The invention relates to a distance measuring device and a method for determining a distance with the distance measuring device .

Distances can be measured between a measuring device and an obj ect without a physical contact between the device and the obj ect by optical methods . In these methods, the obj ect is illuminated by the device and the light back reflected from the obj ect is then captured by a light detector of the device .

Distances can for example be determined by periodically

modulating the light intensity which is emitted from the device and by measuring the phase difference between the emitted light and the back reflected light arriving on the detector . However, due to the periodicity of the light intensity this method results in an ambiguous distance measurement . Unambiguous distance measurements can be determined by measuring the time of flight between the emission of light and the arrival of the back reflected light on the detector .

Conventional distance measurements are carried out by measuring a property of the light, in particular the intensity, as a function of time . Then a plot of the property versus the time is processed in order to obtain the time of flight . This processing can be computationally complicated and can therefore require a long time to be performed . If a distance measurement needs a long time to be performed this can cause a reduction of the repetition rate for taking the distance measurements .

The precision of the conventional distance measurements is limited by the size of the time steps , with which the property of the light is measured. Also for the conventional distance measurement, different reflectivities of the obj ect can lead to different shapes of the plot . When processing a different plot , this can lead to a different distance, so that distance depends on the reflectivity of the obj ect , which further decreases the precision for the conventional distance measurements.

It is an obj ect of the invention to provide a dis tance

measuring device and a method for measuring a distance with the distance measuring device, wherein the distance measurement is simple to perform whilst remaining precise .

The distance measuring device according to the invention for measuring a distance between the distance measuring device and an obj ect comprises a light source adapted to illuminate the obj ect with light pulses having different durations , at least one photo element adapted to capture the light pulses after being back reflected from the obj ect , a trigger generator for controlling the emission of the light pulses and for activating the photo element during a temporal integration gate having an integration start point in time Δ _s and an integration end point in time Δ _e , wherein the photo element is adapted to output a signal value U at the end of the integration gate with the signal value U depending on the energy of the light arriving on the photo element during its activation and wherein the trigger generator stores a trigger scheme to activate the photo element and to control the emission of the light pulses such that at least one short Light pulse with a duration T _p,s and a plurality of long light pulses with a duration T _p,1 being longer than T _p,s are emitted, that an invariable delay between the emission start point in time of the short light pulse and the

integration gate is such that Δ _tof and Δ _tof +Tp, _s are between Δ _s and Δ _e to output a reference signal value U _ref , with Δ _tof being the first point in time when the light pulse arrives on the photo element, and that for each long light pulse a respective variable delay τ between the emission start point in time of the long light pulses and the integration gate is such that the variable delays τ are different from each other in order to form a convolution function f _c := U ( τ ) out of the intensity of the light arriving on the photo element and the integration gate, and a processing unit adapted to identify the delay τ _c in the convolution funct ion which corresponds to U _ref and to calculate the distance by using τ _c . The method according to the invention for determining a

distance between the distance measuring device and an obj ect by means of the distance measuring device comprises the steps : a ) illuminating the obj ect with at least one short light pulse with the duration T _p,s ; b) illuminating the obj ect with a plurality of long light pulses with the duration T _p,1 ; c) outputting a signal value U _ref at the end of the integration gate, wherein an invariable delay between the emission start point in time of the short light pulse and the integration gate is such that Δ _tof and Δ _tof +T _p,s are between Δ _s and Δ _e ; d) forming a convolution function f _c : = U ( τ ) out of the intensity of the light arriving on the photo element and the integration gate with a respective variable delay i for each long light pulse between the emission start point in time of the long light pulses and the integration gate, wherein the variable delays are different from each other in order to form the convolution funct ion f _c ; e ) identifying the delay τ _c in the convolution function which corresponds to U _ref ; f ) calculating the distance by using the delay τ _c in the convo Lution function as identified in step f ) .

