DE19704496C2 - Method and device for determining the phase and / or amplitude information of an electromagnetic wave - Google Patents

Description

The invention relates to a method and an apparatus for loading tuning the phase and amplitude information of an electroma gnetic wave.

The term phase here generally stands for phase runtime and for each Term used also according to signal form.

The following is a light wave instead of an electroma gnetic wave spoken. However, this means no restrictions only on the spectral range of the visible electromagnetic Waves, but only serves to simplify.

For measuring frequency components by amplitude and phase in broadband and high-frequency signals are used in the electro African measuring technology and communications technology often phase detection ren used the unknown signal with a sine wave multiply or mix and the DC component that is present a signal component of the same frequency through integration or Low pass filtering arises, determine.

This process creates the correlation function of the unknown Si gnals with the mixed signal for a certain, adjustable relative Phase position. This can be done by changing the mixing frequency (wobble) unknown signal can be broken down into its spectral components. By at least 3 phase positions of the mixed frequency can Change amplitude and phase of the unknown frequency component same frequency can be determined.

Examination of appropriate optical signals that keep you awake gaining importance in measurement technology and communications engineering have happened today i. a. via broadband photodetectors as elec trooptic converters with subsequent electronic measured values mood - as previously described for electrical signals.  

Because of the high cost, these processes and the ent speaking measuring instruments usually only one or two channels. In the case of optical signals, however, a large number are often parallel All channels - especially entire picture sequences - with high frequency share to measure.

In addition to the spectral modulation properties of two-dimensional light waves are increasingly interested in the fast course of the Envelopes in space and time. You also want to be quick and exactly 3D objects z. B. measured via optical radar, what very fast detection due to the speed of light of the echo signals gates in the subnanosecond range. At the same time, they should be as Detector array are present when looking at a time consuming scan who wants to do without active or passive glowing 3D objects.

In the published patent application DE 44 39 298 A1, of which the present Er such a 3D camera is proposed.

Fig. 10 is intended to serve to illustrate this 3D camera or on the Echolaufzeit- phase delay method is based. The RF-modulated light wave 101 emitted by a modulated light transmitter 107 and 103 and reflected by the 3D object 100 contains the entire depth information in the delay of the phase front. If the incident light wave in the receiving aperture 102 is again modulated with a two-dimensional optical mixer 104 of the same frequency, which corresponds to a homodyne mixing or demodulation process, a stationary high-frequency interferogram arises.

This HF interferogram can be recorded with a conventional CCD camera 105 and further processed with an image processing 106 . The integration of the direct component of the mixed product in the CCD photo charge corresponds to the formation of the correlation function of the two mixed signals. The distance-related phase delays due to the echo propagation times and the amplitudes can be pixel by pixel from three or more interferograms by different phases of the demodulating mixed frequency, e.g. B. 0 °, 120 ° and 240 ° or 0 °, 90 °, 180 ° and 270 ° and thus the 3D depth image can be re-constructed.

The two-dimensional optical mixer 103 or 104 , which is also referred to as a spatial light modulator (spatial light modulator SLM), consists, for example, of a Pockels cell which has a number of serious disadvantages described in the literature.

LCD windows offer further implementation options lig, but with respect to the desired bandwidth by about a factor of 1000 lie low.

The use of a so-called Mi is also expensive and complex Cro-channel plate as used in image intensifiers. By Mon dulation of the acceleration chip applied to the microchannels voltage that affects the secondary electron emission in the microchannels influences, the gain can be modulated.

Furthermore, a proposal of a 2D correla is made in the prior art made on the basis of a CCD photodetector array: "The Lock- In CCD-Two Dimensional Synchronous Detection of Light "by Spi rig, Seitz et. al., published in the IEEE Journal of Quantum Electronics, Vol. 31, No. 9, Sept. 1995, pages 1705-1708. There is a photopixel queried via 4 transfer gates to the phase of sine-modulated light to determine. With each of the 4 transfer gates, there is one for each sine period Equidistant sample taken, which makes the phase easy to calculate leaves. This process is too for the problems presented slowly, since the harmonic light signal initially during a Bandwidth significantly limiting sampling time is integrated. Only then is the stored charge accepted as Ab sample the desired mixing process.

The invention is therefore based on the technical problem, a Ver drive and a device for determining the phase and / or Amplitude information and thus the envelope of one Specify light wave that is a simpler, broadband and less expensive correlator concept and a predefinable one Illumination enables fast 3D object measurement.  

The technical problem outlined above is now solved by the procedure ren according to claim 1 and by the photonic mixing element Claim 14, by the mixing element arrangement according to claim 20 and solved by the device according to claim 23.

The principle according to the invention is based on a drift generated by the modulation photogate voltage and separation of the minority charge carriers photogenerated by the light wave in the material below at least two adjacent light-sensitive modulation photogates. These charge carriers drift under the influence of the modulation photo voltages applied to the modulation photogates U _am (t) and U _bm (t), depending on the polarity or phase, to those with preferably twice the DC voltage U _a and U _b before tensioned accumulation gates. The modulation photogates voltages U _am (t) and U _bm (t) are preferably complementary and are preferably composed of a bias voltage U ₀ and the modulation voltage + U _m (t) or -U _m (t), which is superimposed in the clock cycle men. The two modulation photogates preferably form a square surface. A pixel with only two modulation photogates can also be referred to as a double pixel.

This principle according to the invention sets the photoelectric quan effect caused by electromagnetic waves. Nevertheless - without this being understood as a limitation - always talked about light waves.

The actual mixing or multiplication process consists in the modulation voltage-dependent or phase-dependent drift of the photo-generated charge carriers to the right or to the left of the modulation photogates (“charge swing”). The charge difference between the charge carriers thus separated, collected under the accumulation gates and forwarded to the readout electronics, taking into account integration in a predetermined time, is a measure of the correlation function of the envelope of the incident modulated light signal and the modulation voltage U _m (t) .

At the same time, the total charge remains with the accumulations gates drifted and forwarded load carriers from the position the charge swing is unaffected and stands as a corresponding pi xel intensity or as a pixel gray value.

