DE19846145A1 - Three-dimensional imaging device for shape measurement has transmitter array whose elements move in straight, parallel lines - Google Patents

Abstract

The imaging device has a movement system, the components of which are arranged so that a transmitter array moves in a direction parallel to a straight line (gA), while a receiver array picks up an object. The elements of the transmitter array move on straight lines parallel to the line gA, which intersects the focal point of an illumination lens in the region of the transmitter array. An Independent claim is given for a three-dimensional imaging method.

Description

Technical application area

The technical field of application is the determination of the 3D shape or 3D shape of bodies or objects in space or even complete scenes with at least one image sensor in an optical recording system. The one necessary arrangement is further referred to as a 3D recording arrangement and that The process is referred to as the 3D recording process.

This 3D recording arrangement and this method are used firstly in the sense of 3D measurement technology. As a result of the calculation it becomes optical signals obtained the 3D point cloud of an object or a scene with Be train determined to a zero point of the 3D recording arrangement. The calculated one Point cloud can be processed further in a CAD system, for example.

The second area of application is the acquisition of 3D images of objects or scenes for 3D playback devices and 3D playback methods. This mainly affects the multimedia and television sector.

State of the art

The 3D shape of surfaces is often used with strip triangulation methods measure up. The object or scene is usually under a large one Illuminated angle of incidence, for example between 30 ° and 60 °. However, this leads Frequently disturbing shadowing of the object or details in the scene.

A variety of 3D measuring arrangements with coherent light are known. The influence of the speckle phenomenon, however, limits the resolution. This was in the works "Automation of Optical Quality Inspection" by H. J. Tiziani in Technical Messen, 55th year, issue 12/1988, pp. 481-491 on page 488.

Interference fringe fields caused by the superimposition of coherent plane waves be produced, also have speckle effects, s. Technical measurement, 62. Volume 9/1995, pp. 328-330. These speckle effects limit the sensible Elevation resolution often less than 1/50 of the effective wavelength.

For security reasons, the admission of people is prohibited directional laser radiation.

In German patent DE 44 13 758 C2, published on September 17, 1998, FIGS . 4-7 show a possibility of obtaining interference fringes for shape measurement by means of a two-beam interferometer with roof edge reflectors through lateral shear of two coherent light beams. Elimination of the lateral shear results in an interference fringe field with few stripes. However, a beam splitter system must be used to eliminate the lateral shear. This leads to a rather large-volume arrangement with high demands on the quality of the lenses used.

In conventional optical 3D measurement, discontinuous surfaces are often a problem. For example, larger paragraphs or levels in cause a violation of the sampling theorem on the surface of the object. This is remedied by the Gray code method, in which a series of Binary images are projected on and then - especially at higher accuracy requirements - sinusoidal grids are mapped onto the object surface and the known phase shift method is used. This is an example Modular optical 3D measuring system opto TOP from Breuckmann GmbH in D-88709 Meersburg. Furthermore, the COMET-500 system from Steinbichler Opto technik GmbH in D-83115 Neubeuern.

The company ABW in D-72636 Frickenhausen offers programmable line projectors are available with up to 1280 lines for the absolutely measuring Moire method. Projecto LCD-based devices still work relatively slowly and the strip to be observed contrast on the object is worse than that of projectors with lattice structures Glass plates. The company Gottfried Frankowski Meßtechnik in D-14513 Teltow offers digital light projection based on micromirrors, digital light devices, also known as DMD. Grid images with a repetition rate of about 10 Hz are generated and read. However, this frequency is for the High-speed image acquisition is not yet sufficient.

To achieve a high depth of field when measuring deep objects, will generally bleed to a greater or lesser extent in all of the methods mentioned det, usually with small lens openings, for example with a relative Luminous intensity of 1: 8, worked. This requires strong light sources, for example in Form of flash lamps or strong halogen lamps, or it is just the illumination comparatively small fields possible, for example 200 mm × 200 mm to 500 mm × 500 mm.

At the 2nd ABW workshop 3-D BV / TAE from January 25-26, 1996, R. Lampal zer, G. Häusler and Schielzeth on the advantage of a large lighting aperture Reduce speckle noise by gaining spatial incoherence pointed out.

The state of the art in the acquisition of analog 3D images is based on the Usually on the use of two cameras as with two-eyed vision. this will not considered further, since this involves the acquisition of 3D data in digital Shape goes.

In the past, using two cameras caused problems with the lateral accuracy when generating the 3D point cloud. New Developments, for example from the company gom in D-38106 in Braunschweig, leading to a series of sensors based on the projection of different Stripe patterns and capturing with two cameras from different perspectives angle. However, the sometimes considerable dimensions of the measurement are disadvantageous systems. That is why the call for miniaturization is highly accurate 3D measurement technology, the one-camera technology and the one-projector technology apparently basically an advantage.

A new process with three cameras with optical axes arranged in parallel and a suitable base distance solves the problem of capturing the 3D scene  by evaluating stereoscopic images using a high-performance computer, s. VDI News No. 21 of May 22, 1998, page 15, "Journey into the Radiant Rui ne ". This 3D information is used to control a robot, of the robot "Pioneer", in the Chernobyl ruin and for 3D acquisition of the bauli conditions and the environment. Only with high performance Computing technology enables real-time 3D image reconstruction. The Bauvolu This arrangement is comparatively large due to the three cameras.

In the conference proceedings "Optical shape detection" GMA report 30, DGZfP-VDI / VDE- GMA conference 28.129. April 1997, Langen, Federal Republic of Germany, p. 199-209, is by the authors, R. Schwarte, H. Heinol, z. Xu, J. Olk and W. Tai da pointed out that for the rapid shape detection in the approx. 20 cm to 50 m No precise, flexible and cost-effective procedure is available. This view applies particularly to the area from 20 cm to 2 m, since here too the runtime methods only have a depth measurement accuracy of 2 mm to 20 mm sen. The above mentioned technical solution presented on the The basis of a photonic mixer detector PMD is currently not considered suitable seen to realize a high depth accuracy in the mentioned close range. The technical effort z. Currently rated as quite high.

In the above-mentioned conference volume "Optical shape detection" on pages 211-222, by the authors W. Schreiber, V. Kirchner and G. Notni, the concept for a self-measuring, optical 3D measuring system based on structured lighting is presented. Based on the experience in photogrammetry, it allows the simultaneous determination of system parameters and coordinates from the measured values. An extraordinarily high measuring accuracy for the object coordinates is achieved with a relative error of up to 10 ^-5 . To do this, however, the measured value must be recorded several times, using Gray code sequences and strips with a sinusoidal profile in conjunction with phase shifting techniques. This means that a considerable amount of time is required to carry out the measurement due to the need to use several gratings in one measurement sequence. The object must also be illuminated in succession from several projectors in different positions or from the same projector in different directions. This also does not allow highly dynamic measurement or real-time 3D detection of objects.

The main features of this new and very advantageous optical process are below the title "Optical three coordinate measurement with structured lighting" in Technical measuring, 62nd year, issue 9/1995, pp. 321-329 by W. Schreiber, J. Gerber and R. Kowarschik. The reference point for the coordinate system stem lies in the area of the object.

Figure 2 of this latter publication shows an arrangement with parallel Axes for the lighting and imaging lens and a two-sided zen illustrated perspective view.

However, it can be estimated that due to the specified Appa rates and the mathematical model the implementation of the evaluation algorithms in straight forward algorithms very difficult, so that a real-time Evaluation using special high-speed processors in this way is hardly possible.  

Since the camera and the projector are spatially separated, the realization is one compact 3D measuring module not possible.

In 1997, Daimler-Benz Aerospace installed the 2nd generation of a laser camera the basis of the time-of-flight measurement of the light. The measurement accuracy is only at et wa 2π of the measuring range and is thus for metrological applications in the near area considered unsuitable. Obtaining the color information from the Scene is considered practically excluded.

In the description of the invention P 197 49 974.0 "Process and apparatus for generating 3D point cloud for topometry or 3D vision for multimedia Applications "by the authors K. Körner and L. Nyársik will be a solution for the High-speed image acquisition specified. However, not in detail shows which special arrangement features for a 3D recording Arrangement and which process steps to follow for a 3D recording process are.

In the thesis "General approach for the description of optical 3D-measuring system" by the authors P. Andrä, W. Jüptner, W. Kebbel and W. Osten in SPIE Vol. 3174, S. 207-215 is a general description of optical methods for the Given 3D coordinate measurement. Deriving suitable algorithms for Here is the high-speed evaluation of image series for 3D measurement not given, but was also not the goal of this work.

A 3D real-time camera was featured in Poster P 28 of the 99th conference of the German Ge society for applied optics from 2.6 to 6.6.1998 in Bad Nenndorf by G. Bohn, L. Cloutot, G. Häusler, M. Hernandez, C. Horneber, R. Lampalzer, S. Seeger below Use of a ferroelectric display presented. This ferroelectric display can be switched in 0.1 ms and is used in conjunction with an astigmatic image used to generate sinusoidal strips of fixed spatial frequency. The Measuring time of the prototype is currently 320 ms. An absolute coding of the Rau So far, however, mes using structured light has not been possible. The reachable Re accuracy is currently limited to about 1/2000 of the measuring range.

Object achieved with the invention

The invention solves the task of areal inspection of the 3D shape techni and natural surfaces of objects in space and scenes with one absolute reference to the 3D recording arrangement. The shape of objects in space and scenes can be measured with high accuracy in depth and with high speed be measured.

Furthermore, the light output required for illuminating the scene is partially very much reduced, or it can reduce the image recording frequency with high light output increase. The evaluation process results in a resolution of up to 1/2000 the effective wavelength in depth. However, this is not always the case sensible and possible.  