The convolution function f _c can be described by following equation :

wherein I (t ) is the intensity of the light of the long light pulses arriving on the photo element and g (t ) is the temporal integration gate . For early variable delays x with no overlap of the integration gate and the long light pulses arriving on the photo element, the convolution function has a stationary value. The function value begins to change as soon as the delay I is so long that the integration gate and the long light pulses begin to overlap . The convolution function comprises an extreme value at delays τ with a maximum overlap of the

integration gate and the long light pulses . The extreme value in the convolution function is a single point if the long light pulses and the integration gate have the same durations and is a plateau that becomes broader for an increasing difference in the durations of the long light pulses and the integration gate . By increasing the delay τ from the extreme value further, the function value develops back to the stationary value . The delay τ _c in convolution function f _c which corresponds to the reference signal value U _ref is the intersection of the

convolution function f _c and the function U=U _ref . The

intersection can for example be identified by forming the inverse function τ (U) of the convolution function f _c and then forming τ _c (U _ref ) , which is a mathematically simple method .

Alternatively, the intersection can be identified by

parametrizing the convolution function prior to measurement and performing fits to the measured data, before extracting the actual intersection analytically from the fitted convolution function and the function U=U _ref · By performing the fit it is possible to asses time steps between the measured data points, which provides an increased precision in the measurement of the delay τ _c . By obtaining the increased precision for x _c one also obtains an increasing precision for the distance . By

identifying the intersection of the function U=U _ref and the convolution function it is also achieved that different

reflectivities of the obj ect are compensated. The convolution function has two delays i _c , at which f _c = U _ref , one on each side of the extreme value . It is conceivable to form the convolution function only on one side of the extreme value and to identify only one delay τ _c or it is conceivable to form the convolution function on both sides of the extreme value and to identify both delays τ _c . If both delays are identified it is then possible to calculate a distance for each delay τ _c and it is then possible to form the average of both distances , advantageously increasing the accuracy of the distance measurement .

In order to arrange the integration gates with respect to the emission start point in time a distance range in which the obj ect can be located is predetermined . From the distance range , invariable delay can be chosen such that Δ _tof and Δ _tof +T _p,s are between Δ _s and Δ _e for all possible distances of the distance range. Also the invariable delays can be chosen such that the convolution function is formed.

It is preferred that the light source comprises light emitting diodes, VCSELs (vertical-cavity surface-emitt ing laser) and/or lasers that are in particular adapted to emit in the visible and/or infrared spectral region . The distance measuring device preferably comprises a CCD chip with an image intensifier and/or a CMOS chip that comprise the at least one photo

element .

It is preferred that the trigger scheme is adapted to control the emission of the light pulses such that the obj ect is illuminated alternating with the short light pulses and the long light pulses . Since the short 1 ight pulses are used for the reference signal value U _ref a possible long time drift in laser intensity would affect both the convolution function f _c and U _ref in the same manner, so that the long time drift would be compensated by the alternating short light pulses and long light pulses . The ratio of the number of the short light pulses to the number of the long light pulses is preferably from 0.2 to 0.4. Surprisingly, experimental results showed that this ratio resulted in the highest precision for the distances .

The trigger scheme is preferably adapted to control the

emission of the light pulses such that the intensity of the l ight pulses rises from an intensity I ₁ to an intensity I ₂ being higher than I ₁ at the emission start point in time and drops back to I ₁ after the durations T _p,s , and T _p,1 from the emission start point in time, respectively, wherein T _p,s , and T _p,1 are in order of tens of nanoseconds . Here, the extreme value of the convolution function is a maximum. Alternatively, the trigger scheme is preferably adapted to control the