To the relative phase or time delay of the incident light to determine the wave, it is necessary - as described above - three Measurements for the three sizes of DC voltage and AC voltage proportion and relative phase. Hence an out design of the pixel of the photonic mixing element with three light sensitive modulation photogates possible with the modulation photogate voltages are applied to the three different Pha shifts to the light wave emitted by the transmitter sen.

To determine the phase of the received signal at each pixel of the photonic mixing element from the resulting correlation However, amplitudes are expediently four different ones Measurements for four different phases of the mixer signal drawn. This gives you an overdetermination through which the Noise can be significantly reduced.

Due to the push-pull arrangement of the modulation photogates voltages on two modulation photogates per pixel, two of these measurements are carried out simultaneously. It is therefore sufficient, for example, for an RF modulation, two measurements shifted by 90 ° at 0 ° / 180 ° as well as with a 90 ° / 270 ° phase difference of the modulation photo voltages U _am (t) or U _bm (t) to carry out the phase of the incident light in order to obtain the necessary four different measured values.

An arrangement is therefore particularly preferred in which the respective one Pixel-forming photonic mixing element consisting of four symmetrically arranged ordered modulation photogates, two each opposite modulation photogates with push-pull or 180 ° phase-shifted modulation photogate voltages are, the two related to the double pixel too before described measurements shifted by 90 °  0 ° / 180 ° as well as 90 ° / 270 ° phase difference of the modulation photo In this case, gate voltages are carried out simultaneously. A such a pixel can also be called a quadruple pixel.

For a calibration of the phase shift of the modulation photogate voltages U _am (t) and U _bm (t), it is also possible in a preferred manner to reference part of the light wave emitted by the transmitter directly to at least one of several pixels of an arrangement of a plurality of photonic mixing elements. The phase and amplitude information obtained from this directly irradiated pixel can then be used for the calibration or for adjusting the phase shift to a predetermined value.

Conversely, with externally excited unknown modulation, the from incident light emitted by an active object at least one photonic mixing element with the light wave the known high resolution of a lock-in amplifier will be. For this purpose, the photonic mixing element forms together with a tunable Mo replacing the transmitter dulationsgenerator a phase locked loop. Furthermore, both in the case of the lock-in amplification, the phase lock loop for, for example, one HF modulation as well as the delay lock loop for a digital module tion application.

For the measurement of passive objects, the modulation of the emitted light and the corresponding modulation of the modulation photogate voltages U _am (t) or U _bm (t) can be carried out in various ways. First of all, continuous RF modulation can be carried out, the charge differences and the charge sums for evaluating the phase and amplitude information of the light wave being read out repeatedly at time intervals which can be influenced retrospectively by the pixel intensity.

An intermittent mode of operation with a pulse is advantageous RF modulation and lighting, e.g. B. a disturbing rear briefly surpassing basic lighting. This will only  the photo-generated charges inte during the RF pulse and then evaluated.

When determining in particular the phase or runtime information tion of reflected light waves can be used to increase the phase or runtime resolution the HF-Im known from radar technology pulse compression method with narrow correlation functions, e.g. B. the chirp technique can be used. Both the modu tion signal of the individual photonic mixing element as well with a given phase relation illuminating light wave of the sen ders and thus also the reflek with the desired phase relationship repeated modulated light wave with a chirp. Through the Chirp modulation can be done appropriately by inserting a adjustable delay between the modulation photo chip voltage of the photonic mixing element and that of the transmitter emitted light multiple targets resolved or annoying multiple reflections xions of an illuminated scene suppressed.

Another modulation is the pseudo- Noise modulation (pseudo-noise (PN) modulation) both as Ba sisband-PN and HF-PN modulation are available. A Sampling mode with sampling and holding processes (sample and hold) with repeating light signals is a special case of mixing and Correlation with needle impulses. For this as well as for other applicants The photo according to the invention can be used with pulsed modulation African mixing element can be used advantageously.

The types of modulation listed are all from the State of the art known.

The charges drifting to the accumulation gates can now be in processed in various ways. For one thing, it can photonic mixing element can be realized in CCD technology, wherein then the charges are collected below the accumulation gates or integrated and then in a conventional manner up to CCD readout circuit z. B. shifted in the three-phase shift cycle and can be read out via p or n diffusion.  

On the other hand, the photonic mixing element in CMOS technology can be implemented as an active pixel element with pixel-specific readout and signal preprocessing electronics. Practically the read-out circuit customary in CCD technology is introduced to the modulation photogates on both sides up to immediately. The accumulation gates are preferably designed as blocked low-capacitance pn diodes and pass the incoming photogenerated charges preferably directly via the electrodes G _a and G _b to the pixel readout and signal preprocessing electronics for storage and processing there.

In the latter case, the two parts of the charge become the charge swing read continuously and can z. B. with a La tion amplifier practically without feedback on one downstream connection capacity can be saved.

It is state of the art that those involved before each new measurement and charged capacities by electronic reset switch ent will be loaded and that expediently those measured in the reset state Fault voltages are used to correct the actual measured values will. This application of pixel-wise non-reactive off Reading has the advantage that the entire dynamic of the photonic Mixing element and thus the measuring method significantly compared the realization in CCD technology can be increased.

In a further preferred manner, it is possible to determine the phase and amplitude deninformation in a pixel readout and signal preprocessing electronics preferably to be calculated directly as on-chip integration. Such an application specific optoelectronic chip (ASOC) or such an active pixel sensor (APS) increases the measurement rate and he enables pixel-by-pixel preprocessing of the phases and / or Am flounder.

An important advantage of the present invention is that the Modulation simultaneously with charge generation and separation follows. In other words, find the detection and the mixture at the same time and without additional noise and band-limiting Intermediate stages instead. Therefore, in the prior art  other occurring temporal drift errors prevented, which at a modulation separated in time and space from the detection and Integration of the charges inevitably occur and not too under are pressing.

Another advantage of the present invention is the high Cutoff frequency of the photonic mixing element. The cutoff frequency of the charge transfer by the push-pull modulation voltage regarding the maximum drift length or transfer distance, i.e. the Total length of the modulation photogates, with the cutoff frequency corresponding MOS transistors comparable and thus achieves that GHz range. Furthermore, due to the antisymmetric charge carrier separation and difference formation disturbing common mode signals un oppressed. Any interference not correlating with the modulation signal signal, e.g. B. the backlight, is in the charge difference suppressed, which leads to a high signal-to-noise ratio. Furthermore, there is only a slight time drift due to the summary detection, mixing and charge carrier integration and -difference formation on the same chip. In addition, a Zu Summary of practically all measuring functions within one possible semiconductor structure.

