JPH0412310A - Automatic focusing device - Google Patents

Abstract

PURPOSE:To evaluate the extent of focusing at each image pickup position by small edge quantity and to shorten a focusing operation time by counting the number of edge points which are detected by an edge detecting means, finding plural extremal values from an edge quantity-image pickup means position curve, and selecting the best position among them. CONSTITUTION:An image pickup means 2 is moved within a preset range, the edge detecting means 4 detects edges of a picked-up image, and an edge counting means 5 counts the number of edges to find the distribution of the number of edges. An extremal value position detecting means 6a finds plural extremal value positions from its distribution configuration and an optimum position selecting means 6b selects the optimum position which is set previously among them to obtain a focusing position. Consequently, even a subject whose focusing position is hard to decide such as a wafer where a multi-layered pattern is formed can be focused with high reliability and the extent of focusing can be evaluated at each image pickup position by small edge quantity counting to shorten the focusing operation time.

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention is based on a plurality of image information obtained by continuously photographing the surface condition of a wafer with an ITV (industrial television) camera while moving. In a device that automatically inspects the wafer, it is necessary to correct the defocus at each position on the wafer due to uneven wafer thickness or uneven flatness of the inspection stage, and capture images sequentially. The present invention relates to an automatic focusing device that realizes focusing in time.

[Conventional technology]

Regarding conventional automatic focusing devices, [Electronic Materials 198
The explanation will be based on the contents of "Reticle quality evaluation device using silicon wafer method" published in February 1999 issue (Kogyo Kenkyukai) J.

FIG. 6 shows the relationship between the boundary data detected from the image, that is, the number of edge points and the position of the imaging means. The in-focus position is the imaging means position where the number of edges is maximum.

To find the in-focus position using this principle, for example,
The position of the imaging means is moved several steps at a predetermined pitch, and the number of edges is counted each time to find the position of the imaging means with the maximum number of edges.

[Problem to be solved by the invention]

However, in the method of determining the in-focus position using the reticle quality evaluation apparatus using the silicon wafer method, for example, in the image of sea urchin A on which a multilayer pattern is formed, there are multiple in-focus positions. The position curve will not be as shown in FIG. 6, but will be a curve as shown in FIG. 7, which is a combination of several curves in FIG. 6 with some deviations in the direction of the imaging means position.

In such a case, the first problem is that when determining the in-focus position, it is not possible to simply determine the position with the maximum number of edge points.

In addition, in the quality evaluation device using the silicon wafer method described above,
Based on the above principle, in order to find the in-focus position, even if the focus tends to be out of focus globally due to factors such as uneven thickness, such as with wafers, the imaging means position must be determined in advance. The second problem is that it is necessary to move several steps at a predetermined pitch, count the number of edges each time, and determine the focusing position from the form of the edge point-imaging means position curve, which takes a long focusing time. was there.

The invention as claimed in claim 1 has been made to solve the above-mentioned problems, and can be applied to wafers on which multilayer patterns are formed, etc.
To provide an automatic focusing device that enables highly reliable focusing even for a subject whose focusing position is difficult to determine.

In addition, in order to solve the above problem, the invention of claim 2 makes it possible to evaluate the degree of focus by counting a small number of edges for each imaging position when there is a global trend in defocus. An object of the present invention is to obtain an automatic focusing device that can shorten operation time.

[Means to solve the problem]

The automatic focusing device according to claim 1 includes an edge counting means for counting the number of edge points detected by the edge detecting means, and a plurality of extreme values obtained from the edge point number-imaging means position curve. A means is provided for selecting the optimum position from among them.

Further, the automatic focusing device according to the second aspect of the present invention is provided with a control means for predicting the current direction of defocus by using the previous focusing information and determining the focusing position.

[For production]

The edge counting means in the invention of claim 1 counts the number of edge points detected by the edge detecting means, and calculates a plurality of extreme values from the edge point-imaging means position curve within a preset movement range of the imaging means from the counted number of edge points. is determined, the optimal position is selected from these multiple extreme values, and the optimum position is set as the in-focus position.

Further, the control means in the invention of claim 2 first moves in the same direction as the previous movement direction of the imaging means in order to predict the current position on the edge point number-imaging means position curve from the current number of edge points. When the number of edge points obtained by moving the imaging means is compared with the number of edge points obtained by moving the imaging means for the second time, if the result increases, it is determined to be in focus, and if it decreases, it is determined to be out of focus, and the in-focus position is determined. decide.

〔Example〕

Embodiments of the automatic focusing device of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of one embodiment. In FIG. 1, reference numeral 1 denotes a camera as an imaging means having an imaging element.

This camera 1 is configured to focus on the subject 3 by moving along the optical axis using a Z table 2, approaching or moving away from the subject 3.

In the embodiment shown in FIG. 1, the object 3 is a wafer, and in the following description, the object 3 will be described as the wafer 3.

