WO2005019901A2

WO2005019901A2 - Method and device for aligning and interferometer

Info

Publication number: WO2005019901A2
Application number: PCT/NL2004/000557
Authority: WO
Inventors: Petrus Arnoldus Andreas Maria Somers
Original assignee: Technische Universiteit Delft; Stichting voor de Technische Wetenschappen STW
Current assignee: Technische Universiteit Delft; Stichting voor de Technische Wetenschappen STW
Priority date: 2003-08-07
Filing date: 2004-08-06
Publication date: 2005-03-03
Anticipated expiration: 2006-02-07
Also published as: WO2005019901A3; NL1024070C2

Abstract

Method and device for the alignment of an interferometer arrangement, which comprises an object beam part (4), a beam splitting part (5) and a beam combination part (6). A detector unit (2) is arranged to detect an interference pattern for two beams that can be differentiated, via each of n optical paths, and the interferometer is aligned by setting at least n-1 of mirroring elements (16, 17) in the beam combination part (6). The at least n-1 mirroring elements (16, 17) in the beam combination part are aligned depending on an alignment signal derived from the detected interference patterns.

Description

Method and device for aligning an interferometer

Field of the invention The present invention relates to a method and device for aligning an interferometer arrangement, such as a speckle interferometer, wherein the interferometer arrangement comprises an object beam part for obtaining two beams that can be differentiated (such as a first and second object beam that are circularly polarised in opposing directions or that have a linear polarisation that can be differentiated), at least a first beam of which that can be differentiated originates from an object to be tested. Furthermore the interferometer arrangement comprises a beam splitting part for splitting the two beams that can be differentiated between n optical paths, where n is an integer and n > 2, and a beam combination part that is arranged to receive the two beams that can be differentiated, to apply a phase step in the mutual phase difference present between the two beams that can be differentiated in n-1 of the n optical paths and to direct the two beams that can be differentiated and travel over each of the n optical paths to a detector unit. The detector unit is arranged to detect an interference pattern for the two beams that can be differentiated, via each of the π optical paths, and the interferometer is aligned by setting at least n-1 of mirroring elements (such as mirrors) in the beam combination part in order to shift the interference pattern on the detector unit belonging to at least n-1 of the optical paths. The invention can be employed in so-called 'shearing* speckle interferometers, where the object beams have a different viewing direction towards the object ('shearography'), but also in other interferometers, including speckle interferometers where one of the object beams (in this case termed reference beam) originates directly from the light source. These techniques are known to those skilled in the art and require no further explanation here.

State of the art Such a method and device are disclosed in P.A.A.M. Somers and H. van Brug, "A single camera, dual image real-time-phase-stepped shearing speckle interferometer", Fringe 2001, proceedings pp. 573-580. The alignment of systems with multiple optical paths is highly critical, the phase- stepped interference images must correspond to one another at sub-pixel level. Stable alignment is readily possible under laboratory conditions, but for industrial application (where the interferometer also has to be transported) a correct alignment that does not drift is difficult to achieve. Under such conditions limited mechanical stability, rigidity and temperature-dependent behaviour have an adverse effect on the alignment. An interferometer of this type that can be used industrially and that must be able to be used (in a mobile manner) under various conditions must therefore be equipped with heavy and expensive mechanical alignment and fixing means, such as stands and base plates. The stability of mirror holders, lenses and the use of (expensive) temperature-stable materials are important in this context.

