JP5620289B2 - Test surface shape measuring method, test surface shape measuring apparatus, and test surface shape measuring program - Google Patents

Description

  The present invention relates to a test surface shape measuring method, a test surface shape measuring apparatus, and a test surface that obtain data of the actual surface shape of a test surface using a share image of the test surface output from an image sensor of an interferometer. The present invention relates to a surface shape measurement program.

  As an interferometer used for observing the condition of the test surface, for example, as disclosed in Patent Document 1, a wavefront generated from the reference surface of the reference lens and a wavefront generated from the test surface are directly used. And detecting the interference fringes by the interference and, as disclosed in Patent Document 2, irradiate the test surface with the laterally shifted light without using the reference lens, and synthesize the reflected light. Thus, it is known that a shared image is detected.

When the configuration disclosed in Patent Document 1 is applied, if the test surface is a spherical surface, the shape error of the test surface can be obtained directly from the wavefront aberration distribution using a simple spherical reference lens. However, if the surface to be tested is aspheric, partial measurement of the reference sphere and data synthesis are required, and enormous image synthesis processing is required after detecting the interference image. There were practical problems in terms of processing speed and the like.
Of course, even with such a configuration, if the reference aspheric surface is created with a module that generates a physical aspheric surface with high precision as the reference surface and the corresponding wavefront, partial measurement and data of the reference sphere However, this method is not practical because it requires manufacturing of a different physical reference aspheric surface for each object to be examined, which requires a lot of cost and preparation time.
Whatever measures are taken, the two inspection light paths required for forming the interference image are separated from each other, so the characteristics that are easily affected by the measurement environment such as vibration and air fluctuation are Inconveniences that cannot be resolved and require a large amount of capital investment for environmental improvement of the measurement room and the like remain.

On the other hand, when the configuration disclosed in Patent Document 2 is applied, a wavefront aberration distribution is obtained by performing a fringe scanning process or a phase unwrapping process based on the detected shear image, and this wavefront aberration distribution is obtained. In general, the condition of the surface to be inspected is estimated from the above.
However, the wavefront aberration distribution obtained in this way is not the result of a direct comparison between the wavefront generated from the reference surface and the wavefront from the test surface, so the absolute shape of the test surface must be specified. That is, there is an inconvenience that it is difficult to directly grasp the actual surface shape of the test surface.

Japanese Patent Application No. 2007-557049 Japanese Patent Application No. 2006-503124

  Therefore, an object of the present invention is to use a shared image of the test surface output from the image sensor of the interferometer, and to detect the actual surface shape directly regardless of the shape of the test surface. The object is to provide a surface shape measuring method, a surface shape measuring apparatus, and a surface shape measuring program.

The test surface shape measurement method of the present invention is a test surface shape measurement for directly grasping the actual surface shape of a test surface using a share image of the test surface output from the image sensor of the interferometer. In order to achieve the object,
A reference spherical surface for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data obtained by processing the image data of the share image of the test surface output from the image sensor of the interferometer in the logical space. Virtual,
After generating wavefront distribution surface data based on the reference spherical surface based on the reference spherical surface and the wavefront aberration distribution data,
A wavefront distribution side method that represents the reflected light of each minute cell by obtaining a normal passing through the center of each minute cell obtained by dividing the wavefront distribution surface represented by the data of the wavefront distribution surface. Each line obtained by dividing an illumination light wavefront having a concentric circular wavefront distribution that is configured to follow an optical path converged to one point by an optical system after being emitted from a light source during measurement For each micro cell, the normal passing through the center of the micro cell is obtained and set as the illumination light wavefront side normal for each micro cell,
Based on the fact that each of the wavefront distribution surface side normals follows the optical path of any reflected light of the incident light incident on the test surface following the optical path that converges to one point when measured by the interferometer, For each minute cell of the wavefront distribution surface represented by the data of the wavefront distribution surface, the wavefront distribution surface side normal representing the reflected light that is reflected by the test surface and reaches the minute cell, and the incident light corresponding thereto Identify the illumination wavefront normal that represents the optical path of
For each minute cell of the wavefront distribution surface, the intersection position between the corresponding wavefront distribution surface side normal and the illumination light wavefront normal is obtained and used as actual surface shape data of the surface to be measured. Have

First, the reference spherical surface for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data obtained by processing the image data of the share image of the test surface output from the image sensor of the interferometer with a computer is represented in a logical space. The data of the wavefront distribution surface based on the reference spherical surface is generated based on the data of the reference spherical surface and the wavefront aberration distribution.
If the test surface is a perfect spherical surface, the optical path that converges (images) at one point during the measurement by the interferometer, that is, the incident light that is emitted from the illumination light optical system and enters the test surface is the illumination light. Reflected by the test surface consisting of a sphere centered at the same center as the light, that is, the convergence position of the illumination light, becomes reflected light, and always returns to the original direction by following the same optical path as when incident, but the test surface is completely If it is not a spherical surface, for example, the shape of the test surface that should be a spherical surface is no longer a spherical surface due to abnormalities in processing, damage, or aging, or the design shape of the original test surface is aspheric. If the shape is properly reproduced, the incident light incident on the test surface is reflected by the test surface to become reflected light, and then follows an optical path different from the incident time. Return to the illumination light optical system.
In either case, these reflected lights are finally detected as image elements that form a shared image of the test surface by the image sensor of the interferometer. Is completely spherical, the optical path of the reflected light and the optical path of the incident light are completely coincident. However, if the test surface is not a perfect spherical surface, the optical path of the reflected light is the optical path of the incident light. Does not match at all. That is, the reflected light detected by the image sensor of the interferometer when the test surface is not a perfect spherical surface is generated by reflecting another incident light that does not match the optical path of the reflected light. means.
In short, in the conventional interferometer and the processing device connected to the interferometer, the optical path of the reflected light detected by the image sensor is the same as the optical path of the incident light, or does not match the optical path of the reflected light. The wavefront aberration distribution is finalized on the assumption that the optical path of the reflected light detected by the image sensor matches the optical path of the incident light without distinguishing whether the light is generated by reflecting another incident light. Since it is the structure which outputs as a result, the actual surface shape of a to-be-tested surface cannot be grasped | ascertained directly.
Therefore, in the present invention, even if the reflected light detected by the image sensor of the interferometer is reflected by a perfect spherical surface or reflected by an aspherical surface, at least the reflected light is reflected. The incident light that is the cause of the generation of light is one of the incident light that has entered the test surface following one of the normal paths on the imaginary illumination light wavefront, that is, the optical path that converges to one point when measured by the interferometer. On the basis of the fact that there is no difference, the optical path of the incident light corresponding to this reflected light is determined, and the intersection position of these two optical paths is obtained to obtain the actual surface shape data of the test surface. Use as a basic principle of action.
More specifically, for each micro cell obtained by dividing the wave front distribution surface represented by the data of the wave front distribution surface, a normal passing through the center of the micro cell is obtained, and the wave front distribution side of each micro cell is obtained. In addition to the normal line, a normal line passing through the center of the minute cell is obtained for each minute cell obtained by dividing the illumination light wavefront, and is defined as the illumination light wavefront side normal line for each minute cell. Among these, the illumination light wavefront side normal line corresponds to the optical path of incident light incident on the test surface by following any normal line on the virtual illumination light wavefront, and the wavefront distribution surface has a test signal. Since the reflected light reflected from the surface is always incident at right angles, the wavefront distribution surface normal of each microcell obtained by dividing the wavefront distribution surface is in the optical path of the reflected light reflected from the test surface. It will be equivalent.
Therefore, utilizing that each of the normals on the wavefront distribution surface side coincides with the optical path of any reflected light of the illumination light incident on the test surface, following the optical path that converges to one point at the time of measurement by the interferometer, For each microcell on the wavefront distribution plane represented by the data on the wavefront distribution plane, specify the wavefront distribution plane normal to the microcell and the optical path of incident light corresponding to this normal cell, and each microcell on the wavefront distribution plane If the intersection position between the corresponding wavefront distribution surface side normal and the illumination light wavefront normal is obtained for each time, data of the actual surface shape of the test surface can be obtained.

  More specifically, for example, when the test surface is a convex surface and the test surface is a perfect spherical surface, the test is performed from the distance from the illumination light wavefront to the convergence position of the incident light that converges to the one point. If the distance obtained by subtracting the paraxial radius of curvature of the surface is the same as the distance from the illumination light wavefront to the test surface, and the test surface is a concave surface, and the test surface is a perfect sphere, In accordance with the principle that the distance obtained by adding the paraxial radius of curvature of the test surface to the distance to the convergence position of the incident light that converges at the one point coincides with the distance from the illumination light wavefront to the test surface, the wavefront distribution plane side method The illumination light wavefront normal intersecting the line and the intersections of the illumination light wavefront normal and the wavefront distribution surface normal are all obtained, and when the test surface is a convex surface, the wavefront distribution surface and the intersection The wavefront distribution side normal length between and the illumination light wavefront between the illumination light wavefront and the intersection The sum of the normal lengths, that is, the sum of the optical path lengths of the incident light and the reflected light is a single point at the separation distance between the convergence position of the illumination light (incident light) that converges to the one point and the illumination light wavefront. Add the convergence position of the convergent illumination light and the distance to the reference spherical surface used to calculate the wavefront distribution surface, and subtract the value twice the paraxial radius of curvature of the test surface design from this value. While approximating the illuminating light wavefront normal to be approximated, if the test surface is a concave surface, the length of the wavefront distribution side normal between the wavefront distribution plane and the intersection and the illuminating light wavefront and the intersection The distance between the convergence position of the illumination light (incident light) where the sum of the lengths of the illumination light wavefront side normals, that is, the sum of the optical path lengths of the incident light and the reflected light converges at the one point, and the illumination light wavefront Is added to the distance from the convergence position of the illumination light that converges to one point to the reference sphere used to calculate the wavefront distribution surface. By calculating the illumination light wavefront side normal closest to the distance obtained by adding twice the design paraxial radius of curvature of the test surface, the wavefront distribution surface side method corresponding to this illumination light wavefront normal is obtained. Lines can be identified.

  Also, when reflecting on an arbitrary surface placed in a uniform medium, when considering a tangent plane that exists on that surface and reflects the light beam, it is perpendicular to the tangent plane from that point. In accordance with the principle that the incident angle and the reflection angle of light are equal to the normal set in (Snell's law), all wavefront surface normals that intersect the wavefront surface side are obtained, and the wavefront surface side The angle formed by the normal of the test surface assumed at the position where the normal and each illumination light wavefront normal intersect, the normal of the wavefront distribution surface, the normal of the test surface, and the obtained illumination light wave The illumination light wavefront side normal that minimizes the difference in angle between the surface side normal and the illumination light wavefront side normal is identified as the illumination light wavefront side normal corresponding to the wavefront distribution surface side normal. May be.