Compared to the prior art of DE 44 39 298 A1 with the Ver The use of Pockels cells as modulators are only minor moduls Lation voltages in the 1 instead of 1000 volt range necessary. In addition is achieved by a 2D arrangement of photoni according to the invention mixing elements have a large aperture on the receiver side ensures.

For the determination of the phase and / or amplitude information no coherent or polarized light is required. This allows further specific properties of the incident Light waves by inserting selective filters z. B. regarding Po larization and wavelength of light can be used. In addition are high sensitivity and high signal-to-noise ratio nis by the elimination of the used according to the prior art broadband photo detector amplifier and electronic mixer given.  

The spectral optical bandwidth of the light waves to be measured is due to the spectral photosensitivity of the in the Raumla zone determined under the material used by Photogates, d. H. e.g. B. in silicon about the wavelength range 0.3 to 1.1 microns at InGaAs about 0.8 to 1.6 µm and for InSb about 1 to 5.5 µm.

The photonic mixing elements can be in any zero, one or two-dimensional arrangement can be arranged and bie thus a wide range of application geometries. Here can create several 100,000 photonic mixing elements in parallel with egg ner modulation bandwidth of z. B. 10-1000 MHz are operated, so that z. B. a camera shot of a 3D scene with determination of Realize distance information in each pixel extremely quickly is cash. About the charge differences of the accumulation gates flowing and read out charges becomes pixel by pixel bild ϕ (x, y) or - in the case of modulated lighting - the distance image or depth image with the radius vector or voxel distance R (x, y) certainly. The corresponding charge sums result in the conv tional pixel gray value A (x, y). Both can be scaled gray value image or to the 3D image A (x, y, z) can be combined.

The 3D image repetition rate is in the range from about 10 Hz to over 1000 Hz and depends on the number of photoni used mixing elements and the light intensity. By additional Color filters make it possible to use the usual color values red (x, y), green (x, y) and To win blue (x, y) of the distance image R (x, y).

Thanks to the integrated structure of the mixture and the charge carrier ink Last but not least, a simple structure of the photonic Mixing element reached. Finally, no special effort is required in the receiving channel, because a conventional image Optical optics are sufficient for imaging the incident, possibly reflective th light wave, provided a one- or two-dimensional scene and not just one point should be included. By synchronous Zooming the transmitting and receiving optics is the measuring device at un Different 3D scenes can be flexibly adapted.

The invention is described below using exemplary embodiments explained in more detail, reference being made to the drawing. In the drawing shows

Fig. 1a) in cross-section a pixel of a first Ausführungsbei game of an inventive photonic Mischele mentes in CCD technology and b) -f), the Potentialver distribution U _S (t) for the different phases or times of the two complementary modulation photogate voltages U _am ( t) and U _bm (t),

Fig. 2 is a block diagram representation of two linearly arranged pixels in CCD technology including a portion of an interline transfer-out apparatus,

Fig. 3 in the diagram the intensity distribution of the incident light and the potential curves of the voltages U _sep (t), U _a (t), U _am (t), U _bm (t) and U _b (t) in the case of an RF module tion,

Fig. 4 in the diagram shows the characteristic of the mixing and correlation result of the photonic mixing element in the form of the averaged to the accumulation gates drifting tendency photogenerated charge carrier currents i _a and i _b with an RF modulation depending on the relative phase or transit time shift ϕ _opt = ω _m τ,

Fig. 5 in the diagram for a PN modulation a) the modulation signal, b) the characteristics of the mixing and correlation ons Result both for a double pixel (only i _a and i _b ) and for a quadruple pixel with i _c and i _d a delay in the modulation signal for the 3rd and 4th modulation gates cm and dm of T _B and c) the difference values Δi _ab + Δi _cd = i _a - i _b + (i _c - i _d ) and Δi _ab - relevant for the distance evaluation Δi _cd = i _a - i _b - (i _c - i _d ),

Fig. 6a) in cross-section a pixel of a second execution example in CCD technology of a photonic mixing element according to the invention having an average modulating photo gate G ₀ and the potential distributions among the modulating photo gates and accumulation gates b) for a positive and c) for a negative modulating voltage U _m (t ),

Fig. 7a), in cross section, a pixel of a third Ausführungsbei game of an inventive photonic Mischele mentes and b) -f), the potential distributions for the various phases ver analogous to FIG. 1,

Figure a top view of a pixel of a fourth embodiment of an inventive photonic execution mixing element accumulation gates. 8 with four modulating photo gates and four Ac, referred to as quad pixel,

Fig. In a plan view a pixel of a fifth embodiment of an inventive photonic execution mixing element accumulation gates 9 with four modulating photo gates and four Ac and a central symmetrical middle gate G _0,

Fig. 10 is a schematic representation of a known prior art device for determining the phase and amplitude information of a light wave,

Fig. 11 is a schematic representation of a device according to the invention for determining the phase and Amplitu deninformation a light wave for RF modulation,

Fig. 12 is a schematic representation of a device according to the invention for determining the phase and on the amplitude information of a light wave z. B. for PN modulation or rectangular modulation,

FIG. 13a) in cross-section a pixel of a sixth execution example of a photonic mixer according to the invention element with Pixelauslese- and -vorverarbeitungselek electronics in CMOS technology, as well as b) and c) the Potenti alverteilung analogous to FIG. 6 for two phases or polarities of the modulating photo gate voltage and

Fig. 14 in a plan view of a pixel of a seventh embodiment example of a photonic mixing element according to the invention with four modulation photogates, four accumulation gates and a cross-shaped central gate G ₀ , preferably for digital modulation.

FIG. 1a shows the cross section of a single pixel 1 of a photonic rule mixing element the example of a CCD structure. In addition to the pixel 1, the photonic mixing element comprises the structures necessary for the voltage supply and the signal derivatives. The outer gates G _sep only serve to electrically delimit this pixel from neighboring structures.

The embodiment shown in FIG. 1 is carried out on a p-doped silicon substrate 2 . The mixing or multiplication process of the proposed concept is initially considered for pure CW radio frequency modulation.