Summary of the invention The present invention aims to provide a method for the alignment of a speckle interferometer and an alignment system for a speckle interferometer, without the need for heavy mechanical alignment and fixing means, that can be employed industrially in various locations. According to the present invention, in a first aspect a method according to the type defined in the preamble is provided wherein the at least n-1 mirroring elements in the beam combination part are set depending on an alignment signal derived from the detected interference patterns. Such a method ensures that the interferometer can be aligned in a rapid and efficient manner. The interferometer can consequently be of relatively simple and lightweight construction, as a result of which this is easily transportable and can be deployed rapidly. With the present method the alignment of a speckle interferometer is not a limiting factor when making the interferometer operational. In one embodiment the alignment signal can be displayed on a monitor, by which means the at least n-1 mirroring elements can be set interactively by a user of the interferometer. As an alternative, in a further embodiment the at least n-1 mirroring elements are set with the aid of a control signal, derived from the alignment signal, to an associated mirror controller. The mirror controller can be, for example, a micromanipulator with a high resolution, for example a mirror controller that can be set electrically with the aid of a piezo element or an electrostrictive element. This makes an automated alignment process possible, which takes place such that it is highly transparent to a user. The alignment signal can be derived in a number of ways, each of which leads to a differing degree of accuracy. Firstly, the alignment signal can be a difference signal from image data for the interference patterns belonging to the at least two optical paths. Especially in the case of interactive alignment by a user, this can give a good result. In a further embodiment the alignment signal is obtained by a correlation of image data for the interference patterns belonging to the at least two optical paths. The correlation can be calculated for a number of relative pixel positions in the images concerned. The highest value for the correlation indicates the pixel position with the best alignment. Sub- pixel resolution of the alignment can, for example, be derived by interpolation or another mathematical method and an alignment signal can thus be determined. In a specific embodiment the alignment signal corresponds to the maximum correlation signal that is determined from image data for a series of interference patterns that are recorded by the detector unit at different settings of the at least n-1 mirroring elements. The correct setting of the at least n-1 mirrors can be determined by interpolation or another mathematical processing operation. hi yet a further embodiment the alignment signal is derived from a cross-correlation of image data, for example from sections (such as the mid sections), for the interference patterns belonging to the at least two optical paths. By, in addition, for example, viewing all portions of the image data that contain interesting information (most usable, most reliable, most relevant for alignment, ..., such as, for example, the central region), the requisite computing capacity can be lower. This can be used to make the method more rapid and/or more accurate. The alignment signal can also be derived from a cross-correlation of image data from a plurality of sections of the interference patterns belonging to the at least two optical paths, for example distributed over the entire image. This gives a large amount of data, as a result of which very accurate alignment can take place. Moreover, in this case rotation can also be adjusted during the alignment. The alignment signal can be representative of a rotational shift, which, for example, can be caused by manufacturing tolerances of the (optical) elements used in the interferometer. These errors can be corrected by correct setting of the at least n-1 mirrors (in a preferred embodiment there are two mirrors in the beam combination part). In yet a further embodiment one of the two object beams that can be differentiated is blocked for the alignment process. As a result it is possible to prevent modulation effects occurring in the speckle interferometer described, which can lead to a weaker or absent alignment signal, so as nevertheless to obtain an accurate alignment. In a further aspect the present invention relates to an alignment device for aligning an interferometer arrangement, wherein the interferometer arrangement comprises an object beam part for obtaining two beams that can be differentiated, at least a first beam of which that can be differentiated originates from an object to be tested, a beam splitting part for splitting the two beams that can be differentiated between n optical paths, where n is an integer and n > 2, and a beam combination part that is arranged to receive the two beams that can be differentiated, to apply a phase step in the mutual phase difference present between the two beams that can be differentiated in n-1 of the n optical paths and to direct the two beams that can be differentiated and travel over each of the n optical paths to a detector unit. The detector unit detects an interference pattern for the two beams that can be differentiated, via each of the n optical paths, and the interferometer arrangement is aligned by setting at least n-1 of mirroring elements in the beam combination part in order to shift the interference pattern on the detector unit belonging to at least n-1 of the optical paths. The alignment device comprises a processing device that is connected to the detector unit and at least n-1 mirror controllers for setting the at least n-1 mirroring elements in the beam combination part, where the processing device is arranged to drive the at least n- 1 mirror controllers depending on an alignment signal that is obtained from the interference patterns received by the detector unit. In one embodiment the mirror controller is provided with an electrically controllable fine mechanical adjustment element, such as piezo-actuator or an electrostrictive element. For the embodiment with interactive alignment by a user, which has been discussed above, the processing device is connected to a display screen and input means. hi further embodiments the processing device is arranged to perform one or more of the methods discussed above.

Brief description of the drawings The present invention will now be discussed in more detail on the basis of a number of illustrative embodiments, with reference to the appended drawings, in which Fig. I shows, diagrammatically, one example of a speckle interferometer; Fig. 2 shows a block diagram of one embodiment of the alignment device according to the present invention.

