} + {S ₀ + (s ₅ −s ₀ )} − 6v _x = 0 (17) s ₆ + {s ₆ + (s ₇ −s ₆ )} + ｛s ₆ + (s _{8 -s 6)} + {s 6} + (s 9 -s 6)} + {s 6 + (s 10 -s 6)} + {s 6 + (s 11 -s 6)} - 6v y = 0 ·・ ・ ・ ・ ・ (18)

In the X-axis direction, a displacement v _x of the rotation center RC _{X in} the X-axis direction with respect to the designated center DC is calculated based on the equation (17), and the rotation center RC _X is obtained based on the calculation. In the Y-axis direction, the rotation center R is calculated based on equation (18).
A position shift v _y in the Y-axis direction with respect to the designated center DC of C _Y is calculated, and a rotation center RC _y is obtained based on the calculated position shift v _y .
The intersection of the straight line passing through the rotation center RC _X and parallel to the Y axis of the seek line setting coordinate and the straight line passing through the rotation center RC _Y and parallel to the X axis is the rotation center R of the image processing object shown in FIG.
C.

FIG. 43 shows another example of an image processing object in which seek lines are set in two directions perpendicular to each other. Equations (19) and (20) hold for this image processing object, and the shifts v _x and v _y between the rotation center RC and the designated center DC in the X-axis direction and the Y-axis direction are respectively (21) It is calculated based on the equation and the equation (22). However, in the image processing object shown in FIG. 43, there is no deviation between the rotation center RC and the designated center DC, v _x = v _y = 0, and the distance between the seek line and the rotation center RC and the designated center DC is It is illustrated as being the same, and therefore equation (19) and (2
The same reference numerals are used in the expression (0) and the expressions (21) and (22). s ₀ + s ₁ + s ₂ + s ₃ = 0 (19) s ₄ + s ₅ + s ₆ + s ₇ = 0 (20) s ₀ + ｛s ₀ + (s ₁ -s ₀ )｝ + ｛s ₀ + (s ₂ −s ₀ )｝ + ｛s ₀ + (s ₃ −s ₀ )｝ − 4v _x = 0 (21) s ₄ + ｛s ₄ + (S ₅ −s ₄ )} + {s ₄ + (s ₆ −s ₄ )} + {s ₄ + (s ₇ −s ₄ )} − 4 v _y = 0 (22)

In order to calculate the rotation angle, an angle factor shown in FIG. 44 is set in advance. The angle factor is
It is determined by the position and polarity of the seek line. FIG. 45 shows the angle factors shown in FIG. 44 on the XY coordinates.

FIG. 46 shows an example of calculating the position and the rotation angle of the image processing object using the rotation center RC and the angle factor. The rotation angle is calculated according to equations (23) and (24). The sign of the sign of d ₀ t ₀ to d ₄ v _{1 in} the equation (23) is determined by the angle factor.
The rotation angle calculated by Expression (23) is the rotation angle of the image processing target object with respect to the measurement template coordinate plane. Also,
The displacement Δx in the X-axis direction and the displacement Δy in the Y-axis direction of the rotation center RC are calculated by equations (25) and (26), respectively. These positional deviations Δx and Δy are deviation amounts of the actual position of the rotation center RC from the original position. Rotation angle _{= (- d 0 t 0 -d} 1 t 1 + d 2 t 2 + d 3 v 0 -d 4 v 1) / t a · ···· (23) t a = t 0 2 + t 1 2 + t 2 ² + v ₀ ² + v ₁ ² (24) Δx = (d ₃ + d ₄ ) / 2 (25) Δy = (d ₀ + d ₁ + d ₂ ) / 3 (26) where d ₀ , d ₁ , d ₂ , d ₃ , and d ₄ are each 5
This is the shift amount of the edge point from the size point in the book seek line.

The rotation angle and rotation center when the image processing object is a rectangle and a seek line is set on each of two sides parallel to the X axis and the Y axis of the rectangular image 258 as shown in FIG. The calculation of the RC displacements Δx and Δy are shown in equations (27) to (29). Rotation angle = (− d ₀ t ₀ + d ₁ t ₁ + d ₂ t ₂ −d ₃ t ₃ + d ₄ v ₀ −d ₅ v ₁ −d ₆ v ₂ + d ₇ v ₃ ) / (t ₀ ² + t ₁ ² + t) ₂ ² + t ₃ ² + v ₀ ² + v ₁ ² + v ₂ ² + v ₃ ² ) (27) Δx = (d ₄ + d ₅ + d ₆ + d ₇ ) / 4 (28) Δy = (D ₀ + d ₁ + d ₂ + d ₃ ) / 4 (29) d _{0 to} d ₃ are edge points and size points on the seek line whose distance from the rotation center RC is t ₀ to t _3. And d _{4 to} d ₇ are distances from the rotation center RC of v
A shift amount of the edge point and size point of ₀ to v ₃ on the search lines.

Each of the above equations is equivalent to the rotation angle and the displacement Δx,
Mathematically proves that it is a valid expression for Δy
Instead of clarifying, justify the specific values and justify them.
You. The image processing object having the shape shown in FIG.
While rotating -0.1 (radian) around C,
Rotation center RC is shifted 5mm in Y axis direction and 0mm in X axis direction
Assume the case. In this case, t₀ = 30mm, t₁ = 2
0_mm, T_Two = 50_mm, V₀ = 30_mm, V₁ = 30_mmToss
Then, the rotation angle is so small that it can be considered that sin θ = θ
As long as d₀ = + 8mm, d₁ = + 7mm, d_Two = 0mm, d_Three 
= -3mm, d _Four = + 3 mm. Also, t_a
= T₀ ^Two+ T₁ ^Two+ T_Two ^Two+ V₀ ^Two+ V₁ ^TwoIs calculated to be 5600
(Mm^Two ) Is obtained. These values are calculated by (30), (31), and (32).
To the rotation angle and the rotation center R
The displacements Δx and Δy of C are -0.1 radia, respectively.
, 5mm and 0mm, and the validity of the equation was confirmed.
You. Rotation angle = {− (8 × 30) − (7 × 20) + (0 × 50) + (− 3 × 30) − (+ 3 × 30)} / 5600 = −0.1 (30) Δx = (− 3 + 3) / 2 = 0 (31) Δy = (8 + 7 + 0) / 3 = 5 (32)

Although the rotation angle and the displacement of the rotation center RC are calculated as described above, when the image processing object is the electric component 208, the electric component
Since the rotation of the designated center DC and the positional deviation are necessary for the execution of the mounting work on the circuit substrate of No. 8, the rotation center R
It is necessary to calculate those of the designated center DC from the rotation angle and the displacement of C. However, since the rotation angle is the same for the rotation center RC and the designated center DC, there is no need to perform the calculation, and the calculation may be performed only for the displacement. If there is no rotation angle error in the image processing object and only the displacements Δx ₁ and Δy ₁ are generated in the X-axis direction and the Y-axis direction, the displacement of the rotation center RC is also the displacement of the designated center DC. Are also Δx ₁ and Δy ₁ . But,
When the rotation angle error Δθ occurs in the image processing object,
As shown in FIG. 49, the X-axis direction and Y
The positional deviations of the designated center DC located at positions separated by v _x and v _{y in the} axial direction are (Δx ₁ −Δθ × v _y ) and (Δ
y ₁ + Δθ × v _x ).

