TIME OF FLIGHT SENSOR

Abstract

A method of measuring a distance using a time of flight sensor comprising a substantially transparent cover covering a light emitter and one or more photodetectors. The method comprises emitting a series of pulses of light from the light emitter; and using the one or more photodetectors to obtain a distribution of times at which at least one photodetector of the one or more photodetectors detected photons after each emission of the series of pulses of light. If the distribution of times comprises only a single peak, the method further comprises analysing the single peak to determine if the single peak includes counts of photons reflected from a target. If the single peak includes counts of photons reflected from a target, the method further comprises measuring the separation between a reference time and a point of the single peak.

Claims

1. A method of measuring a distance using a time of flight sensor comprising a substantially transparent cover covering a light emitter and one or more photodetectors, the method comprising: emitting a series of pulses of light from the light emitter; using the one or more photodetectors to obtain a distribution of times at which at least one photodetector of the one or more photodetectors detected photons after each emission of the series of pulses of light; and if the distribution of times comprises only a single peak: analysing the single peak to determine if the single peak includes counts of photons reflected from a target, and if the single peak includes counts of photons reflected from a target, measuring the separation between a reference time and a point of the single peak.

2. A method according to claim 1 wherein analysing the single peak to determine if it includes counts of photons reflected from a target comprises comparing one or more parameters of the peak to one or more parameters of a reference peak of a reference distribution of times obtained by: emitting a series of pulses of light from the light emitter with no objects within a range and field of view of the at least one photodetectors; and using the one or more photodetectors to obtain the reference distribution of times at which the at least one photodetector of the one or more photodetectors detected photons after each emission of the series of pulses of light.

3. A method according to claim 2 wherein the one or more parameters of the reference peak include the total number of photons detected in the reference peak and the time in the distribution at which the reference peak is located.

4. A method according to claim 3 wherein a background or noise level of the distribution comprising the single peak is added to the reference peak or subtracted from the single peak prior to said comparing.

5. A method according to claim 1 wherein the distribution of times is a distribution of times at which at least one photodetector of the one or more photodetectors detected photons before, during and after each emission of the series of pulses of light.

6. A method according to claim 1 wherein the times at which the at least one photodetector detected photons are relative to an emission of one of the series of pulses of light during or immediately preceding a detection period in which the photon was detected.

7. A method according to claim 1 wherein the distribution is obtained in a plurality of periods each containing the emission of one of the pulses of light, each period being divided into an identical series of time intervals relative to the emission contained therein.

8. A method according to claim 7 wherein the distributions of times is obtained by counting a total number of photons detected by the at least one photodetectors during all of the of time intervals at each of a plurality of times relative to the emission during the detection period comprising that time interval.

9. A method according to claim 8 wherein a counted number of photons detected by the one or more photodetectors during all of the time intervals at each of the plurality of times are or contribute to values of the distribution for time intervals at the plurality of times.

10. A method according to claim 1 wherein the series of pulses of light comprises at least 75,000 pulses of light.

11. A method according to claim 1 wherein the frequency distribution is a histogram.

12. A method according to claim 1 wherein peaks are parts of the distribution above a predetermined threshold.

13. A method according to claim 1, further comprising, if the distribution comprises two or more separate peaks, measuring the separation in time between the earliest peak and at least one of the one or more other peaks.

14. A method according to claim 1 wherein the light emitter is a vertical-cavity surface-emitting laser.

15. A method according to claim 1 wherein each of the one or more photodetectors is a single photon avalanche diode.

16. A method according to claim 1 wherein the one or more photodetectors are a plurality of photodetectors, and are used to obtain a plurality of distributions, each distribution being the distribution of times at which one or more photodetectors of the plurality of photodetectors detected photons after each emission of the emissions of the pulses of light.

17. A method according to claim 1, wherein the light emitter forms part of an optical stack and where the method calibrates the time of flight sensor in conjunction with the optical stack.