Detailed description of illustrative embodiments In Fig. 1 a speckle interferometer 1 that is suitable for carrying out phase-stepped speckle interferometry in real time is shown diagrammatically. It will be clear to those skilled in the art that in speckle interferometry use is made of a (coherent) light source (not shown in Fig. I) the light of which is scattered by an object 3 so that speckles are produced. In such a speckle interferometer 1 there is more than one beam pair, each consisting of a first and a second object beam, each beam pair travelling along its own optical path. In one of the beam pairs an additional phase difference is applied between the associated first and second object beam, whilst no modifications are made in the other beam pair. The difference in optical wavelength, expressed as phase difference between the first and second object beam, can be determined quantitatively. The interferometer 1 shown is a so- called 'shearing' speckle interferometer where the two object beams originate from the object 3 at a different viewing angle. The invention can, however, also be employed in a conventional speckle interferometer, where one of the object beams comes directly from the light source (and is then termed reference beam). There are various types of phase-stepped speckle interferometers 1, depending on how the object beams are generated. Phase-steps between different object beam pairs can be generated sequentially, where a single object beam pair is sequentially temporally modified, or can be generated spatially separated, where a different optical path is needed for each object beam pair. In the case of spatially separated object beams all information that is needed to determine the phase of the object at a specific point in time is available at a specific point in time, as a result of which this implementation is preferred when measuring dynamic events. In Fig. 1 a phase-stepped speckle interferometer 1 is shown that generates an interference pattern for each optical path, the interference patterns being detected by a single camera 2 (for example a CCD camera). This interferometer 1 is also referred to as a '2-bucket, 1 -camera, real time phase stepping shearing speckle interferometer'. In this case there are two relative phase differences between the first and second object beam. Each second object beam travels together with a first object beam, originating from an object 3 to be tested, along a different optical path, as a result of which it is possible, per optical path, to apply a phase step suitable for that path with the aid of, for example, polarisation-sensitive optics. The interferometer 1 can be arranged to generate more than two object beams. The interferometer 1 can be arranged with a detector (camera 2) per path or, as in the embodiment indicated in Fig. 1, with a camera 2 for the detection of two sub- images. If there are several sub-images it is also possible that there are one or more cameras that each record several sub-images. Using the interferometer 1 shown in Fig. 1 all phase-stepped speckle patterns that are generated and that belong to a specific state of the object 3 to be studied can be recorded at the same point in time. The interferometer 1 comprises three parts, that is to say an object beam part 4 that supplies two beams with different polarisation states, one of which has shearing, a beam splitting part 5 that produces two identical paths and a combination part 6 that is arranged to apply the phase step in one of the paths. In the embodiment shown the object beam part 4 has a so-called Michelson configuration. It is also possible to use a Mach-Zehnder configuration, which is known to those skilled in the art, or yet other configurations. A beam of light originating from the object 3 is received via a field lens 7 and transfer lens 8 on an input face of a polarising beam splitting element 9. The beam splitting element 9 is provided on three sides with λ/4 plates 10 that are oriented at π/4, so that light with a circular polarisation can emerge from these sides. A normal mirror 12 creates a normal beam and a shearing mirror 11 creates a beam with shearing, both of which are reflected to the polarising beam splitting element 9. Linearly polarised light that enters the object beam part 4 emerges as two circularly polarised beams, with opposing directions