The calculated rotation angle Δθ and displacement (Δx ₁ −Δθ × v _y ), (Δy ₁ + Δθ × v _x )
Is the rotation angle and the displacement of the designated center DC of the image processing object with respect to the measurement template coordinate plane, and the measurement template coordinate plane itself has the rotation angle θ and the displacement Δx _{2 with} respect to the reference coordinate plane as shown in FIG. , Δy ₂ . And the reference coordinates are generally C
The optical axis of the CD camera 78, that is, a coordinate plane having the origin at the center of the field of view, is set, and the position of the suction nozzle 206 is positioned so as to coincide with the optical axis of the CCD camera 78 at the component posture detection position. In this case,
The rotation angle θ and the displacement Δx ₂ , Δy ₂ of the measurement template coordinate plane with respect to the reference coordinate plane are the rotation angle and the displacement of the measurement template coordinate plane with respect to the axis of the suction nozzle 206. Therefore, the rotation angle and the displacement of the designated center DC of the electric component as the image processing object with respect to the axis of the suction nozzle 206 are θ + Δθ, Δx ₂ + (Δx ₁ −Δθ × v _y ) cos, respectively.
θ, Δy ₂ + (Δy ₁ + Δθ × v _x ) sin θ.

If there is a failure, the coordinates of the rotation center,
Values such as the distance from the rotation center to each seek line and the displacement between the rotation center and the designated center are calculated as described above. If there is no failure, these values are used for the respective image processing objects and the template. Since fixed values are determined for each combination, these values are stored in the memory card as default values, and the above calculations are performed using the default values. If the calculation of the dimensions is not specified, all the seek lines are paired, and there is no failure, the calculation of the size point is omitted, and the edge point and the ideal point are replaced with the size point. The difference from the point may be calculated, and the position and the rotation angle may be calculated based on the calculation result.
In this case, even if there is an error in the size, the influence of the dimensional error is canceled by the two seek lines constituting the pair seek line, and the position calculation result is not affected. Further, even when there is a failure, if the calculation of the displacement is not specified, the size point is unnecessary, and the calculation may be omitted.

Next, a case will be described in which the object to be image-processed has a circular shape or a partially cut-out circular shape such as a reference mark on a printed circuit board. The reference mark is attached to the printed circuit board, and a position error and a rotation angle error of the printed circuit board are calculated based on the image of the reference mark.
Therefore, there is no need to calculate the rotation angle of the reference mark. In addition, the calculation of the positional deviation of an image processing object in which a circle or a part of a circle is cut out has a lot in common with the calculation of the positional deviation of a rectangular image processing object. is there. Hereinafter, this point will be described.

First, the case where the seek line of the circular image 360 has a failure will be described with reference to FIGS. Here, as shown in FIG. 51, a seek line of 0 degree indicated by a broken line is a seek line of failure, and
It is assumed that the circular image 360 is larger than its original size (the size indicated by the two-dot chain line in the figure). The size calculation is also performed on the circular image 360 first. In the size calculation, the size factor is calculated in the same manner as in the case of the rectangular image 358. Assuming that eight seek lines are radially set at intervals of 45 degrees as shown in FIG. 51, the size factors sizeFX and sizeFY in the X-axis direction and the Y-axis direction are respectively size
After the initial settings of XM = sizeYM = 0 and baseXM = baseYM = 0, the calculation is performed by the equations (33) to (35). for (i = 0; i <n; i ++) ｛sizeXM + = (measurement span [i] / original span [i]) * | cos angle [i] |; baseXM + = | cos angle [i] |; sizeYM + = (measurement span [i] / original span [i]) * | sin angle [i] |; baseYM + = | sin angle [i] |｝; (33) sizeFX = (sizeXM) / baseXM ····· (34) sizeFY = (sizeYM) / baseYM; ····· (35) However, expression (33) is described in C language and for (i =
0; i <n; i ++) means the sum of the results of performing the operation in {} on the i-th seek line while changing i from 0 to n-1 by one. Also, n is the number of pair seek lines, which is 4 in the example of FIG.
In the calculation of the equation (33), i is sequentially changed from 0 to 3. When i becomes a value designating a pair seek line including a fail seek line,
The value can be changed to a value that specifies the next pair seek line without performing the operation. Therefore, FIG.
In the example of (1), the calculation of the expression (33) is performed for three pairs of pair seek lines other than the pair seek line including the 0 ° seek line.

In the case of a circle, since the seek line is set radially, there are seek lines inclined with respect to the X axis and the Y axis, and these seek lines have excessive dimensions in both the X axis direction and the Y axis direction. It has a rate component. Therefore, in the equations (33) to (35), the X-axis direction component and the Y-axis direction component of the oversize ratio on each seek line are obtained, and each contributes to the determination of the size factor according to the component ratio. It is made to be made. The sizeFX and sizeFY obtained by this calculation are the dimensional excess rates in the X-axis direction and the Y-axis direction, respectively. In order to cope with the case where the dimensional error occurs at a different ratio between the X-axis direction and the Y-axis direction, the size factor is separately calculated in both directions.

Next, a size point is calculated. First, as shown in FIG. 53, the pairDiff, which is the difference between the ideal point and the size point, is calculated by the equations (36) to (43) ((36), (3
Equation 7) is described in C language), the size point is calculated from the obtained pairDiff and the ideal point, and the deviation Diff between the size point and the edge point is calculated using the edge point. It is. _{ΔL x = pairRadius * cos θ *} (sizeFX -1); ·· (36) ΔL y = pairRadius * sin θ * (sizeFY -1); ·· (37) when the [Delta] L _x ≧ 0 and ΔL _y ≧ 0 pairDiff = {(ΔL _x ) ² + (ΔL _y ) ² };... (38) When ΔL _x <0 and ΔL _y <0, pairDiff = −√ ｛(ΔL _x ) ² + (ΔL _y ) ² }; (39) When ΔL _x <0 and ΔL _y ≧ 0 and | ΔL _x | ≧ | ΔL _y | pairDiff = −√ ｛(ΔL _x ) ² − (ΔL _y ) ² ｝; ) ΔL _x <0 and ΔL _y ≧ 0 and | ΔL _x | <| ΔL _y | pairDiff = iff− (ΔL _x ) ² + (ΔL _y ) ² ｝;... (41) ΔL _x ≧ 0 and ΔL _y <0 and | ΔL _x | ≧ | ΔL _y | pairDiff = {(ΔL _x ) ² − (ΔL _y ) ² }; (42) ΔL _x ≧ 0 and ΔL _y < When 0 and | ΔL _x | <| ΔL _y |, pairDiff = −√ ｛− (ΔL _x ) ² + (ΔL _y ) ² ｝;

