JP4718231B2 - Shape measuring apparatus and program - Google Patents

Description

  The present invention relates to a technique for measuring the shape of an observation object from at least two differential interference images of the observation object with different polarization component retardation amounts.

  In the differential interference microscope, the light applied to the object is divided into two, and the phase difference and interference color caused by the interference of these lights can be visualized. Therefore, there is an advantage that the shape can be observed without staining even a living body or a transparent sample. Accordingly, it is widely used for observing an object having a fine structure such as a living body or an IC pattern. In addition to observation, it is also used for measuring the phase distribution and fine shape of an object. The shape measuring apparatus performs measurement using a differential interference image obtained by a differential interference microscope. Examples of conventional shape measuring apparatuses include those described in Patent Documents 1 and 2 or Non-Patent Document 1.

  For example, Patent Document 1 describes a shape measuring apparatus that can be applied to phase shift reticle defect detection and phase detection, assuming that the differential interference microscope is a shearing interferometer and a Mach-Zehnder interferometer. However, the shape measuring apparatus applies a known interference measuring technique, and does not consider the influence of light diffraction generated on the observation object surface. Also, no consideration is given to the influence on the change in light intensity due to the change in the reflectance and transmittance of light in the observation object. For these reasons, the measurement accuracy is low for an observation object whose shape is not smooth.

  For example, Patent Document 2 or Non-Patent Document 1 discloses a conventional shape measuring apparatus that takes into account the effects of light diffraction and intensity change on the observation object surface. The shape measuring apparatus extracts the phase information of the observation object from the image obtained by the differential interference microscope as follows. Here, in order to simplify the explanation, it is assumed that an image of one-dimensional data is assumed.

  If the relative retardation (phase difference) amount of two deflection components in the differential interference microscope is θ, the intensity distribution I (x, θ) of the one-dimensional image in the differential interference microscope can be expressed as follows.

Here, Φ (f) is the Fourier coefficient of the phase distribution Φ (x) of the observation object, C is the transmittance or reflectance at the point x, f is the spatial frequency, and Δ is the shear amount of the Nomarski prism. ing. Each of the first to third terms of equation (1) right-hand side, the transmitted light component without being diffracted, the differential value component of the phase distribution, and refracted light component and multiple non-linear image component of the frequency of the observation object surface, the It is table. If the pupil function of the imaging optical system mounted on the differential interference microscope is P (ξ), and the intensity distribution function of the pupil of the illumination optical system is S (ξ), M (f) in equation (1), m0 (f) and md (f) can be expressed as follows.

  By using the difference image information and the sum image information of two differential interference images in which the polarization component retardation amounts are θ and −θ, the phase information of the observation object can be separated and extracted. Specifically, the difference image information and the sum image information can be approximated by the following expressions (2) and (3) using Expression (1), respectively.

  The phase distribution θe (x) of the stepped portion of the observation object can be obtained by Expression (4) derived using Expression (2) and Expression (3). The process for obtaining the phase distribution θe (x) is called a deconvolution process.

Therefore, the phase distribution θe (x) of the observation object can be obtained by performing the deconvolution process using the optical response characteristic of the differential interference microscope for the differential interference image. When the observation object has a structure having an edge such as a step, an accurate phase amount can be measured in consideration of the effects of light diffraction and scattering at the edge portion.
JP-A-5-149719 JP-A-9-15504 “High-precision step measurement using a phase modulation differential interference microscope”, Oplus E, Vol. 23, no. 12

  By the way, when the spatial frequency is near zero (plane), the value of the frequency response function (OTF) of the differential microscope is close to zero. Therefore, the calculation result of 1 / (sin (πΔf) M (f)) becomes unstable when the phase distribution θe (x) of the portion near the plane on the observation object is obtained by the deconvolution process. As a result, the deconvolution process cannot be applied to that portion.

  As a solution, a method of increasing the amount of data by increasing the number of significant digits and performing the deconvolution process can be considered. However, even if such a method is adopted, the instability cannot be completely eliminated. In addition, other problems such as a longer processing time and a larger amount of memory are required for the processing. Therefore, it can be said that the adoption of this method is not desirable.