18. A time of flight sensor configured to perform a method according to claim 1.

19. A device comprising a sensor according to claim 18.

20. A non-transitory storage medium comprising computer instructions executable by one or more processors comprised by or in communication with a time of flight sensor comprising a light emitter and one or more photodetectors, the computer instructions when executed by the one or more processors causing the time of flight sensor to perform a method according to claim 1.

Description

[0128] Embodiments will now be described by way of example with reference to the figures.

[0129] FIG. 1A is a sectional view of a time of flight sensor in which a light emitter and photodetector array are located in a shared cavity beneath a cover;

[0130] FIG. 1B is a sectional view of a time of flight sensor in which a light emitter and photodetector array are located in separate cavities beneath a shared cover;

[0131] FIG. 2 is a diagrammatic view of a time of flight sensor;

[0132] FIG. 3 is a flow chart of a method for measuring distances using a time of flight sensor;

[0133] FIG. 4A is a histogram obtained by a time of flight sensor with no target present;

[0134] FIG. 4B is a histogram obtained by a time of flight sensor with a target close to the cover;

[0135] FIG. 4C is a histogram obtained by a time of flight sensor with a target distant from the cover; and

[0136] FIG. 5 is graph showing a pair of reference distributions and thresholds for target detections derived therefrom.

DETAILED DESCRIPTION

[0137] FIGS. 1A and 1B show respective sectional views of the optical components of two embodiments of time of flight sensors 100, 150. Each of the sensors 100, 150 comprises a light emitter 102, 152 for emitting pulses of light, a photodetector array 104, 154 arranged to detect the reflections of the pulses of light, and a transparent cover 108, 158 arranged to cover the light emitter 102, 152 and the photodetector array 104, 154.

[0138] The light emitters 102, 152 and the photodetector arrays 104, 154 are located in cavities 106, 156, 157 underneath the covers 108, 158. In the first time of flight sensor 100, the light emitter 102 and the photodetector array 104 are both arranged within a shared cavity 106. In the second time of flight sensor 150, the light emitter 152 is located within a first cavity 156 and the photodetector 154 is located within a second separate cavity 157. In one embodiment both of the two cavities 156, 157 of the second sensor 150 are covered by the same cover 158, although it is envisaged that, in another embodiment two separate covers may be used to respectively cover the two cavities 156 and 157.

[0139] In some time of flight sensors, the cavities 106, 156, 157 may be defined by apertures through a spacer wafer or spacer member arranged between the cover 108, 158 and a substrate on which the light emitter 102, 152 and photodetector array 104, 154 are mounted.

[0140] In use, pulses of light are emitted by the light emitters 102, 152 and the photodetector arrays 104, 154 detect photons incident thereon. The times at which photodetectors of the photodetector arrays 104, 154 detect photons are measured relative to the times at which the light emitters 102, 152 emitted pulses of light, in order to determine the travel time of photons emitted by the light emitters 102, 152 and reflected off targets 110, 160 onto the photodetector arrays 104, 154. These measured times allow the distance between the sensors 100, 150 and the targets 110, 160 to be determined.

[0141] However, in addition to light 120, 170 which is emitted by the light emitters 102, 152 and reflected off targets 110, 160, the photodetector arrays 104, 154 also detect light from other sources which may be referred to as optical crosstalk. Optical crosstalk includes light which is not emitted by the light emitters 102, 152, as well as light 122, 124, 172 which is emitted by the light emitters 102, 152 and reflected off the covers 108, 158.

[0142] In the first sensor 100, light 124 is reflected off the exterior of the surface of the cover 108 adjacent to the cavity 106 and light 122 is totally internally reflected of the interior of the surface of the cover 108 distal from the cavity 106. In the second sensor 150, as the light emitter 152 and photodetector array 154 are located in separate cavities 156, 157, light reflected off the exterior of the surface of the cover 158 will not reach the photodetector array 154. Light 172 totally internally reflected of the interior of the surfaces of the cover 158 travel from the first cavity 156 to the second cavity 157 and thereby from the light emitter 152 to the photodetector array 154. Light may also be reflected off material, such as dirt or other contaminants, on the cover.