The relationship between the distance between the wafer 3 and the imaging means, that is, the camera, and the number of edge points is shown in FIG. In FIG. 2, reference numeral 21 represents the edge score--imaging means position curve of the first image taken by the camera 1.

In the first focusing operation, an image of the subject 3 is captured by the camera 1, and the captured image is sent to the edge detection means 4.

This edge detecting means 4 is for detecting the contour part of the imaging target of the input image, that is, edges, and the number of edge points detected by this edge detecting means 4 is counted by an edge point counter as an edge counting means. 5, and the counting result is sent to the first focus control means 6.

Based on this counting result, the first focus control means 6:
Focusing on the subject 3 is performed. This first focus control means 6 includes an extreme position detection means 6a and an optimum position selection means 6b.
It is a combination of

In FIG. 2, 23 is edge point data measured by the first focus control means 6, the number of edge points at 20 set points is counted, and the edge point number-imaging means position curve in the first imaging screen is calculated. 21, the extreme value position detection means 6a detects a plurality of extreme values in this edge point number-imaging means position curve 21, and the optimum position selection means 6b selects the optimum position from the plurality of extreme values. The arbitrary focus position 2 determined by the first focus control means 6
Determine 5.

In addition, when creating the edge point number-imaging means position curve 22 of the second captured image, as shown by the edge point data 24 measured by the second focus control means 7, the first focus control Focus position 25 determined by means 6
Based on this, the number of edge points at the determined neighboring points is counted, the focus shift is corrected, and the correct focus position 26 determined by the second focus control means 7 is obtained.

Next, a second embodiment of this invention (invention according to claim 2)
I will explain about it. FIG. 3 is a block diagram showing the configuration of this second embodiment.

In FIG. 3, reference numeral 41 denotes an imaging means having a moving mechanism movable in the optical axis direction. This imaging means 41 is composed of a camera 41a and a Z table 41b.

Edges in the image taken of the subject 45 by the camera 41a of the imaging means 41 are detected by an edge detection means 42, and the number of edges detected by the edge detection means 42 is counted by an edge counting means 43. It looks like this.

Based on the edge points counted by the edge counting means 43,
The control means 44 outputs a movement command signal to the Z table 41b.

FIG. 4 is a conceptual diagram for explaining the second embodiment and a 1□ diagram showing a curve of the number of edge points and the position of the imaging means.

311 to 313 in the figure indicate the positions of the imaging means 41 at consecutive times, that is, the (n-1)th, nth, and (n+1)th imaging times.

In addition, 321 to 323 are each time, that is, (
The position of the subject 45 (observation point on the wafer, hereinafter referred to as observation point) in the (n−1)th, nth, and (n+1)th imaging is shown.

Furthermore, 331 to 333 are the (n-1)th and n
Edge point number-imaging means position curves 341 to 343 for the (n-1)th, n-th, and (n+1)-th times of imaging respectively indicate the in-focus positions of the (n-1)th, n-th, and (n+1)th times of imaging.

FIG. 5 is a flowchart showing the flow of the focusing operation based on this second embodiment, and the description will be made along the flowchart of FIG. 5.

The wafer 12 as the subject 45 has thickness unevenness, as shown by the hatched area in the center of FIG. The imaging means 4 in the previous time, that is, the (n-1)th imaging
At the observation point 321 in the (n-1)th imaging at the position 311 of 1, focus correction in the FAR direction (direction in which the imaging means 41 is moved away from the subject 41) is performed.

Focusing is performed at the current observation point 322 (ON point) observed by the imaging means 41 during the n-th imaging.

In this case, first, in step S1, an image of the subject 45 is captured by the imaging means 41, in step S2, edges of the image are detected by the edge detection means 42, and in step S3, the number of edge points N+ detected by the edge detection means 42 is detected by the edge detection means 42. The counting means 43 counts.

This is the edge point-imaging means position curve 332 (Cn) corresponding to the imaging position 312 of the imaging means 41 for the nth time (this time), and the counted edge point number Nl is the in-focus position 342 for the nth imaging. .

Next, in step S4, the direction in which the imaging means 41 is moved along the optical axis is determined. This direction is FAR (direction that moves the imaging means 41 away from the subject 45) or NEAR.
(a direction in which the imaging means 41 is brought closer to the subject 45), but when the imaging means 41 was at the position 311 in the previous time, that is, the (n-1)th imaging, it moved in the FAR direction and focused. Therefore, this time too, it is predicted to be FAR, and the direction is determined to be FAR.

Next, the process proceeds to step S5, and in step S5, the imaging means 41 is moved in the direction D, that is, in the direction of FAR. This amount of movement is a preset amount ΔP (movement pitch).

Subsequently, a new image is taken in step S6, the edges of the image taken in step S7 are detected by the edge detection means 42, and the edges of the image taken in step S7 are detected by the edge counting means 43 in step S8.
The number of edge points N2 detected is measured. The count value is Hn at the focus position 343 at the n-th imaging.

