OPTICAL SENSOR

Description

[0001] The invention relates to an optical sensor comprising a transmitter unit, a receiver unit, and a control and evaluation unit. Furthermore, the invention relates to a sensor system and a method for calibrating an optical sensor.

[0002] When evaluating measurement signals, for example from a sensor, it is generally necessary to distinguish genuine measurement signals, which represent the actual measured quantity, from signals caused by random events, outliers, or disturbances, but not by the quantity actually being measured. Filtering methods are known to remove so-called false-positive signals from the acquired measurement signals in order to separate them from the genuine measurement signals, also known as true-positive signals.

[0003] A lidar system can, for example, measure the distance between an object and the system by determining the travel time of laser pulses emitted by the lidar system. To measure the travel time, the laser pulses reflected by the object are detected by a receiver unit of the lidar system. However, such a receiver unit is exposed to environmental conditions, including ambient light such as sunlight, which introduces a certain background level or offset into the lidar system's measurement signals. Furthermore, the measurement signals output by the lidar system's receiver unit generally contain a certain amount of noise, the intensity of which depends, for example, on the ambient temperature.

[0004] When established filtering methods for eliminating false-positive signals are applied to such a lidar system, a mean value is often calculated across the measured signals. This mean value corresponds to a background level and thus includes, for example, the contribution of ambient light. A difference is then defined by which a measured signal must exceed this mean to be recognized as a true positive signal. Such a suitable difference is defined, for example, as a predefined multiple of the standard deviation of the acquired signals. Alternatively, a Mahalanobis distance can be used as the difference from the mean.

[0005] However, such filtering methods, which use a predefined distance between a real measurement signal and a mean value, rely explicitly or implicitly on the assumption that the noise of the measurement signal follows a specific distribution, such as a Poisson distribution, a binomial distribution, or a normal distribution. This assumption is either explicitly present in the noise modeling or is implicitly taken into account through the calculation of the standard deviation, which presupposes a normal distribution.

[0006] In In practice, however, the noise of the measurement signals does not follow a specific distribution exactly. Therefore, the quality of the filtering method depends on the accuracy of the approximation to the assumed distribution. This can lead to variations in the quality of false-positive filtering under different environmental conditions, such as those found in lidar systems with varying background levels and noise intensities. The assumption that the noise follows a specific distribution is fulfilled to varying degrees in a lidar system for different background levels due to differing environmental conditions. Therefore, the so-called false-positive rate (FPR), i.e., the proportion of false positives, is a significant factor.The probability of undetected false-positive measurement signals, relative to the totality of measurement signals, is not constant for different environmental conditions, for example, of a lidar system.

[0007] To ensure that a specific false-positive rate is achieved regardless of environmental conditions, for example, 1%, known filtering methods often use a rather conservative or relatively large difference from the mean value that must represent a true positive measurement signal. However, this can conversely filter out an unnecessarily large number of true positive measurement signals from the acquired signals, especially if the measurement signal has a relatively high background level and a rather moderate signal-to-noise ratio.

[0008] US patent 2021 / 0088634 A1 describes an optical sensor with single-photon avalanche diodes that operate in two different modes. The first mode operates in a linear range below the breakdown voltage of the avalanche diodes. The intensity level of the incident light is measured by applying a current across the avalanche diodes. Based on this measured intensity, an overvoltage level is selected to be applied to the avalanche diodes in the second mode, where the avalanche diodes are operated above the breakdown voltage in a so-called detection mode.

[0009] US patent 2024 / 0067094 A1 describes a system with a clocked camera and a calibration light source.

[0010] A 3D monitoring sensor is described in EP 3 591 426 A1.

[0011] One object of the invention is to provide an optical sensor, a sensor system and a method for calibrating an optical sensor in which false-positive measurement signals are reliably detected regardless of the environmental conditions of the optical sensor.

[0012] This problem is solved by an optical sensor, a sensor system, and a method with the features of the independent claims. Advantageous embodiments of the invention are specified in the dependent claims, the description, and the drawings.

[0013] The optical sensor comprises a transmitter, a receiver, and a control and evaluation unit. The transmitter is designed to emit an optical signal into the vicinity of the optical sensor, while the receiver is configured to detect a reflected or remitted portion of the signal. The control and evaluation unit is designed to receive and store a calibration data set prior to each operating phase of the optical sensor. The calibration data set includes a variety of signal thresholds, each corresponding to a specific background level and a predetermined light intensity.

[0014] The control and evaluation unit is further configured to receive operational measurement signals from the optical sensor's receiver unit during the operating phase of the optical sensor when the transmitter unit is activated, to determine an instantaneous background level, and to select one of the signal thresholds from the calibration data set based on this instantaneous background level. This enables the control and evaluation unit to identify output signals derived from the operational measurement signals as invalid signals during the operating phase if the respective output signal is lower than the selected signal threshold.

[0015] If the optical sensor is designed, for example, as an optical scanner or laser scanner, objects in the vicinity of the optical sensor can be detected and their respective distances relative to the optical sensor can be determined. In this case, the transmitting unit can include a laser that emits short pulses into the vicinity of the sensor.

[0016] The output signal is derived from the operational measurement signals by, for example, counting these signals, which are assigned to a predetermined interval representing the travel time of the transmitted signal backscattered or reflected by an object until it reaches the receiving unit. These travel time intervals can further correspond to intervals for distances between the optical sensor and the respective object, i.e., so-called distance bins. In this case, the output signal can include one count per distance bin.

[0017] The calibration data set, which is received and stored by the control and evaluation unit, is generated during a calibration phase of the optical sensor. In this phase, for example, the optical sensor's transmitter is deactivated and the receiver receives the respective predetermined light intensity. However, it is not strictly necessary to deactivate the transmitter during the calibration phase. The only requirement for generating the calibration data set is that no light from the transmitter falls directly or indirectly (i.e., through scattering, reflection, or remission) onto the receiver. In other words, the calibration data set is created without any light emitted by the transmitter and scattered, reflected, or remitted in the vicinity of the optical sensor reaching the optical sensor's receiver.This can also be achieved, for example, by appropriately covering, darkening or hiding the transmitting unit during the calibration phase.

[0018] The calibration dataset comprises not just one signal threshold, but a multitude of signal thresholds, each corresponding to a specific background level during the operation of the optical sensor. The specific signal threshold used to determine whether the output signal is invalid or valid (i.e., a false positive or true positive) thus depends on the current background level during the optical sensor's operating phase and is derived from the calibration dataset.