  In the plane of the observation object and a portion close thereto (hereinafter collectively referred to as a “planar portion”), it may be considered that incident light is regularly reflected or transmitted and diffracted light is not generated. It can be said that it can be handled as being formed by interference of only transmitted light. Accordingly, it is considered that the deconvolution process is not necessarily employed in the measurement of the shape of the planar portion existing in the observation object.

  The present invention can measure information indicating a shape such as a shape distribution or a phase distribution with high accuracy regardless of the shape of an observation object, that is, whether or not an edge portion or a planar portion is mixed. It aims at providing a shape measuring device.

The shape measurement apparatus according to the first aspect of the present invention is a shape measurement apparatus that measures the shape of an observation object from at least two differential interference images having different amounts of retardation of the polarization component of the observation object. an image acquisition unit configured to acquire, from at least two differential interference image acquired by the image acquiring means, image forming means for forming an associated component image having image components associated with the differential value of the observation object, an image forming means A region distribution unit that divides the formed related component image into one or more types of regions by paying attention to the contrast, and performs processing according to the type of each region divided by the region division unit to form a phase distribution. Component forming means to perform, and the component forming means picked up a differential interference image in a region of a type having a relatively large contrast change as a process according to the type of the region. Perform deconvolution processing using the separation characteristics of the polarization components in the interference microscope and the response characteristics obtained from the optical characteristics of the differential interference microscope, and in a region where the contrast change is not relatively large, Perform integration using the amount of separation.

In the shape measuring apparatus according to the second aspect of the present invention, in addition to the configuration of the first aspect, the component forming unit forms each region when the region dividing unit divides the related component image into a plurality of types of regions. It performs synthesis of position phase distribution you, synthesizing means for play the position phase distribution of the observation object, further comprising the.

In addition, it is desirable that the synthesizing unit performs the synthesis in consideration of the inclusion relationship between the areas divided by the area dividing unit.
The upper image forming unit desirably subtracts each corresponding pixel from at least two differential interference images acquired by the image acquiring unit to form a related component image having an image component. Alternatively, a difference image obtained by performing subtraction for each corresponding pixel and a sum image obtained by performing addition for each corresponding pixel are formed from at least two differential interference images acquired by the image acquisition unit. A difference image obtained by forming a related component image using the formed difference image and the sum image, or subtracting each corresponding pixel from at least two differential interference images acquired by the image acquisition means It is desirable to form product images obtained by performing multiplication for each corresponding pixel, and to form related component images using the formed difference images and product images.

It is desirable that the region dividing unit divides the related component image into two types of regions according to the magnitude of the contrast change by paying attention to the contrast .

The program of the present invention has a function of acquiring a differential interference image in a computer used as a shape measuring apparatus that measures the shape of the observation object from at least two differential interference images having different retardation amounts of the polarization component of the observation object; , from at least two differential interference image acquired by the function of acquiring a function to form an associated component image having image components associated with the differential value of the phase distribution of the observation object, the associated component image formed by the function of forming A function of dividing the region into one or more types by paying attention to the contrast, and a function of performing processing according to the type of the region for each region divided by the dividing function, and forming a phase distribution; As a process according to the type of the region, the function for performing the phase distribution is used for the region with a relatively large change in contrast. A type in which the deconvolution process using the amount of separation of the polarization component in the differential interference microscope capturing the differential interference image and the response characteristic obtained from the optical characteristics of the differential interference microscope is performed, and the change in contrast is not relatively large In this area, integration processing using the separation amount of the polarization component is performed.

  According to the present invention, the information indicating the shape such as the shape distribution or the phase distribution is obtained regardless of the shape of the observation object, that is, whether or not the edge (step) portion or the planar portion is mixed. It can measure with high accuracy.

Embodiments of the present invention will be described below with reference to the drawings.
<First Embodiment>
FIG. 1 is a configuration diagram of a shape measurement system in which the shape measurement apparatus according to the first embodiment is employed.

  As shown in FIG. 1, the system includes a light source 1, an illumination optical system, and an imaging optical system, and a differential interference microscope 2 that forms a differential interference image of the observation object SM by the imaging optical system. The CCD camera 3 for capturing the differential interference image formed by the imaging optical system and the shape measuring device 4 according to the present embodiment are provided. Thereby, the shape measuring apparatus according to the present embodiment is realized as a device that measures the shape of the observation object SM using the differential interference image captured by the camera 3.