  Alternatively, according to the principle that light is incident and reflected symmetrically with respect to the normal of the test surface, all the illumination light wavefront normals intersecting the wavefront distribution surface normal are obtained, and the wavefront distribution surface normal and each Illumination in which the normal of the test surface assumed at the position where the illumination light wavefront side normal intersects, the wavefront distribution surface side normal, and the obtained illumination light wavefront normal are on the same plane The light wave front normal may be obtained, and the illumination light wave front normal may be specified as the illumination light wave front normal corresponding to the wave distribution surface normal.

  In particular, when it is assumed that an aspheric lens or the like is used as the test surface, the test surface is assumed to be a position where the wavefront distribution surface side normal and each illumination light wavefront normal intersect. It is reasonable to use the surface of a sphere centered on the convergence position of incident light that converges at the one point.

The test surface shape measuring apparatus of the present invention achieves the same object as described above,
A reference spherical surface for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data obtained by processing the image data of the shear image of the test surface output from the image sensor of the interferometer is hypothesized in the logical space. A reference spherical surface generating means;
Wavefront distribution surface data generation means for generating data of a wavefront distribution surface based on the reference spherical surface based on the reference spherical surface and the data of the wavefront aberration distribution virtualized by the reference spherical surface generation means;
Storage means for temporarily storing calculation data;
For each minute cell, a normal passing through the center of the minute cell is obtained for each minute cell obtained by partitioning the wavefront distribution surface represented by the wavefront distribution surface data generated by the wavefront distribution surface data generating means. Is stored in the storage means as a wavefront distribution plane-side normal representing the reflected light, and a normal passing through the center of the microcell is obtained for each microcell obtained by dividing the illumination light wavefront. The normal data extraction means to be stored in the storage means as the illumination light wavefront normal for each cell, and each of the wavefront distribution plane normals stored in the storage means converges to one point when measured by the interferometer Corresponding to each microcell of the wavefront distribution surface represented by the data of the wavefront distribution surface, based on the fact that it matches the optical path of any reflected light of the incident light incident on the test surface following the optical path Wavefront distribution representing reflected light stored in the storage means An illumination light path specifying means for specifying the illumination optical wavefront side normal line representing the optical path of the incident light corresponding to the Re side normal to 此,
For each microcell on the wavefront distribution surface, the actual surface of the surface to be measured is obtained by calculating the intersection position of the wavefront distribution surface side normal and the illumination light wavefront side normal whose correspondence is specified by the illumination light optical path specifying means It has a configuration characterized by comprising actual surface shape data generating means for making shape data.

  The principle of operation of this test surface shape measuring apparatus is as already described.

The test surface shape measurement program of the present invention is a test surface shape measurement program for controlling a computer that processes image data of a share image of a test surface output from an image sensor of an interferometer, To achieve the same purpose as
Computer
A reference spherical surface for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data obtained by processing the image data of the shear image of the test surface output from the image sensor of the interferometer is hypothesized in the logical space. Reference spherical surface generation means,
Wavefront distribution surface data generation means for generating data of a wavefront distribution surface based on the reference spherical surface based on the reference spherical surface and the wavefront aberration distribution data hypothesized by the reference spherical surface generation means;
For each minute cell, a normal passing through the center of the minute cell is obtained for each minute cell obtained by partitioning the wavefront distribution surface represented by the wavefront distribution surface data generated by the wavefront distribution surface data generating means. Is stored in the storage device of the computer as a wavefront distribution plane side normal representing the reflected light of the light source, and a normal passing through the center of the microcell is obtained for each microcell obtained by dividing the illumination light wavefront. Normal data extraction means for storing in the storage device of the computer as the illumination light wavefront normal for each cell;
Each of the wavefront distribution plane normals stored in the storage device of the computer follows an optical path that converges to one point when measured by the interferometer, and is an optical path of any reflected light incident on the test surface. The wavefront distribution surface side normal representing the reflected light stored in the storage device of the computer corresponding to each minute cell of the wavefront distribution surface represented by the data of the wavefront distribution surface Illumination light optical path specifying means for specifying an illumination light wavefront side normal representing the optical path of incident light corresponding thereto, and
For each microcell on the wavefront distribution surface, the actual surface of the surface to be measured is obtained by calculating the intersection position of the wavefront distribution surface side normal and the illumination light wavefront side normal whose correspondence is specified by the illumination light optical path specifying means It has a configuration characterized by functioning as actual surface shape data generating means for shape data.

  The principle of operation of the computer in which the test surface shape measurement program is installed is as described above.

  The test surface shape measuring method, the test surface shape measuring apparatus, and the test surface shape measuring program according to the present invention determine the optical path of incident light corresponding to the reflected light from the test surface, and determine the intersection position of these two optical paths. Since the actual surface shape data of the surface to be measured is obtained, even if the optical path of the reflected light detected by the image sensor of the interferometer matches the optical path of the incident light, the incident light If the optical path of the test surface does not match the optical path of the reflected light, that is, if the shape of the test surface that should be a spherical surface is no longer spherical due to abnormalities in processing, damage, or secular change, or the design shape of the original test surface Even if is an aspherical surface, the actual surface shape of the test surface can be directly grasped.

  Accordingly, when the test surface is not a perfect spherical surface, for example, when the shape of the test surface that should be a spherical surface is no longer spherical due to processing abnormality, damage, secular change, or the like, or the original test surface Even when the design shape is an aspherical surface, the actual surface shape of the test surface can be easily measured without requiring a special aspherical surface functioning as a reference surface or a module that generates the corresponding wavefront. can do.

It is the functional block diagram which simplified and showed the structure of the to-be-tested surface shape measuring apparatus of one Embodiment in connection with this invention. It is the block diagram shown about the optical system of the interferometer employ | adopted in the embodiment. It is the block diagram which simplified and showed the hardware constitutions of the to-be-tested surface shape measuring apparatus of the embodiment. It is the flowchart shown about the outline of the test surface shape measurement program run by CPU of the test surface shape measuring apparatus of the embodiment. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is a continuation of the flowchart shown about the outline of the test surface shape measurement program. It is the conceptual diagram which showed the structural example of the data table of the memory | storage means which memorize | stores an illumination light wave front side normal and a wave front distribution surface side normal. It is the conceptual diagram illustrated about several optical paths of the 1st, 2nd test | inspection light irradiated and reflected on the to-be-tested surface of a subject. It is the conceptual diagram which showed an example of the correspondence of the wave front of illumination light, and a wave front distribution surface.

  Next, an example is given about the form for implementing this invention, and it demonstrates concretely with reference to drawings.

  FIG. 1 is a simplified functional block diagram showing a configuration of a test surface shape measuring apparatus according to an embodiment of the present invention, and FIG. 2 is a configuration diagram showing an optical system of an interferometer employed in this embodiment. FIG. 3 is a simplified block diagram showing the hardware configuration of the surface shape measuring apparatus.

  First, the configuration of the interferometer 1 that functions as a kind of sharing interferometer will be briefly described with reference to FIG.

  The main part of the interferometer 1 includes a laser light source 2 composed of a He-Ne laser or the like, a polarizing plate 3 functioning as a polarizer for linearly polarized light, a beam expander 4, a condensing lens 5, and a beam splitter 6. The birefringent optical element 7 is connected to a collimator lens 8 and an objective lens 8 ′, a support 9 for setting an object 13 such as an aspherical lens, and a polarizing plate 10 functioning as an analyzer for linearly polarized light. An image lens 11 and an imaging element 12 such as a CCD functioning as a light receiving surface are included.

  Among these, the laser light source 2, the polarizing plate 3, the beam expander 4, the condensing lens 5, the beam splitter 6, the birefringent optical element 7, the collimator lens 8, the objective lens 8 ′, and the support 9 are single in this order. The other polarizing plate 10, the imaging lens 11, and the image pickup device 12 are provided on the side of the beam splitter 6 in such a manner that the optical path is divided from the optical path L1 to the side. .

  The polarizing plate 3 includes a grating in a direction perpendicular to the paper surface of FIG. 2, and linearly polarizes light emitted from the laser light source 2 into polarized light having only vibration in a direction perpendicular to the paper surface of FIG.

  The birefringent optical element 7 includes a Wollaston prism or a Nomarski prism. The birefringent optical element 7 emits light S0 that has been emitted from the laser light source 2 and linearly polarized by the polarizing plate 3 and further expanded in diameter by the beam expander 4 and transmitted through the condenser lens 5 and the beam splitter 6, and the optical axis L1. In addition, they are arranged in a direction that divides into two light beams S1 and S2 whose emission angles are slightly different from each other in a direction perpendicular to the plane of the drawing of FIG. One of the two light beams S1 and S2 is an ordinary ray and the other is an extraordinary ray.

  The collimator lens 8 and the objective lens 8 ′ convert the light beams S1 and S2 emitted from the birefringent optical element 7 with slightly different emission angles into parallel light beams, and then convert these light beams S1 and S2 into the first light beams. This is for irradiating the test surface 13a of the subject 13 such as an aspherical lens as the first and second inspection light.

  The birefringent optical element 7 is provided with an adjuster 14 for moving and positioning and fixing the birefringent element 7 along the optical axis L1 while maintaining the posture of the birefringent optical element 7.

  The support 9 for setting the subject 13 includes a linear movement along the optical axis L1, a linear movement in each axial direction of two axes orthogonal to the optical axis L1, and a two-axis orthogonal to the optical axis L1. The support is allowed to swing around each axis. That is, the support 9 is a 5-axis support having 3 degrees of freedom regarding linear motion and 2 degrees of freedom regarding rotational motion, and the center axis of the subject 13 such as an aspheric lens can be reliably secured by an appropriate operation. The optical axis L1 can be matched.

  The structures of the adjustment tool 14 and the support tool 9 are known as a telescopic mechanism, a tilt / shift mechanism, and the like in an optical device. The adjustment tool 14 is necessary for feeding π / 2 required for the fringe scanning process. A feeding mechanism having a high accuracy is provided.

  The polarizing plate 10 can be rotated, and the arrangement direction of the grating used for linearly polarized light can be adjusted as necessary.

  In the above configuration, the laser beam emitted from the laser light source 2 is first linearly polarized by the polarizing plate 3 in the direction perpendicular to the plane of FIG. 2 and expanded in diameter by the beam expander 4. The birefringent optical element 7 is irradiated through the splitter 6.