Based on the cross-section shown in FIG. 1b-f schematically shows the potential distributions for different phases of the mixing process. The middle modulation photogates G _am and G _bm represent the light-sensitive part and are in the state of inversion. In addition to a positive bias voltage U ₀ on the conductive but optically partially transparent upper cover z. B. made of polysilicon they are operated with the superimposed push-pull voltages U _m (t). The modulation voltages U _am (t) = U ₀ + U _m (t) or U _bm (t) = U ₀ - U _m (t) result.

These cause multiplicatively a separation of the minority charge carriers generated by the Pho tons of the incident light wave in the space charge zone immediately below the insulator layer 3 , for. B. of silicon oxide or silicon nitride. These charge carriers (electrons in the example) drift under the influence of the modulating push-pull voltage to the closely adjacent positive accumulation gates G _a or G _b and are integrated there while the majo rity charge carriers or holes flow to the ground connection of the p-Si substrate. Back lighting is also possible.

Fig. 2 shows a top view of two pixels 1 of the inventive photonic mixing element including part of an interline transfer readout device 7 in the form of a 3-phase CCD shift register, at one end of which the readout electronics with a diffusion transition for the serial Further processing of the charge values obtained by the correlation is located. After a predeterminable time T for the charge accumulation under all accumulation gates of the row, for. B. at pixel number n the charges q _a and q _b un under Ga and G _b on the transfer gate TG _a and TG _{b are given} to the 3-phase read shift register. The limiting separation gates G _sep shield the correlation pixel against undesirable external influences and are preferably at ground potential.

In Fig. 3, belonging to FIG. 1, voltage profiles are shown. The modulation photogates G _am and G _bm are driven by means of the modulation photogate voltages shown in FIG. 3, which contain a counter-modulating HF modulation voltage U _m (t), which are described as follows:

U _am = U ₀ + U _m cos (ω _m t) (1a)

and

U _bm = U ₀ + U _m cos (ω _m t - 180 °) = U ₀ - U _m cos (ω _m t) (1b)

In Fig. 1b-f is the potential distribution U _S (s) in the space charge zone over the spatial extent s of a representative pixel 1 for all gates of this pixel involved in the time sequence from t ₀ to t ₈ for the duration of a period T _m des RF modulation signal shown clearly. At the accumulation gates G _a and G _b , a relatively high positive voltage ensures the accumulation of the photogenerated charge carriers, after they are predominantly to the left or to the right, depending on the polarity of the modulation photogate voltages U _am (t) and U _bm (t) Side of the pixel 1 shown in cross section in FIG. 1 are drifted. This process has a special effect if the light modulation and the modulation photogate voltage U _am (t) have the same frequency. Then, depending on the phase difference ϕ _{opt, there is} an average preferred direction of the charge carrier drift to the accumulation gates G _a and G _b . The associated averaged currents are described by i _a and i _b .

The underlying correlation process can be described mathematically as follows: In the receiving plane of the 2D array of photonic mixing elements in the most general case, z = 0 and the incident modulated light wave is generally described there by P _opt (x, y, t - τ). Here it is described via the photogenerated charge carriers with the push-pull modulation signal acting there, in general form by U _m (x, y, t), with respect to the charge differences of the two accumulation gates, in an almost multiplicative and integrative manner. The corresponding correlation function ϕ _{Um, Popt} (x, y, t) is z. B. for all averaged differences in the charge carrier drifts Δq _ab / T = Δi _ab = i _a - i _b (with T = integration time) to the accumulation gates G _a and G _b in the most general case location-dependent as a triple fold:

ϕ _{Um, Popt} (x, y, τ) = k ₁ .U _am (-x, -y, -τ) *** P _opt (x, y, τ) = k ₂ .Δi _ab (x, y, τ) (2)

with the transit time difference τ = ϕ _opt / ω _m , the modulation angular frequency w _m and the structure-dependent constants k ₁ and k _2, which are not essential for the functional principle.

The photonic mixing element according to the invention solves this task with high spatial and time resolution through the rapid, separating charge transport of the photoelectrons and their push-pull storage and evaluation of differences and sums. By forming the averaged drift currents Δi _ab (t) = i _a (t) - i _b (t), which are time-dependent for non-stationary light waves, all disturbing offset components are suppressed and at the same time the desired correlation function of the light signal P _opt (t - τ) with the modulation voltage U _m (t).

This process will be described in more detail. The HF drift field caused by U _am (t) and U _bm (t) causes the electrons to drift to the respective positive side. During e.g. B. the positive half-wave of the modulation photogate voltage U _am (t) = U ₀ + U _m (t), ie, during the negative half-wave of U _bm (t) = U ₀ - U _m (t), the photogenerated charge carriers become the accumulation gate G _a drift and accumulate there as a charge q _a or forwarded (compare the two upper modulation photogate voltage distributions in Fig. 1b and c). In FIG. 3, the optical power per pixel is shown as in the case of stationary, harmonically modulated lighting

P _opt (t - τ) = P ₀ + P _m cos (ωt - ϕ _opt ) (3)

where P _{0 represents} the mean value including the backlight, P _m the modulation amplitude, ω _m the RF modulation frequency, ϕ _opt the phase delay and τ = ϕ _opt / ω _m the corresponding delay time of the incident light wave compared to the modulation phase at G _am . The total photocurrent generated per pixel is

i (t) = S _λ .P _opt (t - τ) = S _λ . [P ₀ + P _m .cos (ω _m t - ϕ _opt )] (4)

i (t) = I ₀ + I _m .cos (ω _m t - ϕ _opt ) (5)

with the quantities i (t) = i _a (t) + i _b (t), I ₀ = mean value of the pixel photocurrent according to P ₀ , I _m = alternating amplitude of the modulated photocurrent according to P _m , and S _λ = spectral sensitivity. This total photo current per pixel is divided into two parts, namely in the current i _a (t) of the accumulation gate G _a and in the current i _b (t) of the accumulation gate G _b . Since these values are integrated - in CCD technology under the respective accumulation gates G _a and G _b and in the case of pixel-by-pixel reading CMOS technology, preferably in the readout electronics - it is sufficient to add the mean values i _a and i _{b of} these currents below consider. The maximum charge separation is achieved for the angle ϕ _opt = 0 or τ = 0. This case is shown in Fig. 3.