  The differential interference microscope 2 includes a retardation changing device 21 for changing the amount of retardation, a lens 22, a diaphragm 23 and a lens 24 arranged on the light incident side of the changing device 21, and a changing device therefor. A half mirror 25 that reflects the light linearly polarized by the light 21 toward the observation object SM, a Nomarski prism 26 that divides the light reflected by the mirror 25 into two linearly polarized lights orthogonal to each other, and the prism 26 transmitted therethrough. An objective lens 27 that condenses the light on the observation object SM, an analyzer 28 on which reflected light from the observation object SM enters through the objective lens 27, the Nomarski prism 26, and the half mirror 25, and the analyzer 28. An imaging lens 29 that focuses the light after the two polarized light beams interfere with each other and forms a differential interference image of the observation object SM; It is configured Te.

  In the differential interference microscope 2, the light emitted from the light source 1 enters the retardation changing device 21 through the lens 22, the diaphragm 23, and the lens 24. The changing device 21 is for giving an arbitrary amount of retardation to the polarized light incident on the Nomarski prism 26. For example, as shown in FIG. 3, a polarizer 21a and a λ / 4 plate 21b arranged around the optical axis indicated by the alternate long and short dash line in FIG. 3 and the polarizer 21a are rotated around the optical axis. The motor 21c is provided. With such a configuration, the retardation amount can be arbitrarily changed through adjustment (change) of the rotation angle of the polarizer 21a by the motor 21c. The motor 21 c is driven by the shape measuring device 4. Note that the retardation amount may be changed by rotating the λ / 4 plate 21b instead of the polarizer 21a.

  The light linearly polarized by the retardation changing device 21 is reflected by the half mirror 25 and enters the Nomarski prism 26. Thereby, the light is divided into two linearly polarized lights orthogonal to each other by the prism 26, and the divided linearly polarized lights are respectively condensed on the observation object SM by the objective lens 27. Those polarized lights reflected on the observation object SM are again transmitted through the objective lens 27 and then combined by the Nomarski prism 26. The combined polarized light passes through the half mirror 25 and then passes through the analyzer 28 to interfere with each other. By making the combined and interfered polarized light incident on the imaging lens 29, a differential interference image of the observation object SM is formed on the CCD camera 3. The shape measuring device 4 measures the shape of the observation object SM using the differential interference image formed as such and imaged by the CCD camera 3.

  FIG. 2 is a diagram for explaining the functional configuration of the shape measuring apparatus 4. As shown in FIG. 2, an image acquisition unit 41 that acquires a differential interference image captured by the CCD camera 3 and converted into an electrical signal, and an arithmetic process that performs arithmetic processing on the differential interference image acquired by the acquisition unit 41. Unit 42, motor drive unit 43 for driving motor 21c constituting retardation changing device 21, input unit 44 for a user to make various inputs, and display unit for displaying an image on a display device 45, a control unit 46 that controls the entire apparatus 4, and a storage unit 47 that can store various data.

  The shape measuring apparatus 4 having the above-described functional configuration is realized by loading a program that realizes the units 41 to 44 and 46 into a computer (data processing apparatus), for example. The storage unit 47 is, for example, a hard disk device, and the program is stored in the storage unit 47. The arithmetic processing unit 42 and the control unit 46 are realized by the CPU executing the program stored in the storage unit 47 while using the resources installed in the computer as needed. Each of the image acquisition unit 41 and the motor drive unit 43 is an interface for connecting to an external device, for example, and each of the input unit 44 and the display unit 45 is, for example, such an interface or one or more devices. . For example, the input unit 44 is a pointing device such as an interface, a keyboard, and a mouse. Here, for convenience, the following description will be made assuming that the input unit 44 and the display unit 45 each include the latter, that is, the input unit 44, including a device that is a target of operation by the user, its interface, and the like.

The operation of the above configuration will be described.
The user operates the input unit 44 to instruct the shape measurement of the observation object SM. When the instruction is given, the control unit 46 instructs the motor driving unit 43 to drive the motor 21c, and rotates the polarizer 21a so as to obtain a predetermined retardation amount. After the rotation, the image acquisition unit 41 is instructed to take a differential interference image by the CCD camera 3, and the image is acquired. After acquiring one differential interference image in this manner, the motor drive unit 43 is driven so that another retardation amount is obtained, and the differential interference image is captured with the retardation amount. .