of rotation. Incoming light is preferably polarised at 45°, as a result of which the beam splitting element 9 divides the beam into two beams of the same amplitude and with different directions of polarisation. Unpolarised light is also divided in the same way. The beam splitting part 5 splits these two beams between two paths with the aid of a normal beam splitting element and directs them towards the beam combination part 6, for example by means of a prism as shown, or using a mirror. The beam combination part 6 comprises a 45° mirror 16 and a normally oriented mirror 17. In one path the first and second object beam are subjected to a phase step with the aid of a polarisation plane rotation element 13; in the other path the reference beam can pass through normally. It is also possible to subject only one of the first and second object beam to a phase step, for example with the aid of an element that retards light with a specific polarisation direction compared with light with a different polarisation direction (for example materials that have a different refractive index in one direction to that in another direction (perpendicular thereto), such as quartz). Two circularly polarised beams with opposing directions of rotation will interfere with one another when they are incident on a detector, such as via transfer lens 18 on camera 2 in Fig. 1, after they have passed through a polariser. Rotation of the direction of polarisation of incident light by polarisation plane rotation element 13 through an angle φ will give rise to a phase step of 2 , so that a phase step between the two paths can be introduced by placing a polarisation plane rotation element 13 in one path. The beam combination part 6 comprises a polarising beam splitter 14, with a λ/4 plate 15 on one side. The beam combination part 6 receives circularly polarised light from the two paths at a different input face. The polarisation plane rotation element 13 is fixed to one input face. After passing through the polarising beam splitting element 14, the circularly polarised beams have become horizontally polarised beams. The component perpendicular to the polarisation direction of the polarising element 14 has disappeared from the beam. After passing through the beam splitting part 5 and the 45° mirror 16, the beams in one path are allowed through directly to one half (image part) of the camera 2, where interference takes place. Circularly polarised light from the other path is also allowed through as horizontally polarised light, but is incident on a mirror 17 after passing through the λ/4 plate 15 that is oriented at π/4, as a result of which the polarisation state again changes to a circular polarisation. It is then reflected back by mirror 17 and again passes through the λ/4 plate 15 and polarising beam splitting element 14. Because of this second path through the λ/4 plate 15 the polarisation state becomes vertical, as a result of which the polarising beam splitting element 14 reflects the beam to the other half of the camera 2, where once again interference takes place. The two phase-stepped speckle patterns are thus available at the same time, as a result of which a good basis is obtained for the calculation of phase differences for a specific state of the object 3. In an alternative embodiment of the beam combination part 6 the polarisation plane rotating element 13 has been replaced by a λ/2 plate with an orientation of 22.5° with respect to the horizontal axis. Such a λ/2 plate with such an orientation also gives rise to a rotation of 45° and thus to the desired phase step of 90°. In yet a further embodiment the polarisation plane rotating element 13 on one of the input faces of the beam splitting element 14 is replaced by a polariser on both input faces of the beam splitting element 14. One of the polarisers has an orientation of +22.5° with respect to the horizontal axis and the other an orientation of -22.5° with respect to the horizontal axis. Here again the phase-step is twice the difference in orientation, that is to say 90°. This embodiment gives a low light loss, but is less dependent on the angle of incidence and on the wavelength. The phase-step can also be accurately adjusted during assembly (with the suitable adjustment facilities). In the speckle interferometer 1 with two images (interference patterns) on the camera