Technically, the examination of the surface shape of workpieces of the machine construction, vehicle construction and also aerospace. The application is now in automated production, in robotics, under construction beings, in medicine, for example in the geometric measurement of men human, biological objects and also in the artistic field of plastics or Given stucco.

Furthermore, the task of cavity and interior inspection is solved. La byrinthe in different sizes and also contaminated or contaminated inside spaces to humans due to serious health risks or are not accessible due to the small dimensions, can by Ro bots are committed using one or more 3D imaging arrangements are equipped. It can be rolling, striding, sucking, swimming Acting, flying, crawling and crawling robots. Also is the sub Use of water for shipwreck inspection with the 3D image arrangement at least possible.

Also the electronic support of visually impaired people for orientation Space in the room is possible with a miniaturized 3D recording arrangement. At Recycling processes is one way of automatically dismantling Altge councils, old bodies and the separation of waste products also in a ge given dangerous environment.

The different reflectivity of technical surfaces, which for example caused by a pronounced texture can be compensated become.

Improvements and advantages achieved compared to the prior art

The invention enables the rapid acquisition and testing of the 3D shape of Bodies and scenes with great depth, preferably with dimensions in the area above one millimeter. By lighting the object under egg nem comparatively small triangulation angle, for example by 10 ° or smaller, can also smooth surfaces with a low spreading capacity without Preparation of the surface - as is often still the case in practice today - recorded men.

In addition, the spatial structure of non-technical objects can be captured the, the space in which the objects are located in different The depths are illuminated in a structured manner one after the other. With a correspondingly high Dy The namics of the components used for 3D imaging can also be used objects and scenes including people and animals the. The real-time capability of the process is basically given. Are there basically several 3D recording arrangements can be used, and the color be processed the object points when using color image sensors become.

The shadows of light that interfere with the detection of spatial structures due to a Relatively large triangulation angles from 30 ° to 60 °, for example, are widely used avoided.  

A further improvement is the significant increase in the evaluation speed 3D acquisition. There is a real technical possibility that 3D point cloud of the object or scene is available in video clock put.

The body to be captured in its 3D form or the entire scene are shown preferably perpendicular to the optical axis of the taking lens. That allows due to the optimal imaging conditions a high lateral resolution and the use of tomographic or deep-scanning methods. The 3D point cloud has an absolute geometric reference to a zero point 3D recording arrangement.

The application of the invention enables the almost complete utilization of the high performance optical imaging systems for testing the 3D surface design.

The realization of a high lighting aperture ensures a low speckle Noise in the pixels of the camera.

On the other hand, complete scenes with moving objects can be viewed as real-time 3D scenes can be recorded. The limit for the detectable volume is the available light energy, the light intensity of the lenses used and the photo represents the metric sensitivity of the image receivers used tangible space can be up to 20 m. The limit is ultimately by the signal-to-noise ratio in the light receiver or image receiver.

Basics of the solution

The basic features of the method and the arrangement are described. Means a transmissive or reflective, gridded structure with several maxima and minima of transmission or reflection and a light source becomes a struktu generated, luminous surface with local extremes of luminance. This straight Sterte structure is hereinafter referred to as a transmitter array. The transmitter array is at least one lighting lens that is as open as possible with a positive one Subordinate focal length for mapping this transmitter array. So created by the Image of the transmitter array in the space of the object on its surface structured lighting.

In principle, the transmitter array can also be self-luminous and local extremes the luminance.

For the inventive purpose, the transmitter array is depicted in depth the area of the object is as small as possible ches realized so that the spatially structured lighting is more in one disc-shaped volume is located. There is at least one for the lighting lens Shooting lens arranged to depict the object or the scenes. Illumination and shooting lens are preferably of the same type and thus have the same focal length. The focal planes also fall the lenses in the space of the object preferably together and the two  Lenses are preferably arranged at a short distance from one another with front preferably parallel axes. If necessary, the lenses can even in a ge shared housing.

The lenses advantageously have a large image field. The correction in the Depth should be good in a wide range of depths. Due to the coincidence of the Focal planes in the space of the object, the lenses can always be the same Object and image levels are set and then form due to the low Ab the lens axes stood in one area, preferably at a distance Lich below the hundredfold focal length of the lighting lens - typical is three to six times - at least an identical part of the object field in the space of the object, usually scaling down, into the space of the arrays. In There are separate picture fields in this room, but preferably in a common one seed plane, which is preferably perpendicular to the optical axis. The Focus areas of the two lenses in the space of the object preferably fall into the largest possible area.

The imaging lens has a preferably telecentric in the space of the arrays beam path in the strict sense, d. H. the exit pupil is very far away from the lens, for example by 100 m. The lighting lens points in Space of the arrays preferably a likewise well corrected telecentric Beam path on. However, it is also possible in principle that this lighting lens in the space of the arrays a parallel beam path with a decent entry pupil.

In the image plane of the taking lens there is another array, the receiver Array arranged. This array is also preferably located in a telecentri beam path. This receiver array can be a rasterized image receiver, for example, a receiver matrix or a micro-optical array. The micro-optical array in the image plane of the lighting lens is preferably two se a microlens array of diffractive or refractive elements.

The telecentricity of the imaging lens in the space of the arrays ensures that when the receiver array is moved parallel to the axis of the imaging lens, i.e. in the usual notation in the z direction - here the z _A direction, the imaging rays have a fixed position to the elements of the Maintain receiver arrays, i.e. do not laterally migrate the imaging beams. In contrast, in the space of the object, the illumination and the imaging lens have a preferably central perspective beam path in order to be able to capture a large spatial area.

In the case of using a microlens array in the imaging beam path as Receiver arrays are further components of the microlens array, such as an image recipient, subordinate.

By moving the transmitter array and the receiver array together by means of a movement system, parallel to the optical axes of the lighting objective, i.e. in the z _A direction, all areas of the object or scene are gradually reduced in depth the respective focus area is illuminated and displayed. The movement system is preferably designed as a translatory movement system. Structured lighting is created in each case in exactly the plane of the object or the scene by imaging the transmitter array, which is sharply imaged by the recording lens on the receiver array. The known decrease in the illuminance on the object surface is compensated for as a function of its distance by adapting the luminance of the light source to the respective distance of the focus area from the 3D recording arrangement.

The following movement regime is implemented for the transmitter array and the receiver array: The amounts of movement of the two arrays parallel to the optical axis of the illumination and imaging lens, the z _A direction, are included in an illumination and recording lens same focal length the same.

The transmitter array additionally and preferably guides time by means of a linear guide equal to the movement parallel to the optical axis of the lighting lens lateral movement, that is perpendicular to the optical axis of the Lighting lens.

In the case of an electronically controllable array, the ele are shifted laterally elements of the same luminance, or the displacement of the locations of the local extremes the luminance preferably at the same time as the movement parallel to the optical Axis of the lighting lens through electronic control of the transmis sion, the reflection - or the luminance with a self-illuminating array - each because in the elements of the array. Be on the electronically controllable grid the local extrema of the luminance by local extrema of the transmission, reflection or by directly controlling the luminance of yourself generated luminous array.

For the grid elements or the elements of the same luminance or the locations of the local extremes of the luminance of the transmitter array or also the locations of the same phase of the luminance distribution, a linear movement is thus generated which is aligned parallel to a straight line g _A. This straight line g _A is defined such that it always intersects the focal point of the illumination lens in the space of the arrays and the rise, defined as the quotient of the focal length of the illumination lens in the space of the arrays and the distance of the focal point of the imaging lens in the space of the object from the axis of the illumination lens , has, to which the straight line g _{A is} based on an axis perpendicular to the axis of the illumination objective, which lies for example in its focal plane in the space of the arrays.

In the case of the parallel position of the two axes of illumination and imaging objectively, the distance between the two axes corresponds to the lighting and image The objective of course is the distance of the focal point of the image object jective in the space of the object from the axis of the lighting lens.

The straight line g _A - for a given arrangement with two parallel axes of the illumination and imaging lens and coincident focal and main planes of the two lenses - always intersects both the focal point of the lighting lens on the side of the transmitter array and on the same side the main point of the imaging lens. In this case, points along this straight line are mapped according to the known Scheimpflug condition onto a straight line g _O parallel to the objective axis of the illumination objective.

The images of the straight lines parallel to the straight line g _A , which contain the locations of the elements of the transmitter array, form a bundle of straight lines in the space of the object with the straight line g _O , the intersection of all straight lines of this bundle always being at the focal point of the arrangement described Imaging lens should be in the space of the object. At the same time, all imaging rays of the imaging lens in telecentricity in the space of the arrays also intersect the focal point of the imaging lens in the space of the object and thus form a bundle of rays. The lines of the bundle of straight lines and the rays of the bundle of rays coincide. Thus, with an object surface perpendicular to the axis in the area of the arrays and a movement of the transmitter array in the direction of the straight line g _A, the pixels belonging to the respective imaging beams on the receiver array each detect the same amount of phase shift. This phase shift corresponds to the shift in the elements of the transmitter array or the shift in the luminance distribution on the transmitter array. A parallel shift of this object surface leads to the same amount of phase shift in the signals, which can be detected in all pixels. This means that the depth sensitivity of the described 3D recording arrangement in a plane perpendicular to the axis in the space of the object is a constant, that is to say has no lateral dependency. The depth sensitivity can be described by the effective wavelength of the 3D recording arrangement. The effective wavelength is defined here as the amount of displacement Δz _{2π of} an object point on an imaging beam, in which the phase has changed by exactly 2π in the same imaging beam, with Z _OB as the coordinate in the space of the object that is parallel to the axis of the illumination lens .

On the basis of the depth sensitivity that is constant in a plane perpendicular to the axis the further processing of calculated phase values 3D point cloud greatly simplified. In this case, the calculations can done quickly. Straight formrard algorithms can be used men. This represents a great advantage of the described 3D recording arrangement represents.