[0019] The calibration data set can also be determined by defining or predetermining light intensities that correspond to or are assigned to the respective background levels. The predetermined light intensities can be selected in accordance with expected background levels or those relevant to the operating phase in order to illuminate the receiving unit with these predetermined light intensities and to determine the signal thresholds based on the signals detected by the receiving unit, for example, with the transmitting unit deactivated.

[0020] The signal thresholds ultimately serve to filter the output signal for false positives that are lower than the respective signal threshold for the current background level. Output signals identified as invalid can either be removed or marked before output.

[0021] The current background level can be determined either from instantaneous operational measurement signals with the transmitter activated, for example, by calculating a mean or median across all distance bins or a subset of distance bins where there is a high probability of no signal or echo being generated by an actual object. Alternatively, a second measurement can be taken immediately after each measurement with the transmitter activated, with the transmitter deactivated, to determine the instantaneous background level based on this second measurement. In this case, the operational measurement signals encompass both the first measurement with the transmitter activated and the second measurement with the transmitter deactivated, since both measurements are performed during the optical sensor's operating phase.The control and evaluation unit thus determines the current background level in both cases based on operational measurement signals that are recorded during the operating phase.

[0022] The respective signal thresholds can be set in such a way that a defined and constant false positive rate (FPR) is achieved, independent of the background level or the environmental conditions of the optical sensor. Conversely, a desired FPR can first be defined, for example, 1%, from which the respective thresholds for the predetermined light intensities or the corresponding background levels are then determined. A predetermined or desired FPR can thus define the conditions for determining the thresholds.

[0023] The respective signal thresholds are therefore not based on assumptions regarding the distribution of noise in the calibration measurement signals. Instead, while the signal thresholds do depend on the actual noise distribution, their determination requires no knowledge of the shape of the actual noise distribution. In other words, no models of the distribution, shape, and / or time course of the calibration measurement signals are necessary to determine the signal thresholds.

[0024] During the operational phase of the optical sensor, systematic errors can therefore be avoided. These errors can occur when the false positive rate varies depending on the background level, for example, when using a constant multiple of the standard deviation of the operational measurement signals or an output signal derived from them. Instead, an output signal can be reliably identified as an invalid or false positive, i.e., with a constant false positive rate independent of the background level or environmental conditions. This improves the overall reliability of the output signal filtering.Conversely, there is no unnecessary filtering out of true positive output signals, since the respective signal thresholds are adapted to the respective background levels and the appropriate signal threshold is selected that corresponds to the current background level.

[0025] Furthermore, the optical sensor is configured to be connected, during the calibration phase prior to the operational phase, to a calibration unit that includes an illumination unit illuminating the receiving unit with the respective predetermined light intensity corresponding to the respective background level. The control and evaluation unit is further configured to acquire calibration measurement signals from the receiving unit for each predetermined light intensity during the calibration phase, enabling a processing unit to generate the calibration data set based on these signals.

[0026] The processing unit can be part of the calibration device, or it can be integrated into the control and evaluation unit of the optical sensor. In this case, the processing unit is only active during the calibration phase of the optical sensor.

[0027] During the calibration phase, the calibration device for the optical sensor's receiver unit uses the illumination unit to simulate the respective background levels and corresponding noise generated during operation of the optical sensor by ambient light and other influences in the respective environmental conditions, based on predetermined light intensities. To determine the predetermined light intensities, relevant background levels for the operation of the optical sensor can first be identified. These levels depend, for example, on the intended operating environment of the optical sensor and can cover a possible measurement range. In other words, all relevant background levels for the operation of the optical sensor can be determined beforehand.

[0028] Subsequently, the predetermined light intensities for the optical sensor's calibration phase are set to correspond to the determined relevant background levels. Because the optical sensor's transmitter is deactivated during the calibration phase, the calibration measurement signals represent only the respective background level for each predetermined light intensity and the corresponding noise, i.e., fluctuations around the respective background level.

[0029] Signals can be derived from the calibration measurement signals in a similar manner to how output signals derived from operational measurement signals were described above. Specifically, calibration measurement signals that fall within a certain amplitude range can be counted, and these counts can be assigned to the respective distance bins. The signal thresholds can then be determined based on these derived signals.

[0030] A specific signal threshold for each predetermined light intensity can be defined, for example, based on one or more maxima of the signals derived from the calibration measurement signals. The output signal derived from the operational measurement signals is only identified as a valid signal if, for the corresponding instantaneous background level, it is greater than or equal to these maxima of the calibration measurement signals, i.e., greater than or equal to the corresponding signal threshold for the instantaneous background level.

[0031] Furthermore, the optical sensor is designed for time-correlated single photon counting (TCSPC). The control and evaluation unit is also designed to generate histograms of the time-correlated single photon count (TCSPC histograms) for both the calibration measurement signals and the operational measurement signals.

[0032] An optical sensor with time-correlated single-photon counting, for example, periodically emits light pulses, typically a few nanoseconds long, which define the start time of each measurement. During the time interval until the next light pulse, the light reflected or backscattered by an object in the vicinity of the optical sensor is detected by the sensor's receiver. The time interval between two light pulses emitted by the transmitter can be divided into a multitude of short time segments, such as 500 ps long. Each time segment can be assigned a point in time corresponding to the time interval from the start time of the measurement, at which the last light pulse was emitted by the transmitter.

[0033] Depending on the distance to the object in the vicinity of the optical sensor, the reflected or backscattered light pulse reaches the optical sensor's receiver at different times. In other words, the emitted light pulse has different travel times depending on the object's distance from the receiver. Upon detecting each reflected or backscattered light pulse, the receiver can generate an electrical signal, which can be assigned to one of the time intervals between two light pulses using a time-to-digital converter (TDC).

[0034] When the optical sensor's transmitter emits a large number of light pulses, the electrical signals or events generated in the receiver by reflected or backscattered light pulses, and assigned to a specific time interval, can be counted for the multitude of emitted light pulses. The counted electrical signals or events over the respective time intervals form a time-correlated single-photon count histogram (TCSPC histogram), which can be represented, for example, by digital signals in a memory of the control and evaluation unit.Since the time intervals of the TCSPC histogram are assigned to different travel times of the light pulses to an object in the vicinity of the optical sensor, and thus to different distances relative to the optical sensor, the time intervals that divide the time between each pair of light pulses from the transmitter unit correspond to respective distances relative to the optical sensor. Therefore, these time intervals are also referred to as distance bins, and counts can be represented in a TCSPC histogram using distance bins.