As described above, it is not necessary to calculate the rotation angle of the circular image 360, and only the position is calculated. In calculating the position, first, the position factor (posFactor) is calculated according to the equation (44). posFactorM [i] = 1 / Σcos ² (angle R [i] −angle [n]) (44) where i is a value specifying a seek line and is from 0 to 7. Further, the value of n is also increased by one from 1 to 8, which is the number of seek lines. Σcos ² (angle R
The calculation of [i] -angle [n]) calculates the sum of the squares of the cosine of the difference between the angle R [i] of the i-th seek line and the angles R [n] of all the seek lines without fail. It is. Also in this calculation, if i becomes a value specifying the seek line of the fail, the value can be changed to a value specifying the next seek line without performing the calculation. In, the calculation is not performed on the 0 ° seek line, and the position factor is calculated on each of the 45 ° to 315 ° seek lines.

Next, a circular image 360 is obtained using the position factor.
Is calculated according to equation (45) (described in C language). for (x = 0, y = 0, i = 0; i <n; i ++) ｛x = x + cos (i) * posFactorM [i] * ss [i]; y = y + sin (i) * posFactorM [i ] * Ss [i];｝ ... (45) ss [i] is the shift amount Diff between the size point and the edge point for each seek line previously calculated, and For all of them, the position factor is added to the shift amount ss [i] of the seek line.
posFactorM [i] is multiplied, and a component parallel to the X-axis and the Y-axis is calculated and added for each of the X-coordinate and the Y-coordinate to obtain the center position (x, y). The coordinates of the center position indicate the displacement of the center of the circular image 360 from the origin of the measurement template coordinate plane in the X-axis direction and the Y-axis direction. As described above, if there is a failure in the seek line, the rotation center RC of the image processing object is shifted from the position where there is no failure (the position of the normal rotation center). Is calculated by using the position factor, and the position shift is calculated by the equation (45) using the position factor. Thus, even if there is a failure in the seek line, the normalization of the circular image 360 is performed in the same manner as when there is no failure. Of the center corresponding to the center of rotation is calculated. Therefore, a value obtained by adding the positional deviation of the measurement template coordinate plane to the reference coordinate plane to the obtained positional deviation represents the positional deviation of the circular image 360 on the reference coordinate plane.

If the object to be processed is circular and there is no failure, calculation is performed using default values for the position factor and the like. This is the same as the above-described case where the image processing target is the rectangular image 358. Although the term “position factor” is not used in the above description in the case where the image processing target is the line segment in FIG. 38 or the rectangle in FIG. 47, in the actual embodiment, 44)
The calculation of the position factor is performed using the equation,
1/3 in equation (8) and 1 in equations (28) and (29)
/ 4 or the like corresponds to the position factor. This can be easily confirmed by substituting the actual value into the equation (44). Equation (44) can be used generally regardless of the figure. The reason why equation (44) is not used in the description of FIGS. 38 and 47 is to make it intuitively understandable. This is to confirm the validity of the expression (44) by comparing the result calculated using the expression (44). Of course, the values of 1/3, 1/4, etc. will change if a failure occurs in the seek line. In addition, for example, not only can the above-described calculation for the circular image object be applied to an octagon in which a rectangular corner is cut off obliquely at 45 degrees, but also for an image processing object generally having inclined sides. Calculations of dimensions and positions can be performed similarly.

The case where the image processing object is rectangular, such as a square chip having no leads, and the case where the image processing object is circular, such as a reference mark, have been described. Some have a plurality of lead wires, such as a quad flat package type electric component. When such an electrical component is an image processing target, image processing is performed by a pattern matching manager, which is a combination of pattern matching processes.
The pattern matching manager performs image processing by combining the pattern matching processes a plurality of times.

For example, in the case of an image 370 of a QFP (quad flat package type electric part) shown in FIG. 54, image processing is performed not based on the outline of the entire QFP but on the image of the lead. This is because in the QFP, the error between the lead and the pattern of the board can be minimized if the mounting is performed based on the data of the position of the lead.
The size, position, rotation angle, and the like of the QFP are calculated by a combination of pattern matching using each lead as an image processing target.

At the time of image processing, first, the lead image 372 is read.
Is searched. Therefore, a search window 37 sized appropriately to include one lead image 372
6, a lead image 37 is set using a search template including a plurality of point pairs set in advance.
2 is searched.

A full set of pattern matching of the search step, the re-search step, the measurement step, and the re-measurement step is performed, and the position and the rotation angle of the lead image 372 are measured. When the position and the rotation angle of the lead image 372 are known, the lead image 372 of the entire side is searched by pattern matching of a subset including a search step and a re-search step. Subsequently, the position and the rotation angle of the image 372 of the adjacent lead are measured. The pitch of the lead is known in advance, and the position and the rotation angle of the coordinate plane of the search template are determined in the search step for the next image 372 of the lead. Since the position and the rotation angle of the lead image 372 obtained by the re-search step immediately before that and the pitch between the leads are predicted fairly accurately, sufficient accuracy can be obtained without performing full set pattern matching. The next lead image 372 can be searched well.

When pattern matching is performed on the image 372 of all leads on one side and the positions of the images 372 of all leads are calculated, the X coordinate value and the Y coordinate value of the center of the image 372 of these leads are added. Is divided by the number of leads, and the center coordinates of one side are calculated. Similarly, 3
The image 372 of the lead of the side is searched by the pattern matching of the subset. The positional relationship between the sides is known in advance,
The position and the rotation angle of the search template are set based on the position and the rotation angle of the re-search template coordinate plane of the previously performed pattern matching. If the respective center coordinates are calculated for all four sides, the intersection of two straight lines a and b connecting the centers of the two sides parallel to each other is calculated, and is set as the center coordinate of the electric component. The rotation angle of the QFP image 370 is (4
It can be obtained by equation (6). (Slope of straight line a−90 degrees + slope of straight line b) / 2 (46)

As described above, according to the pattern matching or the pattern matching manager, it is possible to recognize an image of an inspection object having almost any shape or an electric component to be mounted on a circuit substrate. Although it is necessary to create a master search template and a master measurement template, image processing programs using these templates can be shared, and compared to creating an image processing program for each type of inspection object or electrical component, the program The time required for creation is short.