emission of the light pulses such that the intensity of the light pulses drops from an intensity I ₂ to an intensity I ₁ being lower than I ₂ at the emission start point in time and rises back to I ₂ after the durations T _p,s , and T _p,1 from the emission start point in time, respectively, wherein T _p,s , and T _p,1 are in order of tens of nanoseconds . Here, the extreme value of the convolution function is a minimum. By using the light pulses that comprise the intensity drop at the emission start point in time, it is advantageously possible with the distance measuring device to both measure a distance and to illuminate the obj ect . The illumination of the obj ect can be such that the obj ect becomes visible for a human eye or for another vision system. Furthermore, it is not required to use an additional illumination system that would interfere with the distance measurement, whereby the precision for the distance measurement is high. It is preferred that in steps a ) and b) the obj ect is

illuminated alternating with the short light pulses and the long light pulses . The ratio of the number of the short light pulses to the number of the long light pulses is preferably from 0.2 to 0.4.

It is preferred that in step d) the convolution function is fitted to the plot of the signal values U _n versus the variable delay τ , wherein the convolution function f _c comprises in particular a linear function . By using the fit the convolution function f _c can be determined with an arbitrary step size, advantageously increasing the precision of the distance

measurement , independent on the number of different delays i between the emission start point in time of the long light pulses and the integration gate. Therefore, also the distance can be determined with an arbitrary step size . Since the delay T _c in the convolution function which corresponds to U _ref is identified, it is advantageously suff icient to fit only one linear function to the plot , which is computationally simple . This is not the case if for example an extreme value of the convolution function is identified. For identifying the extreme a respective linear function on both sides of the extreme value has to be fitted to the plot and the intersection of both linear functions has to be calculated,, which is computationally difficult .

It is preferred that in step d) the convolution function f _c is formed by first forming a coarse convolution function f _{c, coarse} with coarse steps of the different variable delays T _coarse , subsequently identifying in the coarse convolution function fc, coarse a coarse delay τ _c , coarse that corresponds to U _ref and the two variable delays τ _1,coarse and τ _r,coarse neighbouring τ _c,coarse , and then forming the convolution function f _c between τ _1,coarse and τ _{r, coarse} with fine steps having a shorter step size than the coarse steps. This provides an efficient method for determining the distance with a high precision .

The intensity of the light pulses preferably rises from an intensity I ₁ to an intensity I ₂ being higher than I ₁ at the emission start point in time and drop back to Ι ₁ after the durations T _p,s , and T _p,1 from the emission start point in time, respectively, wherein T _p,s , and T _p,1 are in the order of tens of nanoseconds . Alternatively, the intensity of the light pulses preferably drops from an intensity I ₂ to an intensity Ι ₁ being lower than I ₂ at the emission start point in time and rises back to I ₂ after the durations T _p,s , and T _p,1 from the emission start point in time, respectively, wherein T _p,s , and T _p,1 are in the order of tens of nanoseconds .

It is preferred that in step e) the average over a plural ity of reference signal values U _ref is used for identifying T _c , in particular over all the signal values U _ref . This results in a high precision for the signal values U _ref and therefore also in a high precision for the distance .

In the following the invention is explained on the basis of schematic drawings .

Figure 1 shows a schematic cross section through a distance measuring device, Figure 2 shows temporal profile diagrams with integration gates and intensities of light pulses , and

Figure 3 shows a section of a convolution function .

As it can be seen in Fi gure 1 a distance measuring device 1 comprises a light source 2 , a photo element 3, a trigger generator 4, a memory unit 5 and a processing unit 6. The 1 ight source 2 comprises light emitting diodes, VCSELs ( vertical- cavity surface-emitt ing laser) and/or lasers, wherein the light emitting diodes, VCSELs and/or the lasers are adapted to emit in the visible and/or infrared spectral region . The distance measuring device 1 comprises a CCD chip with an image

intensifier and/or a CMOS chip that comprise the at least one photo element 3 , wherein the CMOS chip comprises at least one condenser that can be discharged via a photodiode . The trigger generator 4 provides an activation signal 12 for controlling the emission of the light source 2 and an activation signal 13 for activating the photo element 3 during a temporal