[0035] The respective TCSPC histograms are generated not only during the operating phase of the optical sensor, but also during the calibration phase using the calibration measurement signals, i.e., without the optical sensor's transmitting unit being actively involved in the measurement. During the calibration phase, the TCSPC histograms can be generated using the calibration measurement signals even without emitted light pulses from the transmitting unit, because the illumination unit of the calibration device can illuminate the optical sensor's receiving unit with a constant light intensity during the calibration phase. This intensity can, for example, correspond to the expected ambient light under specific conditions during the operating phase.

[0036] The TCSPC histograms thus represent, for the calibration phase, the signals derived from the calibration measurement signals, which are used to determine the respective signal thresholds, and for the operational phase, the output signals derived from the operational measurement signals. These output signals are subsequently compared with the respective signal thresholds to identify them as valid or invalid signals. The valid output signals can also be referred to as echoes, since they are ultimately caused by actual reflection or backscattering from an object.

[0037] In During the calibration phase, the background and noise components of the respective TCSPC histograms can be simulated using the respective predetermined light intensity for the corresponding background level. The respective signal thresholds for the operating phase can thus be directly determined from the TCSPC histograms of the calibration phase and essentially read directly from the histograms. To define the respective signal thresholds, for example, one or more maxima of the TCSPC histograms generated during the calibration phase for the respective predetermined light intensity or the corresponding background level can be determined.

[0038] The control and evaluation unit is further configured to generate a multitude of TCSPC histograms for each predetermined light intensity during the calibration phase of the optical sensor. Furthermore, the processing unit is configured to determine a maximum count value or histogram value for each of the TCSPC histograms associated with the respective predetermined light intensity during the calibration phase, in order to create a statistical distribution of the maximum count values ​​for the respective predetermined light intensity and to determine the signal thresholds for the respective predetermined light intensities based on this statistical distribution.

[0039] Since the respective signal thresholds depend on the statistical distribution of the maximum count values ​​from a multitude of histograms, the evaluation of the operational measurement signals during the optical sensor's operating phase can be carried out reliably, as a valid signal must be greater than the respective signal threshold and thus greater than a certain proportion of the maximum count values ​​in the statistical distribution. Instead of a single maximum count value, multiple count values ​​can also be used, corresponding, for example, to the highest, second highest, third highest, etc., count value.

[0040] The processing unit can further be configured to determine, during the calibration phase of the optical sensor, a cumulative relative frequency of the maximum count values ​​for each predetermined light intensity based on the statistical distribution, and to determine the respective signal thresholds for the predetermined light intensities based on the respective cumulative relative frequency. The cumulative relative frequency can, for example, be displayed as a function of the maximum count values ​​and can thus represent an efficiently usable transformation of the statistical distribution.

[0041] From such a representation of the cumulative relative frequency, the respective signal thresholds can be determined, for example, as those maximum count values ​​at which the respective cumulative relative frequency reaches a certain value, for example 99%, so that conversely, the cumulative relative frequency for invalid output signals that are not detected in the operating phase is only 1 minus the cumulative relative frequency at the maximum count value, for example 1%.

[0042] The processing unit can further be configured to use at least a predetermined percentage of false-positive or invalid output signals during the optical sensor's calibration phase to determine the respective signal thresholds for the associated background level based on their respective cumulative relative frequencies. In such an embodiment, the false-positive rate can thus be predetermined based on the specified percentage of false-positive output signals, so that the respective signal thresholds for all background levels correspond to the predetermined false-positive rate.In other words, the respective signal thresholds can be determined based on the respective cumulative relative frequency of the maximum count values ​​in such a way that the proportion of filtered false-positive output signals, which are thus recognized as invalid output signals, is constant regardless of the assigned background levels.

[0043] Furthermore, the processing unit can be additionally configured to use several predefined percentages—that is, more than one predefined percentage—for false-positive output signals during the optical sensor's calibration phase. This allows for the determination of a set of corresponding signal thresholds for each predefined percentage, based on the respective cumulative relative frequency of each false-positive output signal. For example, it can be specified that 0.1%, 1%, 5%, and 10% of the operational measurement signals must not be detected as false-positive or invalid output signals during the operational phase. Therefore, the predefined percentage for false-positive output signals to be filtered would be 99.9%, 99%, 95%, and 90%, respectively. Corresponding sets of signal thresholds can then be determined based on the respective cumulative relative frequency of each associated background level.as the maximum count value with this respective percentage share from the distribution of cumulative relative frequencies.

[0044] During the operating phase of the optical sensor, the operating signals can be validated and marked based on whether they are above, for example, one, two, three, or all signal thresholds associated with the current background level during the operating phase. This can enable flexible further processing of the optical sensor's output signals, for example, when algorithms that process the optical sensor's output signals require varying degrees of filtering for false-positive output signals.

[0045] The processing unit can further be configured to generate the calibration data set during the optical sensor's calibration phase in the form of at least one lookup table containing the respective signal thresholds and their associated background levels. In other words, the result of the calibration phase can be presented as a lookup table that can be directly used to evaluate the current operational measurement signals during the optical sensor's operation phase. If several percentages of false-positive output signals are specified to determine respective sets of signal thresholds for the associated background levels, the control and evaluation unit can generate multiple lookup tables during the calibration phase—that is, one lookup table for each of the specified percentages.Representing the signal threshold values ​​using at least one lookup table can require minimal computer effort, for example, little storage space and uncomplicated computational processing.

[0046] Additionally, the processing unit can be configured to approximate the lookup table using an analytical function. This can further reduce the computational effort required during the optical sensor's operational phase to filter out invalid output signals from the operational measurement signals. The analytical function for approximating the lookup table can include one or more parameters and be represented, for example, as a polynomial or Laurent series. Approximating the lookup table using the analytical function can be particularly relevant when considering a large number of possible background levels, which can increase the required memory for the lookup table and its processing time. Using the approximated analytical function, the information regarding signal thresholds obtained during the calibration phase can be stored in a compressed format.

[0047] According to another embodiment, the optical sensor can be configured as a lidar sensor with time-correlated single-photon counting (TCSPC). In this embodiment, the transmitting unit can include a laser that emits short light pulses into the vicinity of the optical sensor and can scan this environment as a scanning laser. The TCSPC histograms generated by the control and evaluation unit of such a lidar sensor include the aforementioned distance bins, which correspond to a respective transit-time interval for the light pulses emitted by the lidar sensor's laser. In other words, the TCSPC histograms of such a lidar sensor can include respective count values ​​for each distance bin.