  In this way, at least two differential interference images having different retardation amounts are acquired. Here, for convenience, it is assumed that two differential interference images are acquired and the retardation amounts are θ and −θ. The acquired two differential interference images are sent to the arithmetic processing unit 42.

  The arithmetic processing unit 42 has a functional configuration including an image forming unit 51, a region dividing unit 52, a phase unit 53, a combining unit 54, and a converting unit 55. Each part 51-55 performs the following processes, respectively.

  The image forming unit 51 performs a difference operation (subtraction) for each corresponding pixel from the two differential interference images (image data) acquired by the image acquisition unit 41, and uses the difference therebetween as pixel data. An image D is formed, and an image having an image component related to the differential value of the phase distribution of the observation object SM (hereinafter referred to as “related component image”) is formed using the formed difference image D. The related component image is specifically formed as follows.

As described in Patent Document 2, by forming the difference image D, the phase components can be separated as in the conventional fringe scanning method. First, this will be briefly described.
Patent Document 1 describes a method of measuring the phase distribution of an observation object by adopting a fringe scanning method used for interference measurement in a differential interference microscope. Normally, when the fringe scanning method is adopted in the differential interference microscope, four differential interference images having polarization component retardation amounts of 0, π / 2, π, and 3π / 2 are captured, and each of these four differential interference images is captured. The differential value of the phase distribution of the observation object is obtained using the pixel data as follows.

  When a differential interference image having a polarization component retardation amount of, for example, ± π / 2 is captured and a difference image is formed, information corresponding to I (x, π / 2) −I (x, 3π / 2) is obtained from the difference image. Is obtained. The same applies to any retardation amount. For this reason, by forming the difference image D, it is possible to form an image component related to the differential value of the phase distribution of the observation object SM.

  Further, by adding each corresponding pixel from the two differential interference images whose acquired retardation amounts are θ and −θ, and forming a sum image S using the sum between them as pixel data, , The strength component can be separated. Further, by calculating ((1−cos θ) / sin θ) · (D / S), it is possible to form an image component related to the differential value of the phase distribution of the observation object SM that is not affected by the fluctuation of the intensity component. it can. Therefore, in the present embodiment, as the related component image, an image component having an image component formed by the calculation result of ((1−cos θ) / sin θ) · (D / S) from the difference image D and the sum image is used. Forming. The difference image D may be adopted as the related component image.

  The area dividing unit 52 receives the related component image from the image forming unit 51 and performs area division. The region division is performed by performing processing for extracting an edge existing in the related component image and binarizing the result.

  Edge extraction can be performed by a first-order differential operation, but may be performed by a second-order differential filter such as a SOBEL filter or a Laplacian filter. The binarization is performed based on whether or not the threshold value is equal to or greater than a predetermined threshold value. In order to remove (cut) the noise component, expansion processing and contraction processing are performed on the binarized image.

  By the binarization as described above, the related component image is divided into two types of regions, that is, a portion where the image contrast change is large (1-pixel portion) and a portion where the contrast change is not large (small) (0-pixel portion). Is done. Due to the characteristics of the differential interference microscope 2, a portion where the image contrast change is large (1-pixel portion) corresponds to a step portion on the observation object SM, and a portion where the image contrast change is small (0-pixel portion) is on the observation object SM. Corresponds to a planar (plane or near plane) portion. The phase calculation unit 53 receives the related component image and the binarization (region division) result.

  The phase calculation unit 53 calculates a phase distribution from the related component image, and includes first and second phase distribution calculation units 53a and 53b. The first phase distribution calculation unit 53a targets the 1-pixel portion (stepped portion) as a deconvolution process (formula (4)) using a response characteristic OTF obtained from the polarization component separation amount and optical characteristics of the differential interference microscope 2. To obtain a phase distribution φe. The second phase distribution calculation unit 53b calculates the phase distribution φs of the 0-pixel portion as follows.