An arrangement with two rigidly arranged lenses is technically very large accuracy regarding the parallelism of the axes and the position of the focal planes reali sable and also comparatively economical to produce and is therefore favors.

However, the 3D recording arrangement can also be constructed as follows: the lighting lens is assigned a transmitter array for the purpose of imaging, which can be moved at an oblique angle to the axis of the lighting lens. The rectilinear direction of movement of the array defines the position of a straight line g _A and is made structurally unchangeable with respect to the position of the axis of the illumination lens. The arrangement is so constructed that the line g _A intersects the focal point in the array as well as the main plane of the illumination lens at a finite distance from its optical axis, namely where the axis of the imaging lens should lie. The real or mental mapping of the straight line _A through the lighting lens into the space of the object creates the straight line g _O parallel to the axis of the lighting _lens . This straight line g _O penetrates the focal plane of the lighting lens in the space of the object. The object-side focal point of the imaging lens is adjusted as precisely as possible in the point of intersection of the straight line g _O through the focal plane of the lighting lens, the axis of the imaging lens preferably being aligned parallel to the axis of the lighting lens, so that the axis of the imaging lens with the position of the straight line g _O collapses. The straight line g _A has, more or less well approximated, the increase with the amount from the quotient of the focal length of the lighting lens and the distance of the focal point of the imaging lens in the space of the object from the axis of the lighting lens, which increased to a perpendicular straight line is related to the axis of the lighting lens.

The movement regime for the 3D recording arrangement can be realized on the one hand by a system with two linear guides, the axes of which are preferably aligned perpendicular to one another. The moving components of these linear guides, also referred to as slides, can be controlled independently of one another in the linear movements. Miniaturization of these components is possible. Linear direct drives can be used, for example electrodynamic moving coil systems or linear motors. The direction of movement of the moving component of the first linear guide is parallel to the direction of the optical axis, the z _A coordinate, that of the movement component of the second linear guide in a direction perpendicular to it, the x _A direction. These two linear guides thus testify the movement of the elements of the transmitter array parallel to the line g _A. The receiver array is assigned to the first linear guide, which works in the direction of the optical axis. The slide of the first linear guide thus carries the receiver array and preferably the second linear guide and the slide this second linear guide carries the transmitter array.

On the other hand, a single linear guide can also be used with a slide which carries the transmitter array and the receiver array simultaneously on this slide. The direction of movement of this slide is parallel to the straight line g _A - not perpendicular to the axis, but at an oblique angle to the lens axes. In the latter case, however, there is an undesirable shift of the image to the pixels of the receiver array on the receiver array, corresponding to the carriage movement. This shift can be calculated back pixelwise in the image evaluation algorithm. A high computing speed can be achieved if the carriage movement and the recording of images are synchronized and images are only recorded depending on the carriage position if the displacement of the image on the receiver array occurred at least one whole pixel or an integral multiple thereof is. The storage of the image data in the data set is always taking into account the image shift that occurs, so that the result is a compensation of the displacement of the image using numerical methods. For example, a certain real imaging beam is tracked in the object space by assigning the stored data in the image data set, which consists of several layers or data fields. Each layer corresponds to a picture taken. In this way, a real imaging beam emanating from a specific element of the transmitter array always remains assigned to an element with the same indices, that is to say the same position, in each layer or depending on the data field of the entire image data set in the entire movement process of the two arrays.

However, the method with only one linear guide effectively reduces it grasped object width by the lateral displacement of the receiver array, so that a longer receiver array is required to capture a standard format image to be able to. The additional required length of the receiver array is solely from depends on the length of the shift, which results from the number of images taken which results per movement of the slide, also referred to as a scan.

For a motion system with two linear guides, proceed as follows:
The lateral and linear movement of the transmitter array takes place by means of the carriage of the second linear guide and preferably begins at a highly stable zero point. The transmitter array can represent a line grid, the translation of which can be detected by a highly precise length measuring system. Preferably, however, the lattice movement is detected in terms of phase with a high degree of accuracy by means of a second line lattice, so that the absolute lattice phase based on the zero point is also known in fractions of 2π. High-resolution interpolation techniques are known from the incremental length measurement technique. The lattice phase itself can thus represent the reference for scanning the optical signals picked up from the object surface from the object space. So the phase of these signals can be determined highly precisely. This is very important, since the phase of the optical signals and their location in connection with the geometry parameters of the arrangement contain the information about the respective object position.

In the case of lenses of the same type, the area of the receiver array is preferably in the same plane as the area of the transmitter array. In addition, the transmitter array, here the line grid, by the movement system or, in the case of an electronically controllable array, the local extremes of the luminance, in addition to the movement in depth, perform a lateral movement, so that the resulting movement of the line grid or the local extremes Luminance is a linear movement parallel to the line g _A already mentioned. This movement changes the phase of a signal which can be observed in the pixel of an object point.

On the other hand, through this movement regime, the geometry of the arrangement and the telecentricity of the lenses in the space of the arrays that illuminates beam of a transmitting element of a certain phase and an image beam that intersect in any focus, in all Cut sharpening areas. So it is also given that when moving the two ar rays the images of the sending elements of the same phase and those of the receiving ones Elements of the two arrays with an ideal arrangement in the entire object space always collapse. Lenses with low distortion and very ge wrestle errors and the smallest possible focal length difference between the two objects ve and an adjustment that is as precise as possible is an advantage and reduces the numerical correction effort considerably. The deviations from the ideal state can only be tolerated within certain limits.

Due to the movement of the line grid parallel to the line g _A and the "fixed coupling development of the focus areas" of the two lenses in the space of the object to a common focus area, the "carrying the phase in the imaging beam" takes place. This means that the phase remains constant in every imaging beam that reaches the imaging objective from a point on the object surface that is moving in the common focus area, even if the position of this focus area changes in the depth of the space of the arrays. An imaginary object point running continuously in the imaging beam in the focus area would consequently experience no phase change in the detectable signal when the line grid is moved, since, for example, the same section of the strip is always detected in the pixel associated with the imaging beam.

In contrast, a periodic signal can always be observed in the pixel of a fixed object point when moving the line grid. In this periodic signal itself and due to its position in relation to the phase of the line grid, the information about the current z _OB position of this object point in the space of the object is clearly contained.

The goal is to vary the position of the focus areas in depth up to every object point of the object and a small amount beyond, in order to be able to detect the signal in the vicinity of the object point, the object phase to determine each plane perpendicular to an object point men. This is done by detecting a modulated periodic signal a modulation maximum in each depicted object point, that is, in the associated one towards the pixel. The periodic signal occurs when the two pass coinciding areas of focus through the plane of the respective object point and the maximum modulation corresponds to the position of the coinciding focus areas. A certain deviation from the coincidence of the focus areas leads to a effective focus area and is tolerable within certain limits. There is one Reduction of the maximum modulation in the detected signal.

It is possible to couple the focus areas of the two lenses in the space of the Execute object so that a certain phase value with respect to the Modulation maximum in the observed signal by the lateral fine adjustment of the Sender arrays, or the line grid, results.

The periodic signal in the pixels is preferably more constant in steps Phase change sampled. Steps of constant phase change arise in the Ob jektraum through constant steps of a moving grid, which is a line grid with a constant grating period, in the space of the arrays. Under the Arrays Room the space is always understood where the receiver array and the transmitter Array, here the line grid.

It is assumed that the z _OB position can be described as a distance from the focal plane of the illumination lens in the space of the object of each perpendicular plane in the object space by a laterally invariant, absolute _object phase ϕ _Obj , so that it is perpendicular to the plane of the constant phase in the space Object there.

The laterally invariant, absolute _object phase ϕ _Obj at a point of the object is only dependent on the position z _{Obj of the} same and the geometry of the optical arrangement and therefore has no lateral dependency with ideal imaging conditions and adjustments. Furthermore, the _object phase ϕ _Obj is usually spoken of. The geometry of the optical arrangement is described by the focal length f _{B of} the illumination lens, the distance d between the two mutually parallel lens axes and the grating constant p of the transmitter array, here a line grating. The straight line g _O , the image of the straight line g _A , coincides with the axis of the imaging lens, so that the size d also describes the distance of the straight lines g _O from the axis of the illumination lens.

The following geometry model is assumed: The laterally invariant, absolute _object phase ϕ _Obj for a plane perpendicular to the axis with the distance z _OB always corresponds to the number of strip orders plus the two strip fractions that are on the path between a point on this plane the optical axis of the imaging lens and a point on this plane lie on the straight line g _O.

In the infinite the object phase is therefore zero due to the infinite stripe width and in the focal plane of the object space the object phase approaches the infinite due to the stripe spacing zero. The _object phase ϕ _Obj usually has negative values for the chosen notation. For (-) z _OB = f _B the object phase ϕ _Obj = ϕ _fB and it applies

with d as the distance of the straight line g _O from the axis of the illumination lens, or here also the distance between the parallel axes of the two lenses, and with p as the lattice constant of the line grid which is used as a transmitter array.

With a flat and axis-perpendicular reference plate, the object phase can be determined in the space of the object as the number of stripe orders between the straight line g _O and the axis of the lighting objective, or here between the two optical axes, by counting stripes or more precisely using known phase shifting technology. The problem here is firstly the knowledge of the exact position of the two optical axes in the arrangement. The highly precise determination of the number of strips between the optical axes can be done in several steps when the reference plate is moved in parallel - both as a criterion for checking the deviation from the parallelism of the optical axes of the two lenses and the position of the reference plate perpendicular to the axis. With a calibrated arrangement with a known position of the optical axes with respect to the transmitter and receiver array, the reference phase can be determined experimentally.