[0048] For this embodiment, the processing unit can further be configured to determine a set of signal thresholds for several distance ranges within the lidar sensor's range during the calibration phase. The different sets of signal thresholds for these multiple distance ranges may be due to the fact that the lidar sensor with time-correlated single-photon counting exhibits a distance-dependent noise characteristic. The noise and the corresponding background level can therefore have different values ​​for different distances to objects from which the lidar sensor's light pulses can be reflected. It can therefore be advantageous to divide the lidar sensor's range into several distance ranges or groups of adjacent distance bins and to determine a set of signal thresholds for each of these.These multiple sets of signal threshold values ​​can then be represented in a two-dimensional lookup table and stored in a memory of the optical sensor for use during its operational phase.

[0049] According to a further embodiment, the receiving unit of the optical sensor can comprise several groups of pixels, each of which may exhibit different noise characteristics. In this case, the processing unit can further be configured to determine a respective set of signal thresholds for each of these groups of pixels during the calibration phase of the optical sensor.

[0050] For this purpose, during the calibration phase of the optical sensor, the processing unit can control the illumination unit of the calibration device with the optical sensor's transmitter unit deactivated, such that the illumination unit illuminates the receiver unit with several predetermined light intensities and, for each predetermined light intensity, captures calibration measurement signals from the optical sensor's receiver unit for each group of pixels in order to determine a respective signal threshold value for each pixel group based on the respective calibration measurement signals and to assign the respective signal threshold value for each pixel group to one of the background levels for the optical sensor.

[0051] The result of such a calibration phase, i.e., the respective sets of signal threshold values ​​per pixel group, can in turn be created as a two-dimensional lookup table and stored as a calibration data set in the optical sensor's memory. In the extreme case, calibration can be performed per pixel, meaning that in such a case, each group of pixels can comprise only a single pixel.

[0052] According to a further embodiment, the control and evaluation unit can also be configured to determine the instantaneous background level as a mean or median over at least a predetermined proportion of the operational measurement signals during the operating phase of the optical sensor. If the optical sensor is configured as a time-correlated single-photon counting sensor, the mean can be determined over an entire TCSPC histogram acquired during the operating phase of the optical sensor. The instantaneous background level can then be determined with minimal effort using such a mean. Alternatively, the instantaneous background level can also be determined in another way, for example, by separately acquiring ambient light when the transmitter is briefly deactivated.

[0053] A further aspect of the invention is a sensor system comprising an optical sensor as described above and a calibration device. The calibration device communicates with the optical sensor during a calibration phase and is configured to generate a calibration data set for the optical sensor. The calibration data set comprises a plurality of signal thresholds, each corresponding to a specific background level and a predetermined light intensity.

[0054] The sensor system thus comprises the optical sensor and the calibration device described above. The descriptions of the optical sensor and the calibration device apply accordingly to the sensor system, particularly with regard to advantages and preferred embodiments.

[0055] A further aspect of the invention is a method for calibrating an optical sensor comprising a transmitter and a receiver unit, which is connected to a calibration device during a calibration phase. The method includes varying the light intensity of an illumination unit of the calibration device, for example, with the transmitter unit of the optical sensor deactivated, such that the receiver unit is illuminated with several predetermined light intensities. For each predetermined light intensity, calibration measurement signals from the receiver unit are acquired, a respective signal threshold is determined based on these calibration measurement signals, and the respective signal threshold is assigned to a background level for the optical sensor that corresponds to the respective predetermined light intensity.The signal thresholds are provided as a calibration data set for an operating phase of the optical sensor with the transmitter unit activated, so that a control and evaluation unit of the optical sensor is able during the operating phase to select one of the signal thresholds of the calibration data set based on an instantaneous background level determined from operational measurement signals acquired by the optical sensor's receiver unit during the operating phase, and then to identify output signals derived from the operational measurement signals with the transmitter unit activated as invalid signals if the respective output signal is smaller than the selected signal threshold.

[0056] The method can be carried out using the sensor system described above. The descriptions of the sensor system, the optical sensor, and the calibration device apply accordingly to the method, particularly with regard to advantages and preferred embodiments. Furthermore, it is understood that all features mentioned herein are combinable unless explicitly stated otherwise.

[0057] As explained above, the instantaneous background level during the operating phase can be determined either directly from measurement signals when the optical sensor's transmitter is activated, or by immediately following each measurement with the transmitter activated with a second measurement taken with the transmitter deactivated, thus establishing the instantaneous background level. In both cases, the instantaneous background level is therefore determined from operational measurement signals acquired during the operating phase.

[0058] Furthermore, the optical sensor is designed for time-correlated single-photon counting (TCSPC). Histograms of the time-correlated single-photon count (TCSPC histograms) are generated for both the calibration measurement signals and the operational measurement signals.

[0059] For the calibration measurement signals, a large number of TCSPC histograms are generated at each predetermined light intensity. For each of these TCSPC histograms, which are assigned to the respective predetermined light intensity, a maximum count value is determined in order to establish a statistical distribution of the maximum count values ​​for the respective predetermined light intensity. The signal thresholds for the respective predetermined light intensities are then determined based on this statistical distribution.

[0060] Furthermore, the statistical distribution allows for the determination of a cumulative relative frequency of the maximum count values ​​for each predetermined light intensity. The respective signal thresholds for the predetermined light intensities can also be determined based on these cumulative relative frequencies.

[0061] The invention is described below by way of example with reference to an advantageous embodiment and the accompanying figures. These show, schematically: Fig. 1 shows an optical sensor according to the invention in an operating phase and in a calibration phase, Fig. 2 shows two different scenarios in which there is an almost identical background level, but different noise characteristics of the optical sensor, and Fig. 3 shows several diagrams illustrating the calibration phase of the sensor system.

[0062] Fig. 1 Figure 1 shows a schematic representation of an optical sensor 110. The optical sensor 110 is in Fig. 1A in an operational phase and in Fig. 1B shown in a calibration phase in which the optical sensor 100 is integrated into a sensor system 100 which additionally includes a calibration device 150.