2 that has been described above, the two images are mirrored with respect to one another. If two non-mirrored images are compared with one another during an alignment procedure the possibility could arise that geometric aberrations have an influence on the correct alignment. In the case of non-mirrored images the alignment can be correct but pixels at the edges will correspond poorly as a consequence of symmetrical aberrations (for example barrel- or cushion-shaped deformation), whilst pixels in the centre can be well aligned. As a result of the use of two mirrored images this arises to a much lesser extent, as a result of which the system is much less sensitive to symmetrical aberrations. The position of each of the two images can be influenced by changing the angular position (in two directions) of the 45° mirror 16 or the mirror 17. After setting the 45° mirror 16 to a correct angular position, the mirror 17 can be manipulated manually until the other image has approximately the correct position. This can be checked by subtracting the two images from one another after one of the images has been mirrored. The resulting alignment is not accurate and will not be maintained for a prolonged period if the components of the speckle interferometer 1 are not fixed in an extremely stable position. The present invention solves the problem of correct alignment by making use of a control device 20, one illustrative embodiment of which is shown in the form of a block diagram in Fig. 2. The control device 20 comprises a processing device 21 (such as a system based on a (microprocessor, or other processing means) that is connected to the recording device 2 (CCD camera). The processing device is also connected to a first mirror controller 23 and a second mirror controller 24 for controlling the angular position of the 45° mirror 16 and mirror 17, respectively. The mirror controllers 23, 24 are conventional mechanical fine adjustment elements (such as micrometer adjustment elements) combined with electrically driven adjustment elements with high resolution, such as with piezo- electric or electrostrictive elements, which are mechanically connected to the 45° mirror 16 and mirror 17. Such mirror controllers provide an adequate resolution for setting the angular position of the 45° mirror 16 and/or mirror 17 and can, for example, be controlled by a computer (processing device 21). The mirror controllers 23, 24 are arranged to adjust the angular position of the mirrors 16, 17 in two dimensions. The mirrors are thus rotatable with an angle α or β, respectively, as indicated in Fig. 1 about an axis perpendicular to the plane of the drawing in Fig. 1. In addition, the mirrors can also be rotated about an axis through the respective mirrors 16, 17 and the plane of the drawing in Fig. 1. After an initial alignment, a mirror controller for one of the mirrors (preferably mirror 17) is sufficient to perform the alignment. In a first variant the alignment is carried out in an open-loop controller. For this purpose the control device 20 is provided with a display screen 25 connected to the processing device 21 and an input unit 26 (such as a keyboard, mouse or other input means) connected to the processing device 21. By comparing the two sub-images detected by the camera 2 pixel by pixel and displaying a difference signal (for example a difference image) on the display screen 25, the mirror controllers 23, 24 can be controlled via the input unit 26 in order to obtain as small as possible a difference between the two sub-images. Of course, it is also possible to set the (accurate) mirror controllers 23, 24 manually. In a second variant the alignment is carried out in a closed-loop controller. For this purpose the processing device 21 carries out the comparison of the two sub-images by calculating a difference signal and driving the mirror controllers depending on the difference signal. The pixel by pixel comparison of the two sub-images can be implemented by the processing device 21 in a number of different ways. The simplest is the pixel by pixel subtraction of the two sub-images (after mirroring one of the sub-images). Another method is to calculate, by means of the processing device 21 , a correlation coefficient that indicates to what extent the two sub-images are similar to one another. If this is employed on the two interference images as obtained by the camera 2, an indicator is obtained that can be used to quantify the degree of correspondence and also the relative position of the images. The correlation can be calculated for the complete sub-images, in which case the correlation is then calculated for a series of positions of the mirrors 16, 17 about their initial positions. The mirrors 16, 17 (or one of the mirrors 16, 17) can be accurately set in two axes down to sub-pixel level by means of their respective mirror controllers 23, 24. A curve can be fitted to the series of correlations and the correct mirror position found at the maximum correlation value of the curve. The processing device 21 then drives the mirror controllers 23, 24 to take up this correct mirror position. This can be implemented as a completely closed control loop in the control device 20 (automatic alignment). In a further embodiment it is not the complete sub-images that are used but only sections thereof, such as the mid sections. With the mirrors 16, 17 in a fixed position (not necessarily the optimum aligned state) a series of correlation calculations for the mutually shifted sections of the sub-images is performed. The mutual correlation is calculated for each position (variation over both the x and y direction). Furthermore, a cross-correlation can be calculated, implemented by, for example, a calculation of a series of correlations for different mutual positions of the images to be correlated, or via techniques in the frequency domain, which leads to a two-dimensional cross-correlation graph. Each x-y position in this graph gives a correlation value for a specific mutual shift of the two sub-images. If the sub- images are correctly aligned, the highest cross-correlation value will occur in the centre of the x-y graph (0, 0). If the sub-images are not correctly aligned, the cross-correlation peak indicates the shift between the images. The x-y graph of correlation values or the location of the cross-correlation peak can be displayed on the display screen 25, for example in colour coding or brightness coding. This enables interactive adjustment of the mirrors 16, 17 until the best alignment position has been achieved. This can, of course, also again be implemented completely as closed loop control. The highest correlation value can be found with sub-pixel accuracy from the x-y graph of cross-correlations