integration gate 21. The CCD chip is activated by switching on the image intensifier and the CMOS chip is activated by closing a switch in the circuit of the condenser and the photodiode which allows that the condenser is discharged via the

photodiode . The photo element 3 is adapted to output a signal value U at the end of the integration gate 21 , wherein the signal value U depends on the energy of the light arriving on the photo element during its activation. The signal value U is readout in a readout operation 14 and stored in the memory unit 5. The memory unit 5 is adapted to store a multitude of signal values U . The multitude of the signal values U can then be processed by the processing unit 6 in a processing operation 15 in order to determine a distance between the distance measuring device 1 and the obj ect 9. Detection optics 8 are arranged in front of the photo element 3 in order to image a field of view 11 onto the photo element 3. I l l uminat ion optics 7 are arranged in front of the light source 2 in order to shape the light emitted by the light source 2 such that an illumination area 10 can be illuminated by the light source 2. The illumination area 10 and the field of view

11 are shaped such that the field of view 11 is substantially completely covered by the illumination area 10. The distance measuring device 1 is adapted such that the light emitted by the light source 2 impinges onto the obj ect 9 located within the fie Id of view 11 , and arrives on the photo element 3 after being back reflected from the obj ect 9. The il Lumination optics 7 and the detection optics 8 are preferably a respective lens . It is also possible to use a single lens for the both the illumination optics 7 and the detection optics 8.

In Figure 2 three temporal profile diagrams are shown, wherein an intensity 16 and a gate 17 is plotted versus time 18. The first temporal profile diagram is a plot of the intensity of the emitted light pulses 19 versus the time 18 , the second temporal profile time diagram is a plot of the intensity of the light pulses 20 arriving on the photo element 3 versus the time 18 , and the third temporal profile diagram is a plot of the integration gate 21 versus the time 18. The trigger generator 4 controls the emission of the light source 2 such that a

plurality of short light pulses 23 having a duration T _p,s and a plurality of long light pulses 24 having a duration T _p,1 is emitted, wherein T _p,s < T _p,1 . The light pulses 23, 24 in Figure 2 are such that at an emission start point in time of the 1 ight pulses 23, 24 their intensity is switched from I ₁ to I ₂ , wherein I ₂ > I ₁ . After the durations T _p,s , and T _p,1 from the emission start point in time, respectively, the intensity of the light pulses 23, 24 is switched back to I ₁ . In another embodiment the intensity of the light pulses 23, 24 is switched at the emission start point in time of the light pulses 23, 24 from I ₂ to I ₁ and after the durations T _p,s , and T _p,1 from the emission start point in time, respectively, back to I ₂ , wherein I ₂ > I ₁ . In both embodiments, I ₁ , and I ₂ , respectively, are the same for the short light pulses 23 and the long light pulses

24. Figure 2 shows that one respective short light pulse 23 and one respective long light pulse 24 are emitted alternating. In another preferred embodiment one respective short light pulse 23 and three respective consecutive long light pulses 24 are emitted alternating, so that the ratio of the number of the short light pulses 23 to the number of the long light pulses 24 is 1/3. As it can be seen in Figure 2 , after a duration Δ _tof from the emission start point in time the light pulses 23, 24 begin to arrive on the photo element 3. The integration gates 21 have an invariable delay to each of the emission start points in time of the short light pulses 23 , wherein the invariable delay is chosen such that the short light pulses 23 arriving on the photo element 3 are completely within the integration gate 21. The integration gates 21 have a variable delay τ to each of the emission start points in time of the long light pulses 24 , wherein the variable delay x is varied such that a convolution function f _c := U ( τ ) is formed out of the intensity of the light of the long light pulses 20 , 24 arriving on the photo element 3 and the integration gate 21. The convolution function f _c can mathematically be described by equation 1.