The image processing time becomes longer as the offset amount of the center of the image processing object from the center of the search window becomes larger. Therefore, it is desirable to correct the position of the center of the search window based on the position of the search template coordinate plane where the image processing object has been searched. For example, in the case of a QFP image 370, after image processing for a plurality of QFPs, an average deviation from the center of the search window 376 at the center of the lead image 372 is determined, and the average deviation is set to zero. Then, the center position of the search window is corrected.

In the present embodiment, pattern matching or a pattern matching manager and an operation of an object vector are performed on images of a plurality of electric components 208 which are sequentially captured, and identification, position error and rotation angle error of the electric components are performed. Is performed. As shown in the time chart of FIG. 55, during the mounting cycle time Tr of the electric component 208, the imaging by the CCD camera 78 is performed while the electric component 208 is stopped at the imaging position. After the image is captured, the image data is transferred from the CCD camera 78 to the frame grabber memory 264, where the image processing is performed. The transfer of the image data and the image processing are performed in parallel.

The frame grabber memory 264 is capable of storing image data of four electric components 208 in parallel, so that the image processing is performed for the electric components 208 whose processing time is longer than the mounting cycle time. It can be carried out. The image processing may be completed before the image processing result is used. If a plurality of other electric components 208 are mounted between the time when one electric component 208 is captured and the time when the image processing result is used. , 1 mounting cycle time Tr
It is not essential to end the image processing during this period, and it is sufficient that the image processing of the same number of image processing objects as the number of cycles is performed during the multiple mounting cycle time. In the present embodiment, the mounting of the other three electric components 208 is performed during the period from imaging to mounting of one electric component 208,
The image processing time can be exchanged among the four electric components 208 in total.

For example, some electric components have a simple shape without a lead like a square chip and have a short required image processing time, while others have a large number of leads like a QFP and have a required image processing time. Some electric components have a long time, and the surplus time of an electric component having a short required image processing time can be used for processing an electric component having a long required image processing time.
Assuming that the image processing of all electrical components must be completed within one mounting cycle time, the mounting cycle time should be set to the longest of the multiple types of electrical components scheduled for image processing. And the mounting efficiency is kept low. On the other hand, in the electric component mounting apparatus, four image data are stored in the frame grabber memory 264 in parallel, and
Since the processing of the image data is performed during the mounting cycle time, the sum of the image processing times of the four electrical components may be within the mounting cycle time which is several times the electrical component for which the image processing is performed. In other words, the average of the image processing times of the plurality of electric components only needs to be within one mounting cycle time, and the mounting cycle time can be reduced.

An example is shown in FIG. As apparent from the figure, although the one of the image processing time T _e1 through T _e7 required each electrical component has length difference, respectively total T _t1 through T _t4 of the image processing time of four by four, mounting cycle Time T
This is a time shorter than four times r, and the mounting cycle time Tr can be shortened accordingly.

The same effect can be obtained when a plurality of types of inspection objects 180 to be inspected sequentially flow. When the time required for the inspection of the inspection object 180 to be sequentially inspected is different from each other, the frame grabber memory 264
Inspection can be performed with a cycle time close to the average value of the time required for each of the inspection objects 180 to be sequentially inspected, which is the same as the number of inspection objects 180.

Further, the inspection object 180 to be inspected sequentially
Is one type and the time required for the inspection is the same, the effect of the plurality of frame grabber memories 264 can be obtained if the imaging time intervals of the inspection object 180 are different. This is because the image processing time can be exchanged between a plurality of continuous inspection objects 180.

Finally, the details of the inspection step will be described. For example, when it is necessary to inspect whether or not the notch 411, which is a kind of defect, exists on three sides adjacent to each other in the component having the shape shown in FIG. Negative seek lines 412, 41 parallel to each side are located at positions within a small distance of three sides of 410).
3,414 are set. A master inspection template set in advance based on the component 410 having the reference position and the reference rotation angle and stored in the DRAM 256 is read out, and the master inspection template is used to determine the actual position of the component 410 measured before the inspection. The negative seek lines 412 and 4 are converted by the coordinates according to the rotation angle.
An actual inspection template 415 including 13, 414 is set. Inspection template 415 set
Is also stored in the DRAM 256, and the DRAM 256
Functions as an inspection template data storage unit.

In the example of FIG. 58, it is assumed that a silhouette image of the component 410 is obtained, and the notch 411 is formed as a bright image. Negative seek line 412, 413, 41
4 is set inside the dark image of the component 410 and originally does not intersect the edge, but if the notch 411 exists, the boundary between the component 410 and the notch 411 becomes the edge.
Therefore, the result of the determination of “does not intersect with the edge” performed on the negative seek lines 412 and 413 is YES, but the result of the determination performed on the negative seek line 414 is NO. If the determination result is NO for at least one of all the negative seek lines belonging to the inspection template 415, it is displayed on the monitor television 14 that the component 410 is rejected as having the notch 411 defect.

The negative seek lines 412 and 41
By changing the distance of 3,414 from each side of component 410, the size of the detectable notch can be changed.
Even if there is a notch, if the part is acceptable if it is small, the negative seek lines 412, 413, 414 may be set at positions that do not cross the small notch. In addition, if a plurality of negative seek lines having different distances from each other are set for each side, only a larger notch is detected for a negative seek line having a larger distance from the side. You can also see how many notches are present in each.

In the case where it is necessary to inspect the part 410 for the presence of a burr 416 as a kind of defect as shown in FIG. 3 negative seek lines 4 extending in parallel
An inspection template 422 including 17, 418, and 420 is set. The actual inspection template 422 is set by the coordinate conversion of the master inspection template stored in the DRAM 256. In the case shown in the drawing, the result of the determination of “does not intersect with the edge” for the negative seek line 420 is NO, and the component 410 has a defect of the flash 416 and is determined to be rejected. The judgment result is the monitor TV 1
4 is displayed.

If it is necessary to check whether or not the image of the component 410 has the flaw 426 shown in FIG. 60, a large number of negative seek lines 428 to 438 are provided inside the outline of the component 410. Is set. At this time, a hole 4 is placed inside the outline of the part 410.
If there are 40, 442, etc., the negative seek line will be the hole 44, as shown at 436, 438.
0, 442, etc. are set. If the negative seek line intersects the edge of the hole 440, 442 or the like, the result of the determination of "whether the edge does not intersect with the edge" is NO, and an incorrect inspection result is output. In the illustrated example, if the image of the scratch 426 is darker than the image of the surface of the component 410, the negative seek lines 432 and 434, which should not originally intersect the edge, intersect the edge. Based on this, it is determined that the component 410 has a flaw 426 and is rejected.