In the case of harmonic modulation, ideal conditions such as a suitable modulation amplitude, negligible drift propagation times, 100% modulation depth with P _m = P _{0 result} for the average photocurrents i _a and i _b

In FIG. 4, the course is this idealized average pixel currents shows ge. They represent the antiphase correlation functions that result from the RF-modulated received light and the RF-modulation-photogate voltages applied to the modulation photogates G _am and G _bm . With I _0, their sum corresponds to the mean pixel light output P ₀ . The total amount of charge that is accumulated over time T = NT _m (ie, over N periods T _{m of} the RF modulation voltage) results in

with a transit time corresponding to the phase delay τ = ϕ _opt / ω _m . In the following, q _{aT is} only referred to as q _a . The totality of the charges of the accumulation gates G _a and G _{b of} all pixels 1 forms two spatially discrete HF interferograms, the a interferogram and the b interferogram shifted by 180 ° with respect to the a interferogram, from which the runtime determined and sought by difference Differential RF interferogram is formed, which is described by equation (2).

In Fig. 11 the diagram of an inventive 3D camera is shown schelemente direct mixture on the basis of an array of photonic Mi uses. Compared with the 3D camera concept known from the prior art that is shown in FIG. 10, the modulation of a transmitter 4 for lighting optically passive 3D objects is realized in FIG. 11 by direct modulation of the current of a laser diode. The modulation is generated by an RF generator 13 . For larger distances z. B. the use of a high-performance laser diode array with preferably common modulation current and - for eye safety - with different wavelengths advantageous.

A first optics 5 images the light wave onto the surface of an object 6 . The light wave reflected by the object 6 is then imaged by a second optical system 7 on the surface of a photonic mixing elementary rays 8 .

The photonic mixing element array 8 is also controlled by the HF generator 13 , the control for different phase shifts to the phase of the emitted light wave being carried out by the HF generator 13 . The signals of the photonic mixing element array 8 are, if not already done on-chip, finally evaluated by an evaluation unit 9 .

Due to the measuring device according to the invention is for the pre striking 3D camera concept next to the photo according to the invention African mixing element array with no additional optical modulator high aperture necessary, resulting in an economically advantageous Solution leads.

To determine the pixel phase ϕ _opt from the resulting correlation amplitudes, a total of four different interferograms are used for four different phases of the mixer signal, as previously indicated. The four phases of the mixer signal result in the event that the modulation _{photogate voltages} U _am and U _{bm are switched} from the state of the phase relationship 0 ° / 180 ° to the state 90 ° / 270 ° or delayed by 90 °. In this way, the two associated imaginary or quadrature components for the real or in-phase components are obtained, from which the searched pixel phase can be calculated according to equation (10) described below.

This procedure also enables the elimination of disturbing offset voltages caused by the background brightness and generated by the mixing process.

In addition to the measurement process of CW-modulated 3D light waves described by way of example, by 2D correlation with a modulation voltage U _m (x, y, t), preferably the same frequency in the plane of the photonic mixing element array, the measuring device according to the invention can also be used advantageously with pulsed modulation signals will.

For tasks of high-precision transit time measurement of 3D light waves, pseudo-noise modulation of the light is particularly advantageous. An exemplary embodiment for the measurement of optically passive 3D objects is shown in FIG. 12. Similar to the exemplary embodiment with harmonic modulation in FIG. 11, the device according to the invention has a corresponding lighting device which illuminates the 3D objects 6 with the intensity PN (pseudo Noise) -modulated light be illuminated and the reflected and received light the correlation process with preferably the corresponding PN modulation signal, which is generated by the generator 13 , undergoes.

Since the correlation of such PN signals with increasing word length T _W = T _B (2 ^N - 1) resembles a triangular needle pulse with a half-width equal to the bit width T _B , it must be illuminated for clear and complete measurement of the entire light volume or the whole Space a relative delay T _D between the light-modulating PN signal and the demodulating PN counter clock voltage U _m (t) of the same signal shape at the modulation pho togates at least once continuously or step by step through the entire delay range of the maximum echo delay time in T _B steps. The delay element 11, which is adjustable by the control and evaluation unit 9 with respect to the delay T _D , is used for this purpose.

In Fig. 5a the example is a 15-bit PN sequence shown rectangular, the modulation signal U _m (t). The result of the correlation by the photonic mixing element are the averaged drift currents i _a and i _b shown in FIG. 5b over the relative delay τ.

In the quadruple pixel described later according to FIG. 8, FIG. 9 and FIG. 14, the push-pull modulation photo voltages applied to the modulation photogates G _cm and G _dm and superimposed on the bias voltage U _{0 are} preferably by T _B compared to the modulation photo gates Ga and Gb applied push-pull modulation pho togate voltages delayed, ie U _cm (t) = U ₀ + U _m (t - T _B ) and U _dm (t) = U ₀ - U _m (t - T _B ), which results in very advantageous amplitude and runtime measurements.

Except for a predefinable delay T _{D of} the modulation voltages, the light intensity emitted by the transmitter 4 has constant. * P _opt (t) the same PN signal structure. The reflection reaches the photonic mixing element after the echo delay. The correlation with the push-pull modulation voltages leads, depending on the relative Laufzeitver τ delay 0 in the ideal case without background brightness in the two-pixel to in Fig. 5b shown mean pixel currents i _a is T _D = and i _b and in the four-pixel with said T _B -Zeitversatz zusätz Lich on the average pixel currents i _c and i _d . This correlation characteristic initially reveals that several object reflections on the same radius vector can be distinguished, e.g. B. for differentiating several consecutive partially transparent objects or for eliminating multiple reflections.

In addition, in succession in the two-pixel, and in the four-pixel same time, preferably in the respective corresponding Pi xelauslese- and Signalvorverarbeitungselektronik 15 which is asked in Fig. 5c formed sum and difference of the average drift current differences. They allow highly sensitive measurements since only in the T _B to _B 2T wide measurement window signal values equal to zero appear. The relevance of a measurement is determined on the basis of a minimum amplitude by evaluating the sum. The difference shows a steep linear curve in the usable T _B -wide measuring window, which allows a runtime determination with high resolution. For the example idealized here

The block diagram of a corresponding measuring device for optically measuring 3D objects with PN modulation on the basis of the proposed correlation photodetector array is characterized by a particularly simple structure, as can be seen in FIG. 12. In addition to the generator 10 and the delay element 11 , the same structure is given as in FIG. 11.