in manners known per se, for example by interpolation or other mathematical methods. With the aid of a predetermined (and possibly pre-stored) ratio between requisite control voltages for a specific angular displacement of each of the mirrors 16, 17 and the alignment shifts in the x and y direction, the processing device 21 can then execute the control. In yet a further variant cross-correlations are calculated not for only the mid section of each sub-image, but for a plurality of sub-sections of each sub-image, as a result of which local correlation values and positions of cross-correlation peaks are obtained that cover the entire sub-image (interference pattern). The local positions of all cross- correlation peaks can be obtained, for example, by interpolation or other mathematical methods and once again displayed in x-y graphs. Such an x-y graph can be displayed on the display screen 25 to show the complete alignment characteristics of the entire image. By averaging the local data (peak positions), shifts in x and y direction, which represent alignment shifts, can be calculated and corrected interactively or using a closed loop control. Such alignment characteristics can also be used to detect any rotational faults present. Such rotational faults can be caused by deviations in the 45° reflection faces in the speckle interferometer 1. The rotation occurring can be calculated from the calculated correlation peaks (sub-pixel accuracy by interpolation) of cross-correlations for a number of subsections and corrected, interactively or with the aid of a closed loop control, by adjusting the angular position of the mirror 17 and the 45° mirror 16. By rotating the 45° mirror 16 about the horizontal axis it is possible to compensate for a rotational fault in the remainder of the system. This has the side effect of a vertical translational movement of one of the sub-images, which can be corrected by also rotating mirror 17 about its horizontal axis. If necessary, both beams must be positioned at the correct height for the camera again by subsequent adjustment of the mirrors in the object beam part. Compensation for rotational faults can be used in particular after assembly of the various elements of the speckle interferometer 1 and as a rule does not have to be repeated any more. With the aid of the embodiments of the alignment device and alignment method according to the present invention which have been discussed, rapid and efficient alignment of a speckle interferometer 1 is possible without a heavy and expensive mechanical alignment system being needed. The present invention is advantageous in particular for phase-stepped speckle interferometers 1 that record two or more phase- stepped images simultaneously. The alignment according to the present invention can be employed easily and rapidly in the field (outside the laboratory), both for initial alignment and for periodic calibration. In the case of existing speckle interferometers 1 problems with regard to the alignment can arise by modulation of the interfering images. The alignment technique is based on the availability of two sub-images that display sufficient correspondence. In the case of speckle patterns that overlap one another the intended interference occurs, but not in all positions. Many positions in a speckle pattern are dark and no modulation then occurs if a light speckle from the one pattern is coincident with a dark speckle from the other pattern. This applies to both optical paths and the intensities for pixels where no modulation occurs are then identical for both optical paths. These pixels are outstandingly suitable for inclusion in alignment. For other pixels where both speckle patterns do have a certain intensity, modulation does occur and since a phase difference (the phase step) has been applied in one of the two optical paths the resulting intensity after interference for these pixels is on the contrary different for the two paths. These pixels are thus not suitable for alignment (the correlation between images with a phase step of π/2 is zero). Under most conditions there are sufficient pixels that have not been modulated, so that alignment can nevertheless take place. However, if a flat or spherical reference beam is used as one of the two beams that can be differentiated, all speckles are modulated. The reference beam then has to be blocked for the alignment, so that the two sub-images (originating from the two optical paths) are in principle identical to one another. In the case of shearography as well, where two speckle patterns interfere with one another, it is advantageous to prevent modulation, so that as large a number of usable speckles as possible can be available in the alignment procedure. This can be implemented in various ways. Mirror 11 or mirror 12 can be blocked, as a result of which no further interference takes place. Furthermore, the polarisation of the light entering via the lens 7 can be changed into exclusively horizontal or exclusively vertical, as a result of which the mirror 11 or the mirror 12 receives no further light. This can be achieved using a (rotatable) polariser (not shown) or a combination of a fixed polariser with a rotatable λ/2 plate. These elements can be positioned before or after the lens 7, lens 8 or before the beam splitter 9. Temporary positioning (during the alignment) before lens 7 is operationally advantageous and adequate. Positioning before beam splitter 9 is optically better when the elements concerned are permanently fitted. In this case a control device for the polariser or the λ/2 plate does have to be provided which sets the correct polarisation. A simple implementation is constituted by a controllable delay element based on liquid crystal, because this contains no mechanically moving parts, which is beneficial for the reliability. As an alternative, an optical shutter based on an LCD element with polansers can be used, by means of which light is or is not allowed through, depending on a control voltage. This optical shutter can be placed between mirror 11 or 12 and the beam splitter 9. A further alternative is offered by the use of a spatial light modulator (SLM), which offers the additional advantage that the speckle patterns can be further manipulated so that a number of independent patterns can be obtained by means of which the signal/noise ratio of the speckle interferometer can be improved. The embodiments described above are intended solely as explanation and not as a restriction on the scope of protection, which is defined by the appended claims. Thus, for example, the orientation of the parts 4, 5 and 6 of the speckle interferometer 1 (and thus the mutual orientation of object 3 and camera 2) can be changed.