The negative seek line 412, FIG.
413, 414 and the negative seek line 41 in FIG.
It is also possible to set an inspection template including both 7, 418 and 420, and in this case, it is possible to inspect both the notch 411 and the burr 416 at once. Further, if the negative seek lines 428 to 438 in FIG. 60 are set together, the inspection of the notch 411, the burrs 416 and the scratches 426 can be performed at one time.

According to the present optical inspection apparatus, not only the whole component but also a part of the component can be inspected.
An example is shown in FIG. The parts shown are solar cells 450
And the inspection object is its grid pattern 452. The grid pattern 452 includes a large number of grids 454, 456, 458, etc. extending parallel to each other. If a discontinuous portion 460 exists in these grids, the grid pattern 452 cannot be used. In addition, if dirt 462 exists between the grids, there is a possibility of a malfunction, so that the inspection is also required. Therefore, each grid 4
Negative seek lines 466, 468, 470 and the like for detecting a discontinuous portion are provided inside the contour line such as 54, and negative seek lines 472 and 472 for detecting a stain are provided between grids.
474 and the like are set. Each of these seek lines 466 and the like is formed so as to extend in parallel with the longitudinal direction of the grid 454 and the like, and forms an inspection template 478.

If any of the negative seek lines of the inspection template 478 passes through the discontinuous portion 460 like the negative seek line 468, the negative seek line 468 intersects the edge and the solar cell 450 It is assumed that the grid pattern 452 of FIG.
A rejection is determined. Also, if there is something that passes through the dirt 462 like the negative seek line 474, the negative seek line 474 will intersect with the edge, and it is assumed that the solar cell 450 has a dirt which is a kind of defect. A rejection is determined.
For this purpose, a negative seek line 466 for detecting a discontinuous portion and a negative seek line 47 for detecting dirt are used.
2 etc. are distinguished in the inspection template.

In the inspection template described above, if at least one of the negative seek lines intersects with the edge, the inspection object is rejected. However, the performance of the inspection object such as small scratches and dirt is affected. If there is only one that does not exist, it may be determined to be a pass. For example, the size of a defect can be known based on how many consecutive negative seek lines have a negative determination result. A pass judgment is made, or a reject judgment is made when the number of negative seek lines for which the judgment result is NO is greater than or equal to the set number.
It is desirable that the monitor television 14 display not only the pass / fail judgment result but also the type, size, number, and the like of the defects.

In the above description, the positive seek line is used for measuring and inspecting the position, rotation angle, dimensions, etc., and the negative seek line is used for defect inspection. However, these can be used for other purposes. It is. An example is shown in FIG. This example is an example in which a positive seek line is used for defect inspection. In order to check whether or not the discontinuous portion 460 exists in the grid 454 or the like of the solar cell 450, a large number of positive seek lines 480 intersecting (or orthogonal in the illustrated example) with the longitudinal direction of the grid 454 or the like are set. I have. A large number of positive seek lines 480 are set at a pitch smaller than the size of the discontinuity 460, and together form the inspection template 482. The positive seek line 480 normally intersects with the edge, but does not intersect with the edge in the discontinuous portion 460. Therefore, the result of the determination of "whether to intersect with the edge" is NO, and the solar cell 450 is a rejected product. Is determined.

FIG. 63 shows an example in which a negative seek line is used for searching for an object. In this example, the part 4
In this example, when a large number of square-shaped bright pads 492 are formed on the dark surface of 90, the upper right corner pad 492 is searched. Therefore, a large number of positive seek lines 494 intersecting with the outline of the pad 492 and a plurality of (four in the illustrated example) negative seek lines 496 are provided.
Are set. Then, the determination of “whether to intersect with the edge” is made for all the positive seek lines 494, and the determination of “whether or not to cross the edge” is made for all the negative seek lines 496, and all the determination results become YES. For example, it is determined that the pad 492 located at the position where the positive seek line 494 is set is the upper right corner pad. For example, if the same inspection template 498 is set for the pad 492 located on the left side of the pad 492 in the upper right corner, the negative seek line 496 extending in the vertical direction intersects the pad 492 in the upper right corner, and If the same inspection template 498 is set for the pad 492 located at the right side, the negative seek line 496 extending in the lateral direction is
Intersect with Therefore, in any case, there is a negative seek line 496 for which the result of the determination of “does not intersect with the edge” is NO, and the inspection template 498 is actually present.
It is found that the pad 492 in which is set is not the pad in the upper right corner, and the above condition is satisfied only in the case of the illustration.

In the above example, all the seek lines are set with finite straight lines, that is, line segments at both ends, but this is not essential. For example, as shown in FIG. 64, when the image processing target 504 is circular, it is convenient to set circles concentric with the image processing target 504 as the negative seek lines 506 and 508. There is. The negative seek line 506 is for notch detection, and the negative seek line 508 is for burr detection. Also, a large number of concentric circles may be set inside the outline of the image processing object 504, and these may be used as an inspection seek line (negative seek line) for inspecting the presence or absence of internal area defects such as scratches and dirt. It is possible.

In the example shown in FIG. 65, since the image processing object 512 has a sector shape, an inspection template 522 including arc negative seek lines 514 and 516 and line segment negative seek lines 518 and 520 is set. I have. The inspection template 522 is for notch detection, and other than that, it can be set in a similar manner for burr detection, internal area defect detection, and the like.

After the presence or absence of a defect is detected as described above, it is determined based on the detection result whether the image processing object is a pass or reject. The simplest determination is a rejection if there is at least one defect, but it is rejected if there is a defect exceeding a predetermined standard such as a size and number exceeding a defect inspection standard. It is also possible to make it. This defect inspection reference is also created in advance for each image processing object and stored in the memory card, and is read into the DRAM 256 of the image processing apparatus 12 and used.

As described above, in the image processing apparatus 12, the data of the image of the inspection object 180, the electric component 208, the reference mark, and the like captured by the CCD camera 78 is processed by the pattern matching or the pattern matching manager. However, the pattern matching and the pattern matching manager search the image processing object using the search template, the re-search template, the measurement template, the re-measurement template, and the inspection template, and calculate the edge points. ,
Only the part where the template is set is searched, measured and inspected. Only the part that requires image processing is processed, and the part that does not require image processing is not processed even if image data is obtained. Therefore, image processing can be performed in a short time, and most image noises (white spots, black spots, spots, etc.) unless the image processing object is directly touched.
Image processing can be performed without being affected by the above.

Further, in the search step, it is determined whether or not the state is in conformity with the state of the difference in optical characteristic values such as the luminance difference between the two point pair constituent points, and the re-search step, the measurement step, and the re-measurement step are performed. In the inspection step, an edge point is determined by a change gradient of an optical characteristic value such as luminance on a seek line. Therefore, even when the image processing target is relatively large and the illumination tends to be uneven, as in the case of the QFP, the comparison of the optical characteristic values is performed between the portions having almost no difference in the illumination. Image processing can be accurately performed without being affected by the image processing.

