JP2010133711A - Method and device of measuring temporal variation of position of object - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that a method of detecting the position of a mark by image processing cannot grasp high-frequency positional variation due to the time required in the processing. <P>SOLUTION: (a) The moving object is imaged and acceleration of the object is measured. (b) The position of the object at the point of time when the image data are obtained is obtained on the basis of the image data obtained by imaging the object. (c) The temporal variation of the position of the object is obtained on the basis of the positional information obtained on the basis of the acceleration of the object and the position of the object at the point of time when the image data are obtained. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a method and an apparatus for imaging a target and measuring a temporal change in its position.

  In a semiconductor manufacturing process or the like, a method is widely employed in which an alignment mark attached to an observation object is imaged by an imaging device and the position of the mark is detected from the obtained image data (for example, Patent Document 1).

  An image sensor provided on the lower surface of the trolley that is attached to the structure and moves in the lateral direction detects a mark provided on the hanger, and detects deflection displacement and speed of the hanger (Patent Document 2). . An acceleration sensor attached to the structure detects vibrations (displacement, etc.) of the structure. The speed control of the trolley is performed based on the displacement of the structure and the swing displacement of the suspension.

JP 2004-264404 A Japanese Patent Laid-Open No. 8-290892

  The method of detecting the position of the mark by image processing requires time for processing, and thus cannot detect high-frequency position fluctuations. The position can be calculated by second-order integration of the acceleration obtained from the acceleration sensor. However, since the second-order integral value of acceleration is accompanied by a drift, the position calculated based on the acceleration gradually shifts from the true position.

According to one aspect of the invention,
(A) imaging a moving target and measuring the acceleration of the target;
(B) determining the position of the target at the time when the image data was acquired based on image data obtained by imaging the target;
(C) obtaining a time change of the position of the target based on position information obtained based on the acceleration of the target and the position of the target at the time when the image data was acquired. A method for measuring time change of the position of a target is provided.

According to another aspect of the invention,
An imaging device for imaging a target;
An acceleration sensor for measuring the acceleration of the target;
A control device to which image data captured by the imaging device and acceleration data measured by the acceleration sensor are input, the control device: (a) a moving target with a first time interval And measuring the acceleration of the target with a second time step shorter than the first time step, and
(B) determining the position of the target at the time when the image data was acquired based on image data obtained by imaging the target;
(C) obtaining a time change of the position of the target based on position information obtained based on the acceleration of the target and the position of the target at the time when the image data was acquired. A time change measuring device for the position of a target is provided.

  The position of the time zone when image data is not acquired can be determined based on the acceleration. In order to determine the target based on both the position obtained based on the acceleration and the position obtained based on the image data, it is possible to suppress accumulation of errors caused by noise of the acceleration sensor. Can do.

  Embodiments will be described with reference to the drawings.

  FIG. 1A is a schematic diagram of a position time change measuring apparatus according to the first embodiment. A workpiece 20 and a reference member 15 are disposed on a base 10 that defines a plane. The workpiece 20 is a moving body of a stage device, for example, and translates on the base 10 in a two-dimensional direction. The reference member 15 is fixed to the base 10. Reference marks Ro, Rx, and Ry are displayed on the upper surface of the reference member 15. XY orthogonal coordinate system (reference coordinate system) in which the reference mark Ro is the origin, the direction from the reference mark Ro to Rx is the positive direction of the X axis, and the direction from the reference mark Ro to Ry is the positive direction of the Y axis Is defined. A target (target) 11 is fixed to the workpiece 20. A target mark P is displayed on the upper surface of the target 11.

  As shown in FIG. 1B, an acceleration sensor 12 is stored in the target 11. The acceleration sensor 12 measures the acceleration in the X direction and the Y direction, and inputs the measurement result to the control device 40.

  An imaging device 30 is disposed above the base 10. The imaging device 30 includes a light receiving unit 32 and a telecentric lens system 33. Illumination light emitted from the light source 35 is guided to the telecentric lens system 33 via the optical fiber 36 and irradiated to the base 10. A UV orthogonal coordinate system (camera coordinate system) is defined on the image receiving surface of the light receiving unit 32. The reference marks Ro, Rx, Ry and the target mark P on the base 10 are projected on the image receiving surface. The image projected on the image receiving surface is input to the control device 40 as image data.

  FIG. 2 shows a flowchart of a method for measuring time change in position according to the embodiment. First, the target 11 is attached and fixed to the workpiece 20.

  In step ST1, the image on the base 10 is acquired by the imaging device 30. In the UV camera coordinate system, the positions of the reference marks Ro, Rx, Ry are detected. Thereby, the correspondence between the XY reference coordinate system and the UV camera coordinate system is obtained.

  Hereinafter, an example of a method for detecting the position of the reference mark Ro will be described with reference to FIGS. 3A and 3B. The position detection of the other reference marks Rx, Ry and the target mark P is performed in the same manner.