To determine the distance quickly with a lower resolution, a simple rectangular modulation of the transmitter 4 by the generator 10 with the period T and preferably the same pulse and pause duration T _{B is also} used according to the invention. The runtime is determined according to equation (9). The resolution is gradually increased by the period T decreasing by a factor of 2, with the first measuring step being followed by a second with the same period but a time shift T _D = T / 4.

The cross section of the pixel 1 of the photonic mixing element according to the invention, which is shown by way of example in FIG. 1, can be optimized with regard to its cutoff frequency by a suitable design of the potential gradient caused by the clock modulation voltage. 6 this is shown in Fig. An embodiment in which a mitt Leres gate G _{0 bm} between the modulating photo gates G _am and G is arranged on, preferably on the bias voltage U ₀ is located, and which together with the modulating photo gates G _am and G _bm three Poten forms stages. What is desired is a potential gradient that is as uniform as possible or a modulation drift field that is as constant as possible, which is achieved by increasing the number of stages from two to three or even more. In the photosensitive space charge zone, the shape of the steps decreases with the distance from the insulating layer 3 . This effect is used in a further embodiment according to the invention, namely by using a so-called "buried channel", a lightly doped n-channel a few micrometers away from the insulating layer and slightly deeper in the p-substrate under the modulation photogates. Furthermore, shading 12 is provided for the accumulation gates G _a and G _b so that they are not illuminated by the light wave and additional charge carriers are generated.

Fig. 7 shows a special embodiment and connection of photonic mixing elements, in which compared to the two modulation onsphotogates in Fig. 1 are separated only by a common accumulation gate G _{s, n} , whereby a higher filling efficiency is achieved. Here too, shading 12 of the accumulation ges G _a and G _{b is} provided. The polarity of the push-pull modulation voltages or the order of G _{am, n} and G _{bm, n changes} from pixel to pixel. This three-period period of the gates is also suitable for direct readout by operation as a three-phase shift register. A disadvantage that is tolerable in certain applications is the charge distribution also to the neighboring pixels, which leads to an apparent pixel enlargement and lower spatial resolution in the relevant direction.

A calculation of these relationships shows that compared to a 100% payload when evaluating the charge differences central, considered pixels only receives 50% and the two neighboring pixels each received 25%.

To illustrate the charge distribution, the different phases of the potential distribution for CW modulation are shown in FIG. 7 analogously to FIG. 1.

In FIG. 8, a further advantageous embodiment of the design is shown of a Pi xels a photonic mixing element that requires at CW-Mo no IQ (in-phase, quadrature phase) dulation measurement switching between the I and Q-states. Instead of the double pixel described above, a quadruple pixel with the modulation photogates G _am , G _bm , G _cm and G _dm and the associated accumulation gates G _a , G _b , G _c and G _{d is} proposed, which enables the correlation for four phase positions at the same time, since the Push-pull modulation photogate voltages U _am (t) and U _bm (t) or U _cm (t) and U _dm (t), in particular in the case of HF modulation, are shifted by 90 °.

In orthogonal arrangement to the described modulation photo gates G _am with ϕ _am = 0 ° and G _bm with ϕ _bm = 180 ° there are therefore two further modulation photogates G _cm with ϕ _cm = 90 ° and G _dm with ϕ symmetrically integrated within the pixel _dm = 270 °, which work on the same principle. In this way, a four-phase charge accumulation with the individual charges q _a , q _b , q _c and q _{d occurs} under the associated accumulation gates G _a , G _b , G _c and G _d or in the associated readout electronics, using a simple arithmetic operation the associated phase ϕ _opt is calculated directly as follows:

For the simple gray value determination of a single pixel, the individual charges of all accumulation gates of a pixel are summed up: q _pixels = q _a + q _b + q _c + q _d . In this case, the reading process of the four charges is expediently carried out by an active pixel design using CMOS technology with pixel-integrated signal preprocessing.

Fig. 9 8, a four pixel points, as well as Fig. Photonic Mi schelementes, but with a corresponding to FIG. 6 smoothed potential gradient with the aid of the central, preferably on the poten tial U lying ₀ square central gate G _0.

FIG. 14, like FIG. 9, shows a quadruple pixel of a photonic mixing element with a structure optimized for digital modulation signals. The middle gate G ₀ arranged between the preferably square modulation photogates serves, similarly as in FIG. 9, to smooth the potential gradient generated by the modulation photogate voltage.

Fig. 13, finally, a further preferred embodiment shows ei nes pixel 1, approximately examples in contrast to the previously-indicated exporting not in CCD technology, but in CMOS Techno logy with a pixel by pixel readout and 15 Signalvorverarbeitungselektronik realized. The functioning of the modulation voltage-dependent drifting of the charge carriers on the charge swing is the same as in the previously shown exemplary embodiments. Among the kinds of differently processing is le in the illustrated in Fig. 13 embodiment diglich the tes to the Akkumulationsga G _a and G _b drifted charges q _a and q _b.

The accumulation gates G _a and G _b are formed in the present exemplary embodiment as blocked pn diodes. On a preferably weakly doped p-Si substrate 3 in Fig. 13, the positively biased accumulation gates G _a and G _{b are formed} by n ₊ -doped electrodes. In so-called "floating diffusion" mode or in the high-resistance voltage reading mode, as in CCD technology, the charges q _a and q _{b are integrated} on the capacitances of the accumulation gates G _a and G _b and read out as voltage values with high resistance.

A current reading mode can also advantageously be used in which the photogenerated charge carriers are not integrated in the potentiometer pot, but are continuously passed on via an output diffusion via suitable current reading circuits connected to the accumulation gates G _a or G _b . Then these charges are each integrated on an external capacity, for example.

A readout circuit in the current readout mode, which keeps the accumulation gate voltage virtually constant by amplifier feedback, advantageously prevents the amount of accumulated charges q _a and q _b from becoming ineffective or even overflowing when the pixel is intensively irradiated Potential pot leads. This significantly improves the dynamics of the photonic mixing element. Here, too, improvements, including an increase in the cutoff frequency, are achieved by the technique of a weakly doped n-channel (“buried layer”) below the insulating layer of the modulation gates.