  The light S0 irradiated to the birefringent optical element 7 includes a direction crossing the polarization plane of linearly polarized light by the polarizing plate 3 in the process of passing through the birefringent optical element 7, that is, including the optical axis L1 in FIG. The light beams are divided into two light beams S1 and S2 having slightly different emission angles in a direction intersecting with a plane perpendicular to the paper surface, and are further converted into parallel light beams in the process of passing through the collimator lens 8 and the objective lens 8 ′. Thus, the test surface 13a of the subject 13 such as an aspheric lens set on the support 9 is irradiated perpendicularly as the first and second inspection lights.

The first and second inspection lights (light fluxes S1 and S2) irradiated perpendicularly to the test surface 13a are reflected by the test surface 13a of the test subject 13, and the objective lens is oriented in a reverse direction along the path at the time of irradiation. 8 ', the collimator lens 8 and the birefringent optical element 7 are transmitted again to be overlapped, reflected to the side of the optical axis L1 by the beam splitter 6, and disposed at a position where the optical path is split from the optical axis L1 to the side. After passing through the polarizing plate 10 and linearly polarized, it is projected onto the image sensor 12 through the imaging lens 11 to form interference fringes.
However, when the test surface 13a is not a perfect spherical surface, the incident light incident on the test surface 13a is reflected by the test surface 13a and reflected along an optical path different from that at the time of incidence, so that the objective lens 8 ', The light is projected onto the image sensor 12 through the collimator lens 8, the birefringent optical element 7, the beam splitter 6, the polarizing plate 10, and the imaging lens 11.

FIG. 14 is a conceptual diagram illustrating several optical paths of the first and second inspection lights that are irradiated and reflected on the test surface 13a of the subject 13. FIG. l ′ _{1 to} l ′ ₅ are examples of the optical paths of the first and second inspection lights emitted from the objective lens 8 ′ shown in FIG. 2, and as shown in FIG. It enters perpendicularly to 13a, in other words, converges with the sphere center O of the subject 13 as one point. Here, if the full spherical test surface 13a, for example, normal m tangent plane 31 of the sphere as the incident light l _'1 is the reflected light follow exactly the same optical path as the incoming k' ₁ If the test surface 13a is not a perfect spherical surface, the optical path of the reflected light of the incident light l' ₁ is, for example, k " ₁ and the optical path l 'upon incidence is reflected. This is different from _1. This is because the tangent plane 32 itself at the reflection position of the incident light l ′ ₁ is inclined because the surface 13a to be measured is an aspherical surface. If the normal line m ″ ₁ of the tangential plane 32 is taken as a reference, naturally, the angle formed by the incident light l ′ ₁ and the normal line m ″ ₁ of the tangential plane 32 and the reflected light k ″ ₁ and the tangential plane are determined from the relation of incident reflection. 32 normal m for "becomes equal to the _{angle 1} is formed, also light normal m of the test surface 13a" or be symmetrically incident reflective to ₁ Also, it is clear that "the _first reflective light k" optical path and the normal m of the incident light l _'1 and the optical path of ₁ are coplanar.

  In addition, when adjusting the phase of the interference fringe projected on the image pick-up element 12, the adjustment tool 14 is operated and the deployment position of the birefringent optical element 7 is moved along the perpendicular direction of the optical axis L1.

  When the center of the interference fringe is deviated from the center of the image sensor 12, it means that the center axis of the subject 13 such as an aspherical lens does not coincide with the optical axis L1, and thus the support 9 is operated. By adjusting the position and posture of the subject 13, the central axis of the subject 13 is made to coincide with the optical axis L1.

  Further, when no interference fringes are confirmed, it means that the first and second inspection lights are not irradiated perpendicularly to the test surface 13a of the subject 13, so that the support 9 is operated. The distance between the subject 13 and the objective lens 8 ′ is adjusted so that the first and second inspection lights are irradiated perpendicularly to the test surface 13a. The brightness of the interference fringes can be adjusted by rotating the polarizing plate 10.

  As shown in FIG. 2, the light S <b> 0 emitted from the laser light source 2 and transmitted through the polarizing plate 3, the beam expander 4, the condenser lens 5, and the beam splitter 6 is divided into two light beams having slightly different emission angles by the birefringent optical element 7. The light beams are split into light beams S1 and S2, and these lights are made parallel to each other by the collimator lens 8, and irradiated to the test surface 13a of the subject 13 by the objective lens 8 ′ as the first and second test lights. Since the reflected light is projected onto the image sensor 12 through the objective lens 8 ', the collimator lens 8, the birefringent optical element 7, the beam splitter 6, the polarizing plate 10, and the imaging lens 11, the first inspection light and its reflection Regardless of whether the light, the second inspection light, or its reflected light, the optical system through which these light beams pass is completely the same, and the theoretical measurement accuracy and resistance to disturbance are the same as those of conventional Fizeau interferometers. Compare High.

  In addition, regardless of whether the first inspection light and its reflected light or the second inspection light and its reflected light, the optical system through which these light beams pass is completely the same. Even if it acts, the disturbance will act equally on the optical path of the first and second inspection light and the reflected light, and in particular, it has excellent resistance to vibration and generates interference fringes. And can be maintained easily.

  In addition, since both the first and second inspection lights are reflected by the test surface 13a of the subject 13, a minute positional shift between the optical components due to vibration or the like is not necessarily a serious problem. It is possible to maintain interference fringes against this.

  As a result, it is possible to simplify the anti-vibration measures required for the installation of the interferometer 1, and it is not necessary to use a reference lens as a base, so that a complicated shape such as an aspheric lens is used. It is possible to introduce an interferometer that measures the test surface of the subject at low cost.

  Polarizers using various coating techniques on the surface of the beam splitter 6 on the side receiving the light from the laser light source 2 (the left end surface of the beam splitter 6 in FIG. 2) instead of the polarizing plate 3 installed on the optical axis L1. Is formed on the surface of the beam splitter 6 facing the image sensor 12 (the upper end surface of the beam splitter 6 in FIG. 2) instead of the polarizing plate 10 disposed at the position where the optical path is divided from the optical axis L1 to the side. The analyzer may be formed using various coating techniques. Thereby, the interferometer 1 can be further downsized.

  The laser light source 2, the polarizing plate 3, the beam expander 4, the condensing lens 5, the beam splitter 6, the birefringent optical element 7, the collimator lens 8, the objective lens 8 ′, and the support 9 are on a single optical axis L1. Although it is necessary to arrange, the optical axis L1 itself does not need to be along a single straight line.

  For example, in a situation where the total length of the interferometer 1 needs to be shortened, the relative distance between the beam expander 4 and the condenser lens 5 and between the birefringent optical element 7 and the collimator lens 8 is relatively large. By installing one reflection mirror each having an angle of 90 °, the entire optical axis L1 can be refracted into a substantially U shape, and the overall length of the apparatus can be shortened.

  As shown in FIG. 1, the main part of the test surface shape measurement apparatus 16 of this embodiment is obtained by processing image data of a share image of the test surface output from the image sensor 12 of the interferometer 1. The wavefront aberration distribution represented by the wavefront aberration distribution data, the reference spherical surface generation means a for virtualizing the reference spherical surface for calculating the wavefront distribution in the logical space, the reference spherical surface and the wavefront virtualized by the reference spherical surface generation means a Wavefront distribution surface data generating means b for generating wavefront distribution surface data based on the reference spherical surface based on the aberration distribution data, storage means d for temporarily storing calculation data, and wavefront distribution surface data generation A normal passing through the center of the micro cell is obtained for each micro cell obtained by dividing the wave front distribution surface represented by the data of the wave front distribution surface generated by the means b, and the reflected light for each micro cell is represented. Wavefront distribution side normal For each micro cell obtained by partitioning the illumination light wavefront, the normal passing through the center of the micro cell is obtained and stored in the memory d as the illumination light wave front side normal for each micro cell. The normal data extracting means c to be stored and the reflected light reflected by the test surface 13a are always incident on the wavefront distribution surface at right angles, that is, the wavefront distribution surface side normal stored in the storage means d. Based on the fact that each follows the optical path converged to one point during measurement by the interferometer 1 and coincides with the optical path of any reflected light incident on the test surface 13a, it is represented by data on the wavefront distribution plane. The wavefront distribution plane normal representing the reflected light stored in the storage means d corresponding to each small cell of the wavefront distribution plane to be generated and the illumination light wavefront normal representing the optical path of the incident light corresponding thereto are specified. Illumination light optical path specifying means e and illumination light for each micro cell on the wavefront distribution plane Actual surface shape data generating means f that obtains the position of the intersection of the wavefront distribution plane side normal and the illumination light wavefront normal whose correspondence has been specified by the path specifying means e and makes the actual surface shape data of the test surface 13a. Consists of.

  The illumination light optical path specifying means e intersects the wavefront distribution surface side normal line stored in the storage means d and the intersection of the wavefront distribution surface side normal line and the illumination light wavefront side normal line. All are obtained, and the sum of the length of the wavefront distribution plane side normal between the wavefront distribution plane and the intersection and the length of the illumination light wavefront normal between the illumination light wavefront and the intersection converges to the one point. Add the distance from the convergence position of the illumination light that converges to one point to the reference spherical surface used to calculate the wavefront distribution surface to the separation distance between the convergence position of the illumination light (incident light) and the illumination light wavefront, The illumination light wavefront side normal that approximates the distance obtained by subtracting a value twice the design paraxial radius of curvature of the test surface 13a from this value is obtained, and this illumination light wavefront side normal is calculated on the wavefront distribution surface side. An optical path length condition corresponding normal specifying function that specifies the illumination light wavefront side normal corresponding to the normal, and stored in the storage means d All the illumination light wavefront normals intersecting with the wavefront distribution plane side normal are obtained, and the normal of the test surface 13a assumed at the position where the wavefront distribution plane normal and each illumination light wavefront normal intersect An illumination light wavefront normal that minimizes the difference between the angle formed by the wavefront distribution plane side normal and the normal of the test surface 13a and the calculated illumination light wavefront normal is obtained. An incident reflection angle condition corresponding normal specifying function for specifying an optical wavefront normal as an illumination light wavefront normal corresponding to the wavefront distribution plane normal, and a wavefront distribution plane normal stored in the storage means d; All intersecting illumination light wavefront side normals are obtained, and the normal of the test surface 13a assumed at the position where the wavefront distribution surface side normal intersects with each illumination light wavefront side normal, and the wavefront distribution surface side method. The illumination light wavefront side normal line is obtained by locating the illumination light wavefront side normal line and the obtained illumination light wavefront side normal line. Comprising a coplanar condition corresponding normal specific function of specifying the optical wavefront side normal as illumination light wave surface normal corresponding to normal the wavefront distribution side.