  As shown in FIG. 3A, images of the reference member 15, the reference marks Ro, Rx, Ry, the workpiece 20, the target object 11, and the target mark P are projected on the image receiving surface on which the UV camera coordinates are defined. First, a square or rectangular analysis target region 50 including the reference mark Ro is cut out from the image data. The position of the analysis target area 50 in the UV camera coordinates, for example, the UV coordinate of the lower left vertex 51 in FIG. 3A is stored.

  FIG. 3B shows a plan view of the analysis target region 50. Noise is removed by performing low-pass filtering or contraction / expansion operation on the image data in the analysis target region 50. Based on the image data from which the noise has been removed, the position of the reference mark Ro in the analysis target region 50 is detected by performing centroid calculation or peak value detection. When performing the centroid calculation, a threshold is set for the image signal intensity, and an image signal equal to or higher than the threshold is set as the centroid calculation target. Since image signals less than the threshold value are not subject to calculation, they are less affected by the background.

  The reference mark Ro or the like may be intentionally out of focus. If the focus is shifted, it is less likely to be affected by minute fluctuations in the shape of the outer peripheral line such as the reference mark Ro and the fine gradation of the background.

  Based on the position of the analysis target area 50 in the camera coordinates and the position of the reference mark Ro in the analysis target area 50, the UV coordinates of the reference mark Ro in the camera coordinate system are calculated.

  Next, in step ST2 of FIG. 2, the acceleration sensor 12 is calibrated. Specifically, the workpiece 20 is moved by a predetermined distance, and the movement distance is calculated from the output of the acceleration sensor 12. The actually moved distance can be calculated from the image data from the imaging device 30. The acceleration sensor 12 is calibrated so that the distance actually moved is equal to the movement distance calculated from the output of the acceleration sensor 12. The moving distance at this time is set to such an extent that the error of the moving distance caused by accumulating the noise components of the acceleration measured by the acceleration sensor 12 can be ignored.

  In step ST <b> 3, while moving the workpiece 20, image data is collected by the imaging device 30 and acceleration data is collected by the acceleration sensor 12. Image data is collected at a time step size Δtc, and acceleration data is collected at a time step size Δta. The time interval Δta for collecting acceleration data is shorter than the time interval Δtc for collecting image data.

FIG. 4 shows a block diagram of a method for measuring time change of position in the first embodiment. In step ST3, the imaging device 30 generates image data I (t) including the target mark P with a time interval Δtc. The control device 40 performs image processing on the image data I (t), and obtains the camera coordinates (Uc, Vc) of the target mark P at the time when the image data is acquired. The acceleration sensor 12 outputs acceleration (a _x , a _y ) with a time interval Δta. Here, a _x and a _y are accelerations in the X direction and the Y direction, respectively.

  In step ST4, as shown in FIG. 4, the position (Uc, Vc) of the target mark P in the camera coordinates is converted into a position (Xc, Yc) in the reference coordinate system.

FIG. 5A shows an example of a temporal change in the position (Xc, Yc) of the target mark P. The actual trajectory of the target mark P is indicated by a broken line. The black circle symbols shown at times t ₀ , t ₁ , t ₂ ,... In FIG. 5A indicate the positions (Xc, Yc) obtained based on the image data. The time increments t ₁ -t ₀ , t ₂ -t ₁ , t ₃ -t ₂ ,... Are equal to Δtc. Due to the restriction of the shutter speed of the imaging device 30, image data cannot be acquired with a step size shorter than the time step size Δtc.

In step ST5, as shown in FIG. 4, the acceleration _(a x, _{a y)} by second-order integration time change (time-series signal), calculates the position of the target mark P (Xa, Ya).

  FIG. 5B shows a time change (time-series signal) of the position (Xa, Ya) of the target mark P calculated based on the acceleration. Actually, the position (Xa, Ya) is calculated for each time step size Δta. However, since the time step size Δta is sufficiently shorter than the time step size Δtc, the time at the position (Xa, Ya) is shown in FIG. 5B. The change is represented by a continuous line. The actual trajectory of the target mark P is indicated by a broken line.

  Since the error due to the noise of the acceleration sensor is accumulated at the position obtained from the acceleration, the difference between the calculated position and the actual position increases with time.

  In step ST6, as shown in FIG. 4, the position (Xp, Yp) of the target mark P is determined based on the position (Xc, Yc) obtained from the image data and the position (Xa, Ya) obtained from the acceleration. Find time change. Hereinafter, a method for obtaining the time change of the position (Xp, Yp) of the target mark P will be described.