[0063] The optical sensor 110 is designed as a lidar sensor that uses the measurement principle of time-correlated single photon counting (TCSPC). The optical sensor / lidar sensor 110 therefore comprises a transmitter unit 112 with a laser that periodically emits light pulses into the vicinity of the lidar sensor 110. The periodic light pulses of the laser in the transmitter unit 112 are in Fig. 1A schematically represented as transmission signal 124. The periodic light pulses are a few nanoseconds long, i.e., shorter than, for example, 10 ns, and define a start time of each measurement by the lidar sensor 110.

[0064] The lidar sensor 110 further comprises a receiver unit 114, which is equipped with an array of multiple pixels. The receiver unit 114 detects those light pulses 124 that are reflected or backscattered by an object 130 in the vicinity of the lidar sensor 110 and accordingly arrive at one of the pixels of the receiver unit 114 as reflected or backscattered light pulses 126. The lidar sensor 110 also comprises a control and evaluation unit 116, which controls the transmitter unit 112 and its laser by means of control signals 122 and also acquires operating measurement signals 128 from the receiver unit 114. From the operating measurement signals 128, the evaluation unit 116 derives output signals 160 of the lidar sensor 110, as explained in detail below.

[0065] The transmitter unit 112 of the lidar sensor 110 is further designed to deflect the emitted laser beam or the transmission signal 124 such that the transmission signal 124 covers and scans a certain spatial area. For this purpose, the control and evaluation unit 116 outputs a corresponding control signal, which is contained in the control signals 122 and is received by the transmitter unit 112. The deflection of the transmission signal 124 and the corresponding scanning of a spatial area are described in Fig. 1A schematically represented by arrow 115.

[0066] The time interval between two light pulses emitted by the laser of transmitter unit 112 can be subdivided into a multitude of short time segments, each, for example, 500 ps long. Each of these time segments can also be assigned a point in time that corresponds to a time interval from the start time of the respective measurement, at which the last light pulse was emitted by the laser of transmitter unit 112.

[0067] Depending on the distance of the object 130 relative to the lidar sensor 110, the reflected or backscattered light pulse 126 reaches the receiver 114 of the lidar sensor 110 at different times, since the light pulses 124 and 126 have different travel times from the start of the measurement until their detection in the receiver 114, depending on the distance of the object 130. Upon detection of each reflected or backscattered light pulse 126, the pixels of the receiver 114 generate an electrical signal as the respective operating measurement signal 128, which is assigned to one of the time segments within the time interval between two light pulses by means of a time-to-digital converter (TDC) in the control and evaluation unit 116.

[0068] During the operating phase of the optical sensor or lidar sensor 110, the transmitter unit 112 emits a multitude of light pulses by means of its laser. The corresponding electrical signals or events, generated in the receiver unit 114 by reflected or backscattered light pulses 126 and assigned to a respective time interval between two light pulses, are counted for this multitude of emitted light pulses 124. The assignment of the counted electrical signals or events, which are output as an operating measurement signal 128 by the receiver unit 114 and recorded by the control and evaluation unit 116, takes place in the control and evaluation unit 116 such that the respective count value per time interval is internally represented in the form of a histogram of the time-correlated single-photon count (TCSPC histogram).

[0069] Since the time intervals of the TCSPC histogram correspond to different travel times of the light pulses to object 130 and back to the lidar sensor 110, and thus to different distances relative to the lidar sensor 110, the respective time intervals into which the time interval between two light pulses of the laser of the transmitter unit 112 is divided can be assigned to corresponding distances relative to the lidar sensor 110. Because the different travel times of the light pulses 124 and 126 can thus be converted into corresponding distances, and the time intervals correspond to these distances, the count values ​​of the TCSPC histogram are usually plotted for respective so-called distance bins, each of which covers a corresponding distance interval that corresponds to a respective travel time interval of the light pulses 124 and 126.

[0070] Two examples of TCSPC histograms are shown in the two diagrams above. Fig. 2 schematically represented, which with Fig. 2A und Fig. 2B are designated. In the two TCSPC histograms of Fig. 2A und Fig. 2B The respective count values ​​(210) on the y-axis are displayed over a multitude of distance bins (220) on the x-axis. Both histograms of Fig. 2A und 2B The histograms show a background level or mean noise value, which is approximately 10 for both histograms. This background level exists due to the environmental conditions of the lidar sensor 110, for example, due to ambient light from an ambient light source 140 (see...). Fig. 1 ) is emitted and detected by the pixels of the receiving unit 114. The background level is therefore a specific count value 210, which is valid for all distance bins and to which a specific intensity of the ambient light is assigned.

[0071] The histograms of Fig. 2A und Fig 2B These represent two different scenarios in which different internal properties or configurations of the optical sensor 110 lead to significantly different noise characteristics. At the same background level or mean noise value, the two histograms of Fig. 2A und Fig. 2B Therefore, significantly different noise levels are observed, i.e., strongly different fluctuations relative to the background level or the mean noise value. As explained in detail below, this is therefore important for the histograms of Fig. 2A und Fig. 2B It is of varying difficulty to decide whether a particular count value 210 is generated by a reflection from an object and is therefore a true positive echo, or whether this count value 210 was caused by extraneous light and is therefore a false positive echo that is filtered out or included in the output signal 160 (cf. Fig. 1 ) should be marked.

[0072] This is also reflected in the two lower diagrams of Fig. 2 against, those with Fig. 2C und 2D are labelled and in which the respective relative frequencies 230 are plotted on the y-axis against the respective count values ​​210 on the x-axis. Fig. 2C This thus represents the relative frequencies of the count values ​​210 for the histogram of Fig. 2A , while Fig. 2D the respective relative frequencies for the count values ​​210 of Fig. 2B represented.

[0073] The distributions of the count values ​​210 show in the representations of Fig. 2C or 2D each reached approximately the same maximum value of around 10, corresponding to the background level or average noise level of Fig. 2A und 2B . According to the strongly different noise levels in the histograms of Fig. 2A und 2B The two distributions of Fig. 2C und 2D However, they have a very different width.

[0074] When TCSPC histograms are evaluated to identify objects 130 and determine their distance relative to the lidar sensor 110, it must be possible to distinguish whether a specific count value of a distance bin actually originates from a light pulse 124, 126 emitted by the laser of the transmitting unit 112 and reflected or backscattered as a light pulse 126 by an object 130, or whether the corresponding count value of the distance bin is merely caused by another event, such as ambient light. The quality of the evaluation depends strongly on the signal-to-noise ratio in the respective histograms, as described in Fig. 2A und 2B This illustrates the point.

