Determining the distance of an object

Abstract

An optoelectronic sensor for determining the distance of an object in a monitoring area has a light transmitter for transmitting transmitted light, a light receiver for generating a received signal from remitted light remitted by the object, and a control and evaluation unit configured to modulate the transmitted light with at least a first frequency and a second frequency, to determine a phase offset between transmitted light and remitted light for the first frequency and the second frequency, and to determine a light time of flight. The control and evaluation unit is configured to determine a first amplitude and a second amplitude for the first frequency and the second frequency from the received signal and to detect whether the transmitted light impinges on an edge in the monitoring area on the basis of an evaluation of the first amplitude and the second amplitude.

Claims

1. An optoelectronic sensor (10) for determining the distance of an object (20) in a monitoring area (18), comprising: the sensor (10) having a light transmitter (12) for transmitting transmitted light; a light receiver (26) for generating a received signal from remitted light (22) remitted by the object (20); and a control and evaluation unit (28) configured to modulate the transmitted light (16) with at least a first frequency and a second frequency, to determine a phase offset between transmitted light (16) and remitted light (22) for the first frequency and the second frequency, and to determine a light time of flight from the phase offset between transmitted light and remitted light for the first frequency and the second frequency, wherein the control and evaluation unit (28) is furthermore configured to determine a first amplitude and a second amplitude for the first frequency and the second frequency from the received signal and to detect whether the transmitted light impinges on an edge in the monitoring area (18) on the basis of an evaluation of the first amplitude and the second amplitude, and wherein the control and evaluation unit (28) is configured to locate an object (20) in the monitoring area (18) means of a detected edge.

2. The sensor (10) according to claim 1, wherein the control and evaluation unit (28) is configured to evaluate at least one of a difference of the first amplitude and the second amplitude with a tolerance threshold and the ratio of the first amplitude and the second amplitude.

3. The sensor (10) according to claim 1, wherein the control and evaluation unit (28) is configured to evaluate a quotient of standard deviation and mean value of the amplitudes.

4. The sensor (10) according to claim 1, wherein the control and evaluation unit (28) is configured to determine a significance value for the detection of whether the transmitted light impinges on an edge in the monitoring area (18).

5. The sensor (10) according to claim 1, wherein the control and evaluation unit (28) is configured to discard distance values where the transmitted light impinges on an edge in the monitoring area (18).

6. The sensor (10) according to claim 1, wherein the control and evaluation unit (28) is configured to determine phase offsets and amplitudes in-phase and quadrature (IQ) demodulation.

7. The sensor (10) according to claim 1, wherein the control and evaluation unit (28) is configured to modulate with an artificial first frequency and second frequency independent of the carrier frequency of the transmitted light (16).

8. The sensor (10) according to claim 1, which is configured as a laser scanner having a movable deflection unit for periodically scanning the monitoring area (18) with the transmitted light (16) and an angle detection unit for determining the respective angular position of the deflection unit.

9. The sensor (10) according to claim 1, wherein the control and evaluation unit (28) is configured to output the distance values with additional information.

10. The sensor (10) according to claim 9, wherein the additional information is a binary edge flag or a significance value.

11. A method for determining the distance of an object (20) in a monitoring area (18), the method comprising: modulating transmitted light (16) with at least a first frequency and a second frequency prior to transmission into the monitoring area (18); receiving a signal generated from remitted light (22) remitted by the object (20), for the first frequency and the second frequency, determining a respective phase offset between transmitted light (16) and remitted light (22) and determining a light time of flight, from the phase offset between transmitted light and remitted light, for the first frequency and the second frequency; determining a first amplitude and a second amplitude for the first frequency and the second frequency from the received signal, and detecting whether the transmitted light impinges on an edge in the monitoring area (18) on the basis of an evaluation of the first amplitude and the second amplitude; and locating an object (20) in the monitoring area (18) by means of detected edge.

12. An optoelectronic sensor (10) for determining the distance of an object (20) in a monitoring area (18), comprising: the sensor (10) having a light transmitter (12) for transmitting transmitted light; a light receiver (26) for generating a received signal from remitted light (22) remitted by the object (20); and a control and evaluation unit (28) configured to modulate the transmitted light (16) with at least a first frequency and a second frequency, to determine a phase offset between transmitted light (16) and remitted light (22) for the first frequency and the second frequency, and to determine a light time of flight from the phase offset between transmitted light and remitted light for the first frequency and the second frequency, wherein the control and evaluation unit (28) is furthermore configured to determine a first amplitude and a second amplitude for the first frequency and the second frequency from the received signal and to detect whether the transmitted light impinges on an edge in the monitoring area (18) on the basis of an evaluation of the first amplitude and the second amplitude, wherein the control and evaluation unit (28) is configured to output the distance values with additional information, and wherein the additional information is a binary edge flag or a significance value.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention will be explained in the following also with respect to further advantages and features with reference to exemplary embodiments and the enclosed drawing. The Figures of the drawing show in:

(2) FIG. 1 a schematic diagram of an optoelectronic sensor;

(3) FIG. 2 a schematic representation of transmitted light of an optoelectronic sensor impinging on an edge;

(4) FIG. 3a a representation of a vector in the complex plane to explain the determination of phase and amplitude of a signal;

(5) FIG. 3b a representation similar to FIG. 3a to explain the superimposition of two signals;

(6) FIG. 4a-b a representation of the superimposition of the received signals in the case of an edge for two different frequencies f.sub.0, f.sub.1; and

(7) FIG. 5 a schematic representation of the measuring points of a laser scanner at object edges for comparative explanation of the measurement of the object with and without detection of edges.

DETAILED DESCRIPTION

(8) FIG. 1 shows a schematic sectional view of an optoelectronic sensor 10. A light transmitter 12, for example a laser or LED, transmits transmitted light 16 into a monitoring area 18 via transmission optics 14. The light transmitter 12 preferably has a laser light source, in particular a semiconductor laser in the form of a VCSEL laser or edge emitter, but other light sources such as LEDs are also conceivable.

(9) When the transmitted light 16 impinges on an object 20 in the monitoring area 18, part of the light returns to the sensor 10 as remitted light 22, where it is guided by receiving optics 24 onto a light receiver 26, for example a PIN diode, an APD (avalanche photodiode) or a single-photon APD (SPAD, avalanche photodiode in Geiger mode) or a multiple arrangement thereof.

(10) A control and evaluation unit 28 controls the light transmitter 12 and evaluates the received signal of the light receiver 26. The light time of flight is measured using a phase method (CW method, iToF), wherein, in order to increase the unambiguity range, the transmitted light 16 is modulated with at least two frequencies f.sub.0, f.sub.1, and the phase offset of the received signal is determined in both frequencies. The modulation is preferably carried out with an artificial frequency, for example in the range between 0.3 and 300 MHz, wherein a sinusoidal, rectangular and in principle any other periodic modulation is conceivable.

(11) According to the invention, not only the phase but also the amplitude of the received signal is determined for each of the modulation frequencies. Based on these amplitudes, the control and evaluation unit can detect 28 edges, i.e. where the light spot generated by the transmitted light 16 impinges on an edge in the monitoring area 18. This is explained in more detail below with reference to FIGS. 2 to 5.

(12) The basic configuration of sensor 10 according to FIG. 1 is only exemplary. Other arrangements are conceivable, for example a coaxial instead of a biaxial configuration, and also sensor types other than a one-dimensional scanning sensor, in particular a laser scanner. In a laser scanner, the transmitted light 16 is periodically deflected as a scanning beam with the aid of a rotating mirror and thus scans a scanning plane in the monitoring area 18. Alternatively, a measuring head with light transmitter 12 and light receiver 26 rotates. The respective angular position of the rotating mirror or the measuring head is determined with an encoder or the like, so that from the angle and measured distance measuring points in polar coordinates result. Instead of the single beam of transmitted light 16 as shown, multi-beam systems are also possible. In the case of a laser scanner, several scanning planes are thus generated. It is also conceivable to configure the sensor 10 as a time-of-flight camera with a pixel matrix as light receiver 26, wherein the monitoring area 18 is illuminated in an entire area or pixel by pixel, using a large number of simultaneous light beams or one controlled light beam after the other.

(13) FIG. 2 schematically illustrates an edge measurement where transmitted light impinges on an edge in the monitoring area 18. This means that the transmitted light 16 impinges on two object sections or objects 20a-b simultaneously during the same measurement, namely a closer object 20a and a farther object 20b. The transmitted light 16 is kind of divided, with one partial cross-section 16a impinging on the nearer object 20a and another partial cross-section 16b on the farther object 20b. Accordingly, the received signal is generated by a superposition of remitted light 22 from the closer object 20a and the farther object 20b. Such edge measurements regularly result in incorrect measurement values, which complicates subsequent data processing and affects the final measurement results. According to the invention, edge measurements should be detected in a phase method with several frequencies.

(14) FIG. 3a shows the representation of a sinusoidal signal in a vector diagram in the complex plane in order to explain the general mathematical background. The two axes are labeled in accordance with an IQ modulation. A signal oscillating at a certain frequency f.sub.i can be characterized by its amplitude a and its phase ϕ. The respective proportion I.sub.fi and Q.sub.fi in the two axes corresponds to a sin ϕ and a cos ϕ.