As shown in FIG. 5C, at time t ₀ , the position (Xa ₀ , Ya ₀ ) obtained from the acceleration is obtained from the acceleration so that it matches the position (Xc ₀ , Yc ₀ ) obtained from the image data. It was position (Xa, Ya) to (increase or decrease the X coordinate and Y coordinate) of adding the correction _{a 0.} The corrected position (Xa, Ya) is adopted as the position (Xp, Yp) of the target mark P from time t ₀ to t ₁ . The corrected position (Xa ₁ , Ya ₁ ) at time t ₁ does not match the position (Xc ₁ , Yc ₁ ) obtained from the image data due to accumulation of noise components of the acceleration sensor.

At time t ₁ , the position (Xa ₁ , Ya ₁ ) obtained from the acceleration is corrected so that the position (Xc ₁ , Yc ₁ ) obtained from the image data matches the position (Xc ₁ , Yc ₁ ) obtained from the image data. xa, the Ya), further adding the correction _{a 1.} Position of the target mark P from time _{t 1} to _{t 2} as (Xp, Yp), to adopt a position after the addition of the correction _{A 1} (Xa, Ya). After time t ₂ , the position (Xp, Ya) similarly obtained from the acceleration is sequentially corrected based on the position (Xc, Yc) obtained from the image data, so that the position (Xp, Yp) is determined.

In general, acceleration data can be acquired at a shorter cycle than image data. By using the acceleration and the image data together, the target mark position P can be obtained even during a period when the image data cannot be obtained. Further, the position (Xa, Ya) obtained from the acceleration is corrected so as to coincide with the position (Xc, Yc) obtained from the image data at the time t ₀ , t ₁ , t ₂ ,. By doing so, it is possible to suppress accumulation of errors caused by accumulating noise components of the acceleration sensor.

In the first embodiment, the position of the target mark P changes discontinuously at times t ₁ , t ₂ , t ₃ . Position at time _t 1 -0 (Xp, Yp) is the position at time _t 1 +0 (Xp, Yp) to match the position at time _t 0 -0 (Xp, Yp) may be modified. At this time, the position (Xp, Yp) at an intermediate time between the times t ₀ and t ₁ may be corrected by interpolation calculation.

  Next, a method for measuring a time change of a position according to the second embodiment will be described with reference to FIG. The flowchart of the method according to the second embodiment is the same as the flowchart of the method according to the first embodiment shown in FIG. 2, and the specific execution method of step ST5 is different between the first and second embodiments.

  FIG. 6 shows a block diagram of a method for measuring time change of position according to the second embodiment. The processing for obtaining the position (Xc, Yc) based on the image data and the processing for obtaining the position (Xa, Ya) based on the acceleration are the same as those in the first embodiment.

In the second embodiment, the first low-pass filter LPF1 is applied to the position (Xc, Yc) obtained from the image data. The transfer function _{F L} of the first low-pass filter LPF1 is as follows.

A high-pass filter HPF is applied to the position (Xa, Ya) obtained from the acceleration. The transfer function F _H of the high pass filter HPF is as follows.

  Here, s represents a Laplace operator (differential operator), ω represents a cutoff angular frequency, and ζ represents an attenuation rate. The cut-off angular frequency ω is set lower than the imaging frequency of the imaging device 30. The attenuation rate ζ is 0.8 as an example.

The acceleration second-order integration process (transfer function 1 / s ² ) and the high-pass filter HPF can be replaced with one second low-pass filter LPF2. The transfer function _{F L2} of the second low-pass filter LPF2 is as follows.

Adding the position (F _L2 a _X , F _L2 a _Y ) after applying the second low-pass filter LPF2 to the position (F _L Xc, F _L Yc) after applying the first low-pass filter LPF1 Thus, the position (Xp, Yp) of the target mark P is calculated.

Assuming that the actual position of the target mark P is (X, Y), the output (a _X , a _Y ) of the acceleration sensor 12 has the noise components of the acceleration sensor 12 as (Δa _X , Δa _Y ) as follows: expressed.

  Further, it is assumed that noise components (ΔXc, ΔYc) are superimposed on the position (Xc, Yc) obtained from the image data. The position (Xp, Yp) of the target mark P is expressed as follows.

In the above equation, the error (ΔXc, ΔYc) included in the position obtained from the image data is attenuated by the low-pass filter process. As a result, low-pass filter processing is applied to the noise components (Δa _X , Δa _Y ) of the acceleration sensor, so that they do not accumulate and cause a large error.

  In this way, the high-frequency component of the position time-series signal is acquired from the acceleration, the low-frequency component of the position time-series signal is acquired from the image data, and both are superimposed. By using the acceleration and the image data together, the target mark position P can be obtained even during a period when the image data cannot be obtained. Further, an increase in error due to accumulation of noise components of the acceleration sensor can be prevented.

The transfer functions of the first low-pass filter LPF1 and the second low-pass filter LPF2 are not limited to the above. A transfer function F _L of the first low-pass filter LPF1, as the sum of the transfer function multiplied by the transfer function F _L2 to s ² of the second low-pass filter LPF2 becomes 1, by selecting the transfer functions of both Good.