[0075] In both histograms, the two distance bins with the largest count values ​​are marked with 240, the possible output signals 160 of the lidar sensor 110 (see below). Fig. 1 ). Due to the noise and the relatively high background level, both histograms can be viewed as follows: Fig. 2A und 2B It cannot be readily determined whether the count values ​​at 240 actually originated as an echo of a light pulse 124 at an object 130 or not. However, when evaluating the histograms, it must be possible to ascertain whether, for example, the count values ​​at 240 are genuine output signals 160 (cf. Fig. 1 ) or true positive output signals 160 or invalid output signals 160 or false positive output signals 160. The goal of evaluating such histograms of Fig. 2A und Fig. 2B The goal is to achieve the lowest possible proportion of false-positive output signals 160, for example 1%, that are not identified and filtered out. This means that, on average, only one false-positive output signal 160 is contained in one hundred output signals 160, which are ultimately processed by the control and evaluation unit 116 (see...). Fig. 1 ) will be issued.

[0076] In established evaluation methods, a suitably assumed distance measure of a count value (210) relative to the respective mean of the histogram is usually used to determine whether the respective count value (210) is a false positive or a true positive. Such a distance measure is typically a multiple of the standard deviation relative to the respective mean or background level, or a Mahalanobis distance relative to the background level. However, using the standard deviation or the Mahalanobis distance relative to the mean presupposes that the distribution of the count values ​​follows a specific model or distribution, such as a Poisson distribution, a binomial distribution, or a normal distribution.

[0077] As shown in the diagrams of Fig. 2C und 2D However, it is evident that the distributions of the relative frequencies 230 for the count values, while similar to a normal distribution, are not exactly symmetrical. Consequently, an analytical description of these distributions, for example using a normal distribution, can be flawed. Furthermore, the use of a multiple of the standard deviation, which ultimately presupposes a normal distribution, can lead to varying degrees of filtering of the count values ​​in the histograms at different background levels, since the approximation to a normal distribution can differ depending on the respective background level. This results in a so-called false positive rate (FPR), i.e., the proportion of undetected false-positive output signals 160, which can vary considerably for different background levels.

[0078] This problem can be illustrated by the diagrams of Fig. 2 To illustrate. For example, if a count value of 15 is one of the maximum values ​​in the histograms of Fig. 2A und Fig. 2B To determine whether this count value originates from a real object 130 or not, the control and evaluation unit 116 would have to be able to automatically decide. In the example of Fig. 2A und 2C A value of 15 already lies in an outlier or, so to speak, outside the distribution of relative frequencies 230, while this value is for the histogram of Fig. 2B and the distribution of Fig. 2D still lies at a relative frequency within the statistical distribution, i.e., not yet far enough into a tail of the distribution. In the example of Fig. 2A und 2C It is therefore relatively likely that a count value of 15 originates from a real object 130 and is thus a true-positive output signal 160, while such a value for the example of Fig. 2B und 2D with a relatively high probability it can be considered a false positive output signal 160.

[0079] The in Fig. 1B The sensor system 100 shown and a corresponding calibration procedure are therefore intended for a calibration phase of the optical sensor or lidar sensor 110. During the calibration phase, a calibration data set 170 is generated for the control and evaluation unit 116 of the optical sensor 110. Using the calibration data set 170, the control and evaluation unit 116 is able to filter the signals of the lidar sensor 110 or the TCSPC histograms with respect to false-positive output signals 160 in such a way that an assumption regarding, for example, a specific distribution relative to a mean value, such as a normal distribution, is not required. Furthermore, it should be possible to specify a nearly constant false-positive rate, independent of the respective background level or the environmental conditions of the lidar sensor 110.

[0080] The sensor system 100 therefore includes, in addition to the lidar sensor 110, a calibration device 150 (see Fig. 1B ), which is connected to the Lidar sensor 110 during the calibration phase. As in Fig. 1B As shown schematically, the transmitting unit 112 of the optical sensor or lidar sensor 110 is deactivated during the calibration phase. Instead, the receiving unit 114 is illuminated by a lighting unit 152 of the calibration device 150, which provides several predetermined light intensities during the calibration phase.

[0081] The lighting unit 152 can either be an external light source that generates the various predetermined light intensities itself. However, the lighting unit 152 can also simply provide a constant light intensity from an external light source, such as the ambient light source 140 (see...). Fig. 1A ), modulate or shade the light source to generate the various predetermined light intensities for the receiving unit 114. For example, the lighting unit 152 can only shade the sunlight, so that it does not have its own light source.

[0082] Despite the laser of the transmitting unit 112 being deactivated, the receiving unit 114 outputs calibration measurement signals 154 during the calibration phase. These signals are determined by the respective light intensities provided by the illumination unit 152. The control and evaluation unit 116 of the lidar sensor 110 acquires the calibration measurement signals 154 for the respective predetermined light intensities of the illumination unit 152 and generates, as in the Fig. 1A The operating phase of the lidar sensor 110 is shown in the corresponding TCSPC histograms.

[0083] A processing unit 156 of the calibration device 150 is communicatively connected to the control and evaluation unit 116 of the lidar sensor 110. During the calibration phase of the lidar sensor 110, the processing unit 156 receives the TCSPC histograms, which the control and evaluation unit 116 generates based on the calibration signals 154, and evaluates them as described below to generate the calibration data set 170, which the lidar sensor 110 uses in the operating phase. At the end of the calibration phase, the processing unit 156 transmits the calibration data set 170 to the control and evaluation unit 116, which stores the calibration data set 170 so that it is available during the operating phase of the lidar sensor 110.

[0084] The processing unit 156 is indeed in Fig. 1B The processing unit 156 is shown as part of the calibration device 150 outside the lidar sensor 110. However, the processing unit 156 can also be integrated into the control and evaluation unit 116 and thus be a component of the lidar sensor 110. In this case, the processing unit 156 within the control and evaluation unit 116 is only activated when the lidar sensor 110 is in the calibration phase.

[0085] The processing unit 156 comprises a control module 158, which communicates with the illumination unit 152 to control the respective light intensity with which the illumination unit 152 illuminates the receiver unit 114 of the lidar sensor 110. The control module 158 thus specifies a respective light intensity for the illumination unit 152 and therefore a corresponding background level. The receiver unit 114 generates the respective calibration measurement signals 154 for this background level, based on which the control and evaluation unit 116 in turn generates corresponding TCSPC histograms (see Figure 1). Fig. 3 ). This is then used by the processing unit 156 to determine the respective signal threshold value that is assigned to the background level as an element of the calibration data set 170.