Claims

1. Method for the alignment of an interferometer arrangement, wherein the interferometer arrangement comprises an object beam part (4) for obtaining two beams that can be differentiated, at least a first beam of which that can be differentiated originates from an object (3) to be tested; a beam splitting part (5) for splitting the two beams that can be differentiated between n optical paths, where n is an integer and n > 2; and a beam combination part (6) that is arranged to receive the two beams that can be differentiated, to apply a phase step in the mutual phase difference present between the two beams that can be differentiated in n-1 of the n optical paths and to direct the two beams that can be differentiated and travel over each of the n optical paths to a detector unit, wherein the detector unit is arranged to detect an interference pattern for the two beams that can be differentiated, via each of the n optical paths, and the interferometer is aligned by setting at least n-1 of mirroring elements in the beam combination part in order to shift the interference pattern on the detector unit belonging to at least n-1 of the optical paths, characterised in that the at least n-1 mirroring elements in the beam combination part are set depending on an alignment signal derived from the detected interference patterns.

2. Method according to Claim 1, wherein the alignment signal is displayed on a screen and the at least n-1 mirroring elements are set interactively.

3. Method according to Claim 1, wherein the at least n-1 mirroring elements are set with the aid of a control signal, derived from the alignment signal, to an associated mirror control.

4. Method according to one of the preceding claims, wherein the alignment signal is a difference signal from image data for the interference patterns belonging to the n optical paths.

5. Method according to one of Claims 1 to 3, wherein the alignment signal is obtained by a correlation of image data for the interference patterns belonging to the n optical paths.

6. Method according to Claim S, wherein the alignment signal corresponds to the maximum correlation signal that is determined from image data for a series of interference patterns that are recorded by the detector unit at different settings of the at least n-1 mirroring elements.

7. Method according to Claim 5 or 6, wherein the alignment signal is derived from a cross-correlation of image data, for example of sections, for the interference patterns belonging to the n optical paths.

8. Method according to Claim 5 or 6, wherein the alignment signal is derived from a cross-correlation of image data from a plurality of sections of the interference patterns belonging to the n optical paths.

9. Method according to one of the preceding claims, wherein the alignment signal is representative of a rotational shift.

10. Method according to one of the preceding claims, wherein one of the two object beams that can be differentiated is blocked.

11. Alignment device for aligning an interferometer arrangement, wherein the interferometer arrangement (1) comprises: an object beam part (4) for obtaining two beams that can be differentiated, at least a first beam of which that can be differentiated originates from an object (3) to be tested; a beam splitting part (5) for splitting the two beams that can be differentiated between n optical paths, where n is an integer and n > 2; and a beam combination part (6) that is arranged to receive the two beams that can be differentiated, to apply a phase step in the mutual phase difference present between the two beams that can be differentiated in n-1 of the n optical paths and to direct the two beams that can be differentiated and travel over each of the n optical paths to a detector unit, wherein the detector unit detects an interference pattern for the two beams that can be differentiated, via each of the n optical paths, and the interferometer arrangement is aligned by setting at least n-1 of mirroring elements in the beam combination part in order to shift the interference pattern on the detector unit belonging to at least n-1 of the optical paths, characterised in that the alignment device comprises a processing device (21) that is connected to the detector unit and at least n-1 mirror controllers (23, 24) for setting the at least n-1 mirroring elements (16, 17) in the beam combination part (6), where the processing device (21) is arranged to drive the at least n-1 mirror controllers (23, 24) depending on an alignment signal that is obtained from the interference patterns received by the detector unit.

12. Alignment device according to Claim 11, wherein the mirror controller (23, 24) is provided with an electrically controllable fine mechanical adjustment element, such as a piezo-actuator or an electrostrictive element.

13. Alignment device according to Claim 11 or 12, wherein the processing device (21) is connected to a display screen (25) and input means (26) for the interactive alignment of the interferometer arrangement.

14. Alignment device according to Claim 11, 12 or 13, wherein the processing device (21) is arranged to calculate difference values from image data for the interference pattern belonging to each of the n optical paths.

15. Alignment device according to Claim 11, 12 or 13, wherein the processing device (21) is arranged to calculate correlation values from image data for the interference pattern belonging to each of the n optical paths.

16. Alignment device according to Claim 15, wherein the processing device (21) is arranged to determine the alignment signal from a series of interference patterns that are recorded by the detector unit at various settings of the at least n-1 mirroring elements (16, 17).

17. Alignment device according to Claim 15 or 16, wherein the processing device (21) is arranged to calculate cross-correlation values from image data for the interference pattern belonging to sections of each of the n optical paths.

18. Alignment device according to one of Claims 15 or 16, wherein the processing device (21) is arranged to calculate cross-correlations from image data, for example from a plurality of sections of the interference patterns belonging to the n optical paths.

19. Alignment device according to one of Claims 11 to 18, wherein the alignment signal is representative of a rotational shift.

20. Alignment device according to one of Claims 11 to 19, further comprising blocking means for blocking one of the two object beams that can be differentiated.

***************