  In the planar portion on the observation object SM, light is regularly reflected or transmitted, so that it is difficult to obtain a diffraction component. Thereby, the intensity distribution I (x, θ) of the one-dimensional image in the differential interference microscope 2 can be approximated as follows.

I (x, θ) ≈ (C / 2) m0 (f ′) (1-cos (2πf ′ + θ)) (6)
Here, f ′ is a specific frequency appearing in the planar portion.
Assuming that the frequency f ′ is sufficiently small, the differential value (∂φs / ∂r) of the phase distribution on the observation object SM surface corresponding to the separation direction r of the two polarization components is expressed by the following equation (7) or It can be approximated as in (8).

∂φs / ∂r≈D (7)
∂φs / ∂r≈ ((1-cos θ) / sin θ) · (D / S) (8)
For this reason, the second phase distribution calculation unit 53b calculates the phase distribution φs by performing an integration process using the polarization component separation amount of the differential interference microscope 2.

  As described above, in the present embodiment, the related component image is divided into two types of regions according to the magnitude of the image contrast change, and processing appropriate to the type is performed for each divided region to obtain the phase distribution. Calculate φ. Therefore, the phase distribution φ can always be calculated (measured) with high accuracy regardless of the shape of the observation object SM, without adopting measures such as increasing the number of significant digits. Therefore, by adopting such a correspondence, it is possible to reliably avoid the processing time becoming longer and the required memory amount becoming larger.

  The calculation of the phase distribution φ is performed from two differential interference images having the same or almost the same retardation amount but different signs. Accordingly, the phase distribution φ is calculated from fewer images compared to the technique described in Patent Document 2. Therefore, the imaging of the observation object SM can be performed in a shorter time, and the phase distribution φ can be obtained more quickly and easily.

  The synthesizer 54 receives from the phase calculator 53 the phase distributions φe and φs calculated by the first and second phase distribution calculators 53a and 53b, respectively, and the binarization result. The phase distributions φe and φs are calculated for each region. Therefore, the synthesizing unit 54 synthesizes the data with reference to the binarization result. More specifically, the synthesis is performed as follows.

  First, referring to the binarization result (binarized image), for example, by examining the connectivity of the 1-pixel portion, all areas formed by 1-pixels are specified, and the specified areas are connected component sets. Summarize as CE elements. Similarly, all the areas formed by 0-pixels are specified, and the specified areas are collected as elements of the connected component set CS. After putting them together, the inclusion relationship between the elements (regions) constituting the connected component sets CE and CS is determined. The phase distributions φe and φs are combined in consideration of the determined inclusion relationship.

  FIG. 6 is a diagram for explaining a data synthesis method by the synthesis unit 54. Here, the data synthesis method will be described more specifically with reference to FIG. In FIG. 6, an image of the observation object SM having regions such as CS0 and CE0 above is shown, and a phase component φ on the line L drawn on the image is shown below. For example, CS0 represents the 0th element belonging to the connected component set CS. The same applies to the other CS1 and CE0. Therefore, in the example shown in FIG. 6, the inclusion relationship between the elements is that the element CS1 is included in the element CE0, and the element CE0 is included in the element CS0. Here, the original phase component (phase value) at the point A on the element CS0 is expressed as φA, and the phase components (phase values) at the other points B to D are also expressed in the same manner. The phase component after synthesis) and the phase value) are indicated with “′”.

  Since the phase component φ is calculated for each region, as shown in FIG. 6, the original phase component φ is discontinuous at the boundary of the element (region). However, such discontinuities do not actually occur. For this reason, data synthesis is performed to eliminate the discontinuity.

  In the present embodiment, discontinuity elimination, that is, data synthesis, is performed by matching the phase component φ of one element forming the boundary with the phase component φ of the other element. In order to eliminate the discontinuity as described above, it is necessary to appropriately select the operation of the phase component φ. Otherwise, the effect of eliminating the discontinuity at a certain boundary will affect other boundaries and the discontinuity may not be resolved. In the example shown in FIG. 6, for example, after the phase component φB at the point B is matched with the phase component φA at the point A, the phase component φC at the point C is matched with the phase component φD at the point D. This is because the phase components φB and φA are not matched. In order to avoid such a situation, the side on which the phase component φ is operated is determined in consideration of the inclusion relationship between the elements.