[0143] Light 122, 124, 172 which is reflected off the cover or material thereon travel shorter distances than light 120, 170 reflected off targets 110, and therefore will be detected at earlier times after the emission of a pulse of light by the light emitter 102, 152.

[0144] FIG. 2 shows a diagrammatic view of a time of flight sensor 200. The sensor 200 comprises a vertical cavity surface emitting laser (VCSEL) 205 for emitting a series of light pulses, a driver 206 for driving the VCSEL 205, a detection module 210 for detecting photons and counting the number of photons detected in each of a plurality of consecutive time intervals which recur with each light pulse emission, a control module 215 for controlling the driver 206 and the detection module 210, a data processing module 220, an optical filter 225 and a pair of passive optical elements 230.

[0145] The detection module 210 comprises an array of single photon avalanche diodes (SPADs), one or more time to digital converters (TDCs) and a memory. SPADs are solid state photodetectors which produce a short avalanche current when they detect a photon. The one or more TDCs measure the time between initial signals from the control module (which are at fixed times relative to the emission of light pulses) and signals created the SPADs when they detect a photon. The memory records the times at which photons are detected by the SPADs.

[0146] The memory may comprise one or more sets of counters, each set of counters being associated with single SPAD or a group of SPADS (which may be arranged together and may together define or act as a pixel of the array of the detection module). Each counter of a set corresponds to one of a plurality of consecutive time intervals into which a recurring detection period is divided. The detection period, and time intervals thereof, repeat with each light pulse emission of the series of light pulse emissions.

[0147] When a SPAD detects a photon, the time within the detection period (and consequently which of the plurality of time intervals) the photon was detected in is determined by the TDC. The counter corresponding to that time interval and associated with that SPAD is then incremented by one. Which of the counters is incremented may be controlled by a shift register.

[0148] Each set of counters may therefore obtain counts of the number of photons detected by a SPAD, or by a group of associated SPADs, in each of a plurality of time intervals relative to the emission of a pulse of light in the detection period in which the photons were detected. The relative counts of each time interval corresponding to the relative probabilities of a photon being detected in each of those time intervals when a light pulse is emitted. The counts are a distribution which may be converted into, or visualised as, a histogram; for example, by the data processing module 220.

[0149] The control module 215, which may comprise driving electronics, provides high speed signals to the driver 206 and the detection module 210. The signals synchronize the emission of the series of pulses of light and the repeating detection period of the detection module, such that one pulse of light is emitted at the same point in time relative to (and preferably within) each repetition of the detection period.

[0150] The data processing module 220 receives data from the detection module. The data may be, or may be derived from, the counted number of photons detected in each of the time intervals by each of the SPADs or groups thereof. The data processing module may determine if the distribution of times at which photons were detected by a SPAD, or group of associated SPADS, over the series of emissions comprises peaks, and may determine whether two or more of those peaks are separate.

[0151] If the distribution comprises two or more separate peaks the data processing module 220 may determine the separation between the peaks. This may be used to determine the distance to a target 250. If the distribution instead comprises only a single peak, the data processing module 220 may analyse the shape and magnitude of the peak to determine if it includes only counts of photons reflected off the cover of the sensor, or whether it also includes counts of photons reflected off a target.

[0152] The passive optical elements 230 are lens elements. One of the passive optical elements 230 is arranged over the VCSEL 205 and the other is arranged over the detection module 210 and is arranged to focus incident light thereon such that a range image may be produced by the SPAD array. The optical filter 235 is arranged between the detection module and the lens element over the detection module. The lens elements may be comprised by a cover over the VCSEL 205 and the SPAD array.