  The test surface shape measuring device 16 of this embodiment is constituted by a general computer such as a workstation or a personal computer. As shown in FIG. 3, the CPU 17 for arithmetic processing and the drive control of the CPU 17 are used. ROM 18 (or hard disk drive 21) storing basic control programs required, nonvolatile memory 19 for storing various parameters, RAM 20 used for temporary storage of operation data, and the like, and A hard disk drive 21 or the like that functions as a large-capacity storage device is provided, and a keyboard 24 and a mouse 25 that function as a man-machine interface and a monitor 23 are connected to the input / output circuit 22 of the CPU 17.

  The laser light source 2 of the interferometer 1 shown in FIG. 2 is on / off controlled by the CPU 17 via the input / output interface 26 and the laser drive circuit 27 on the surface shape measuring apparatus 16 side, and supports the interferometer 1. The feed mechanism 15 provided in the tool 9 is also driven and controlled by the CPU 17 via the input / output interface 26 and the driver 28. It is also possible to activate the pulse generator 29 by manual operation from the keyboard 24 and input a pulse signal to the driver 28 to drive the feed mechanism 15 and apply a fine feed to the adjustment tool 14.

  An interference fringe image picked up by the image pickup device 12 such as a CCD provided in the interferometer 1 is read into the CPU 17 via the CCD control circuit 30 and the input / output interface 26.

  The CPU 17 of the test surface shape measuring apparatus 16 includes a reference spherical surface generation means a, a wavefront distribution surface data generation means b, a normal data extraction means c, an illumination light path specifying means e and an actual surface shown in the functional block diagram of FIG. It functions as shape data generation means f, and the CPU 17 functions as reference spherical surface generation means a, wavefront distribution surface data generation means b, normal data extraction means c, illumination light path specifying means e, and actual surface shape data generation means f. And the CPU 17 as the optical path length condition corresponding normal specifying function realizing means, the incident reflection angle condition corresponding normal specifying function realizing means, and the same plane condition corresponding normal specifying function realizing means in the illumination light optical path specifying means e. A test surface shape measurement program including a program for functioning is installed in the hard disk drive 21 in advance.

  The RAM 20 corresponds to the storage means d shown in the functional block diagram of FIG. 1, and temporarily stores the interference fringe image read from the image sensor 12 by the CPU 17 via the input / output interface 26 and the CCD control circuit 30. In addition to the function as a frame memory and the function of temporarily storing intermediate data, the reference spherical surface generated by the reference spherical surface generation means a, the wavefront distribution surface data generated by the wavefront distribution surface data generation means b, and normal data It has a function of storing data related to the wavefront distribution plane side normal and the illumination light wavefront normal obtained by the extraction means c.

  4 to 12 are flowcharts showing an outline of a test surface shape measuring program executed by the CPU 17 of the test surface shape measuring device 16 comprising a computer.

  At this stage, it is assumed that the subject 13 has already been appropriately set on the support 9 and that the operation of the laser light source 2 has been started. The entire reference spherical surface generation means a, wavefront distribution surface data generation means b, normal data extraction means c, storage means d, illumination light path specifying means e, and actual surface shape data generation means f in the test surface shape measuring apparatus 16. A specific processing operation and a test surface shape measurement program will be specifically described.

  The CPU 17 that has started the measurement process first initializes the value of the index i for counting the number of times the image is captured from the image sensor 12 to 0 (step a1), and then increments the value of the index i by 1 ( Step a2), it is determined whether or not the current value has reached the necessary number of fringe scanning, for example, four (step a3).

  If the current value of the index i has not reached four, which is the required number of fringe scanning, the CPU 17 captures the shared image data from the image sensor 12 and uses a part of the storage area of the RAM 20 to construct a frame. The image pickup device 12 is refreshed after being stored in the memory i (step a4), the feed mechanism 15 is driven via the input / output interface 26 and the driver 28, and the adjustment tool 14 and the birefringent optical element attached to the adjustment tool 14 are used. 7 is multiplied by an amount corresponding to π / 2 in the phase of the wavelength of the laser beam (step a5).

  Then, the CPU 17 repeatedly executes the processing of steps a2 to a5 in the same manner as described above, fetches share images having phases of 0, π / 2, π, and 3π / 2 into each of the frame memories 1 to 4, and step When the determination result of a3 is false and the capture of the required number of four share images is confirmed, phase connection processing by phase unwrapping is executed based on these images in the same manner as in the past, and wavefront aberration is performed. Numerical data of distribution is obtained (step a6), and the contents are temporarily stored in the RAM 20 (step a7).

  Next, the CPU 17 functioning as the reference spherical surface generation means a virtualizes the reference spherical surface B required for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data in the RAM 20 serving as a logical space. The reference spherical surface B may be any spherical surface as long as it has the same spherical center as the spherical center O of the subject 13, but here, as an example, on the optical axis of the illumination optical system of the interferometer 1 A reference spherical surface B having a vertex at the exit pupil position is used. Since the positions of the subject 13 and the spherical center O of the reference spherical surface B and the exit pupil plane of the illumination optical system occupy fixed positions on the machine coordinate system of the interferometer 1, special calculations for virtualizing the reference spherical surface B are performed. There is no particular need for processing, and it is sufficient to read the equation representing the reference spherical surface B from the hard disk drive 21 into the RAM 20. Then, the CPU 17 functioning as the wavefront distribution surface data generating means b uses the reference spherical surface B as a reference based on the numerical data of the wavefront aberration distribution temporarily stored in the RAM 20 and the equation of the reference spherical surface B in step a7. Data representing the distribution plane W is generated (step a8).

An example of the correspondence between the reference spherical surface B and the wavefront distribution surface W is shown in the conceptual diagram of FIG. The reference spherical surface B has the same spherical center as the spherical center O of the subject 13 as described above, and the curvature radius has a length connecting the spherical center O and the exit pupil position of the illumination optical system of the interferometer 1. The wavefront distribution surface W is a virtual spherical surface, and is obtained by comparing the numerical data of the wavefront aberration distribution temporarily stored in the RAM 20 with the reference spherical surface B. The numerical data of the wavefront aberration distribution is a collection of simple deviation data obtained by the phase connection process in step a6. Each deviation data is calculated by comparing each deviation data with the reference spherical surface B, and the deviation of each data is calculated as the reference spherical surface. By reflecting in the normal direction of B, data representing the wavefront distribution plane W is obtained.
More specifically, each numerical data of the wavefront aberration distribution is (y _i , z _i ), the radius of the reference spherical surface B centered on the spherical center O is R1, the wavelength of the laser light source 2 to be used is λ1, and the reference spherical surface is When the refractive index of the existing medium is n1 and the wavefront aberration is w, the distance R2 from the spherical center O to the corresponding position of (y _i , z _i ) on the wavefront distribution plane W is R2 = n1 × w × λ1 + R1. , The angle θ formed by the line segment connecting the corresponding position of (y _i , z _i ) on the wavefront distribution plane W and the sphere center O with the optical axis L1 (see FIG. 14) is θ = cos ⁻¹ [(y _i ) · (Y _i / R2)], the corresponding position z of z _i on the wavefront distribution plane W is z = R2 · sin (θ).

Next, the CPU 17 divides the wavefront distribution plane W represented by the data generated by the wavefront distribution plane data generation means b into C1 sections according to a preset cell division condition C1, that is, a set value of the cell division number. Then, cell data W _{1 to} W _{C1 representing} each minute cell are generated and stored in the RAM 20 (step a9). Similarly, the illumination light wavefront E is divided into C2 sections according to a preset cell division condition C2, that is, a set value of the number of cell divisions, and cell data E _{1 to} E _{C2 representing} each minute cell are divided. It is generated and stored in the RAM 20 (step a10).

Next, the CPU 17 is an index for sequentially selecting the value of the counter Q that counts the total number of data representing the actual surface shape of the test surface 13a and the data W _{1 to} W _C1 of each minute cell constituting the wavefront distribution surface W. After both i values are initialized to 0 (step a11, step a12), the value of the index i is incremented by 1 (step a13), and the current value reaches the total number C1 of the minute cells constituting the wavefront distribution plane W. It is determined whether or not (step a14).

If the value of the index i has not reached the total number C1 of the microcells, the CPU 17 functioning as the normal data extraction means c selects one microcell constituting the wavefront distribution plane W from the RAM 20 based on the current value of the index i. read data _{W i} (step a15), the equation of the normal _{k i} clogging microcells _{W i} wavefront distribution surface normals _{k i} corresponding to the passing through the coordinates and the center _{a i} of the center _{a i} of the micro cell _{W i} It is obtained and stored in the RAM 20 functioning as the storage means d (step a16).

As shown in FIG. 15, since the reflected light reflected by the test surface 13a is always incident orthogonally to the wavefront distribution surface W, this wavefront distribution surface side normal line k _i is reflected by the test surface 13a. It has been can be said to be equivalent to the optical path of the reflected light incident on one micro cell W _i which constitutes the wavefront distribution plane W.

Next, the CPU 17 configures the illumination light wavefront E and the value of the counter n that counts the total number of illumination light wavefront normals that intersect the wavefront distribution plane normal k _i selected as the evaluation target at this stage. After initializing both the values of the index j for sequentially selecting the data E _{1 to} E _C2 of each microcell to 0 (step a17, step a18), the value of the index j is incremented by 1 (step a19), It is determined whether or not the current value has reached the total number C2 of minute cells constituting the illumination light wavefront E (step a20).

If the value of the index j does not reach the total number C2 of the micro cells, the CPU 17 functioning as the normal data extraction means c determines the one micro cell constituting the illumination light wavefront E from the RAM 20 based on the current value of the index j. The data E _j is read (step a21), and the coordinates of the center D _j of the minute cell E _{j and} the normal l _j passing through the center D _j , that is, the illumination light wavefront side normal l _j corresponding to the minute cell data E _j An equation is obtained (step a22).
Here, as an example, the minute cell E ₁ for setting the illumination light wavefront side normals l _{1 to} l _C2 representing the optical path of the incident light by dividing the illumination light wavefront E by the same procedure as in the case of the wavefront distribution plane W. virtually a to E _C2, but so as to make a lighting optical wavefront side normal _l 1 to l _C2 at its center _D 1 _{to D C2,} always to make a lighting optical wavefront side normal _l 1 to l _C2 it is not necessary to divide the illumination optical wavefront E, for example, selected in an arbitrary small increments intervals previously user-specified points D ₁ to D _C2 on the illumination optical wavefront E, the spherical center O and the point D ₁ to D _The illumination light wavefront side normals l _{1 to} l _C2 may be obtained from the relations of _C2 . In this case, since the points D _{1 to} D _C2 on the selected illumination light wavefront E mean the centers of the minute cells E _{1 to} E _C2 , processing for dividing the illumination light wavefront E and the minute cells E _{1 to} E The process of step a10 for calculating the center of _C2 is unnecessary.
Alternatively, the straight lines l _{1 to} l _C2 passing through the spherical center O are selected at an arbitrary minute step angle designated in advance by the user, and each of the straight lines l _{1 to} l _C2 is set as the illumination light wavefront side normal line l _{1 to} l _C2. The same applies to the intersections of the illumination light wavefront normals l _{1 to} l _C2 and the illumination light wavefront E as D _{1 to} D _C2 described above.