The design of the photonic mixing element in CMOS-Tech Technology also enables the use of an active pixel design gns (APS), with which a readout and signal preprocessing for each pixel processing circuit can be integrated into the photonic mixing element can. Pre-processing of the electrical signals is thus direct possible at the pixel before the signals are passed on to an external circuit be directed. In particular, the phase and amplitude deninformation can be calculated directly on the chip, so that the Measurement rate can be increased further.

In a further embodiment of the invention, one is preferred two-dimensional photonic mixing element array for a three-dimensional passive electronic object search and tracking or actively shining objects according to various criteria, such as B. Whether project shape, position, color, polarization, speed vector,  -brightness or a combination of object properties ver turns. Is z. B. when going through different Modulationssi signals (e.g. frequency or code change) during 3D measurement an incident light wave, which may initially be unknown, a local correlation through the criterion of differential drift currents found not equal to zero, then this object can continuously measured specifically with regard to the object properties mentioned and, if necessary, in the case of changes via a control loop, in particular including the depth of the image can be tracked.

The photonic mixing element is used in different modes used, which are shown below.

The total charge at the accumulation gates G _a and G _{b is of} less interest because it always corresponds to the total intensity of the incident light waves, q _a + q _b = const. * P _{opt, ges} * T with T = integration time.

The differential charge Δq _ab = q _a - q _b = i _a .T - i _b .T depends on several factors and can be used in several ways to measure the incident light wave. To this end, an always existing basic brightness P ₀ <= P _m (see FIG. 3a) is taken into account.

Optionally, e.g. B. when measuring an object 6 illuminated by a transmitter 4 with modulated light, the transmission power is switched on or off and thus P _m becomes finite or equal to zero. At the same time, the modulation voltage U _m (t) is either switched to zero or to the curve used in the transmitter or contained in the incident light or to a voltage U _m0 constant during the integration time.

With P ₀ ≠ 0 there are four important operating modes:

• 1. 1.) Δq _ab = 0 for P _m = 0 and U _m = 0.
• 2. 2.) Δq _ab = 0 for finite P _m and with U _m (t) as an HF modulation signal.
• 3. 3.) With finite P _m and a high-frequency modulation voltage, Δq _{ab is} a function of U _m (t), of the relative modulation delay time shift τ and of the incident, modulated light power component P _m (t).
• 4. 4.) If during an integration time T an incident average light intensity P ₀ and a constant modulation voltage U _m0, then the differential charge .DELTA.Q _from a function of U _m0 and the optical mean power P _0th

For light waves that are not intensity-modulated, in a white teren embodiment of the invention ent the photonic mixing element ent speaking of the fourth case of a possible mode of operation z. B. for the 2D image processing used.

Each mixing element can be controlled individually and independently of one another, for. B. by pixel-by-pixel assignment of a quickly _writeable modulation _{voltage word} for U _m0, preferably using a RAM module. Only approximately proportional to U _m0 difference drift currents .DELTA.i are evaluated _from difference preferably or charges T * .DELTA.i _from. The modulation _voltage U _m0 is derived from the modulation _{voltage word} .

Thus U _m (t) is no longer periodic or quasi-periodic as in the previous application examples, but aperiodically z. B. set according to a predetermined or according to the measured image content. For U _m (t) = 0, all differential currents result in zero, so that the associated difference image D (x, y) also appears with zero amplitude or intensity.

The difference image brightness can thus be influenced in a targeted manner by varying U _m (x, y, t). Thus, according to the invention, any, ie also unmodulated, light waves or images can be set via an extremely quickly adjustable weight function G (x, y, t) = k ₁ .U _m (x, y, t) z. B. be accessed via the controllable, pixel-wise allocated memory cells of a versatile image processing, such as. B. the previously listed applications for object search and tracking, but without the aspect of depth information.

Claims (24)

1. Method for determining the phase and / or amplitude formation of an electromagnetic wave

• in which an electromagnetic wave is radiated onto the surface of a photonic mixing element having at least one pixel, the pixel having at least two light-sensitive modulation photogates G _am and G _bm and associated accumulation gates G _a and G _b ,
• - In which the modulation photogates G _am and G _bm modulation photogate voltages U _am (t) and U _bm (t) are applied, which as U _am (t) = U ₀ + U _m (t) and U _bm (t) = U ₀ - U _m (t) are configured,
• a direct voltage is applied to the accumulation gates G _a and G _b , the amount of which is at least as large as the amount of the sum of U ₀ and the amplitude of the modulation voltage U _m (t),
• - In which the charge carriers generated in the space charge zone of the modulation photogates G _am and G _bm from the incident electromagnetic wave, depending on the polarity of the modulation photogate voltages U _am (t) and U _bm (t), are exposed to the potential gradient of a drift field and to the corresponding Accumulation gate G _a or G _b drift and
• - in which the charges q _a and q _b drifted to the accumulation gates G _a and G _b are derived.

2. The method according to claim 1,

• - in which an intensity-modulated electromagnetic wave is emitted by a transmitter,
• in which the electromagnetic wave reflected by an object is radiated onto the surface of the photonic mixing element,
• - in which the modulation _{photogate voltages} U _am (t) and U _bm (t) are in a fixed phase relationship with the phase of the electromagnetic wave emitted by the transmitter and
• - In which the charge carriers generated are additionally exposed to the potential gradient of a drift field as a function of the phase of the push-pull modulation _{photogate voltages} U _am (t) and U _bm (t).

3. The method according to claim 2,

• - In the case of two different phase shifts Δϕ ₁ and Δϕ _{2 of} the modulation photogate voltages U _am (t) and U _bm (t) relative to the phase of the electromagnetic wave emitted by the transmitter, the charges q _a1 and q _b1 as well as q _a2 and q _{b2 are} derived and the charge differences (q _a1 - q _b1 ) and (q _a2 - q _b2 ) are formed and
• - where according to the equation
  the pixel phase ϕ _{opt of} the incident electromagnetic wave relative to the phase of the electromagnetic wave emitted by the transmitter and thus the transit time of the electromagnetic wave received by the pixel is determined.