(15) To recover these values a, ϕ from the received signal per frequency f.sub.i, an IQ method can be used. The received signal is mixed with the corresponding transmitted signal and a transmitted signal shifted by 90°, so that after subsequent lowpass filtering the quantities I.sub.fi and Q.sub.fi are detected. This allows the phases and amplitudes to be calculated as

(16) ${\varphi }_{\mathrm{fi}}={\tan }^{-1}\frac{Q_{f_i}}{I_{f_i}}\mathrm{and}{a}_{\mathrm{fi}}=\sqrt{I_{f_i}^2+Q_{f_i}^2}.$

(17) This determination of phase and amplitude is advantageous, but not the only possibility. For example, the phase and amplitude of the desired frequencies in the received signal can alternatively be calculated with a Fourier transformation after digitization. Furthermore, specific receiving pixels are conceivable, in which the generated photoelectrons are collected via multiple charge storages or taps depending on the current phase of the transmitted light.

(18) FIG. 3b shows, in a diagram corresponding to FIG. 3a, the superimposition of two signals s.sub.1, s.sub.2 of the same frequency f.sub.i but different phase and amplitude to form a sum signal s.sub.1+s.sub.2. From the sum signal s.sub.1+s.sub.2, the phase of the individual signals can no longer be easily reconstructed. If such a sum signal s.sub.1+s.sub.2 is interpreted as a simple sinusoidal signal, a single phase and a single amplitude can be determined, but these generally no longer correspond to the phase and amplitude of one of the two signals s.sub.1, s.sub.2.

(19) FIG. 4a-b shows the corresponding situation in the case of an edge measurement, causing a superimposition of two signals as described with reference to FIG. 3b. As shown in FIG. 2, one part 16a of the transmitted light propagates along a different path than the other part 16b. Accordingly, the received signal s.sub.total is a superposition of a signal S.sub.near and a signal S.sub.far wherein the phase is a mixture of ϕ.sub.near and ϕ.sub.far. For each of the frequencies f.sub.0 and f.sub.1 involved, as shown in FIG. 4a and FIG. 4b, sum signals are generated that no longer enable to reconstruct the values ϕ.sub.near and ϕ.sub.far.

(20) However, in the measurements with the modulation frequencies f.sub.0 and f.sub.1, there are different phase differences ϕ.sub.near and ϕ.sub.far. The fact that there is a different phase difference for the same distance at different measuring frequencies is exactly the reason why measurements with different frequencies are carried out and so that the unambiguity range can be extended. Thus, the vectors representing s.sub.near and s.sub.far in FIGS. 4a-b have different angles depending on the frequency f.sub.0 and f.sub.1. In particular, the difference in angle between the vectors for one frequency f.sub.0 according to FIG. 4a is not the same as the difference in angle between the vectors for the other frequency f.sub.1 according to FIG. 4b. In the vectorial addition, despite individual vectors S.sub.near, S.sub.far of equal length, the resulting vectors s.sub.total have different lengths depending on the frequency f.sub.0 and f.sub.1.

(21) As a result of this observation, it can be stated that in the case of an edge measurement, the modulation amplitudes of the received light are usually different for the two measuring frequencies f.sub.0 and f.sub.1. On the basis of this difference in amplitude, edge measurements can be detected. One advantage of this edge detection is that it also works in cases where a distance value that was incorrectly determined due to the edge does not form a clear outlier within a measurement contour. Such an outlier could also be detected by other means. A moderate measuring error, on the other hand, would remain undetected simply by looking at the distance, whereas the invention nevertheless detects the edge.

(22) Accordingly, the control and evaluation unit 28 compares the amplitudes a.sub.fi with one another that are determined in addition to the phases for the respective frequency f.sub.i for a respective measured distance value. If there is a significant difference, it is an edge measurement. The extent of the difference also allows a statement about the reliability of the detection of the edge measurement.

(23) Instead of directly comparing the amplitudes with one another, for example evaluating their difference with a tolerance threshold, their ratio

(24) $v:=\frac{a_{f0}}{a_{f1}}$
can be considered. The more the amplitude ratio deviates from the value one, the higher the probability of an edge measurement. With the same amplitudes a.sub.f0=a.sub.f1 or the ratio having the value one, it is assumed that there was no signal mixing and thus no edge measurement, where random matches due to noise effects or particularly unfavorable phases are neglected, which theoretically still could occur.

(25) In a phase method, more than two modulation frequencies f.sub.i can be used, and most of the above considerations are already formulated in general terms. However, a more general measure must then be found to evaluate the differences in amplitudes a.sub.fi. Such a more general measure can of course also be applied to only two frequencies.