  In the example shown in FIG. 6, the element CE0 includes the element CS1, and the element CS0 includes the element CE0. Therefore, when only the elements CS0, CS1, and CE0 are targeted, data synthesis is performed in the order between the elements CS0 and CE0 and between the elements CE0 and CS1, for example. Thereby, data composition is performed in a form based on the element CS0.

  When the element CS0 is used as a reference, the phase component φA is treated as φ′A. As a result, an operation is performed on the phase component φ of the element CE0 using the difference (= φ′A−φB) between the phase components φ′A and φB to synthesize them. As a result, the phase components φ′B and φ′C are as follows.

φ′B = φB + | φ′A−φB |, φ′C = φC + | φ′A−φB | (9)
Similarly, the operation on the phase component φ of the element CS1 is performed using the difference between the phase components φ′C and φD (= φ′C−φD). As a result, the phase component φ′D is as follows.

φ′D = φD + | φ′C−φD (10)
By referring to the inclusion relationship between the elements, as described above, data synthesis is performed while changing the element of interest in the order of the element adjacent to the reference element and the element adjacent to the adjacent element. Thereby, the composition is appropriately performed so that the discontinuity of the phase component φ does not occur between the elements. The combined data (phase component) is sent to the conversion unit 55.

The conversion unit 55 converts the phase component φ to the height h as follows.
h = k · φ (11)
K in the equation (11) is a type in which the differential interference microscope 2 reflects and observes the observation object SM when the wavelength of the light emitted from the light source 1 is λ, and therefore k = λ / (4π (n -1)). When the observation object SM is of a type that transmits and observes, k = λ / (2π (n−1)). If it is a type that performs transmission observation, shape measurement can be performed even if the observation object SM is transparent.

  The conversion unit 55 calculates the height h at each point. Thereby, height information for one image is generated and sent to the control unit 46. The control unit 46 stores the height information in the storage unit 47 or displays it on the display unit 45.

  FIG. 4 is a flowchart of the calculation process. The arithmetic processing unit 42 is realized by the CPU executing the arithmetic processing. Next, the arithmetic processing will be described in detail with reference to FIG. A program for realizing the arithmetic processing is stored in the storage unit 47. The program is activated, for example, when the user operates the input unit 44 to instruct the shape measurement of the observation object SM.

  First, in step S1, the polarizer 21a of the retardation changing device 21 is rotated by the motor drive unit 43, and the CCD camera 3 is imaged by the image acquisition unit 41, so that the retardation amounts are 2 of θ and −θ. Two differential interference images are sequentially acquired, a difference image D and a sum image S are formed from the differential interference images, and a related component image is formed using the formed difference image D and sum image S. After the related component image is formed, the process proceeds to step S2. The image forming unit 51 constituting the arithmetic processing unit 42 is realized by executing this step S1.

  In step S2, an edge on the related component image is extracted, and region extraction is performed using the extraction result. After the division, the phase component φe is calculated in step S3 in a region where the image contrast change is relatively large, and the phase component φs is calculated in step S4 in a region where the change is not relatively large (small). Thus, after calculating the phase component φ of the region for each type, the process proceeds to step S5. The area dividing unit 52 constituting the calculation processing unit 42 is realized by executing Step S2, and the phase calculation unit 53 is executed by executing Steps S3 and S4. More specifically, in the phase calculation unit 53, the first phase distribution calculation unit 53a is realized by executing step S3, and the second phase distribution calculation unit 53b is executed by executing step S4.

  In step S5, a data synthesis process for synthesizing the phase distribution φ of the region calculated for each type is performed. In subsequent step S6, a process of converting the phase component φ of the entire differential interference image of the observation object SM obtained by executing the synthesis process into a height is performed (formula (11)). The height information thus obtained is stored in the storage unit 47 or displayed on the display unit 45, and then a series of processing is terminated. The combining unit 54 is realized by executing step S5, and the converting unit 55 is executed by executing step S6.

  Depending on the differential interference image, the entire image may be regarded as a stepped portion or a planar portion. When such an image is acquired, only one of steps S3 and S4 is executed. In the data synthesizing process in step S5, data synthesis is substantially not performed.

  FIG. 5 is a flowchart of the data composition process executed as step S5. Next, the data composition process will be described in detail with reference to FIG. As described above, the data synthesizing process is performed in consideration of the region division result in step S2.