Next, the process proceeds to step S9, where the number of edge points N1 counted in step S3 and the number of edge points N1 counted in step S8 are calculated.
Compare with 2. As a result of this comparison, if the prediction is correct and the preset amount ΔP is appropriate, N t > N
+, the focus is correctly performed, and the flowchart exits to the N0 side of step S9 to end the operation.

Further, as a result of the comparison between the count values N1 and N2 in step S9, if the preset amount ΔP is not appropriate and the prediction is incorrect, N + > N z and Y in step S9 is determined.
Moving from the ES side to step SIO, this step 310
In, position 3 of the imaging means 41 in the n-th imaging
12, in order to return to the position before the movement, that is, the position 311 of the imaging means 41 in the (n-1)th imaging, the direction is changed by the same amount ΔP in the direction NEAR opposite to the direction of FAR.
move only. This ΔP is determined in advance based on the expected defocus range, error tolerance, and imaging means moving mechanism.

Position 31 at the (n+1)th imaging by the imaging means 41
In the case of No. 3 as well, the focusing operation is performed by repeating the above series of processes.

Further, the first focusing operation (n=0) is performed using the method described in the conventional example.

In the second embodiment, a wafer is used as the subject 45, but the same effect can be achieved with other subjects having uneven thickness.

〔Effect of the invention〕

As described above, according to the invention of claim 1, the imaging means is moved within a preset range, edges of the captured image are detected by the edge detection means, and the number of detected edges is counted by the edge counting means. The distribution of the number of edge points is obtained by counting, the extreme value position detecting means calculates a plurality of extreme value positions from the distribution form, and the optimum position set in advance is selected from the plurality of extreme value positions by the optimum position selecting means. Since the lens is configured to be in focus, highly reliable focusing is possible even for objects for which it is difficult to determine the focus position, such as a wafer on which a multilayer pattern is formed.

Further, according to the invention of claim 2, in order to obtain the current number of edge points of the captured image and predict the current position on the edge point number-imaging means position curve, the imaging means is controlled to be the same as the previous one by the control means. The edge points are calculated again by moving the camera in the direction, and the edge point data of these two times are compared.If the result of the comparison is that the edge points increase in the second time, the focus is focused, and if it decreases, the focus operation is performed. Since the system is configured to detect an error, if there is a global trend in defocus, the degree of focus can be evaluated with a small number of edge counts for each imaging position, which has the effect of shortening the focusing operation time. .

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an automatic focusing device according to a first embodiment of the present invention, FIG. 2 is a curve diagram of the number of edges versus the position of an imaging means in the first embodiment, and FIG. FIG. 4 is a block diagram showing the configuration of an automatic focusing device according to the second embodiment, and FIG. 4 is a diagram showing the moving situation of the imaging means and the edge point-imaging means position curve diagram for explaining the second embodiment. 6 is a flowchart showing the operation flow of the second embodiment, FIG. 6 is an edge point-imaging means position curve diagram for explaining a reticle quality evaluation apparatus using the conventional silicon wafer method, and FIG. 7 is a diagram showing the conventional silicon wafer method. FIG. 4 is a curve diagram showing the number of edge points versus the position of an imaging means when a focusing position is obtained using a wafer image on which a multilayer pattern is formed using a reticle quality evaluation apparatus using a wafer method. 1.41a-Camera, 2.41b-Z table, 3.
45...Subject, 4.42...Edge detection means, 5
... Edge point counter, 6... First focus control means, 6a... Extreme position detection means, 6b... Optimum position detection means, 7... Second focus control means, 41. . . . Imaging means, 43 . . . Edge counting means, 44 . . . Control means.

Claims (2)

[Claims] (1) An imaging device having an image sensor that can be moved arbitrarily along the optical axis, an edge detection device that detects the outline of the object to be imaged from an image captured by the imaging device, and an edge detected by the edge detection device an edge counting means for counting the number of edges, and the imaging means are moved within a preset range;
and extreme value position detection means for determining the extreme value position of the occurrence distribution of the number of edge points counted by the edge counting means, and optimal position selection means for selecting the optimal position from the plurality of extreme value positions determined by the extreme value position detection means. and causing the edge counting means to count the number of edge points at a nearby point determined based on the focus position determined by the first focus control device during a second focusing operation. and second focus control means for controlling the position of the imaging means so as to correct the focus shift and obtain a correct focus position. (2) an imaging means having an image sensor that can be moved arbitrarily along the optical axis; an edge detection means for detecting the outline of the object from the image taken by the imaging means; and an edge detected by the edge detection means an edge counting means for counting points; and an edge counting means for moving the imaging means in the same direction as the previous movement direction in order to obtain image focus information from the number of edge points counted by the edge counting means and predict the current position of the imaging means. an automatic focusing device comprising: a control means that compares the number of edge points counted by the edge counting means with the number of edge points counted last time after moving the object; and determines that the current number of edge points is in focus if the number of edge points increases; .