(26) As an exemplary value for three frequencies and, in a completely analogous manner, more than three frequencies, the value

(27) $w:=\frac{std\left(a_{\mathrm{f0}},a_{f1},a_{f2}\right)}{\mathrm{mean}\left(a_{f0},a_{f1},a_{f2}\right)}$
can be calculated, where mean is the arithmetic mean and std is the standard deviation. For similar amplitudes a.sub.fi, the value w is small, and it vanishes for an ideal non-edge measurement, accordingly w increases for amplitude values a.sub.fi different from one another. There are other measure functions than w for evaluating whether the amplitude values a.sub.fi are similar enough or whether there is an edge measurement. Both for the value in the numerator another statistical measure for a spread such as “maximum minus minimum” and for the value in the denominator another statistical measure of reference such as the median could be used.

(28) The respective extent of the deviation from a value for an ideal non-edge measurement, which is evaluated with one of the presented or another measure function, can be used as a quality or reliability criterion for the detection of an edge. Another conceivable criterion for reliability depends on the absolute value of the amplitudes. This is because the phase and amplitude measurement is less accurate at small absolute amplitudes, where deviations can be more likely due to noise.

(29) It is possible to demand a certain reliability and thus to evaluate a measured value as an edge measurement or a non-edge measurement in a binary fashion. However, a graded or continuous reliability value, especially normalized to [0,1], can also be useful. This can for example be used to filter edge measurements by excluding measured values that too likely were edge measurements from further processing. On the other hand, it is also possible to consider edge measurements as measured values, but with less weight according to their reliability. In this way, potentially erroneous measured values have a weaker impact on the final result without being discarded completely.

(30) FIG. 5 schematically shows some successively scanned measuring points 30 of a sensor 10, which for example is configured as a laser scanner. For each measuring point 30, the lateral position, for example as the angle from the encoder of the laser scanner, and the distance determined from the light time of flight using the phase method are known. During the scanning of object 20, the respective distances are measured, so that the contour and in particular the height and width of object 20 can be measured.

(31) At the edges of the object, edge measurements occur, and therefore the measuring points 30a are subject to a systematic measuring error. Conventionally, the contour of object 20 would have been determined using the distance values, as shown with dotted line 32. The measuring points 30a are distorted by the edge, but not so much that they would be detected as being edge measurements for this reason alone. Therefore, they are conventionally included and cause the height and width of object 20 to be overestimated.

(32) According to the invention, the edge measurements 30a are marked as edge measurements on the basis of their amplitudes by the method described above, regardless of the associated distance measurement value. Thus the angular position of object 20 is known without considering the measured distances. The distance values of the edge measurements 30a can be reconstructed from the vicinity, since the distances measured with the edge measurements 30a are not reliable. This results in the contour 34 shown with a solid line, which reproduces object 20 more accurately than conventionally. Accordingly, the height and width estimation of object 20 is also improved.

(33) Up to now, it has been assumed that the modulation frequencies f.sub.i are artificial modulation frequencies independent of the carrier frequency of the transmitted light 16. Alternatively, it is conceivable to directly modulate the light field intensity and thus to operate on the carrier frequency itself. For this purpose, it is necessary that the light transmitter 12 provides at least two different light frequencies f.sub.i. Instead of the conventional phase method described so far, phase and amplitude are now determined directly on the scales of the light frequency in a coherent measuring method, for example in an interferometric setup. The inherent unambiguity range in this case is determined by the very small light wavelength, but this can be compensated for by more frequencies f.sub.i or other means. The advantage is that the measurement accuracy can also reach the scale of light wavelengths.

(34) According to the invention, edges or edge measurements are detected based on the amplitudes. Other or supplementary criteria are conceivable. The measurement at each frequency f.sub.i results in a respective phase and thus a distance d.sub.i with a small unambiguity range. In order to combine a common distance value from this, one usually tries to find integer combinations matching the measured distances d.sub.i: Find d and m, n∈N.sub.0, so that d=d.sub.0+n.Math.λ.sub.0=d.sub.1+m.Math.λ.sub.1, where λ.sub.1 and λ.sub.2 are the wavelengths corresponding to the frequencies f.sub.0 and f.sub.1. In a modification, it is conceivable to try and find half-integer shifts, i.e. m,n∈{x+½|x∈N.sub.0}. Should half-integer results fit better than integer results, this indicates that the measured phase values of f.sub.0 and f.sub.1 are not consistent with each other, which for example could be caused by an edge measurement. These edge measurements could be treated in an analogue way as described above, or this criterion can be used to decide once again whether they are edge measurements.