  First, in step S11, by referring to the division result, the connectivity of the 1-pixel portion is examined to identify all regions formed by 1-pixels, and the identified regions are identified as elements of the connected component set CE. Extract as Similarly, in the subsequent step S12, all the areas formed by the 0-pixels are specified, and the specified areas are extracted as elements of the connected component set CS. Thereafter, the process proceeds to step S13.

  In step S13, the inclusion relation between the elements (regions) constituting the connected component sets CE and CS is determined. In the next step S14, the phase distribution φ calculated in each region is synthesized in consideration of the determined inclusion relation. A series of processing ends after the synthesis.

As described above, it is possible to appropriately synthesize the phase component φ (data) by determining the inclusion relationship and synthesizing the phase distribution φ in consideration thereof.
In the first embodiment, the conversion from the phase distribution φ to the height h is performed after the synthesis, but may be performed before the synthesis.
<Second Embodiment>
In the first embodiment, the related component image is formed using the difference image D and the sum image S. On the other hand, in the second embodiment, the related component image is formed using the difference image D and the product image M. The product image M is obtained by performing product operation (multiplication) for each corresponding pixel from the two differential interference images (image data) acquired by the image acquisition unit 41 and having those multiplication values as pixel data. is there.

  The configuration of the shape measuring apparatus according to the second embodiment is basically the same as that in the first embodiment. The operation is largely the same or basically the same. For this reason, only the parts different from the first embodiment will be described using the reference numerals in the description of the first embodiment as they are.

In the second embodiment, the difference image D and the product image M are formed from two differential interference images having retardation amounts of π / 2 and −π / 2. The image forming unit 51 forms a related component image by calculating tan ⁻¹ [D / (2√M)] using the difference image D and the product image M.

In the first embodiment, the calculation of ((1-cos θ) / sin θ) · (D / S) using the difference image D and the sum image S is performed to form a related component image. By forming the related component image in this way, considerable information can be obtained with the conventional fringe scanning method (formula (5)), but it is not completely consistent. However, when tan ^-1 [D / (2√M)] is calculated, information consistent with the fringe scanning method can be obtained.

  In the second embodiment, the image forming unit 51 forms a related component image as described above. The other units 52 to 55 constituting the arithmetic processing unit 42 perform the same processing as in the first embodiment. For this reason, in the arithmetic processing shown in FIG. 4, there is a difference in the processing content of step S1.

The image component in the planar portion on the related component image formed as described above has the following relationship with the differential value ∂φs / ∂r of the phase component φs.
∂φs / ∂r = tan ⁻¹ [D / (2√M)] (12)
Therefore, integration processing is performed on the planar portion in the same manner as in the first embodiment.

  FIG. 7 is a diagram for explaining various images handled in the second embodiment. FIG. 7A shows a differential interference image with a retardation amount of π / 2, FIG. 7B shows a differential interference image with a retardation amount of −π / 2, and FIG. 7C shows the differential interference image. Further, the imaging result (related component image) of the differential value of the shape or phase distribution, and FIG. 7D are the finally reproduced shape or phase distribution result (three-dimensional diagram thereof).

  FIGS. 7A to 7D show a case where the shape of the observation object SM having a shape in which stepped portions and flat portions are alternately arranged is measured. Even with such an observation object SM, the entire image can be accurately reproduced as is apparent from FIG. The same applies to the first embodiment.

  In this embodiment (the first and second embodiments), the shape of the observation object SM is measured from two differential interference images having different retardation amounts. May be more. For example, shape measurement may be performed using three differential interference images with retardation amounts of 0, π / 2, and 3π / 2. The combination of retardation amounts and the number of images may be different.

  Region segmentation for related component images is performed by focusing on the magnitude of contrast change and dividing it into two types of regions depending on whether it is relatively large or not. You may make it divide an area | region into. For the calculation of the phase component φ in each region, another processing method may be arbitrarily adopted as long as the processing method is appropriate for the type.