In Figure 2 , the variable delay τ is varied by choosing an invariable integrat ion start point in time Δ _s and an invariable integration end point in time Δ _e of the integration gates 21 from a start point in time 22 as well as by choosing a variable delay Δ _n of the emission start point in time of the long 1 ight pulses 24 to the start point in time 22 , wherein Δ _n is the delay for the n-th light pulse 23, 24 and Δ _n is different for each long light pulse 24. It is also conceivable to leave the delay from the emission start point in time of the long light pulses 24 to the start point in time 22 constant as well as to vary the integration start point in time Δ _s and the integration end point in time Δ _e . The start point in time 22 is chosen such that it coincides with Δ _s of the preceding integration gate 21. But it is also conceivable to choose any other point in time for the start point in time 22. In order to achieve that each short light pulse 23 is

completely within the integration gates 21 the invariable delay Δ _r of the emission start point in time of the short light pulses 23 from the start point in time 22 is chosen such that Δ _r +Δ _tof and Δ _r +Δ _tof +T _p,s are between Δ _s and Δ _e . Furthermore, it is required that the duration of the short light pulses T _p,s are shorter than the duration | Δ _e -Δ _s 1 of the integration gates 21 : Tp, _s < I Δ _e -Δ _s I . The duration | Δ _s -Δ _e I of the integration gates 21 is the same for both the short light pulses 23 and the long light pulses 24.

The hatched areas in Figure 2 are proportional to the energy of the light arriving on the photo element 3 during its

activation. A reference signal value U _ref being the average of all the signal values being output at the end of the

integration gates 21 for the short light pulses 23 is

determined. For each of the long light pulses 24 a respective signal value U is determined. Figure 3 illustrates the formation of the convolution function f _c and how the convolution function f _c is evaluated in order to determine the distance . For forming the convolution function f _c of Figure 3 the light pulses 23, 24 and integrations gates 21 according to Figure 2 were used . Figure 4 shows a plot of the signal values U determined at the end of the integration gates 21 for the long light pulses 24 versus the variable delay τ . In this plot the signal values U were taken only for delays i that correspond to shorter delays than a delay τ _max, xherein τ _max corresponds to a delay having a maximum overlap of the long light pulses 24 and the integration gate 21 and corresponds to a maximum in convolution function f _c . The convolution function fc comprises a linear funct ion 29 fitted to the plot of U versus τ . After fitting the linear function 29, a delay τ _c in the linear function 29 is identified which corresponds to U _ref . i _c corresponds to the intersection 30 of the linear function 29 with the function U=U _ref . For τ _c in Figure 3 it is: τ _c + T _p,s = Δ _tof + T _p,1 , whereby

In case that the signal values U were taken for delays x that correspond to longer delays than the delay

whereby

For both cases the distance r between the distance measuring device and the obj ect is then calculated by wherein c is the speed of light in the medium in which the distance measurement is carried out.

It is conceivable that the convolution function f _c is formed by first forming a coarse convolution function with coarse

steps of the different variable delays subsequently identifying in the coarse convolution function f _c ,coarse a coarse delay that corresponds to U _re f and the two variable delays neighbouring and then forming the convolution function f _c between with

fine steps having a shorter step size than the coarse steps .

List of reference signs

1 distance measuring device

2 light source

3 photo element

4 trigger generator

5 memory unit

6 processing unit

7 illumination optics

8 detection optics

9 obj ect

10 illumination area

11 field of view

12 activation signal for light source

13 activation signal for photo element

14 readout operation

15 processing operation

16 intensity

17 gate

18 time

19 intensity of emitted light pulses

20 intensity of light pulses arriving on the photo element

21 temporal integration gates

22 start point in time

23 short light pulse

24 long light pulse

29 linear function

30 intersection

Δ _r fixed delay

Δ ₀ , Δ ₂ , Δ ₄ , Δ ₆ variable delays

Δ _tof time of flight

T _p,s duration of short light pulse

T _p,1 duration of long light pulse

Δ _s integration start point in time

Δ _e integration end point in time

U signal value U _ref reference signal value

I ₁ lower intensity

I ₂ higher intensity