In order to obtain the z _OB coordinate of an object point z _Obj along an imaging beam, the laterally invariant, absolute object phase cpobj is to be determined in each object point, which is referred to below as the object phase. An imaging beam should hit exactly one pixel of the laterally fixed receiving area in the model.

The _object phase ϕ _{Obj of} an object point cannot be determined directly, but only in relation to a reference surface. The absolute, laterally invariant phase of the reference surface, the reference phase ϕ _R , is taken from a distance z _OB = z _{OR of} the reference surface which is assumed to be known and has a sign

calculated, where d is the distance of the straight line g _O from the axis of the lighting lens, or here the distance of the parallel axes of the two lenses from each other, f _{B represent} the focal length of the lighting lens and p the grating constant of the transmitter array. Since z _{OR is} counted in the negative axis direction, there is also a negative value for the reference phase ϕ _R. The distance z _{OR of} the reference surface is determined experimentally as precisely as possible. As a rule, all imaging beams used to capture the object or the scene also hit the reference plate.

The basic idea is to evaluate in each imaging beam the evaluation of the modulated, periodic signal to be observed in the associated pixel of a reference point of the reference plate over the phase of the grating ϕ _grating and the modulated, periodic signal to be observed in one pixel of an object point over the phase of the grating Signals to determine the absolute phase difference Δϕ _grid these two signal layers from the phase of the grid. It is assumed that the periodic signals have a modulation with a maximum modulation. The width of this modulation curve over the phase of the grating ϕ _grating or the assigned displacement path of the grating depends on the center distance d of the two objective axes, the focal lengths of the two objectives and the relative aperture of the two objectives, each described by the f-number k of the two objectives, and the properties of the surface in terms of light scattering.

The absolute phase difference Δϕ _grating can be determined such that the signal curve to be observed in an object point above the phase of the grating is shifted over the phase which is derived from the displacement path of the grating by exactly the amount of phase that this signal curve changes with covers the signal curve to be observed in the associated reference point of the reference plate as precisely as possible, ie the correlation of these two signal curves is as high as possible. This phase amount determined in this way then corresponds to the absolute phase difference Δϕ _grating as the difference between the absolute phase of the grating ϕ _grating associated with the two signal profiles.

For this purpose, first of all, the relative reference phase value ϕ _RR mod 2π is determined from a sampling of the signal over the phase in the pixel of a reference point. Secondly, the relative object phase value ϕ _RObj mod 2π is determined from a sampling of the signal over the phase in the pixel of each object point. The relative reference phase value ϕ _RR and the relative object phase value ϕ _RObj are each assigned to the absolute phase of the grating ϕ _grating and subtracted from this, taking into account the respective sign. This creates the absolute phase values of the grid phase ϕ _GridR for a reference point and ϕ _GridObj for an object point. With the difference

Δϕ _grid = ϕ _gridObj -ϕ _gridR (3)

the absolute phase difference Δϕ _{grating is} determined, which has a positive value when the detected object point is further away from the focal plane than the associated reference point. Due to the signed addition of the absolute phase difference Δϕ _grid to the experimentally or constructively derived reference phase ϕ _R , the absolute _object phase ϕ _Obj then becomes

(-) ϕ _Obj = (-) ϕ _R + Δϕ _grid (4)

certainly. Since the sign of the phase of the reference surface ϕ _{R is} negative and the phase difference Δϕ _lattice is always smaller than the phase of the reference surface ϕ _R , a negative value results for the absolute _object phase ϕ _Obj . The coordinate of an object point z _OB = z _Obj can then with the equation

be determined, which also delivers a negative value for a negative absolute object phase. Here, d represents the distance of the straight line g _O from the axis of the illumination lens, or here the distance of the parallel axes of the two lenses from each other, f _B the focal length of the illumination lens, ϕ _Obj the absolute, laterally invariant object phase and p the lattice constant of the line grid .

The zero-plane for z _OB = 0, the focal plane of the illumination objective in the space of the object, is selected as the zero plane of the 3D recording arrangement. Accordingly, the focal point F _{OB of} the illumination lens in the space of the object represents the zero point of the 3D recording arrangement.

The most accurate determination of the absolute phase difference Δϕ _grating in the phase of the grating Δϕ _grating is of particular importance for the method.

Another possibility is to determine, for example, not the relative phase values in the reference point signal and in the object point signal, but the position of the locations of the same phase in the reference point signal and in the object point signal with respect to the phase of the grating ϕ _grating based on the locations of the same phase positions.

In principle, the spatial difference between the positions of the reference point signal and the object point signal can also be determined using a length measuring system which measures the grating movement with high resolution. From the known grid constant p of the transmitter array, the associated phase difference Δϕ _{grid can} be calculated in a known manner from the measured spatial difference. This is to be used if only a single linear guide with a zero point encoder is used, since the phased scanning of the grating is technically particularly difficult due to the movement of the grating with a then occurring movement component in the direction of the optical axis with a counter grating.

With the preferably high-level reference plate, the 3D recording system must be measured in at least one position perpendicular to the axis. The value of the reference _phase ϕ _{R is} calculated from the experimentally determined value z _ORexp for the position of the reference _plate using the given equation (2). The measured displacement in the z _OB direction is compared with that of the z _OR value determined by the 3D recording arrangement. In this way, the 3D recording arrangement can be checked. In the event of deviations from the highly precisely measured shift from that calculated by calculation, there is an incorrect value for the reference phase ϕ _R for a well-adjusted 3D recording arrangement. The reference phase is changed numerically until there is as good a match as possible with the experimentally determined displacement values.

In summary: For the determination of the _object phase ϕ _Obj in the object space, the following condition is advantageously realized in the overall imaging system, which is derived from the parallel movement of the line grid - or more generally from the parallel movement of the maxima and minima of the luminance on the transmitter array - to the straight line g _A results: When passing through the two coinciding areas of focus of the illumination and imaging lens through the depth of the space of the object, the observed phase at the intersection of a beam of the imaging lens with the common area of focus remains constant due to the additional lateral movement of the grating , so that an imaginary object point running in a beam of the imaging lens in the common focus area does not experience any phase change in the entire depth. This is referred to as "carrying the phase". The additional lateral movement of the grating can take place through a second linear guide or can result from the inclination of a common linear guide for the transmitter and receiver array. On the other hand, this condition can also be realized by an electronically controlled grating, for example a line grating, in addition to the continuous movement of the grating in the z _A direction, the phase position of the grating is also continuously changed by the position of the locations of the local extremes of the luminance is changed on the electronically controllable grid.

The following should be noted in this regard: Optically conjugated points have the same absolute phase in terms of amount in the array space and in the object space. This absolute phase can be understood as a lateral coordinate. The absolute phase is derived from the lighting situation and can be determined in the space of the arrays from the χ _AB position in the grating element G _AB , which moves on a straight line g _A in the recording process and during this movement exactly the array-side focal point F _AB cuts. The absolute phase in the space of the arrays also results from this context

with χ _AB1 as the lateral coordinate of the point of intersection of the straight line g _A through the lattice element G _AB and p as the lattice constant of the line lattice. In the optically conjugated point G _{OB there} is the same absolute phase as in point G _AB , but with the opposite sign.

It is fundamentally possible to design the illumination lens with a central perspective beam path in the space of the object, on the other hand to make the image objective but as a lens with a telecentric beam path in the space of the object. In this case, the receiver array need not be moved in the telecentric area of the imaging lens. This option may be advantageous for smaller objects with a diameter of less than 100 mm, where the comparatively low light intensity of the telecentric imaging lens can also be compensated for without major technical effort. However, the evaluation results in a significantly higher computational effort, since the depth sensitivity, described by the effective wavelength, is not laterally constant in the respective planes perpendicular to the axis, since the imaging rays do not form a bundle of rays which coincides with the bundle of straight lines around the line g _O , but a parallel bundle. This is why this case is not considered further here.

Furthermore, the receiver array does not have to be an image receiver; the prior art as a microlens array to a great depth to be able to achieve resolution on the basis of a high strip density. The Microlens array is located in the space of the arrays in a to the transmitter array op table conjugate level. As is known, the microlens array is another ob jective subordinate. This lens is on the assigned to the microlens array Side as well as possible telecentric. On the second page of the The image sensor is located on the lens.

An electronically controllable, preferably transmissive, array can also be arranged in the plane of the receiver array. The procedure is as follows:
A 3D point cloud is first obtained with, for example, medium transmission in the array elements. The modulation in the elements of the image receiver is then checked and the transmission is adjusted element by element such that there is good modulation in the image elements of an ordered image receiver. The electronically controllable, transmissive array can preferably be assigned to the microlens array. But it is also possible that it is assigned to the image sensor. It can also be assigned to the transmitter array. Ultimately, the assignment only has to be made to an optically conjugate level of the receiver array.

To determine the phase position of the periodic signals in the pixels of the In the simplest case, receiver arrays come with the known phase shift Algorithms with 3 to 5 intensity values for use. The intensity in the pixels of the receiver array according to the phase position of the Sender arrays scanned. Scanning in discrete is possible here 90 ° phase steps. With the read out intensity values, for example also the modulation over the phase in 90 ° steps with the known elements equations can be determined. The phase position can be evaluated in each case 180 ° steps of the phase of the transmitter array, here the line grid, take place.

Better for the accuracy of the phase determination and thus for the depth measurement accuracy, however, are algorithms that use a larger number of intensity values ten, for example 8, 16 or 32, the phase in the signal as well as the location of the Mo. Determine maximum dulations. The entire know-how of signal processing, as is already known in electrical engineering, can be used here. At games for the successful application of such signal evaluations are from the White light interference microscopy known. Generally these will Signal evaluation methods applied to short-coherent methods.