  The differential interference image is acquired directly from the differential interference microscope 2, but may be acquired via a communication network or a portable recording medium. Similarly, a program that realizes all or a part of each of the units 41 to 46 shown in FIG. 2 may be loaded into a computer (data processing apparatus) via a communication network or a recording medium. In such a case, the shape measuring apparatus to which the present invention is applied can be realized by loading the program into a computer or an existing shape measuring apparatus. Various modifications other than those described above may be made without departing from the scope of the present invention.

1 is a configuration diagram of a shape measuring system in which a shape measuring device according to a first embodiment is employed. It is a figure explaining the functional composition of the shape measuring device by a 1st embodiment. It is a figure explaining the structure of a retardation change apparatus. It is a flowchart of a calculation process. It is a flowchart of a data composition process. It is a figure explaining a data composition method. It is a figure explaining the various images handled by 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Light source 2 Differential interference microscope 3 CCD camera 4 Shape measuring device 21 Retardation change device 41 Image acquisition part 42 Operation processing part 43 Motor drive part 44 Input part 45 Display part 46 Control part 47 Storage part 51 Image formation part 52 Area division part 53 Phase calculator 53a First phase distribution calculator 53b Second phase distribution calculator 54 Synthesizer 55 Converter

Claims (8)

A shape measurement device that measures the shape of an observation object from at least two differential interference images having different amounts of retardation of the polarization component of the observation object,
Image acquisition means for acquiring the differential interference image;
From at least two differential interference image acquired by the image acquiring means, image forming means for forming an associated component image having image components associated with the differential value of the phase distribution before Symbol observation object,
A region dividing unit that divides the related component image formed by the image forming unit into one or more types of regions focusing on contrast;
For each region divided by the region dividing unit, a process according to the type of the region is performed, and a component forming unit that forms the phase distribution, and
The component forming means, as a process according to the type of the region, the separation amount of the polarization component in the differential interference microscope that has captured the differential interference image in the region of the type where the change in contrast is relatively large, and the Perform deconvolution processing using response characteristics obtained from the optical characteristics of the differential interference microscope, and perform integration processing using the separation amount of the polarization component in a region where the change in contrast is not relatively large.
A shape measuring apparatus characterized by that. When the region dividing unit divides the related component image into a plurality of types of regions, the component forming unit combines the phase distribution formed for each region and reproduces the phase distribution of the observation object. ,
The shape measuring apparatus according to claim 1, further comprising: The synthesizing means performs the synthesis in consideration of an inclusion relationship between the areas divided by the area dividing means;
The shape measuring apparatus according to claim 2. The image forming unit performs subtraction for each corresponding pixel from at least two differential interference images acquired by the image acquisition unit to form a related component image having the image component.
The shape measuring apparatus according to claim 1. The image forming unit includes a difference image obtained by performing subtraction for each corresponding pixel from at least two differential interference images acquired by the image acquisition unit, and a sum obtained by performing addition for each corresponding pixel. Forming each image, and forming the related component image using the formed difference image and the sum image;
The shape measuring apparatus according to claim 1. The image forming means includes a difference image obtained by subtracting each corresponding pixel from at least two differential interference images obtained by the image obtaining means, and a product obtained by multiplying each corresponding pixel. Forming each image, and forming the related component image using the formed difference image and product image;
The shape measuring apparatus according to claim 1. The region dividing unit divides the related component image into two types of regions according to the magnitude of the contrast change by paying attention to the contrast.
The shape measuring apparatus according to claim 1, 2, or 3. In a computer used as a shape measuring device for measuring the shape of an observation object from at least two differential interference images having different retardation amounts of the polarization component of the observation object,
A function of acquiring the differential interference image;
From at least two differential interference image acquired by the function of the obtaining, the function of forming an associated component image having image components associated with one of the differential value of the phase distribution before Symbol observation object,
A function of dividing the related component image formed by the function to be formed into one or more types of regions by paying attention to contrast;
For each of the areas divided by the function of dividing, the processing according to the type of the area is performed, and the function of forming the phase distribution is realized,
As a process corresponding to the type of the region, the polarization component is separated by a differential interference microscope that captures the differential interference image in a region having a relatively large change in contrast as a process according to the type of the region by the function of forming the phase distribution. And a deconvolution process using a response characteristic obtained from the optical characteristics of the differential interference microscope, and an integration process using a separation amount of the polarization component in a region where the contrast change is not relatively large ,
A program characterized by that.