Further, even if the size of the solid-state image pickup device of the CCD camera changes, it is possible to easily cope with the program without substantially changing the program. Further, since the defect inspection is performed by the same method as the measurement and inspection of the position, the rotation angle, and the size of the image processing object, the defect inspection can be performed in a short time accompanying the inspection of the size and the like. .

Although the case where the origin of the master measurement template coordinate plane is located at the designated center DC of the image processing object has been described above, the origin of the master measurement template coordinate plane may be located at a position other than the designated center DC. Good. In this case, the displacement (or rotation angle) between the designated center DC and the master measurement template coordinate plane is added to the displacement (or rotation angle) between the rotation center RC and the designated center DC, and the value is further added. By adding the displacement (or rotation angle) of the master measurement template coordinate plane with respect to the reference coordinate plane, the displacement (or rotation angle) of the designated center DC on the reference coordinates can be obtained.

In this embodiment, the image processing device 12
Makes use of the fact that the shape and dimensions of the object to be inspected are substantially determined, and uses a search template, a re-search template 328,
Since the dimensional inspection and the defect inspection are performed by special processing using the measurement template 336, the re-measurement template, the inspection template 415, and the like, the processing can be completed in a very short time. Further, the support device for the object to be inspected (which is also the object to be measured) is a transparent flat plate, and the search template 300 is adjusted in accordance with the position and rotation angle of the object to be inspected.
And the like are automatically set, the inspection object may be simply placed on the inspection object support plate 72. Measurement object 1
Even if the position and the rotation angle of 80 are not determined accurately, the image processing device 12 can obtain dimensions and detect defects, and can easily set the inspection object on the imaging device 10. Further, since the processing start command to the image processing apparatus 12 can be issued by the foot switch, both hands can be used for other purposes, and there is an advantage that the usability is good. Further, since it is not necessary to fix the object to be measured and the dimension can be measured in a non-contact manner, the dimension measurement and the dimension inspection of a product made of rubber, a soft synthetic resin or the like can be easily performed. For this reason, for example, it is possible to inspect all the dimensions of a large number of products manufactured by injection molding or press molding. In addition, not only dimensional inspection of a predetermined portion such as an inner diameter of a through hole, a pitch between holes, a pitch between protrusions, an outer diameter of a cylinder, a distance between flanges, and a shaft length, but also an inspection of processing leakage or defect can be performed.

As is clear from the above description, in the present embodiment, the parallel light generator 52 constitutes a parallel light generator. Further, the frame grabber memory 264 of the image processing device 12 constitutes image data storage means,
Reference numeral 156 constitutes search template data storage means, measurement template data storage means, and inspection template data storage means. A part of the image processing apparatus 12 that executes the search step alone or both the search step and the re-search step of the pattern matching program constitutes a search object determination unit. A portion of the image processing apparatus 12 that executes the search step, the re-search step, the measurement step, and the re-measurement step of the pattern matching program constitutes a position and the like measurement means, of which the re-search step, the measurement step, and the re-measurement step are included. The part for calculating the coordinates of the edge point in the above constitutes an edge point coordinate calculating means. In the pattern matching program of the image processing device 12, a portion for specifying a point for which luminance is to be calculated in the search step, the re-search step, the measurement step, and the re-search step constitutes a point specifying means, and the luminance value of the specified point The portion that performs the above calculation constitutes virtual point data calculation means.

Further, the portion of the image processing apparatus 12 which compares the measured dimensions of the image processing object with the reference dimensions to determine whether or not the image processing is acceptable constitutes a dimension inspection means. A portion of the image processing apparatus 12 for setting an inspection template including a negative seek line and a portion for setting a search template including a negative seek line constitute a negative seek line setting means, and include an inspection template and a search template including a positive seek line. Constitutes the positive seek line setting means, and it can be considered that the negative seek line setting means or the positive seek line setting means constitutes the seek line setting means alone or jointly. Further, a portion of the image processing device 12 that determines whether a negative seek line intersects an edge constitutes a determination unit, and a portion that determines whether the image processing target is acceptable or not based on the determination result is the determination unit. And pass / fail judgment means. The seek line setting means and the determination means constitute a defect determination means.

In the above embodiment, the re-search step is performed once, but may be performed two or more times. After the number of failures is equal to or less than the set number in the first re-search step and the edge point of the image processing object is obtained, the re-search template is set and the re-search step is performed. Based on the re-search template used in the re-search step and the calculation result of the edge point, the position and the angle are set such that the deviation from the image processing target is reduced. Experience has shown that making the re-searching step multiple times, rather than just once, reduces the probability of a failure occurring in the next measurement step. The reason for this is not clear, but in the search step, the difference between the image processing target and the search template is large, and if it is determined that the search target part barely exists, the deviation between the re-search template and the image processing target is small. In this case, the displacement between the measurement template and the image processing target may increase in the next measurement step and a failure may occur. In contrast, the re-search step may be performed more than once. If so, it is speculated that the probability may decrease.

If the shape, approximate position, and rotation angle of the image processing object are known in advance, the seek line can be set at a position where the edge point can be calculated without performing a search using a search template. The edge point can be obtained from the beginning by using the measurement template. In the above embodiment, the re-searching step is performed using the re-searching template, and then the measurement step is performed. However, when the dimensional error or the shape defect of the image processing target is small, the searching step is performed. It is also possible to execute the measurement step immediately after. In this case, it is possible to use the re-search template as a measurement template, but it is preferable to use a template having a larger number of seek lines.

Further, in the above-described embodiment, the judgment of the abnormality in the pattern matching is made based on whether or not there is a failure exceeding the allowable number of failures set in the program. A program that determines an abnormality and calculates an object vector only for an image processing target having no failure;
At least one failure may be allowed, and a program for calculating the object vector may be created even if a failure occurs, so that the operator can select the failure. A program that does not allow a failure requires a short time for image processing. Therefore, depending on the situation, select this program to increase the inspection efficiency of the inspection object 180 or reduce the time required for mounting one electric component 208. Can be shortened.

In the above embodiment, in order to search for an edge in an image, that is, a portion where the optical characteristics change suddenly,
First, a portion where the optical characteristic values of the two points forming each point pair of the search template are different from each other by a set value or more is searched, and then the seek line (measure seek line) of the measurement template is searched.
The part where the differential value of the optical characteristic value above was found to be the edge point by searching for the part where the differential value of the optical characteristic value on the seek line of the measurement template was greater than or equal to the set value. It is also possible to search for the part and to make that part an edge. In this case, when points whose differential values are equal to or larger than the set value appear continuously on one seek line, the largest one of them may be set as an edge point.