Next, the CPU 17 functioning as a part of the normal line data extraction unit c and the actual surface shape data generation unit f is selected as the wavefront distribution plane side normal line k _i obtained in the process of step a16, that is, the evaluation target at the present time. It is determined whether or not the wavefront distribution plane side normal line k _i and the illumination light wavefront side normal line l _j intersect (step a23), and the value of the counter n is incremented by 1 only when both intersect (step a24). , the n-th address of the data table g built into the RAM20 which functions as a memory unit d, stores the equations of the illumination optical wavefront side normal l _j as equation l _n, also wavefront distribution surface normals k _i And the coordinates of the intersection of the illumination light wavefront side normal line l _j are stored as the intersection point P _n , and the coordinates of the center D _j of the minute cell E _j are stored as the center D _n (step a25). FIG. 13 shows an example of the logical configuration of the data table g.
In addition, when the determination result of step a23 is false, that is, it is clear that the wavefront distribution plane side normal line k _i currently selected as the evaluation target and the illumination light wavefront side normal line l _i do not intersect. If it does, the processing from step a24 to step a25 is not executed, the value of the counter n is held as it is, and nothing is written in the data table g.

  Then, the CPU 17 again increments the value of the index j by 1 (step a19), determines whether or not the current value has reached the total number C2 of minute cells constituting the illumination light wavefront E (step a20). If not, the process from step a21 to step a25 and the process from step a19 to step a20 are repeated as described above.

Therefore, at the time when the decision result in the step a20 reached the total number of micro cell C2 to the value of the index j constitutes an illumination optical wavefront E is false, the data E ₁ to E of minute cells constituting the illumination optical wavefront E _C2 each wavefront distribution surface method, which is selected as the evaluation target wavefront distribution surface normals k _i clogging moment obtained by the processing in step a16 of the illumination optical wavefront side normal l ₁ to l _C2 corresponding to All the illumination light wavefront normals intersecting the line k _i are extracted, and the equations of the illumination light wavefront normals and the coordinates of the intersection and the center associated therewith are all stored in the data table g. (See FIG. 13).

  The number of data sets stored in the data table g is reflected in the final value of the counter n.

As an example, illumination light waves intersecting one microcell W _i which constitutes the wavefront distribution plane W as an example, in its center A wavefront distribution side through the _i normal k _i and此wavefront distribution surface normals k _i Some examples of surface side normals are shown in FIG.
As already mentioned, the wavefront distribution surface normals k _i is equivalent to the optical path of the reflected light from the test surface 13a that is incident on one micro cell W _i which constitutes the wavefront distribution plane W, illumination wavefront Each of the side normal lines l _{1 to} l ₉ corresponds to an optical path of incident light that has entered one of the test surfaces 13 a following a normal line on the illumination light wavefront E. In the example shown in FIG. 15, the wavefront distribution surface normals k illumination wavefront side normal l ₁ to l ₇ are evaluated at the time of此of the illumination optical wavefront side normal l ₁ to l _{9 Although} it intersects with _i , the illumination light wavefront side normals l _{8 to} l ₉ do not intersect with the wavefront distribution surface side normal line k _i , so these illumination light wavefront side normals l _{1 to} l ₉ are included in the data table g. Only the equations of the illumination light wavefront normals l _{1 to} l ₇ and the coordinates of the intersection and the center related thereto are stored, and all the data related to the illumination light wave front normals l _{8 to} l ₉ are ignored. Will be. In this case, the final value of the counter n is 7.
Further, since the reflected light reflected by the test surface 13a is always incident perpendicular to the wavefront distribution surface W, the wavefront distribution surface side normal line k _i is reflected by the test surface 13a and the minute cell W _i. Obviously, it is equivalent to the optical path of the reflected light incident on.
Even if the reflected light that forms the sharing image detected by the imaging device 12 of the interferometer 1 is reflected by a perfect spherical surface or reflected by an aspherical surface, at least this reflection Incident light that is the cause of light generation is one of incident light that has entered one of the optical paths that converge at one point upon measurement by the interferometer 1, that is, one of the normal lines of the illumination light wavefront E and that has entered the surface 13a to be measured. no room for doubt that is, the optical path equivalent to the wavefront distribution surface normals k _i is reflected by the test surface 13a of the illumination optical wavefront side normal l ₁ to l ₉ shown in FIG. 15 There is a possibility of forming reflected light having incident light that intersects with the wavefront distribution plane normal line k _i corresponding to the reflected light, that is, incident light corresponding to the illumination light wavefront normal lines l _{1 to} l _7. It will be only. Therefore, if this is generalized, the illumination light wavefront normals extracted in step a25 out of all the illumination light wavefront normals l _{1 to} l _C2 , that is, the illumination light wavefront side stored in the data table g Only the normals l _{1 to} l _n are candidates for equations representing incident light that may form reflected light having an optical path equivalent to the wavefront distribution plane normal line k _i .

Next, the CPU 17 functioning as the illumination light optical path specifying unit e specifies the illumination light wavefront side normal representing the illumination light wavefront side normal stored in the data table g, that is, the optical path of the incident light corresponding to the reflected light. Minimum error value storage register D _min. Used for determination of divergence degree _. Configurable maximum value set as an initial value (Step a26), the data associated with each of the illumination optical wavefront side normal line crosses the wavefront distribution surface normals k _i being selected as the evaluation target at this stage After the value of the index j for sequentially reading from the data table g is initialized to 0 (step a27), the value of the index j is incremented by 1 (step a28), and the value of the index j is stored in the data table g. It is determined whether or not the total number n of data sets has been reached (step a29).

If the current value of the index j does not reach the total number n of data sets stored in the data table g, the CPU 17 functioning as the illumination light path identifying unit e determines from the data table g based on the current value of the index j. One set of data, that is, an equation l _{j of} one illumination light wavefront normal intersecting with the wavefront distribution surface normal k _i selected for evaluation at this stage, and this illumination light wavefront normal l The intersection P _j between _j and the wavefront distribution plane side normal line k _i and the center D _j of the minute cell of the illumination light wavefront E corresponding to the illumination light wavefront side normal line l _j are read (step a30), As shown in FIG. 15, a sphere R is assumed assuming a sphere R passing through an intersection P _j between the illumination light wavefront side normal line l _j and the wavefront distribution plane side normal line k _i with the sphere center O of the subject 13 as the center. After obtaining the normal equation m _j at the upper position P _j (step A) The illuminating light path identifying means e is activated, and the illuminating light wavefront normal corresponding to the data set read out in step a30 is selected as the evaluation target at this stage. A process for determining whether or not it is appropriate to pair with k _i is started.

As already described, the illumination light path identifying means e of this embodiment is for identifying the illumination light wavefront normal that is paired with the wavefront distribution plane normal k _i selected as the evaluation target at this stage. There are three functions such as an optical path length condition corresponding normal specifying function, an incident reflection angle condition corresponding normal specifying function, and a coplanar condition corresponding normal specifying function, and a flag switching process according to a request from the user, etc. Thus, any function can be selected to specify the illumination light wavefront side normal paired with the wavefront distribution surface side normal line k _i .

Here, when the optical path length condition corresponding normal function is selected, that is, when the determination result in step a32 is true, the CPU 17 functioning as the illumination light optical path specifying means e is in SUB (A). shifts to the process, first, stored in the RAM20 in the processing of step a16 as the center of the micro cell W _i of the wavefront distribution plane W which corresponds to the wavefront distribution surface normals k _i being selected as the evaluation target at this stage Based on the value of the center A _{i and} the value of the intersection P _j of the illumination light wavefront side normal line l _j and the wavefront distribution plane side normal line k _i read out in the process of step a30, the intersection point with the center A _i distance AP between the P _j, i.e., the wavefront distribution plane W from wavefront distribution surface normals k _i leading to the intersection P _j of the wavefront distribution surface normals k _i the illumination optical wavefront side normal line l _j The length AP is obtained (see FIG. 15), and this value is temporarily stored in the register AP. To 憶.
Similarly, the value of the center D _j of the minute cell of the illumination light wavefront E corresponding to the illumination light wavefront side normal l _j read out in the process of step a30 and the read out in the process of step a30. Based on the value of the intersection P _j between the illumination light wavefront normal line l _j and the wavefront distribution surface side normal line k _i , the separation distance BP between the center D _j and the intersection point P _j , in short, the illumination light wavefront seek length BP of the illumination optical wavefront side normal line l _j, from E the intersection P _j of the wavefront distribution surface normals k _i the illumination optical wavefront side normal line l _j (see FIG. 15), this value register Temporarily store in BP (step b1).

Next, the CPU 17 functioning as the illumination light optical path specifying means e sets the illumination light that converges to one point at the separation distance between the illumination light wavefront E and the convergence position O of the illumination light that converges with the ball center O as one point. The distance from the convergence position O to the known reference spherical surface used for calculating the wavefront distribution surface is added, and a value twice the design paraxial radius of curvature r of the known test surface 13a is subtracted from this value. The difference S = | V− (AP + BP) | between V and AP + BP is obtained by subtracting the added value of the separation distance AP and the separation distance BP from the measured distance value V (hereinafter referred to as an equal optical path length condition). (Step b2), deviation S and minimum error value storage register D _min. Are compared with the current value (step b3).
However, the value of the paraxial curvature radius r is reduced when the test surface 13a is a convex surface. When the test surface 13a is a concave surface, the paraxial curvature radius r converges with the sphere O as one point. Is added to the separation distance between the convergence position and the illumination light wavefront E.

Here, when the determination result of step b3 is false, that is, it is determined that the value of the deviation S obtained by this processing is smaller than the values of all deviations obtained previously. when the CPU 17 that functions as an illumination light path specifying means e is the time being, selecting the incident light corresponding to the illumination optical wavefront side normal line l _j read in the process of this step a30 is as the evaluation target at this stage regarded as most likely to be the ones that constitute the reflecting light and the pair corresponding to the wavefront distribution surface normals k _i being, the illumination optical wavefront side normal l _j and wave distribution surface normals k the intersection point P _j with _i, as well as temporarily stored in the data candidate storage register R as 1 candidate data representing the position of the real surface shape on the test surface 13a corresponding to the micro-cell W _i of the wavefront distribution plane W (step b4), minimum error value storage register _min. Is replaced with the current value of the deviation S, that is, the deviation value S which is the smallest at the present time (step b5).