Since the period length, the lattice constant, of the transmitter array is preferably con is constant, is the phase change speed with a constant movement of the transmitter array, also constant. This also enables the application of Sub-Nyquist sampling, because here the be. Based on the knowledge of the signal frequency known sampling theorem can be violated without disadvantages. This reduces the number number of pictures to be taken quite considerably, so that basically a higher one Speed of movement of the grid can be realized and so one High-speed evaluation becomes feasible. Filter operations are in the  to combine the known type with the phase evaluation methods, or already before the phase evaluation. This fact will not be discussed further here low, although the performance of the optimal design of the algorithms of the entire evaluation process. The execution of the arithmetic operation NEN can be done using special high-speed processors.

It is advantageous if the transmitter array, here the line grating, additionally has a zero point mark and the lateral grating movement is detected in terms of phase with a counter grating. The absolute phase of the zero point is determined by positioning the reference plate in the object space in a known z _OB position z _OB = z _OR . The reference phases mod 2π are also determined and stored as relative reference phases ϕ _RR by evaluating the signal curve in the imaging beams in the area of the sharp image of the reference plate. In the case of an unknown object, the associated phase of the grating ϕ _grating is determined by the signal curve in the focus area of the object point at the point in the signal curve in the area of the maximum of the modulation, which corresponds to the associated reference starting phase in the area of the maximum of the modulation.

The object phase can then be determined for each level at a distance z _OB using the phase derived from the grid shift. The highly stable zero point mark can be adjusted as a starting point for the lateral movement of the grating in such a way that the start of the image recording begins shortly before the sharp image of the reference plate is reached due to the coordinated movement of the grating and the image receiver. Even after removing the reference plate, the position of the reference plate remains as "the reference surface of the 3D recording system" until the new measurement.

In the previous presentation it was assumed that the Lighting lens and the imaging lens are always fixed and in present rigid lenses, i.e. do not have their own focus, or that any existing, objective focusing is not used. The Focal planes of the lenses are therefore fixed in the room. Focusing in In order to change the position of the focus areas in the space of the object in the previous illustration, each time moving the transmitter and Receiver arrays in a component parallel to the optical axis of the lenses.

In the case of changing the position of the focus areas in the space of the object above the focusing of the lens, for example by internal focusing, such as If it corresponds to the state of modern photo lenses, the focus must be be electronically controllable, d. H. the lenses also need a motor Allow focus. Even when moving the entire lens to A motor shift must be ensured for the purpose of focusing. Out Reasons for the high dynamics required in the inventive solution moving components becomes the possibility of moving the whole Objectively not considered here.

Because modern lenses with internal focusing options by moving Low-mass, optical components are generally quite fast Allowing focus, on the other hand, is considered in this case, although currently not yet bright lenses with a telecentric beam path on the image side and a motorized focusing option have become known and also one  Would be an exception. Therefore, in the previous presentation, the Possibility of changing the position of the focus areas in the space of the object clearly favored by moving the sender and receiver arrays.

Basically, however, it is only a matter of achieving a defined relative movement between the focal planes of the lenses and the respectively associated arrays or the locations of the extremes of the luminance on the 3D recording arrangement and the 3D recording method according to the invention. In the case of a cos ² -like luminance distribution on the transmitter array, the phase position of this luminance distribution should experience the defined relative movement. To achieve the relative movement, the focal planes can also be moved in space, for example by internally focusing the lens.

The ultimate goal is the detection of a modulated cos ² -like signal with a modulation maximum in the pixels of the receiver array. As is known, the approximation to the cos ² characteristic is generally supported by the optical modulation transfer function of the lens in the respective imaging situation.

It must be realized that the movement of the point of the transmitter array, which coincides with the focal point of the illumination lens in the focal plane position of the transmitter array, takes place on a straight line g _A. Other points or elements of the transmitter array move on parallel straight lines to straight line g _A. This straight line g _A is defined such that it always intersects the focal point of the lighting objective in the area of the arrays and has the increase, defined as the quotient of the focal length of the lighting objective in the area of the arrays and the distance of the focal point from the imaging lens, this increase being the Line g _{A is} related to a line perpendicular to the axis of the illumination lens. As a result, the straight line g _A moves together with the focal plane of the lighting lens. Since the focal point always moves on the axis of the illumination lens during internal focusing, an additional movement of the transmitter array perpendicular to the optical axis is required to implement the movement of the elements of the transmitter array on a straight line g _A. With an electronically controllable grating, a change in the locations of the local extremes of the luminance is generated. This can also be understood as a movement perpendicular to the optical axis.

It is very important for the accuracy of the measuring method that the described movement or displacement of the focal plane of the lighting lens is very precise, so that the movement on the line g _{A occurs} only with the smallest possible deviations. This can be supported by means of a high-resolution, internal position measuring system for the optical components in the lighting lens that are moved for internal focusing. The guiding accuracy of the miniaturized linear guidance of the moving components in the lens must also be subject to high demands, since their transverse deviations are also harmful.

The internal focusing of the imaging lens with a miniaturized Linear guidance should be done so that the focus of the imaging lens with the focus area in the space of the  Object coincides as well as possible. It is also possible to use a high here resolving, internal measuring system for the purpose of internal focusing in the Ab educational lens moving optical components. There is also a rigid coupling of the moving components of the two lenses is basically possible.

To meet the requirements for the accuracy of the internal focusing of the fig To reduce the lens, the imaging lens can be slightly stronger be dimmed, since the depth of field then increases. in the In extreme cases, such a dimming of the imaging lens is also conceivable that then for an object with a limited depth, both on its internal Focusing as well as on a movement of the receiver array - i.e. on one Focusing in the imaging beam path at all - can be dispensed with. It is however, this causes an undesired broadening of the modulation curve in the signal given which is detected in the pixels of the receiver array, which reduces the depth resolution in 3D imaging is to be expected.

According to the invention, an arrangement is also proposed which works with an electronically controllable grating as a transmitter array, which shifts the locations of the local extremes of the luminance on the receiver array or, in the case of a cos ² -like luminance distribution, on the transmitter array Shift of the phase position, generated, and the associated lighting lens has an internal focusing at the same time. As a result, the locations of the extremes or the locations of the same phase of the luminance distribution on the transmitter array move even with the internal focusing on a straight line g _A moving in space.

In principle, it is possible for the two axes of the illumination lens and the imaging lens to be inclined to one another. However, in this case it is of great advantage if the focal point of the imaging lens lies in the space of the object in the focal plane of the illumination lens. The location of this focal point defines the location of the straight line g _O , which, as an image of the straight line g _A, by definition contains this focal point and must lie parallel to the axis of the lighting objective. This means that the depth sensitivity of the 3D recording arrangement in a plane perpendicular to the axis of the illumination lens in the space of the object is a constant, that is to say the depth sensitivity has no lateral dependence in the plane.

For a given arrangement of illumination and imaging lens, the direction of movement of the elements of the transmitter array must be parallel to the line g _A. As already shown, the straight line g _{A is} defined such that its image in the space of the object, the straight line g _O , intersects the focal point of the imaging lens in the space of the object while at the same time being parallel to the axis of the lighting object. The arrangement with the optical axes of the lenses inclined relative to one another brings advantages if regions of the object are at a particularly short distance from the illumination lens and these can no longer be recorded by the imaging lens with a parallel arrangement of the lenses. It is advantageous here if the imaging lens is rotated out of the parallel position or rotated out in a motor and computer-controlled manner so that it can capture the areas of the object at a particularly short distance. It should be noted that the focus areas only partially coincide and the joint movement of transmitter and receiver array in the sense of a rigid coupling is no longer possible without restrictions. Also when calculating the 3D point cloud, the numerical effort increases due to the consideration of the more complicated imaging relationships. Nevertheless, this variant can be an advantage for special objects, even if the technical effort is considerable. Furthermore, it is possible for the receiver array to be additionally rotatably arranged under computer control in order to meet the Scheimpflug condition, where the focus areas in the space of the object can be reached by the coincidence. It is also possible that the two lenses can have different focal lengths, the imaging lens being designed to have a significantly shorter focal length if only the focal point of the imaging lens is in the focal plane of the lighting lens in the space of the object.

It is also advantageous for measuring and checking the 3D recording arrangement that a transparent plate as a permanent one Reference plate perpendicular to the axis at close range in the object space for self measurement Solution is assigned, with a weak on at least one of the two surfaces light-scattering microstructure is applied. Then the measurement can be carried out as often as required checked and corrected if necessary, for example when Temperature changes. It is also possible to encapsulate the arrangement and ther to stabilize mix by a temperature control to a temperature-related To prevent the phase drifts. The use is also for calibration an arrangement with two transparent plates with an air gap arranged in parallel possible in the space of the object, the air gap being constant and known.

In principle, the relative phase values ϕ _RR measured with a reference plate can be stored in a reference phase matrix and used to calculate the object coordinates, even if the reference plate has already been removed.

It is also possible to get distortion of the lenses, and misalignment of the arrangement to determine and store as phase amounts via the spatial coordinates and at Need to use for correction.

The procedure described basically allows the optimal time Evaluation based on straight formrard algorithms. These algorithms can be implemented in special high-speed processors where through real-time evaluation of moving objects and people.

For an optimal signal-to-noise ratio in the signals, the brightness of the Light source can be controlled depending on the position of the transmitter array. For the adaptation to the laterally different reflectivity in the one individual object points, on the other hand, can be Array the light intensity can be adjusted pixel-by-pixel in an object-oriented manner. It can the electronically controllable transmitter array is moved laterally to phase change, but also laterally fixed and electronically by moving the grid lines produce a phase change, for example with a constant Phase velocity.