In addition, without departing from the scope of the claims, the present invention can be implemented in various modified and improved forms based on the knowledge of those skilled in the art.

[Brief description of the drawings]

FIG. 1 is a front view (with a cover member removed) showing an imaging device of an optical inspection device according to an embodiment of the present invention.

FIG. 2 is a plan view of the image pickup apparatus (with a cover member and a beam splitter removed).

FIG. 3 is a right side view of the imaging device.

FIG. 4 is a side view (partial cross section) showing a periphery of a beam splitter in the imaging apparatus.

FIG. 5 is a front sectional view of an interchangeable lens of the imaging device.

FIG. 6 is a front sectional view of another interchangeable lens of the imaging device.

FIG. 7 is a front sectional view of still another interchangeable lens of the imaging device.

FIG. 8 is a diagram conceptually showing the imaging device.

FIG. 9 is a conceptual diagram of an imaging apparatus according to another embodiment of the present invention.

FIG. 10 is a conceptual diagram of an imaging device according to still another embodiment of the present invention.

FIG. 11 is a conceptual diagram of an imaging apparatus according to still another embodiment of the present invention.

FIG. 12 is a front view showing components other than the imaging device of the optical inspection device according to the embodiment of the present invention.

FIG. 13 is a block diagram of an image processing device of the optical inspection device.

FIG. 14 is a diagram showing a pre-processing program stored in a DRAM of the image processing apparatus.

FIG. 15 is a diagram showing an execution processing program stored in a DRAM of the image processing apparatus.

FIG. 16 is a diagram showing a pattern matching program stored in a DRAM of the image processing apparatus.

FIG. 17 is a diagram showing setting data for executing the pattern matching program for a square image.

18 is a diagram showing a master search template set based on the setting data shown in FIG. 17 together with an image processing object.

19 is a diagram showing a master measurement template set based on the setting data shown in FIG. 17 together with an image processing object.

FIG. 20 is a diagram illustrating generation of a search template based on the master search template.

FIG. 21 is a diagram showing setting data for executing the pattern matching program on a partially cut-out disk.

FIG. 22 is a diagram showing a master search template set based on the setting data shown in FIG. 21 together with a disk.

23 is a diagram showing a master measurement template set based on the setting data shown in FIG. 21 together with a disk.

FIG. 24 is a diagram showing a state in which a search template is superimposed on a virtual screen in a search step of the pattern matching program.

FIG. 25 is a diagram illustrating linear interpolation for calculating the luminance of a designated point on a virtual screen.

FIG. 26 is a diagram conceptually showing a physical screen / virtual screen conversion driver that calculates image data on a virtual screen from image data on a physical screen in the image processing apparatus.

FIG. 27 is a diagram showing a state in which a re-search template in a re-search step of the pattern matching program is overlaid on a virtual screen.

FIG. 28 is a diagram illustrating a relationship between a division point set on a seek line of the re-search template and a solid-state imaging device forming an imaging surface.

FIG. 29 is a table showing luminance calculated for the division points shown in FIG. 28;

FIG. 30 is a diagram showing a difference filter for differentiating the luminance calculated for the division point.

FIG. 31 is a diagram showing another difference filter for differentiating the luminance calculated for the division point.

FIG. 32 is a graph showing the calculation result shown in FIG. 29;

FIG. 33 is a graph showing the result of differentiation performed using the difference filter shown in FIG. 30.

FIG. 34 is a graph showing the result of differentiation performed using the difference filter shown in FIG. 31.

FIG. 35 is a diagram showing a state in which a measurement template is overlaid on a virtual screen in a measurement step of the pattern matching program.

FIG. 36 is a diagram for describing dimension calculation of an image processing object performed after execution of pattern matching, taking a rectangular image as an example.

FIG. 37 is a diagram illustrating a calculation of a difference between a size point and an edge point in the dimension calculation.

FIG. 38 is a diagram illustrating the calculation of the rotation center and the rotation angle of the image processing target object using a straight line as an example.

FIG. 39 is a diagram illustrating a calculation of a deviation of the rotation center of the image processing target from the designated center when there is a failure in the seek line.

FIG. 40 is a diagram illustrating another example of the calculation of the deviation of the rotation center of the image processing target from the designated center when there is a failure in the seek line.

FIG. 41 is a diagram illustrating still another example of the calculation of the deviation of the rotation center of the image processing object from the designated center when a seek line has a failure.

FIG. 42 is a diagram illustrating a rotation center of an image processing object and a calculation of a shift of the rotation center with respect to a specified center when a seek line is set in two directions orthogonal to each other.

FIG. 43 is a diagram illustrating another example of the calculation of the rotation center of the image processing object and the deviation of the rotation center from the designated center when the seek line is set in two directions orthogonal to each other.

FIG. 44 is a chart showing angle factors used for calculating an angle of an image processing object.

FIG. 45 is a diagram showing the angle factor on a coordinate plane.

FIG. 46 is a diagram for explaining calculation of a rotation center and a rotation angle of an image processing object.

FIG. 47 is a diagram illustrating another example of the calculation of the rotation center and the rotation angle of the image processing object.

FIG. 48 is a diagram illustrating still another example of calculation of the rotation center and the rotation angle of the image processing object.

FIG. 49 is a diagram illustrating the calculation of the position of the designated center when the designated center and the rotation center of the image processing object are shifted and the position and angle of the image processing object are shifted.

FIG. 50 is a diagram for explaining calculation of a positional shift amount and an angular shift amount of an image processing target object with respect to reference coordinates.

FIG. 51 is a diagram showing a state in which a seek line is set on a circular image which is a kind of an image processing object.

FIG. 52 is a diagram illustrating the calculation of the size of the circular image.

FIG. 53 is a diagram for explaining the calculation of the size point and the calculation of the difference from the edge point in the size calculation.

FIG. 54 is a diagram illustrating pattern matching performed on a QFP.

FIG. 55 is a time chart showing the relationship between the mounting cycle time in the electric component mounting apparatus and the exposure time and image transfer time of the CCD camera.

FIG. 56 is a time chart for explaining an image processing cycle in the electric component mounting apparatus.

57 is a diagram showing a master inspection template set based on the setting data shown in FIG. 17 together with an image processing target.

FIG. 58 is a diagram showing an example of notch inspection using an inspection template.

FIG. 59 is a diagram showing an example of inspection of burrs using an inspection template.

FIG. 60 is a diagram showing an example of a flaw inspection using an inspection template.

FIG. 61 is a diagram illustrating an example of inspection of a discontinuous portion and dirt using an inspection template.

FIG. 62 is a diagram illustrating an example of inspection of a discontinuous portion using another inspection template.

FIG. 63 is a diagram illustrating an example of a search using a search template including a negative seek line and a positive seek line.