Next, the CPU 17 functioning as the illumination light optical path specifying means e stores data related to each of the illumination light wavefront side normals intersecting with the wavefront distribution surface side normal k _i selected as an evaluation target at this stage. The value of the index j for sequentially reading from g is incremented by 1 again (step a28), and it is determined whether or not the value of the index j has reached the total number n of data sets stored in the data table g (step a29). ).

  If the current value of the index j has not reached the total number n of data sets, the CPU 17 functioning as the illumination light path specifying means e performs steps a30 to a32 and SUB ( The process of A) is repeatedly executed.

In this manner, the value of the index j is sequentially incremented, and each time a data set related to the illumination light wavefront side normal l _j corresponding to the value of the index j is read from the data table g one after another, and step a29. -It is set as the object of processing of step a32 and SUB (A).

Accordingly, when the determination result of the final step a29 becomes false, the illumination optical wavefront side normal value of the deviation S of the illumination optical wavefront side normal l ₁ to l _n of此these is minimized, In short, the illumination light wavefront side considered to be appropriate as the illumination light wavefront side normal representing the optical path of the incident light corresponding to the reflected light represented by the wavefront distribution plane normal normal k _i selected as the evaluation target at this stage A normal line is uniquely specified, and the value of the intersection P _j between the illumination light wavefront side normal line and the wavefront distribution plane side normal line k _i is a data candidate by the CPU 17 functioning as a part of the actual surface shape data generation means f. It is held in the storage register R.

Next, the CPU 17 functioning as a part of the actual surface shape data generating unit f represents an illumination path that represents the optical path of the illumination light corresponding to the reflected light represented by the wavefront distribution plane side normal line k _i selected as the evaluation target at this stage. The value of the deviation S obtained in correspondence with the illumination light wavefront normal specified as the light wavefront normal, that is, the minimum error value storage register D _min. It is determined whether or not the final value is within the allowable value ε (step a34).

Here, the minimum error value storage register D _min. CPU 17 functioning as a part of the actual surface shape data generating means f, if the final value is within the allowable value ε, the D _min. The illumination light wavefront normal having the value of S as the deviation S indicates the light path of the incident light corresponding to the reflected light represented by the wavefront distribution surface normal k _i selected as the evaluation target at the present stage Assuming that the side normal is appropriate, the value of the counter Q for counting the total number of data representing the actual surface shape of the surface 13a to be measured is incremented by 1 (step a35), and the current value of the counter Q is set. The value of the data candidate storage register R is stored as it is in the corresponding actual surface shape data storage register R (Q) (step a36).
Further, the minimum error value storage register D _min. CPU 17 functioning as a part of the actual surface shape data generation means f, if the final value of the value exceeds the range of the allowable value ε, the D _min. The illumination light wavefront normal having the value of S as the deviation S is the illumination light wavefront representing the optical path of the illumination light corresponding to the reflected light represented by the wavefront distribution surface normal k _i selected as the evaluation target at this stage The processing of step a35 to step a36 is not executed assuming that the side normal is inappropriate. Therefore, in this case, the value of the counter Q is not updated, and the value of the data candidate storage register R is not transferred to the shape data storage register.

That is, the loop process and is of the obtained by the process of step a34~ step a36, the wavefront distribution between the wavefront distribution plane W _{(W i)} the intersection _{P j} formed by step a28~ Step a32 and SUB (A) the value of AP + BP is the sum of the length BP of the illumination optical wavefront side normal _{l j} between side normals _{k i} of length AP and the illumination optical wavefront E and _{(E j)} and the intersection point _{P j} is one point Is used to calculate the wavefront distribution plane from the convergence position O of the illumination light that converges to one point at the separation distance between the spherical center O and the illumination light wavefront E that is the convergence position of the illumination light (incident light) that converges to Add the distance to the known reference sphere, and illuminate the light wave that most closely approximates this value by subtracting a value twice the design paraxial radius of curvature r of the known surface 13a. Surface-side normal l _j , this illumination light wave-side normal l _j and wave-front distribution surface-side normal This is the value of the intersection P _j (R (Q)) with k _i .

In this embodiment, since S = | V− (AP + BP) | = 0 when the actual surface shape of the test surface 13a completely matches the design of the test surface 13a, the value of the deviation S an illumination optical wavefront side normal l _j but having the minimum illumination wavefront side method representing the optical path of the incident light corresponding to the reflected light representing the wavefront distribution surface normals k _i being selected as the evaluation target at that stage Judgment is made as the most appropriate line.

Next, the CPU 17 again increments the value of the index i for sequentially selecting the data W _{1 to} W _C1 of each microcell constituting the wavefront distribution plane W by 1 (step a13), and the current value is the wavefront distribution plane W. Is determined (step a14), and if not, the data of one minute cell constituting the wavefront distribution plane W from the RAM 20 is determined based on the current value of the index i. W _i newly reads (step a15), in the same manner as described above, the processing of step a16~ step a27, the loop formed by steps a28~ step a32 and SUB (a), the step a34~ step a36 Repeat the process.

Accordingly, the determination result of the final step a14 is they become false, for example, the wavefront distribution surface normal to _k 1 _{to k C1} corresponding to each micro cell _W 1 _{to W-C1} constituting the wavefront distribution plane W for all, the illumination optical wavefront side normal _l 1 to l _C1 forming each pair of the wavefront distribution surface normal to _k 1 _{to k C1} is identified, the illumination light wave wavefront distribution surface normal to _k 1 _{to k C1} intersection _P 1 _{to P C1} of the side normal _l 1 to l _C1 is determined and that the position data of the intersection point _P 1 _{to P C1} is stored in the shape data storage register R (1) ~R (C1) Become.

  Next, the CPU 17 optimizes the position data R (1) to R (C1) stored in the shape data storage registers R (1) to R (C1) using a known Zernike fitting or a least square approximation method. The image of the actual surface shape of the subject 13 represented by these position data R (1) to R (C1) is output and displayed on, for example, the monitor 23 or the like (step a37), and further necessary. Accordingly, the design data of the subject 13 is read from the hard disk drive 21 or the like, and the actual surface shape of the subject 13 represented by the position data R (1) to R (C1) is compared with the design shape, and the two are compared. The deviation (processing error, etc.) is displayed numerically on the monitor 23 in correspondence with the required location of the actual surface shape of the subject 13 (step a38).

  Then, the CPU 17 determines whether or not the user has input a retry request using the keyboard 24 and the mouse 25 (step a39). If there is a retry request, the cell division condition C1, C1 is determined according to the retry request. After executing the change process of C2 and the flag setting change process required for switching the function for specifying the illumination light wave front side normal (step a40), the process returns to the process of step a9 again to reset the cell. Processing operations equivalent to those described above are repeatedly executed according to the division condition, the function of specifying the reselected illumination light wave front normal, and the like.

  Next, processing when the incident reflection angle condition corresponding normal normal function is selected, that is, processing when the determination result of step a32 is false and the determination result of step a33 is true will be described.

In this case, the CPU 17 functioning as the illumination light optical path specifying means e shifts to the processing of SUB (B). First, in the processing of the wavefront distribution plane side normal line k _i selected as the evaluation target at the present stage, that is, the processing of step a16. A sphere R passing through an intersection P _j between the wavefront distribution plane side normal line k _i and the illumination light wavefront side normal line l _j centered on the spherical center O of the subject 13 and the wavefront distribution plane side normal line k _i stored in the RAM 20. , The angle θ1 formed by the normal m _j obtained at the position P _j on the sphere R is obtained (see step c1 / see FIG. 15), and the illumination light wavefront side method read out in the process of step a30 is obtained. After obtaining the angle θ2 formed by the line l _j and the normal m _j (see step c2 / FIG. 15), the angle deviation S = | θ1-θ2 | between them is obtained (step c3), and the deviation S and the minimum Error value storage register D _min. Are compared with the current value (step c4).

Here, when the determination result of step c4 is false, that is, it is determined that the value of the deviation S obtained by this processing is smaller than the values of all deviations obtained previously. when the CPU 17 that functions as an illumination light path specifying means e is the time being, selecting the incident light corresponding to the illumination optical wavefront side normal line l _j read in the process of this step a30 is as the evaluation target at this stage regarded as most likely to be the ones that constitute the reflecting light and the pair corresponding to the wavefront distribution surface normals k _i being, the illumination optical wavefront side normal l _j and wave distribution surface normals k the intersection point P _j with _i, as well as temporarily stored in the data candidate storage register R as 1 candidate data representing the position of the real surface shape on the test surface 13a corresponding to the micro-cell W _i of the wavefront distribution plane W (step c5), minimum error value storage register _min. Is replaced with the current value of the deviation S, that is, the deviation value S that is the smallest at the present time (step c6).

The subsequent processing is the same as the processing in the case where the optical path length condition corresponding normal specifying function is selected, and when the determination result in step a29 is finally false, the illumination light wavefront side Corresponds to the reflected light represented by the wavefront surface normal k _i selected as the evaluation target at the present stage, which is the illumination light wavefront normal with the smallest deviation S among the normals l _{1 to} l _n The illumination light wavefront side normal that is considered appropriate as the illumination light wavefront side normal representing the optical path of the illumination light (incident light) is uniquely identified. The value of the intersection point P _j with the line k _i is held in the data candidate storage register R by the CPU 17 functioning as a part of the actual surface shape data generation means f.

In this embodiment, according to the principle that the incident angle and the reflection angle of light are equal, the illumination light wavefront side normal l _j that minimizes the value of S = | θ1−θ2 |, that is, the value of θ1 and the value of θ2 are the most approximate. The illumination light wavefront side normal line l _j representing the optical path of the incident light corresponding to the reflected light represented by the wavefront distribution plane side normal line k _i selected as the evaluation target at that stage Judgment is made as the most appropriate line.

The illumination optical wavefront side normal line l _j wavefront around the spherical center O as the actual surface shape of the surface 13a at the intersection P _j of the illumination optical wavefront side normal l _j and wave distribution surface normals k _i Assuming that the subject 13 is the sphere R passing through the intersection P _j with the distribution plane side normal line k _i and assuming that the tangent plane on the sphere R reflects incident light. This is because a spherical lens or an aspheric lens is generally used, and it is assumed that a discontinuous change in inclination does not occur on the surface of the test surface 13a.

  Next, processing when the same plane condition corresponding normal specifying function is selected, that is, processing when both the determination results of step a32 and step a33 are false will be described.