The light intensity of the lenses can also be changed. It can drop off det to the depth of field, for example, at fast  To be able to enlarge orientation measurements. This reduces the number of emergencies agile images, but also reduces the depth measurement accuracy. To the contrary To be able to reduce the depth of field becomes one or two apodizations arranged diaphragm, one in the exit pupil of the lighting lens and one in the entrance pupil of the imaging lens in each case in the space of the object. These diaphragms are designed so that the rays close to the axis are weakened or out be dazzled in order to achieve a stronger effect of the marginal rays. It are the apodization functions known from spectroscopy, for example Rectangle, triangle or Gaussian function can be used.

On the other hand, in certain cases, for example for the particularly fast 3D recording, also a movement system with a rotary drive for con continuous, rotary movement of a transmitter array, preferably in the manner of a radial grid with at least one reference mark, wherein a Detail of the radial grid completely fills the field of the lighting lens. The radial grating is assigned a counter-grating for phase sampling, wel ches contains the counter structure for scanning, as well as the lighting, an Optikmo dul and the evaluation electronics with computer interface. Also on the counter grid there is also a reference mark for winning at least one egg high-precision zero point signal.

Furthermore, this movement system has a linear guide which carries the rotary drive for preferably continuous rotary movement of the transmitter array and the receiver array. The direction of movement of the linear guide is parallel to the optical axis of the imaging lens in the z _A direction. The coupling of the radial grid with a rotary drive ensures that the transmitter array undergoes a continuous run, which can be controlled with high precision by the phase-based scanning. It turns out to be not quite optimal in this solution that a certain variation of the lattice constant along the coordinate χ _A occurs because of the structure of the radial lattice. For radial gratings with a comparatively large diameter to the field of the imaging lens, this fact is acceptable in connection with numerical correction methods. However, this is not shown in detail here.

It is basically possible for the lighting and imaging lens Use zoom lenses, each with the same focal length setting work.

In order to let beams of rays open wide from the transmitter array, this can be done Transmitter array can be coupled to a special microlens array, for example This microlens array can be arranged upstream of the transmitter array in the direction of light. It is also possible to use the transmitter array itself as a high-aperture cylindrical lens array to execute.

The process of taking multiple pictures with moving arrays is also referred to as deep scanning.

The parallelism of the optical axes of the two lenses can be checked with Using the comparison of the phase positions between the phase of the transmitter array taken signal and the phase in the signal in a pixel on the opti rule of the imaging lens can be detected, if a  The reference plate is shifted in depth and the displacement path is ge will measure.

It is possible with an electronically controlled array of transmitters through Telezen drifting error of the illumination lens during phase scanning by stretching or compressing the electronically controlled transmitter array balance.

For measuring and checking the depth sensitivity of the arrangement a pin plate can be used. The pens are very precise and firmly attached to a bracket gerplatte attached in a known grid and point based on the carrier plate two different but known pin lengths. The spacing of the pens are chosen so large that in a certain distance range of the plate from the arrangement on the end faces of the pins no shadowing occurs. So be two levels are shown with a very well-known distance. This distance can be measured with a high-precision coordinate measuring machine. In ver Different distances of the plate from the arrangement can make the depth sensitive speed of this arrangement, but also the distortion of the lenses become.

In addition, a plate with only a single pin length can be produced in this way become. Also the use of a multi-level plate with different pen lengths possible and can be used to measure the arrangement with advantage become.

With a double-spherical pin plate, in which the flat or spherical end faces Chen thin pins represent two concentric spheres, the Formmeßge accuracy of the 3D recording arrangement can be tested at different intervals.

The use of a single-sphere pin plate can also be used to advantage the. This is when using the arrangement in a coordinate Measuring device useful to the lateral coordinates of the 3D recording arrangement to be able to determine. Also arranged in a coordinate measuring device Nete plane plate can be very advantageous for measuring the 3D recording arrangement be clingy.

In order to be able to measure very large objects partially or all around, a A larger number of 3D recording arrangements, for example, a screen similar to a sky in a planetarium. The 3D recording Arrangements then represent 3D measuring modules, which in large numbers as 3D video sensors can be used. It is also possible that this Screen approximately the rough shape of the object to be measured, for example se in the form of a complete automobile body. The captured object 3D imaging arrangements partially overlap. The calibration can be made using several reference bodies, including flat plates, possibly with marks take place simultaneously from two 3D recording arrangements in the covering area be grasped richly.

It is also possible that several 3D recording arrangements next to each other are positioned that an object or scene is complete or in a large size ß area can be detected, the detected areas in the space of the object  partially of the 3D recording arrangements positioned directly next to each other cover up. However, there is no overlap with the next but one 3D recording arrangement taking place. The positioned next to each other 3D recording arrangements illuminate and form the object with one each the colored light source. Then, for example, illuminate the 3D recording arrangements with an even number of the object with red light and the 3D recording arrangements with an odd posi the object with the green light. In the imaging beam path the 3D recording arrangements are correspondingly narrow-band filters that just let the light from your own light source pass through. Several are also conceivable Light colors, making more than two 3D imaging arrangements the same part of the object.

Furthermore, several illumination lenses can be assigned to one imaging lens be ordered, for example two. Each lighting lens can have one the colored light source can be assigned, for example a red and the first the second a green one. The imaging lens is mechanical or electrical nisch switchable filter assigned to only light from a single light source to get to the image receiver.

On the other hand, a color camera can also be arranged to simultaneously to process the structured light from the two lighting lenses separately Different lighting directions in the space of the object can be realized be, for example, direct reflections from a special object detail not to appear at least during a recording. The shadows too the other picture is illuminated and so can these object parts be made visible.

It is also possible to separate the images from different lighting lenses using rotating polarizers and the use of polarized light.

For color recording technology with color-sensitive image sensors, however, fol Possible: Illuminated with white light. Several 3D imaging arrangements are positioned so that an object or scene ne can be recorded completely or in a large area, the recorded Area or angular areas in the space of the object immediately next to each other partially cover the positioned 3D recording arrangements. However, it fin there is no overlap of those with the next but one 3D recording arrangement. Illuminate the 3D recording arrangements positioned next to each other and depict the object at different times. For this purpose all 3D recording arrangements synchronized from a control center, so that the Taking pictures is also done synchronously. Then, for example, capture the 3D recording arrangements with an even number of the object in the direction of the moving arrays and the 3D recording arrangements with an un wheel numbered item number the object in the return of the moving arrays and the Objects are only illuminated when pictures are taken.

When using a color-capable camera as an image sensor are preferred the color-sensitive elements that belong together, also as RGB Known sensors, linear and in the direction of the stripes of the line grid, that is arranged on lines of the same phase. This ensures that there is no color  conditional phase positions in the signals there. In contrast, the phase between the RGB sensors may be quite different depending on the object.

Furthermore, it is also possible for a robot arm to take a single 3D image Arrangement over the object leads and the data are read in successively. By covering the detected areas of the object surface, a "Next" takes place shimmy "over the surface of the object, which is basically without the presence of reference marks on the object surface is possible.

It is feasible for the imaging lens to be fixed and at least one Lighting lens together with the transmitter array and the complete Be lighting arrangement rotatable or pivotable about the imaging lens is arranged, advantageously the axis of rotation of the illumination lens can represent the optical axis of the imaging lens. This is lighting of the object from different directions to eliminate shadows. The rotation is computer controlled.

It is advantageous if at the beginning of a 3D recording process in an image a first data record is obtained and saved and the computed Object points that deviate from the maximum modulation are eliminated have mum. The transmitter array can then be moved by a fraction of the git ter period are postponed and at least one second data record added and is stored, the calculated values also being eliminated, the are not in the immediate position of the maximum modulation. Due to the phases caused by the lattice shift in a fraction of the lattice period Change are the calculated object points of the first measurement with those of the second measurement is not identical, so that a complete detection of the majority the object points is reached.

The displacement of the transmitter array described above preferably corresponds to this Amount of the quarter grid period.

It is also possible that there is a swivel in front of the 3D recording arrangement Plane mirror is located, which can be pivoted into the beam path and so one fixed reference plate is missing and after measuring the plan mirror is pivoted back.

Furthermore, a network is made for a high-performance 3D recording arrangement Sender and receiver arrays with the associated compo already described built up. It is advantageous if, for example, nine receiver arrays in a 3 × 3 matrix arranged and in the field of 3 × 3 matrix four transmitter arrays in egg ner 2 × 2 matrix can be arranged, with the transmitter arrays on the Diagona len of the 3 × 3 matrix. This makes a receiver array as a central receiver arranged. Furthermore, the transmitter arrays are designed as line gratings and the Li lines are preferably parallel to the outer lines of the 3 × 3 matrix arrangement. The transmitter and receiver arrays are the lighting and imaging objects ve assigned in a parallel axis position and the main planes of the lenses are next approximately on a common level. For moving the receiver and Transmitter arrays in the direction of the optical axis is preferably a central first Linear guide assigned. For moving the transmitter arrays in the lateral direction A central, second linear guide can also be arranged on the optical axis  be that makes a movement parallel to the outer edges of the 3 × 3 matrix. In principle, each array can also be assigned its own linear guide system become.

Arrangements based on triangles are also possible Area center of the triangles the transmitters and on the corner points the receiver Arrays. Here too there is a receiver-centered arrangement with ins a total of seven receiver arrays and six transmitter arrays.

In principle, transmitter-centered 3D recording arrangements are also possible. In In all cases, the movements and images are recorded by a control system coordinated and processed by assigned processors. The won Image data massif is finally converted into a 3D point cloud, the Redundancy in the system can be used for improvement through averaging can.

The number of transmitter and receiver arrays used is only due to the techni possibilities and the associated costs. So ver equally large networks with a large number of small 3D image arrangements be built, each with a high depth resolution in a comparative manner have a small object field and thus the object vertically, in the manner of a needle ski sens, detect, with practically no gaps in the detection of the object surface there. Such a network can be in a 3D coordinate measuring machine for the inspection of filigree, relief-like structures, which can also have openings can be used with advantage.