[0086] The predetermined, variable light intensities of the illumination device 152 thus correspond to background levels of the respective histograms that are expected during the operating phase of the lidar sensor 110. For each predetermined light intensity, which therefore corresponds to a specific expected background level, the control and evaluation unit 116 generates a large number of histograms so that sufficient statistics are available during the calibration phase.

[0087] In Fig. 3 Ten such histograms are shown as examples (see below). Fig. 3A ) for a specific background level together with the corresponding statistical analysis (see Fig. 3B und 3C ). The histograms are presented as in the diagrams of Fig. 2A und 2C Count values ​​310 are plotted against respective distance bins 320, however, for the calibration phase with the transmitting unit 112 of the lidar sensor 110 deactivated. The histograms of Fig. 3A With the transmitter unit 112 deactivated, they represent only one of the background levels with the associated noise of the calibration signals 154.

[0088] For each of the histograms of Fig. 3A For each of the histograms generated during the calibration phase of the lidar sensor 110 for a given background level, a maximum count value 315 is determined by the processing unit 156 and marked in the respective histograms. The maxima 315 are used to determine a distribution of their relative frequencies, as described in Fig. 3B is shown. Fig. 3B This is therefore the respective relative frequency (on the y-axis) for the maximum count values ​​from the histograms of Fig. 3A (plotted on the x-axis). Based on the relative frequencies of Fig. 3B A distribution of the cumulative relative frequencies is then determined, which is in Fig. 3C The graph shows the frequency as a function of the maximum count value of 315. As expected, the cumulative relative frequency approaches 1.0 with increasing count values.

[0089] Since the underlying calibration measurement signal 154 of the histograms of Fig. 3A generated solely by the lighting unit 152 to simulate a specific background level with noise, a certain value corresponds to the cumulative relative frequency in Fig. 3C , which is assigned to a specific count value, a probability with which all count values ​​smaller than this specific count value in the operating phase of the lidar sensor 110 are false positives caused only by the background level and noise. Conversely, a specific probability or cumulative relative frequency can be determined by the representation of the cumulative relative frequencies of Fig. 3C directly assign a specific maximum count value, above which all smaller count values ​​are recognized as false positive count values ​​with this probability during the operating phase.

[0090] For example, with a certain count value of 20, which is in Fig. 3 Since C is assigned a cumulative relative frequency of 0.99, the probability that a false-positive output signal 160 will not be detected is only 1%, as the remaining cumulative relative frequency of all larger count values ​​is only 0.01. In other words, a count value of 20 in the operating phase of the lidar sensor 110 is, for the corresponding background level, a false-positive output signal 160 or invalid output signal 160 with a probability of 1%, while the count value of 20 is a true-positive or valid output signal 160 with a probability of 99%.

[0091] In the calibration phase of the lidar sensor 110, a specific value for the false positive rate or probability of a false positive output signal 160 occurring is therefore initially set, for example, 1%. Based on the cumulative relative frequency of Fig. 3C A count value is then determined at a cumulative relative frequency that is complementary to the specified probability, i.e., in this example, at a cumulative relative frequency of 0.99. This count value is used in the operating phase as a signal threshold for the background level under consideration, which is simulated by a predetermined light intensity of the illumination unit 152. The respective signal threshold, together with the associated background level, forms an element of the calibration data set 170 (see...). Fig. 1 ), which is generated by the processing unit 156, to be transmitted by it to the control and evaluation unit 116 of the lidar sensor 110 and stored there. The respective signal threshold ensures that, for the present example, during the operating phase at the considered background level, 99% of the false-positive counts in the output signal 160 are detected or filtered out, and only 1% undetected false-positive counts are to be expected.

[0092] During the calibration phase of the lidar sensor 110, all background levels relevant to the operating phase of the lidar sensor 110 are first determined and assigned to corresponding predetermined light intensities, which are to be provided by the illumination unit 152. The histograms of Fig. 3A as well as the distributions of the relative and cumulative frequencies of Fig. 3B und Fig. 3C are then determined for each relevant background level or predetermined light intensity so that the processing unit can generate the calibration data set 170.

[0093] Based on the respective cumulative relative frequencies (see Fig. 3C A specific count value is then determined for each relevant background level as a signal threshold, corresponding to a predefined probability or false positive rate of, for example, 1%. Thus, a specific signal threshold in the form of a specific count value is assigned to each relevant background level for the lidar sensor 110. The respective pairs of background level and signal threshold values ​​are then stored in a lookup table, which is provided for the operating phase of the lidar sensor 110 and thus represents the calibration data set 170. During the operating phase of the lidar sensor 110, the count value in a TCSPC histogram must therefore reach at least the signal threshold for a specific background level in order to be considered a valid output signal 160 (see...). Fig. 1A ) of the lidar sensor 110 to be identified.

[0094] During the operating phase of the optical sensor or lidar sensor 110, operating measurement signals 128 are generated after the calibration phase (see Fig. 1A ) by means of the receiving unit 114 with the transmitting unit 112 activated and the lighting unit 152 deactivated or not present in the operating phase, and a TCSPC histogram similar to those histograms that are in Fig. 2A und 2C The TCSPC histogram is then used to determine an instantaneous background level, for example, as the mean or median of the histogram. This instantaneous background level is then assigned one of the signal thresholds from the calibration phase of the lidar sensor 110. The respective count values ​​from the TCSPC histogram are subsequently identified as invalid or false-positive signals if the count value is less than the signal threshold assigned to or selected for the instantaneous background level. Conversely, only those count values ​​from the TCSPC histogram, along with the corresponding distance bin value, for which the count value is greater than or equal to the signal threshold for the instantaneous background level are output as a valid output signal 160 of the lidar sensor 110. Bezugszeichenliste

[0095] 100 Sensor system 110 Optical sensor, LiDAR sensor 112 Transmitting unit 114 Receiving unit 115 Scanning range 116 Control and evaluation unit 122 Control signal 124 Transmitting signal 126 Receiving signal 128 Operating measurement signal 130 Object 140 Ambient light source 150 Calibration device 152 Illumination unit 154 Calibration measurement signal 156 Processing unit 158 ​​Control module 160 Output signal 170 Calibration data set 210 Count value 220 Distance bin 230 Relative frequency 240 Maximum count values, identified echoes 310 Count value 315 Maximum count value per histogram 320 Distance bin