In this case, the CPU 17 functioning as the illumination light optical path specifying means e shifts to the processing of SUB (C), and first, in the processing of the wavefront distribution plane side normal line k _i selected as the evaluation target at the present stage, that is, the processing of step a16. The intersection P _j between the direction vector of the wavefront distribution plane side normal line k _i stored in the RAM 20 and the illumination light wavefront side normal line l _j and the wavefront distribution plane side normal line k _i with the sphere center O of the subject 13 as the center. assuming sphere R obtains an outer product V ₁ of the the direction vector of the normal line m _j obtained at the position P _j on the sphere R through (step d1), further, the illumination optical wavefront read in step a30 After obtaining the inner product V ₂ of the direction vector of the side normal l _j and the outer product V ₁ (step d2), the inner product V ₂ and the minimum error value storage register D _min. Are compared with the current value (step d3).

Here, if the determination result in step d3 is false, that is, it is determined that this smaller value of the inner product V ₂ obtained by previously sought than the value of all the inner product has the Konotabi process In this case, the CPU 17 functioning as the illumination light optical path specifying means e is not allowed to evaluate the incident light corresponding to the illumination light wavefront side normal l _j read out in the current step a30 as an evaluation target at this stage. It was regarded as most likely to be the ones that constitute the reflected light and the pair corresponding to the wavefront distribution surface normals k _i being selected, the illumination light wave surface normals l _j and wave distribution surface normal The intersection P _j with k _i is temporarily stored in the data candidate storage register R as one candidate of data representing the position of the actual surface shape on the surface 13a corresponding to the minute cell W _i of the wavefront distribution plane W ( Step d4), minimum error value storage register D _min. Replacing the value of the value of the inner product V ₂ which is the minimum current value, i.e. the moment of the inner product V ₂ (step d5).

The subsequent processing is the same as the processing in the case where the optical path length condition corresponding normal specifying function is selected and the processing in the case where the incident reflection angle condition corresponding normal specifying function is selected. to when the determination result in step a29 becomes false, the illumination optical wavefront side normal value of the inner product V ₂ is the minimum of the illumination optical wavefront side normal l ₁ to l _n, in short, be evaluated at this stage The illumination light wavefront normal that is considered appropriate as the illumination light wavefront normal representing the optical path of the incident light corresponding to the reflected light represented by the wavefront distribution surface normal k _i selected as The value of the intersection point P _j between the illumination light wavefront normal and the wavefront distribution plane normal k _i is held in the data candidate storage register R by the CPU 17 functioning as a part of the actual surface shape data generation means f. It will be.

In this embodiment, according to the principle that light is incident and reflected symmetrically with respect to the normal line m _j of the test surface 13a, the illumination light wave front side normal line l _j that minimizes the value of the inner product V ₂ , that is, the test surface. on condition that 13a and the normal line m _j and the illumination optical wavefront side normal l _j and wave distribution surface normal to k _i of are in the same plane, the wavefront distribution surface that is selected as the evaluation target at that stage The most appropriate illumination light wavefront side normal l _j is specified as the illumination light wavefront side normal representing the optical path of the incident light corresponding to the reflected light represented by the normal k _i .

As already mentioned, the illumination optical wavefront side normal l around the spherical center O as the actual surface shape of the surface 13a at the intersection P _j of the illumination optical wavefront side normal l _j and wave distribution surface normals k _{i It} is the subject 13 that assumes a sphere R passing through the intersection P _j between _j and the wavefront distribution plane side normal line k _i and that the tangent plane on the sphere R reflects incident light. This is because a spherical lens or an aspherical lens is generally used, and it is assumed that a discontinuous change in inclination does not occur on the surface of the test surface 13a.

As described above, in the test surface shape measuring method, the test surface shape measuring device, and the test surface shape measuring program of this embodiment, each microcell W ₁ obtained by partitioning the wavefront distribution surface W. each of the wavefront distribution surface normal to k ₁ to k _C1 of ~W each _C1 is incident, the test surface 13a follows the optical path that converges to a point to be a spherical center O of the object 13 at the time of measurement by the interferometer 1 based on that it matches the optical path of one of the reflected light of the incident light, from the wavefront distribution surface normal to k ₁ to k _C1 that test surface 13a in a micro cell W ₁ to _W-C1 wavefront distribution plane W For each of the wavefront distribution plane normals k _{1 to} k _C1 corresponding to the reflected light, the corresponding illumination light wavefront normals l _{1 to} l _C1, that is, the optical path of the incident light, are determined, and these two optical paths, that is, the wavefront distribution planes intersection R of the side normal to _k 1 _{to k C1} and the illumination optical wavefront side normal _l 1 to l _C1 Since (1) to R (C1) are obtained and used as actual surface shape data of the test surface 13a, the optical path of the reflected light detected by the image sensor 12 of the interferometer 1 is the optical path of the incident light. Even when they match (when the test surface 13a is a spherical surface) or when the optical path of the incident light does not match the optical path of the reflected light (when the test surface 13a is an aspherical surface), It is possible to directly grasp the actual surface shape of the test surface 13a of the subject 13 and accurately display it by display using an image or numerical data.

  Therefore, when the test surface 13a is not a perfect spherical surface, for example, when the shape of the test surface 13a that should be a spherical surface is no longer a spherical surface due to processing abnormality, damage, or secular change, or the original test surface Even when the design shape of the surface 13a is an aspherical surface, the actual surface of the test surface 13a is not required without a physical aspherical surface functioning as a reference surface and a module for generating a wavefront corresponding thereto. The shape can be measured easily and inexpensively, and since the partial measurement of the test surface 13a and the synthesis of data are not required, the time required for the calculation is shortened and non-contact measurement is performed. Therefore, it is possible to inspect all expensive test lenses at the manufacturing site without worrying about damaging the expensive test lenses.

In the embodiment described above, the illumination light optical path specifying means e is one of the optical path length condition corresponding normal specifying function, the incident reflection angle condition corresponding normal specifying function, and the same plane condition corresponding normal specifying function. selected according function to requests from the user, the Re此wavefront distribution surface normals k _i representing the reflected light corresponding to each micro cell W _i of the wavefront distribution plane W based only on one of the function selected Although the example of specifying the illumination light wavefront side normal l representing the corresponding optical path of incident light has been described, the normal specifying function corresponding to the optical path length condition, the normal specifying function corresponding to the incident / reflecting angle condition, the same plane condition compatible normal specific function of a combination of two or more, or a combination of all functions, an illumination optical wavefront side normal line l which represents the optical path of the incident light corresponding to the Re此wavefront distribution surface normals k _i particular You may make it do.

  In the embodiment described above, the case where the test surface shape measuring method, the test surface shape measuring device, and the test surface shape measuring program are applied to the sharing interferometer has been exemplified. The use of the surface shape measuring method, the surface shape measuring apparatus, and the surface shape measuring program is not particularly limited to the sharing interferometer, and is three-dimensional regardless of the presence or absence of a physical reference spherical surface. As long as the interferometer has a function of measuring the wavefront aberration distribution, it is possible to specify the shape of the test surface and measure an error or the like that is the amount of deviation between the test surface and the design value.

DESCRIPTION OF SYMBOLS 1 Interferometer 2 Laser light source 3 Polarizing plate 4 Beam expander 5 Condensing lens 6 Beam splitter 7 Birefringent optical element 8 Collimator lens 8 'Objective lens 9 Support tool 10 Polarizing plate 11 Imaging lens 12 Imaging element 13 Subject 13a Subject Surface inspection 14 Adjustment tool 15 Feed mechanism 16 Surface shape measuring device 17 CPU (reference spherical surface generation means, wavefront distribution surface data generation means, normal data extraction means, illumination light path specifying means, actual surface shape data generation means)
18 ROM
19 Nonvolatile memory 20 RAM (storage means)
21 hard disk drive 22 input / output circuit 23 monitor 24 keyboard 25 mouse 26 input / output interface 27 laser drive circuit 28 driver 29 pulse generator 30 CCD control circuit 31, 32 tangent plane a reference spherical surface generation means b wavefront distribution plane data generation means c method Line data extraction means d Storage means e Illumination light path specifying means f Actual surface shape data generation means g Data table B Reference spherical surface W Wavefront distribution surface E Wavefront of illumination light

Claims (15)