FIG. 64 is a diagram showing another inspection template.

FIG. 65 is a diagram showing still another inspection template.

FIG. 66 is a diagram conceptually showing imaging by a conventional imaging device.

[Explanation of symbols]

10: Imaging device 12: Image processing device 14: Monitor television 16: Input device 18: Keyboard
20: mouse 22: foot switch 30: device main body 50: imaging unit 52: parallel light generation unit 5
8: Imaging object support device 70: Concave mirror 72: Beam splitter 74: Orifice 76: Lens system 78: CCD camera 80: Pinhole 8
2: CCD 108: tilt angle adjusting device 110: position adjusting device
146: light source device 148: concave mirror 150: beam splitter 15
2: Orifice 154: Lens system 180: Object to be imaged 192: Uniform illumination device 200, 204:
Imaging device 212: Fluorescent plate 256: DRAM
264: frame grabber memory 300, 308: master search template 302, 310, 324: point pair 328: re-search template 33
0, 346: seek line 336: measurement template 338, 342: seek line 340, 34
4: Master measurement template 350, 352, 35
4: Negative seek line 411: Notch 41
2,413,414: Negative seek line 41
5,422,439,478,482: Inspection template 480: Positive seek line

Continued on the front page F-term (reference) 2F065 AA03 AA07 AA12 AA21 AA37 AA39 AA65 BB05 DD03 DD06 DD09 FF02 FF04 FF26 FF61 GG02 GG07 GG12 GG15 HH03 HH16 JJ03 JJ26 LL02 LL04 LL19 Q24 Q27 Q28 AA30 BA02 BB04 BC05 BC09 BC11 BC30 5C022 AA01 AB00 AC42 AC51 AC77 AC78

Claims (13)

[Claims] 1. A concave mirror, a beam splitter disposed on an optical axis of the concave mirror so as to be inclined with respect to the optical axis, and one of a transmission direction and a reflection direction of the beam splitter when viewed from the concave mirror side. An imaging object support device disposed on one side and supporting the imaging object substantially at the position of the focal point of the concave mirror; and the beam splitter opposite to the focal point of the concave mirror in the other of the transmission direction and the reflection direction. A lens system that is provided on the side and converts light that spreads radially through the focal point into substantially parallel light; an image sensor that captures an image formed by the parallel light converted by the lens system; An image pickup apparatus, comprising: a limiting unit that limits light incident on the image sensor to light incident on the concave mirror substantially in parallel with an optical axis of the concave mirror. 2. The imaging apparatus according to claim 1, wherein said restricting means includes an orifice having a pinhole located at a focal point of said concave mirror. 3. The apparatus according to claim 1, further comprising a light background forming device provided on a side opposite to the beam splitter with respect to the imaging target object supporting device, for forming a bright and substantially uniform background of the imaging target object. 3. The method according to claim 1, wherein
An imaging device according to claim 1. 4. A beam splitter which is provided on a side opposite to the beam splitter with the imaging object support device interposed therebetween, transmits through the imaging object support device, passes around the imaging object, and enters the beam splitter. The imaging device according to any one of claims 1 to 3, further comprising a parallel light generator that projects parallel light that emits light, wherein the parallel light generator functions as the limiting unit. 5. A parallel beam generator, comprising: a second beam splitter provided on a side opposite to the beam splitter as a first beam splitter with the imaging object support device interposed therebetween; A second concave mirror provided separately from the concave mirror as a first concave mirror in a reflection direction of the second beam splitter as viewed from the apparatus; and a second concave mirror in a transmission direction of the second beam splitter as viewed from the second concave mirror. The second mirror is provided at the focal position of
The imaging device according to claim 4, further comprising: a point light source that emits illumination light that spreads radially toward the beam splitter. 6. A second orifice different from the orifice as a first orifice, wherein the point light source has a pinhole provided at a focal position of the second concave mirror, and a second orifice. A parallel light source that is provided on the opposite side to the second beam splitter and emits substantially parallel illumination light toward the orifice; and the light from the parallel light source passes through the pinhole of the second orifice to form the second beam. The image pickup apparatus according to claim 5, further comprising a second lens system different from the first lens system, which converts the light into illumination light that spreads radially toward the splitter. 7. The imaging object includes the first concave mirror, the first beam splitter, the first orifice, and the first lens system, and the second concave mirror, the second beam splitter, the second orifice, and the second lens system, respectively. The imaging device according to claim 6, wherein the imaging device is arranged symmetrically with respect to a position where the object is supported by the object support device. 8. An image pickup device according to claim 1, wherein the image pickup device processes image data representing an image picked up by the image pickup device to obtain a position, a rotation angle,
An image processing device for acquiring at least one of the dimensions. 9. An image processing apparatus comprising: an image data storage unit configured to store data of an image captured by the imaging device; and a measurement template including a plurality of seek lines connecting two points in a straight line. A measurement template data storage unit for storing data, and a measurement template of the measurement template data storage unit is superimposed on a screen on which an image represented by the image data of the image data storage unit exists, and on each of the plurality of seek lines, 9. The optical measuring apparatus according to claim 8, further comprising edge point coordinate calculating means for calculating coordinates of an edge point at which an optical characteristic changes suddenly. 10. A search template data storage means for storing data of a search template having a plurality of point pairs each including a pair of two points separated by a predetermined distance, the search template data comprising: When the search template of the storage unit is overlaid on a screen on which an image represented by the image data of the image data storage unit is present, the difference state of the optical characteristic value of each pair of the plurality of point pairs is set. If the state is equal to or more than the state, each pair of points is in an adapted state located on both sides of an edge, which is a part where the optical characteristics of the image change rapidly, and a set amount or more of the plurality of point pairs A determination unit that determines that the imaging target object is a search target object that matches the search template if the object is in a matching state. Optical measuring device according to. 11. An image pickup apparatus according to claim 1, wherein the image pickup apparatus detects a defect of the object by processing image data representing an image picked up by the image pickup apparatus. An optical inspection device, comprising: 12. The image processing apparatus, comprising: an image data storage unit configured to store image data that is data of an image captured by the imaging device; and an image represented by the image data stored in the image data storage unit. Negative seek line setting means for setting a negative seek line scheduled not to intersect with an edge which is a portion where the optical characteristics of the image suddenly change; and, when the set negative seek line intersects with the edge, the imaging object The optical inspection apparatus according to claim 11, further comprising: a determination unit configured to determine that the object has a defect. 13. An image capturing apparatus according to claim 1, wherein a position and a rotation angle of the image capturing object are obtained by processing image data representing an image captured by the image capturing apparatus. And an image processing apparatus for detecting a defect of an object to be imaged based on the acquired position and rotation angle data and defect detection data provided in advance.