The networks of transmitter and receiver arrays can be used for neural networks Evaluation of the optical primary data can be assigned.  

Embodiments

In FIG. 1, the arrangement and the method are represented. A distinction is made between the space of the arrays and the space of the object. The following notation is used, which differs from the usual one, but has proven to be advantageous for the description in this optical circuit: The sizes and points of the space of the arrays are first indicated with A and the sizes and points of the space of the object with O. In the second position in the index, the associated lens is identified, namely with B if it belongs to the lighting lens and with A if it belongs to the imaging lens.

In the space of the arrays there is a line grating with the grating constant p and an upstream lighting. The line grating is assigned to the lighting lens with a strictly telecentric beam path in the space of the arrays, perpendicular to the axis and extra focal. The lighting lens reproduces the line grid in the space of the object, creating a structured lighting of the object. To simplify the two main planes of the lighting lens H _AB and H _{OB are merged} in the drawing. In real lenses of this class, the main planes are far apart.

In the area of the arrays, a receiver matrix is assigned to the imaging lens with a likewise strictly telecentric beam path in the area of the arrays, perpendicular to the axis and extrafocal. The imaging lens depicts the object in the space of the arrays. A single imaging beam is shown. To simplify the two main planes of the imaging lens H _AA and H _{OA are} also merged in the drawing.

The illuminating and imaging objectives are arranged with their optical axes parallel to one another with the center distance d. The illumination and the imaging lens have the focal points F _AB and F _{AA on} the array side and the focal points F _OB and F _OA in the space of the object.

The first linear guide of the movement system, not shown here, is with the Receiver matrix rigidly connected and also does not carry a second here put smaller linear guide, which in turn carries the line grid. This line grid is significantly longer than the image field captured by the lighting lens corresponds to always the entire image field in the entire movement process cover up.

The first linear guide is connected to a highly precise length measuring system, which has a highly stable zero point. The movement axis of the first Li near guidance is parallel to the objective axes and the measuring axis of the Measuring system is parallel to the two lens axes. The movement direction The second linear guide is perpendicular to the objective axes.

The line grid on the second linear guide is one with the first linear guide firmly connected counter grid with a complete lighting and receiver Optics assigned in the manner of an incremental length measuring system. The existing The evaluation electronics have an electronic interface to the computer, around the calculated shift of the line grid as phase information in real time  to have in the computer for extinction. At the same time on 16199 00070 552 001000280000000200012000285911608800040 0002019846145 00004 16080 A first reference structure is applied outside the image field used a second reference structure, which is also applied to the counter grid, is optically scanned. The lighting and receiver optics are also included orderly and evaluation electronics available. The second evaluation electronics also has an electronic interface to the computer for the passage The zero point of the line grid is available in real time in the computer to have.

Both linear guides of the motion system start from the zero position. The direction of movement of the first linear guide is parallel to the optical axis of the imaging lens. The movement takes place to the focal points. The smaller second linear guide that supports the line grid is a position Control system assigned to a movement of the line grid with a possible to be able to realize constant phase velocity. The direction of movement the second linear guide is perpendicular to the optical axis of the lighting object jective and takes place after the start in the direction of the lighting lens.

The setpoints for the position of the first linear guide are calculated from the current, absolute actual phase of the line _grid ϕ _grid , which is derived from a zero point. This is done in such a way that the locations of the same phase or brightness on the line grid move parallel to a straight line g _A. This straight line g _A is defined so that it intersects the focal point F _{AB of} the illumination lens and also the main point H _{AA of} the imaging lens. Through this movement regime, who passes through the plane of focus of the objects from the focus surface one after the other by observing in each of these planes, in the presence of an object surface, a stripe pattern sharply imaged by the illumination lens, which is imaged by the imaging lens onto the receiver matrix becomes. By moving the locations of the same phase or brightness on the line grid parallel to a straight line g _{A in this way} , the so-called "carrying along of the phase in the focus area is achieved. This leads to the fact that every sufficiently small object detail, in the space of the object, if it is a modulated periodic signal is generated in the associated pixel ij on the receiver matrix, which contains the information about the absolute phase of the object point ϕ _{Obj_ij} . This absolute phase corresponds, for example, for the object point A _O2 in Fig. 1 n the imaginary number of the strips confining Lich the strip fractions .DELTA.n, between the two lens axes, that is n + .DELTA.n. These strips are in the case of a flat, axially perpendicular plate arranged to observe directly. generally Nevertheless, the thus observing absolute object phase due to the not exactly known assignment of the two optical axes to the pixels of the Receiver matrix cannot be determined.

On the basis of Fig. 1 it can be shown that in the position A _O1 B _O1, the number of Stripes fen n + 1 + .DELTA.n, which can be observed between the two lens axes, n exactly to the number + 1 + .DELTA.n the grating periods of the line grid in the Corresponds to the space of the arrays, which results when counting from the axis of the lighting lens in the χ _AB direction.

Furthermore, the point G _AB precisely defines the location on the grating element of the line grating, which intersects the focal point F _{AB of} the illumination lens when moving on the straight line g _A in the partial region of the grating increment. As a result, the absolute, laterally invariant _object phase ϕ _{Obj of} the plane perpendicular to the axis corresponds to the distance.

With knowledge of the associated point G _AB on the grid and the value χ _AB1 and the grid constant p, the absolute, laterally invariant _object phase ϕ _{Obj of} the plane perpendicular to the axis in the space of the object can basically be determined

calculate. If the real position of the lens axis of the imaging lens matches the straight line g _O , which is by definition parallel to the lens axis of the lighting lens, the relationships shown in equation (7) apply.

Before the measurement, ie the extraction of the point cloud of an unknown object, can take place, the system must be carried out in a position on the right-hand axis using a reference plate. This is advantageously done in the near distance of the 3D recording arrangement, but can also be done at any other distance in the detection area of the 3D recording arrangement. The signal curves resulting in the pixels of the receiver matrix are stored. With the aid of a phase-evaluating algorithm, the relative reference _{phase i RR_ij} in the pixel ij of the receiver array in the region of the maximum modulation is calculated from these signal curves and receives a signal from the reference plate. The relative reference phases mod 2π determined in this way are assigned to the absolute phase ϕ _{grid of} the line _grid , see FIG. Fig. 2, and taking into account the respective sign subtracted from this and stored in a field as phase _values ϕ _{gridR_ij} in the _rule longer term. The relative object phases of the object points ϕ _{RObj_ij} mod 2π are each subtracted from the absolute phase ϕ _{grid of} the line _grid taking into account the respective sign, whereby the phase _values ϕ _{gridObj_ij arise} as a field. The phase differences Δϕ _{lattice_ij are} formed point by point from these fields.

The absolute, laterally invariant phase ϕ _{R of} the reference surface was previously calculated from the distance z _{OR of} the reference plate from the focal plane, which is as well known as possible, using the equation (2) shown here again

determined, this determination can be made several times iteratively to approximate the true value. It represents d the distance between the axes of the two lenses, f _B the focal length of the illumination lens and p the lattice constant of the line grating.

With the relationship according to equation (3), where Δϕ _{grid_ij} for the object point ij results from equation (2),

(-) ϕ _{Obj_ij} = (-) ϕ _R + Δϕ _{grid_ij} (8)

the absolute object phase (-) ϕ _{Obj_ij is obtained} from equation (8) for the object point ij. From the relationship

the z _OB coordinate of the object point z _{Obj_ij} can be calculated in the object space, the object point being optically conjugated with the pixel ij of the receiver matrix.

Through axially parallel displacements Δz _OB of flat plates, the displacements being measured with a precision length measuring system, errors can be estimated by comparing the calculated displacements with the measured ones.

The readjustment of the 3D recording arrangement is carried out iteratively. The remainder are entered into a numerical model, which is not shown here is, and can be used for correction.

The Cartesian coordinates for each object point are determined from the calculation of the current image _scale for each object point with the coordinate z _{Obj_ij} . A new coordinate system with the focal point of the imaging lens can be used as the zero point for the lateral coordinates in the x and y directions. The 3D coordinates of the object are thus available in digital form. Depending on the task, this point cloud is used for measurement applications or tasks with 3D rendering.

When an object point, which is in the A _O1 B _O1 position in FIG. 1, is shifted into the A _O2 B _O2 position along the drawn imaging beam, the signal detected in the image point of this object point undergoes a change in the phase position of 2π. The change in the z _OB coordinate corresponds to the Δz _2π value, that is to say the change in depth, which corresponds to a phase change of 2π. This Δz _2π value is also referred to as the effective wavelength.

Fig. 2, for example, shows the waveforms in a pixel matrix of the receiver with respect to the waveform which can be detected on the grid with the aid of a counter-grid upon movement of the grating. The signal curve in the pixel of an object point and the signal curve in the pixel of a reference point are shown. Here, the reference plate is closer to the focal point F _OB than the object surface. The relative phase ϕ _{RR is} calculated at the sampling point in the area of the modulation _{maximum of} the signal in the pixel of a reference point and the relative phase ϕ _RObj at the sampling point in the area of the modulation _{maximum of} the signal in the pixel of an object point. The absolute phase difference Δϕ _{grid is} calculated using equation (3) and the absolute _object phase ϕ _Obj using equation (4), from which the z _OB coordinate of each object point, namely z _Obj , is determined using equation (5) .

Fig. 3 shows a 3D recording arrangement advantageously with only one linear guide, for example for a 3D recording arrangement for multimedia applications. The slide of the linear guide carries a line grid for structured lighting of the object. The lighting is provided through an opening in the base of the guide. Furthermore, the carriage carries the moving parts for a measuring system for the displacement path s, the measuring system also being assigned a zero point transmitter. In addition, evaluation electronics with an interface to the evaluation computer, not shown here, are available.