A reference spherical surface for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data obtained by processing the image data of the share image of the test surface output from the image sensor of the interferometer in the logical space. Virtual,
After generating wavefront distribution surface data based on the reference spherical surface based on the reference spherical surface and the wavefront aberration distribution data,
A wavefront distribution side method that represents the reflected light of each minute cell by obtaining a normal passing through the center of each minute cell obtained by dividing the wavefront distribution surface represented by the data of the wavefront distribution surface. And a normal passing through the center of the micro cell for each micro cell obtained by dividing the illumination light wave front, and the illumination light wave front side normal for each micro cell,
Each of the wavefront distribution plane side normals follows the optical path of any reflected light of the illumination light (incident light) incident on the surface to be detected, following the optical path that converges to one point when measured by the interferometer. For each minute cell of the wavefront distribution surface represented by the wavefront distribution surface data, the wavefront distribution surface side normal representing the reflected light of the minute cell and the optical path of incident light corresponding thereto are represented. The illumination light wavefront normal is specified, and for each minute cell of the wavefront distribution surface, the intersection point between the corresponding wavefront distribution surface normal and the illumination light wavefront normal is obtained, and the actual shape of the test surface A method for measuring the shape of a surface to be measured, characterized in that When specifying the wavefront distribution surface normal representing the reflected light of the microcell and the illumination light wavefront normal representing the corresponding optical path of the illumination light (incident light) for each microcell on the wavefront distribution surface. ,
Find all the intersections of the illumination light wavefront side normal intersecting with the wavefront distribution plane side normal and the illumination light wavefront side normal and the wavefront distribution plane side normal,
When the test surface is a convex surface, the length of the wavefront distribution surface side normal between the wavefront distribution surface and the intersection and the illumination light wavefront side normal between the illumination light wavefront and the intersection Is the distance between the convergence position of the illumination light (incident light) that converges to the one point and the illumination light wavefront from the convergence position of the illumination light (incident light) that converges to the one point. The method of illuminating light wavefront that most closely approximates the distance obtained by adding the distance to the reference spherical surface used for calculation of the wavefront distribution surface and subtracting a value twice the paraxial radius of curvature of the test surface design from this value While seeking a line
When the test surface is a concave surface, the length of the wavefront distribution surface side normal between the wavefront distribution surface and the intersection and the illumination light wavefront side normal between the illumination light wavefront and the intersection Is the distance between the convergence position of the illumination light (incident light) that converges to the one point and the illumination light wavefront from the convergence position of the illumination light (incident light) that converges to the one point. The method of illuminating light wavefront that most closely approximates the distance obtained by adding the distance to the reference spherical surface used for calculation of the wavefront distribution surface and adding this value to a value twice the paraxial radius of curvature of the test surface design Seeking a line,
The test surface shape measuring method according to claim 1, wherein the obtained illumination light wavefront side normal is specified as an illumination light wavefront normal corresponding to the wavefront distribution surface side normal. When specifying the wavefront distribution surface side normal representing the reflected light of the microcell and the illumination light wavefront normal representing the optical path of the incident light corresponding thereto for each microcell of the wavefront distribution surface,
All the illumination light wavefront normals intersecting the wavefront distribution surface side normal are obtained, and the normal of the test surface assumed at the position where the wavefront distribution surface normal and each illumination light wavefront normal intersect An illumination light wavefront side normal that minimizes a difference between an angle formed by the wavefront distribution surface side normal, the normal of the test surface, and the calculated illumination light wavefront side normal, The test surface shape measuring method according to claim 1, wherein the illumination light wavefront side normal is specified as the illumination light wavefront side normal corresponding to the wavefront distribution surface side normal. When specifying the wavefront distribution surface side normal representing the reflected light of the microcell and the illumination light wavefront normal representing the optical path of the incident light corresponding thereto for each microcell of the wavefront distribution surface,
All the illumination light wavefront normals intersecting the wavefront distribution surface side normal are obtained, and the normal of the test surface assumed at the position where the wavefront distribution surface normal and each illumination light wavefront normal intersect The wavefront distribution surface side normal and the obtained illumination light wavefront side normal are obtained on the same plane, and the illumination light wavefront side normal is obtained on the wavefront distribution surface side. The test surface shape measuring method according to claim 1, wherein the test surface shape is specified as the illumination light wavefront side normal corresponding to the normal.   A surface of a sphere having a convergence position of incident light converged on the one point as a center is used as a test surface assumed at a position where the wavefront distribution plane side normal and each illumination light wavefront normal intersect. The method for measuring a shape of a test surface according to claim 3 or 4, wherein: A reference spherical surface for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data obtained by processing the image data of the shear image of the test surface output from the image sensor of the interferometer is hypothesized in the logical space. A reference spherical surface generating means;
Wavefront distribution surface data generation means for generating data of a wavefront distribution surface based on the reference spherical surface based on the reference spherical surface and the data of the wavefront aberration distribution virtualized by the reference spherical surface generation means;
Storage means for temporarily storing calculation data;
For each minute cell, a normal passing through the center of the minute cell is obtained for each minute cell obtained by partitioning the wavefront distribution surface represented by the wavefront distribution surface data generated by the wavefront distribution surface data generating means. Is stored in the storage means as a wavefront distribution plane-side normal representing the reflected light, and a normal passing through the center of the microcell is obtained for each microcell obtained by dividing the illumination light wavefront. Normal data extraction means to be stored in the storage means as the illumination light wavefront side normal of
Each reflected light of the illumination light (incident light) incident on the surface to be examined following the optical path where each wavefront distribution plane normal stored in the storage means converges to one point at the time of measurement by the interferometer The wavefront distribution surface side normal representing the reflected light stored in the storage means corresponding to each minute cell of the wavefront distribution surface represented by the data of the wavefront distribution surface and Illumination light path identifying means for identifying the illumination light wavefront side normal representing the optical path of incident light corresponding to this,
For each microcell on the wavefront distribution surface, the actual surface of the surface to be measured is obtained by calculating the intersection position of the wavefront distribution surface side normal and the illumination light wavefront side normal whose correspondence is specified by the illumination light optical path specifying means A test surface shape measuring apparatus comprising actual surface shape data generating means for generating shape data. When the test surface is a convex surface, the length of the wavefront distribution surface side normal between the wavefront distribution surface and the intersection and the illumination light wavefront side normal between the illumination light wavefront and the intersection Is the distance between the convergence position of the illumination light (incident light) that converges to the one point and the illumination light wavefront from the convergence position of the illumination light (incident light) that converges to the one point. The method of illuminating light wavefront that most closely approximates the distance obtained by adding the distance to the reference spherical surface used for calculation of the wavefront distribution surface and subtracting a value twice the paraxial radius of curvature of the test surface design from this value While seeking a line
When the test surface is a concave surface, the length of the wavefront distribution surface side normal between the wavefront distribution surface and the intersection and the illumination light wavefront side normal between the illumination light wavefront and the intersection Is the distance between the convergence position of the illumination light (incident light) that converges to the one point and the illumination light wavefront from the convergence position of the illumination light (incident light) that converges to the one point. The method of illuminating light wavefront that most closely approximates the distance obtained by adding the distance to the reference spherical surface used for calculation of the wavefront distribution surface and adding this value to a value twice the paraxial radius of curvature of the test surface design A line is obtained, and the obtained illumination light wavefront side normal is specified as an illumination light wavefront side normal corresponding to the wavefront distribution surface side normal. Test surface shape measuring device.   The illumination light path identifying means obtains all illumination light wavefront normals intersecting with the wavefront distribution plane normal stored in the storage means, and the wavefront distribution plane normal and each illumination light wavefront normal The difference between the angle formed by the normal of the test surface assumed at the position where the crossing and the wavefront distribution surface side normal and the angle formed by the normal of the test surface and the obtained illumination light wavefront normal Is characterized in that the illumination light wavefront side normal is minimized, and the illumination light wavefront normal is specified as the illumination light wavefront normal corresponding to the wavefront distribution surface normal. The test surface shape measuring apparatus according to claim 6.   The illumination light path identifying means obtains all illumination light wavefront normals intersecting with the wavefront distribution plane normal stored in the storage means, and the wavefront distribution plane normal and each illumination light wavefront normal The illumination light wavefront side normal line in which the normal of the test surface assumed at the position where the crossing points, the wavefront distribution plane side normal line, and the obtained illumination light wavefront side normal line are located on the same plane is obtained. 7. The test surface shape measuring apparatus according to claim 6, wherein the illumination light wavefront side normal is specified as an illumination light wavefront side normal corresponding to the wavefront distribution surface side normal. .   The illumination light optical path specifying means has a convergence position of incident light that converges to the one point as a test surface assumed at a position where the wavefront distribution surface side normal and each illumination light wavefront normal intersect. The test surface shape measuring apparatus according to claim 8 or 9, wherein a surface of a sphere is assumed. A test surface shape measurement program for controlling a computer that processes image data of a share image of a test surface output from an image sensor of an interferometer,
The computer,
A reference spherical surface for calculating the wavefront aberration distribution represented by the wavefront aberration distribution data obtained by processing the image data of the shear image of the test surface output from the image sensor of the interferometer is hypothesized in the logical space. Reference spherical surface generation means,
Wavefront distribution surface data generation means for generating data of a wavefront distribution surface based on the reference spherical surface based on the reference spherical surface and the wavefront aberration distribution data hypothesized by the reference spherical surface generation means;
For each minute cell, a normal passing through the center of the minute cell is obtained for each minute cell obtained by partitioning the wavefront distribution surface represented by the wavefront distribution surface data generated by the wavefront distribution surface data generating means. Is stored in the storage device of the computer as a wavefront distribution plane side normal representing the reflected light of the light source, and a normal passing through the center of the microcell is obtained for each microcell obtained by dividing the illumination light wavefront. Normal data extraction means for storing in the storage device of the computer as the illumination light wavefront normal for each cell;
Each of the wavefront distribution plane side normals stored in the storage device of the computer follows an optical path that converges to one point at the time of measurement by the interferometer, and reflects any of the reflected light incident on the test surface. The wavefront distribution plane normal representing the reflected light stored in the storage device of the computer corresponding to each minute cell of the wavefront distribution plane represented by the data of the wavefront distribution plane based on matching with the optical path And illumination light optical path specifying means for specifying the illumination light wavefront side normal representing the optical path of incident light corresponding thereto, and
For each microcell on the wavefront distribution surface, the actual surface of the surface to be measured is obtained by calculating the intersection position of the wavefront distribution surface side normal and the illumination light wavefront side normal whose correspondence is specified by the illumination light optical path specifying means A test surface shape measurement program that functions as actual surface shape data generation means for shape data. When the test surface is a convex surface, the length of the wavefront distribution surface side normal between the wavefront distribution surface and the intersection and the illumination light wavefront side normal between the illumination light wavefront and the intersection Is the distance between the convergence position of the illumination light (incident light) that converges to the one point and the illumination light wavefront from the convergence position of the illumination light (incident light) that converges to the one point. The method of illuminating light wavefront that most closely approximates the distance obtained by adding the distance to the reference spherical surface used for calculation of the wavefront distribution surface and subtracting a value twice the paraxial radius of curvature of the test surface design from this value While seeking a line
When the test surface is a concave surface, the length of the wavefront distribution surface side normal between the wavefront distribution surface and the intersection and the illumination light wavefront side normal between the illumination light wavefront and the intersection Is the distance between the convergence position of the illumination light (incident light) that converges to the one point and the illumination light wavefront from the convergence position of the illumination light (incident light) that converges to the one point. The method of illuminating light wavefront that most closely approximates the distance obtained by adding the distance to the reference spherical surface used for calculation of the wavefront distribution surface and adding this value to a value twice the paraxial radius of curvature of the test surface design 12. The apparatus according to claim 11, wherein a line is obtained, and the obtained illumination light wavefront side normal is specified as an illumination light wavefront side normal corresponding to the wavefront distribution surface side normal. Test surface shape measurement program.   The illumination light optical path specifying means obtains all illumination light wavefront normals intersecting with the wavefront distribution plane normal stored in the storage device of the computer, and the wavefront distribution plane normal and each illumination light wavefront method The angle formed by the normal of the test surface assumed at the position where the line intersects and the wavefront distribution surface side normal, and the angle formed by the normal of the test surface and the obtained illumination light wavefront normal The illumination light wavefront side normal line that minimizes the difference between the illumination light wavefront side normal line and the illumination light wavefront side normal line is specified as the illumination light wavefront side normal line corresponding to the wavefront distribution plane side normal line. 12. The test surface shape measurement program according to claim 11, wherein the test surface shape measurement program is a test program.   The illumination light optical path specifying means obtains all illumination light wavefront normals intersecting with the wavefront distribution plane normal stored in the storage device of the computer, and the wavefront distribution plane normal and each illumination light wavefront method The normal surface of the test surface assumed at the position where the line intersects, the wavefront distribution surface side normal line, and the illumination light wavefront side normal line obtained on the same plane. The test surface shape according to claim 11, wherein the illumination light wavefront side normal is specified as an illumination light wavefront side normal corresponding to the wavefront distribution surface side normal. Measurement program.   The illumination light optical path specifying means has a convergence position of incident light that converges to the one point as a test surface assumed at a position where the wavefront distribution surface side normal and each illumination light wavefront normal intersect. The test surface shape measuring program according to claim 13 or 14, wherein a surface of a sphere is assumed.

Images