In Example 2, after acquiring all the time series data of the coordinates (Xc, Yc) and acceleration (a _X , a _Y ) obtained from the image data, these data are converted into the first low-pass filter LPF1 and the second low-pass filter LPF1. You may input into the low-pass filter LPF2. Further, the time series data being acquired may be input to the first low-pass filter LPF1 and the second low-pass filter LPF2 while the time series data is being acquired, that is, while the work 20 is moving. . In this method, the position can be measured while the workpiece 20 is moving. For this reason, the position measurement result can be fed back to the movement control of the workpiece 20.

  Using the position data measured in the first and second embodiments, the position of the moving body (stage) can be controlled. Further, this stage can be employed in a semiconductor manufacturing apparatus.

  Although the present invention has been described with reference to the embodiments, the present invention is not limited thereto. It will be apparent to those skilled in the art that various modifications, improvements, combinations, and the like can be made.

(1A) is a schematic perspective view of the position time change measuring apparatus according to the first embodiment, and (1B) is a perspective view of the target. 3 is a flowchart of a method for measuring a time change of a position according to the first embodiment. (3A) is a plan view of the reference member and the workpiece, and (3B) is a view showing an image of the reference mark in the image receiving surface. It is a block diagram which shows the time change measuring method of the position by Example 1. FIG. (5A) is a graph showing the time change of the position obtained from the image data, (5B) is a graph showing the time change of the position obtained from the acceleration, and (5C) is a method according to the first embodiment. It is a graph which shows the time change of the position obtained. It is a block diagram which shows the time change measurement method of the position by Example 2. FIG.

Explanation of symbols

10 Base 11 Target (Target)
12 acceleration sensor 15 reference member 20 work 30 imaging device 32 light receiver 33 telecentric lens system 35 light source 36 optical fiber 40 control device 50 analysis target region 51 vertex LPF1 first low-pass filter LPF2 second low-pass filter Ro, Rx, Ry reference Mark P Target mark XY Reference coordinate system UV Camera coordinate system

Claims (8)

(A) imaging a moving target and measuring the acceleration of the target;
(B) determining the position of the target at the time when the image data was acquired based on image data obtained by imaging the target;
(C) obtaining a time change of the position of the target based on position information obtained based on the acceleration of the target and the position of the target at the time when the image data was acquired. Time change measurement method for target position. In the step (a),
Imaging the target at a first time step, measuring an acceleration of the target at a second time step shorter than the first time step;
In the step (c),
At the time when the image data is acquired, the position information obtained from the acceleration is corrected so as to coincide with the position obtained from the image data, and the time until the next image data is obtained The method according to claim 1, wherein the position information obtained from the acceleration is corrected to obtain the position of the target. The step (c)
Applying a first low-pass filter to the time change of the position obtained from the image data to calculate the time change of the first position;
Applying a second low-pass filter to the acceleration time change to calculate a time change of the second position;
The time change measurement of the position of the target according to claim 1, further comprising a step of setting the time change of the position of the target by adding the first time change and the time change of the second position. Method. 4. The target according to claim 3, wherein a sum of a transfer function of the first low-pass filter and a transfer function obtained by multiplying the transfer function of the second low-pass filter by s ² is 1 where Laplace operator is s. Time change measurement method of the position of the. An imaging device for imaging a target;
An acceleration sensor for measuring the acceleration of the target;
A control device to which image data captured by the imaging device and acceleration data measured by the acceleration sensor are input, the control device: (a) a moving target with a first time interval And measuring the acceleration of the target with a second time step shorter than the first time step, and
(B) determining the position of the target at the time when the image data was acquired based on image data obtained by imaging the target;
(C) obtaining a time change of the position of the target based on position information obtained based on the acceleration of the target and the position of the target at the time when the image data was acquired. Time change measurement device for target position. In the step (c),
At the time when the image data is acquired, the position information obtained from the acceleration is corrected so as to coincide with the position obtained from the image data, and the time until the next image data is obtained 6. The apparatus according to claim 5, wherein in the belt, the position information obtained from the acceleration is corrected to obtain the position of the target. The step (c)
Applying a first low-pass filter to the time change of the position obtained from the image data to calculate the time change of the first position;
Applying a second low-pass filter to the time change of the acceleration to calculate the time change of the second position;
The time change measurement of the position of the target according to claim 5, further comprising a step of setting the time change of the position of the target by adding the time change of the first time and the time change of the second position. apparatus. 8. The target according to claim 7, wherein a sum of a transfer function of the first low-pass filter and a transfer function obtained by multiplying the transfer function of the second low-pass filter by s ² is 1, where Laplace operator is s. Time change measuring device of the position of.