The carriage is driven by a linear motor and the guide has one miniaturized precision bearings. Instead of the precision bearing there is also a High-precision parallel spring system with highly elastic springs made of crystalline Si silicon can be used. A color-capable camera can also be used as the receiver matrix are set, the color sensitive pixels each on a line across the Li lines of the line grid are arranged.

After starting the slide in the (-) s direction, the zero point is passed through and the grating through the lighting lens in succession to different depths of the object space in sharp focus. When moving, the object surface is created the sharp image of the same, which is transferred from the imaging lens to the Receiver matrix is shown sharply, since this is in the same plane as the line grid is located. Through the movement of the slide of the guide there is an egg ne lateral movement of the receiver matrix, the image of the object being fixed. In order to compensate for this lateral movement, a fixed assignment of the punk te of the object, i.e. the imaging rays, to the pixels of the receiver matrix to maintain, the image information obtained is pixel by pixel in the image data Massively shifted. This ensures that a real imaging beam each egg a laterally fixed pixel regardless of the movement of the receiver Matrix remains assigned.

From the known grid constant p of the line grid, the actual phase of the line _grid ϕ _grid is obtained from the measured displacement path s by means of the component s _x derived from the inclination of the slide

certainly. From the point-by-point evaluation of a field of absolute phase differences between the phase positions of the signals from the measurement of a reference plate and the signals from the measurement on the object, the coordinates of the object z _{Obj are determined} point by point in the manner already described. In addition, who also determines the lateral coordinates of the object points.

Formula symbols used

ϕ _grid

- absolute actual phase of the line grid
ϕ _{grid_R}

- absolute actual phase of the line grid for a reference point
ϕ _{grid_obj}

- absolute actual phase of the line grid for an object point
Δϕ _grid

- Absolute phase difference between the absolute actual phase of the line grid for a reference point and the absolute actual phase of the line grid for an object point
ϕ _Aabs

- absolute phase in the space of the arrays at the intersection of the straight line g _A

through the line grid
ϕ _Obj

- absolute, laterally invariant object phase
ϕ _fB

- Absolute, laterally invariant object phase in the focal plane of the lighting lens
ϕ _R

- absolute, laterally invariant phase of the reference surface
ϕ _{RR_ij}

- Relative reference phase of a point of the reference surface in the area of the modulation maximum, optically conjugated with the pixel ij of the receiver array
ϕ _{RObj_ij}

- Relative object phase of an object point in the area of the modulation maximum, optically conjugated with the pixel ij of the receiver array
ϕ _{Obj_ij}

- Absolute phase of an object point that is optically conjugated with the pixel ij
d - distance of the focal point of the imaging lens in the space of the object from the axis of the lighting lens, also distance of the straight lines g _O

from the axis of the lighting lens, also the distance between the axes of the lighting and imaging lens when they are parallel to each other
f _A

- Focal length of the imaging lens
f _B

- Focal length of the lighting lens
p - grid constant of the line grid
χ - Cartesian coordinate
χ _A

- Coordinate of the space of the arrays
χ _AB1

- lateral coordinate of the intersection point of the straight line g _A

through the grid element G _AB

in the space of the arrays
χ _OB

- Coordinate of the space of the arrays assigned to the lighting lens
z - Cartesian coordinate
z _A

- Coordinate of the space of the arrays
z _OB

- Coordinate of the space of the object, assigned to the lighting lens, also the distance of an axis-perpendicular plane in the object space in the space of the object from the focal plane of the lighting lens
z _Obj

- Coordinate of an object point in the object space
z _{Obj_ij}

- Coordinate of an object point in the object space, which is optically conjugated with the pixel ij of the receiver matrix
z _OR

- Distance of the reference surface in the space of the object from the focal plane of the lighting lens
z _ORexp

- Experimentally determined distance of the reference surface in the space of the object from the focal plane of the lighting lens
Δz _2π

- Amount of the displacement of an object point on a Imaging beam in which the absolute phase is exactly around 2π changed in the same imaging beam

Claims (16)

1. 3D recording arrangement for 3D recording of objects and scenes with at least one light source, at least one transmitter array, which together form a structured, luminous surface, at least one lighting objective with a positive focal length for imaging the transmitter Arrays, these assemblies forming the lighting device, a further object or a scene, at least one taking lens for imaging the object or the scene, the sharpness area of the lighting lens coinciding at least partially with the sharpness surface of the imaging lens in the space of the object, and at least one receiver Array and also at least one movement system, which is assigned to at least one component of the modules of the lighting device, characterized in that the components of the movement system are arranged in such a way that in the space of the arrays with the focal point of the lighting object as a reference point for there s transmitter array is an overall direction of movement parallel to a straight line g _A in the space of the arrays, so that the elements of the transmitter array move on parallel straight lines to the straight line g _A and this straight line g _{A is} the focal point of the illumination lens in the space of Arrays intersects and the increase with the amount of the quotient of the focal length of the lighting lens and the distance of the focal point of the imaging lens in the space of the object from the axis of the lighting lens, this increase in the straight line g _{A is} based on a perpendicular straight line to the axis of the lighting lens . 2. 3D recording delegation according to 1, characterized in that both Illumination lens in the area of the receiver array as well as the recording object tiv in the room of the transmitter array each a telecentric beam path exhibit. 3. 3D recording arrangement according to 1 and 2, characterized in that the Lenses with their axes parallel to each other at a center distance well below half the focal length of the illumination lens, typically three to six times the focal length, are arranged and the imaging object tiv is arranged so that its focal point in the space of the object in the focal plane of the lighting lens. 4. 3D recording arrangement according to 1 to 3, characterized in that the movement system is constructed from two individual linear guides and the first linear guide is assigned to the receiver array and the direction of movement of this linear guide is adjusted parallel to the optical axis of the imaging lens and the second linear guide is assigned to the first linear guide and the transmitter array and the direction of movement of the second linear guide is adjusted perpendicular to the optical axis of the illumination lens and both linear guides as a result of linear individual movements move the transmitter array parallel to the straight line g _A. 5. 3D recording arrangement according to 1 to 3, characterized in that the Movement system is built up from three individual linear guides and the first one Linear guidance of moving components of the lighting lens to the interior kissing and the second linear guide moving components of the Abbil lens for internal focusing and the third for the transmitter array supply is assigned perpendicular to the optical axis of the lighting objective. 6. 3D recording arrangement according to 1 to 3, characterized in that the movement system is constructed from a linear guide and this is assigned to the receiver array and the direction of movement of this linear guide is adjusted parallel to the optical axis of the imaging lens and a rotary drive Linear guide is assigned and the transmitter array is in turn assigned to the rotary drive and as a result of the individual movements a part of the transmitter array is moved approximately parallel to the line g _A in a time interval. 7. 3D recording arrangement according to 1 to 6, characterized in that the Transmitter array is designed as a line grid and this a counter grid for detection Solution of the phase position is assigned when moving the line grid. 8. 3D recording arrangement according to 1, 2 and 5, characterized in that the Receiver array for deep scanning to meet the Scheimpflug condition is rotatably arranged. 9. 3D recording arrangement according to 1, 2, 5 and 8, characterized in that the imaging lens around its focal point in the space of the object is ordered. 10. 3D recording arrangement according to 9 and 2, characterized in that a Transparent plate of the 3D recording arrangement as a permanently remaining reference renzplatte perpendicular to the near distance in the object space for self-measurement is assigned, one on at least one of the two surfaces of the plate weakly light-scattering microstructure is applied. 11. 3D recording method for 3D recording of objects and scenes with a structured, luminous surface according to the arrangement according to items 1 to 10, characterized in that the locations of the local extremes of the luminance of the structured, luminous surface in the space of the arrays experience a straight-line movement parallel to a straight line g _A , the straight line g _A intersecting the focal point of the lighting lens in the space of the arrays and the increase with the amount from the quotient focal length of the lighting lens and distance of the focal point of the imaging lens in the space of the object from the axis of the illumination lens, this increase in the straight line g _{A being} related to an axis perpendicular to the axis of the illumination lens and this structured, luminous surface being mapped into the space of the object. 12. The method according to 11, characterized in that an absolute phase difference Δϕ _grating in a pixel which belongs to an object point, from two signal profiles, namely that in said object point itself and in the associated reference point in the same imaging beam with evaluation of the modulation is determined in the signal, the calculated absolute phase difference being added with the correct sign to the absolute reference phase ϕ _R , which is determined from the at least one experimentally determined position of the reference plate z _OR in the space of the object and related to the illumination lens and the geometry data of the optical arrangement such as common lens focal length and distance of the lens axes are calculated, resulting in an absolute object phase for each detected object point as a result of this arithmetic operation. 13. The method according to 11 and 12, characterized in that the absolute phases difference is derived from the phase of the transmitter array so that the - from the when moving the transmitter array and receiver array in each pixel of the Receiver arrays resulting signal curves of a reference point and one Object point - calculate in each case in the 2π environment of the maximum modulation ten mod 2π phase positions with correct sign to the continuously measured absolute Pha se of the moving transmitter array. 14. The method according to 11 and 12, characterized in that the absolute phases difference from the phase of the transmitter array is determined so that - from the when moving the transmitter array and receiver array in one pixel of the Signal arrays of a reference point and a resulting receiver array Object point - calculated exactly at the maximum modulation 2π phase positions with correct sign to the continuously measured absolute phase of the can be assigned to movable transmitter arrays. 15. The method according to 11 to 14, characterized in that the target position of the he Most linear guides are always based on the location of the local extremes of luminance the structured, luminous surface in the space of the arrays. 16. The method according to 11, characterized in that when moving together sender and receiver arrays, which are rigidly connected to each other, the image information obtained by means of the receiver array after the recording Each individual image is shifted by at least one pixel in the image data set become.