4. The method according to claim 3,

• - in which, with the aid of four modulation photogates G _am , G _bm , G _cm and G _dm and four associated accumulation gates G _a , G _b , G _c and G _d , for two different phase shifts Δϕ ₁ and Δϕ _{2 of} the modulation photogate voltages U _am ( t) = U ₀ + U _m1 (t) and U _bm (t) = U ₀ - U _m1 (t) as well as U _cm (t) = U ₁ + U _m2 (t) and U _dm (t) = U ₁ - U _m2 (t) relative to the phase of the electromagnetic wave emitted by the transmitter, the charges q _a , q _b , q _c and q _d are simultaneously separated and derived and
• - where according to the equation
  the pixel phase ϕ _{opt of} the incident electromagnetic wave relative to the phase of the electromagnetic wave emitted by the transmitter and thus the transit time of the electromagnetic wave received by the pixel is determined.

5. The method according to any one of the preceding claims,

• in which the photonic mixing element has a plurality of pixels,
• - In which at least one pixel is irradiated with part of the intensity-modulated electromagnetic wave from the transmitter and
• - In which a calibration of the phase shift between the emitted electromagnetic wave and the modulation photogate voltages U _am (t) and U _bm (t) is carried out from the phase shift measured with this pixel.

6. The method according to claim 1,

• in which an electromagnetic wave with externally excited unknown intensity modulation is radiated onto the surface of the photonic mixing element,
• in which the modulation _{photogate voltages} U _am (t) and U _bm (t) are generated by a tunable modulation generator,
• in which the charge carriers generated are additionally exposed to the potential gradient of a drift field as a function of the phase of the push-pull modulation _{photogate voltages} U _am (t) and U _bm (t)
• - In which the photonic mixing element and the modulation generator form at least one phase-locked loop and the electromagnetic wave is measured using the lock-in method.

7. The method according to any one of claims 1 to 6, in which as a periodi modulation a continuous or discontinuous RF Modulation, a pseudo-noise modulation or a chirp mod lation is used. 8. The method according to claim 7, wherein the modulation is an HF modulation and preferably the charges q _a and q _b and, if appropriate. Q _c and q _d for the phase shifts Δϕ = 0 ° / 180 ° and 90 ° / 270 ° be derived. 9. The method of claim 1, in which an aperiodic modulation with modulation _{photogate voltages} U _am = U ₀ + U _m0 and U _bm = U ₀ - U _m0 with a time-constant but variable modulation _voltage U _m0 is used. 10. The method according to any one of claims 1 to 9, in which the charges q _a and q _{b are integrated} below the accumulation gates G _a and G _b and are read out with a multiplex structure, preferably with a CCD structure. 11. The method according to any one of claims 1 to 9, wherein the accumulation mulations gates G _a and G _b as pn diodes, preferably as a blocked low-capacitance pn diodes and preferably in CMOS technology, and in which the charges q _a and q _b can be read out directly as voltage or as current. 12. The method of claim 11, wherein the pixel phase directly with Using an active pixel sensor structure (APS) is calculated.   13. The method according to any one of claims 1 to 12, wherein the Pixel brightness as the sum of the charges of all accumulation gates is evaluated. 14. Photonic mixing element

• - with at least one pixel ( 1 ),
• - The at least two light-sensitive modulation photogates (G _am , G _bm ) and
• - The modulation photogates (G _am , G _bm ) assigned, shaded from the incident electromagnetic wave Akkumulati onsgates (G _a , G _b ).

15. Mixing element according to claim 14, characterized in that a central gate (G ₀ ) is arranged between the modulation photogates (G _am , G _bm ). 16. Mixing element according to claim 14 or 15, characterized in that the pixel ( 1 ) four, preferably symmetrically arranged, modulation photogates (G _am , G _bm , G _cm , G _dm ) and accumulation mulations gates (G _a , G _b , G _c , G _d ). 17. Mixing element according to one of claims 14 to 16, characterized in that the pixel ( 1 ) is executed in MOS technology on a silicon substrate ( 2 ) and can be read out with a multiplex structure, preferably with a CCD structure. 18. Mixing element according to one of claims 14 to 16, characterized in that the accumulation gates (G _a , G _b ) are designed as pn diodes, preferably as blocked, low-capacitance pn diodes and preferably in CMOS technology, and the Charges q _a and q _b can be read out directly as voltage or as current. 19. Mixing element according to claim 18, characterized in that the pixel ( 1 ) is designed as an active pixel sensor structure. 20. Mixing element arrangement with at least two photonic Mi elements according to one of the device claims 14 to 19 , characterized in that the photonic mixing elements are arranged in a one-dimensional or two-dimensional arrangement. 21. Mixing element arrangement according to claim 20, characterized in that in each case two adjacent, different pixels (n, n + 1) assigned modulation photogates (G _{am, n} , G _{am, n + 1} ) or (G _{bm, n} , G _{bm, n + 1} ) each have a common accumulation gate (G _S ) and that the modulation photogates (G _{am, n} , G _{am, n + 1} ) and (G _{bm, n} , G _{bm, n + 1} ) each of the same modulation photo voltages U _am (t) and U _bm (t) are applied. 22. Mixing element arrangement according to claim 20 or 21, characterized in that the transmitter ( 4 ) directly radiates at least one pixel ( 1 ) with part of the intensity-modulated electromagnetic wave. 23. Device for determining the phase information of an elec tromagnetic wave

• with at least one photonic mixing element according to one of the device claims 14 to 19 ,
• - with a modulation generator ( 10 , 13 ),
• - With a transmitter ( 4 ), the emitted electromagnetic wave of the modulation generator ( 10 , 13 ) is intensity modulated in a predetermined manner,
• - The electromagnetic wave reflected by an object ( 6 ) radiates onto the surface of the photonic mixing element and
• - Wherein the modulation generator ( 10 , 13 ) the photonic Mischele element with modulation voltages U _m (t), which are in predetermined phase relation to the phase of the emitted electromagnetic wave of the transmitter.

24. The device according to the preceding device claim, characterized in that an optical system ( 7 ) and possibly a mixing element arrangement according to one of claims 20 to 21 are provided, the optical system ( 7 ) reflecting the reflected electromagnetic wave onto the surface of the mixing element or Mapping element arrangement maps.