Claims

1. An optical sensor (110), comprising: a transmission unit (112) for emitting an optical transmission signal (124) into the environment of the optical sensor (110), a reception unit (114) which is configured to detect a reflected or remitted portion of the transmission signal (124), and a control and evaluation unit (116) which is configured: to receive and store a calibration data set (170) before an operating phase of the optical sensor (110), wherein the calibration data set (170) comprises a plurality of signal threshold values that are associated with a respective background level which corresponds to a respective predetermined light intensity, during the operating phase of the optical sensor (110): to receive operating measurement signals (128) from the reception unit (114) when the transmission unit (112) is activated, to determine a current background level and to select one of the signal threshold values of the calibration data set (170) based on the current background level, and to identify output signals (160) that are derived from the operating measurement signals (128) as an invalid signal if the respective output signal (160) is smaller than the selected signal threshold value, wherein the optical sensor (110) is configured to be in communication with a calibration device (150) during a calibration phase before the operating phase, said calibration device (150) having an illumination unit (152) that illuminates the reception unit (114) with the respective predetermined light intensity to which the respective background level corresponds, the control and evaluation unit (116) is further configured, during the calibration phase, to detect calibration measurement signals (154) of the reception unit (114) for each predetermined light intensity so that a processing unit (156) is able to generate the calibration data set (170) based on the calibration measurement signals (154), the optical sensor (110) is configured for a time-correlated single photon counting, TCSPC, the control and evaluation unit (116) is configured to produce respective histograms of the time-correlated single photon counting, TCSPC histograms, for both the calibration measurement signals (154) and the operating measurement signals (128), the control and evaluation unit (116) is further configured, during the calibration phase of the optical sensor (110), to produce a respective plurality of TCSPC histograms for each predetermined light intensity, and the processing unit (156) is configured, during the calibration phase of the optical sensor (110), to determine a respective maximum count value (315) for each of the TCSPC histograms that are associated with the respective predetermined light intensity in order to produce a statistical distribution of the maximum count values (315) for the respective predetermined light intensity, and to determine the signal threshold values for the respective predetermined light intensities based on the statistical distribution.

2. An optical sensor (110) according to claim 1, wherein the processing unit (156) is further configured, during the calibration phase of the optical sensor (110): to determine a respective cumulative relative frequency of the maximum count values (315) for each predetermined light intensity based on the statistical distribution, and to determine the respective signal threshold values for the predetermined light intensities based on the respective cumulative relative frequency.

3. An optical sensor (110) according to claim 2, wherein the processing unit (156) is further configured, during the calibration phase of the optical sensor (110), to use at least one predefined percentage for false-positive output signals (160) in order to determine the respective signal threshold values for the associated background level based on the respective cumulative relative frequency.

4. An optical sensor (110) according to claim 2 or 3, wherein the processing unit (156) is further configured, during the calibration phase of the optical sensor (110), to use a plurality of predefined percentages for false-positive output signals (160) in order to determine, for each predefined percentage, a set of respective signal threshold values for the associated background level based on the respective cumulative relative frequency.

5. An optical sensor (110) according to any one of the claims 1 to 4, wherein the processing unit (156) is further configured, during the calibration phase of the optical sensor (110), to produce the calibration data set (170) in the form of at least one look-up table in which the respective signal threshold values and the associated background levels are contained.

6. An optical sensor (110) according to any one of the claims 1 to 5, wherein the optical sensor (110) is configured as a lidar sensor with time-correlated single photon counting, TCSPC.

7. An optical sensor (110) according to claim 6, wherein the processing unit (156) is further configured, in the calibration phase of the lidar sensor (110), to determine a respective set of signal threshold values for a plurality of distance ranges within a range of the lidar sensor (110).

8. An optical sensor (110) according to any one of the claims 1 to 7, wherein the reception unit (114) of the optical sensor (110) comprises a plurality of groups of pixels, the groups of pixels each have different characteristics with respect to the noise, and the processing unit (156) is further configured to determine a respective set of signal threshold values for each group of pixels in the calibration phase of the optical sensor (110).

9. A sensor system (100) comprising: an optical sensor (110) according to any one of the preceding claims; and a calibration device (150) which is in communication with the optical sensor (110) during a calibration phase of the optical sensor (110) and which is configured to generate the calibration data set (170) for the optical sensor (110).

10. A method of calibrating an optical sensor (110) which has a transmission unit (112) and a reception unit (114) and which is in communication with a calibration device (150) during a calibration phase, wherein the method comprises that: a light intensity of an illumination unit (152) of the calibration device (150) of the optical sensor (110) is varied such that the reception unit (114) is illuminated with a plurality of predetermined light intensities, calibration measurement signals (154) of the reception unit (114) are acquired for each predetermined light intensity, a respective signal threshold value is determined based on the calibration measurement signals (154), and the respective signal threshold value is associated with a background level for the optical sensor (110), which background level corresponds to the respective predetermined light intensity, and the signal threshold values for an operating phase of the optical sensor (110) when the transmission unit (112) is activated are provided as a calibration data set (170) so that, during the operating phase, a control and evaluation unit (116) of the optical sensor (110) is able: to select one of the signal threshold values of the calibration data set (170) based on a current background level which is determined based on operating measurement signals (128) that are acquired by means of the reception unit (114) of the optical sensor (110) during the operating phase, and to identify output signals (160) that are derived from the operating measurement signals (128) when the transmission unit (112) is activated as an invalid signal if the respective output signal (160) is smaller than the selected signal threshold value, wherein the optical sensor (110) is configured for a time-correlated single photon counting, TCSPC, respective histograms of the time-correlated single photon counting, TCSPC histograms, are produced for both the calibration measurement signals (154) and the operating measurement signals (128), a respective plurality of TCSPC histograms are produced for the calibration measurement signals (154) at each predetermined light intensity, a respective maximum count value (315) is determined for each of the TCSPC histograms that are associated with the respective predetermined light intensity in order to produce a statistical distribution of the maximum count values (315) for the respective predetermined light intensity, and the signal threshold values for the respective predetermined light intensities are determined based on the statistical distribution.

11. A method according to claim 10, wherein a respective cumulative relative frequency of the maximum count values (315) is determined for each predetermined light intensity based on the statistical distribution, and the respective signal threshold values for the predetermined light intensities are determined based on the respective cumulative relative frequency.