[0153] The sensor 200 may operate under computer instructions which may be stored on one or more non transitory storage media which may be comprised by the control module 215, the data processing module 220, the detection module 210, and/or an external computing device in communication with the sensor, and/or modules thereof. The computer instructions may be executed by one or more data processors comprised the control module 215, the data processing module 220, the detection module 210, and/or an external computing device in communication with the sensor, and/or modules thereof, in order to use the sensor 200 to perform a method 300 for measuring optical crosstalk due to reflections of light pulses off the cover of the sensor 200.

[0154] FIG. 3 is a flow diagram of a method 300 for measuring distances using a time of flight sensor with a transparent cover covering a light emitter (such as a VCSEL) and one or more photodetectors (such as a SPAD array).

[0155] The method comprises emitting a series of light pulses, obtaining one or more distributions of the times at which photons are detected by individual photodetectors, or groups of associated photodetectors, analysing the distributions to determine if they include counts of photons reflected off targets, and if so, the times at which photons reflected off targets were detected. The times at which the one or more photodetectors detect photons being times at which photons were detected within and relative to a recurring detection period 315 which recurs with each emission 330 of a light pulse.

[0156] In a first step 310 of the method 300, a detection period 315 is begun. The detection period is divided into a series of consecutive time intervals in which photons may be detected by the one or more photodetectors. The number and duration of the time intervals is pre-set, and the time intervals are preferably of equal duration. For example, the detection period may comprise three hundred and fifty time intervals, each time interval lasting one hundred picoseconds; the detection period therefore lasts thirty-five nanoseconds in total.

[0157] During the detection period 315, in a second step 320, if any of the one or more photodetectors detects a photon during any of the time intervals, a count of photons detected in that time interval by that photodetector, or by a group of associated photodetectors comprising that photodetector, is incremented by one.

[0158] At a pre-set point in time during the detection period 315, in a third step 330, a pulse of light is emitted from the light emitter. The pulse of light may be emitted after a pre-set number of time intervals and preferably has a fixed duration. For example, the light pulse may last five hundred picoseconds and may be emitted beginning immediately after the twentieth time interval.

[0159] In a fourth step 340, the detection period 315 is repeated a predetermined number of times. During each repetition, one light pulse is emitted (at the same point during each repetition of the detection period 315). Each of the time intervals of the detection period 315 is repeated and if a photodetector detects any photons therein, the count of photons detected by that photodetector in that time interval is incremented. The number of repetitions of the detection period may be between 80,000 and 800,000 inclusive; for example, with lower numbers of repetitions being in a low power mode of the sensor, and/or for use with targets with greater reflectivity.

[0160] A distribution is therefore obtained for each of the photodetectors, or groups of associated photodetectors, of the number of photons detected by that photodetector, or group of photodetectors, in each of the time intervals. The distribution comprising counts for each of the time intervals of the number of photons detected by that photodetector over all repetitions of the detection period 315. The distribution may be, or may be visualised as, a histogram. Examples of such distributions 300, 440, 480 are shown in FIGS. 4A to 4C.

[0161] In a fifth step, 350, whether the obtained distributions each comprise a single peak or two or more separate peaks is determined.

[0162] Peaks may be identified as groups of consecutive time intervals in which the number of counted photons exceeds a threshold value, which may be dependent upon the background or noise count level in the distribution. The background or noise count level of a distribution may be determined by averaging the counts per time interval of time intervals before a pre-set time, for example of time intervals before the emission of a light pulse during the detection period.

[0163] Two or more peaks may be considered to be separate if they are separated by a minimum number of time intervals which are not comprised by any peaks. For example, if they are separated by one or more time intervals which are not comprised a peak (such as time intervals which have counts below a threshold value as described above).

[0164] If the distribution comprises two or more peaks, in step 360, the separation in time between the earliest peak and each other peak (for example, between the maxima thereof) is measured. These times may be used to calculate the distance between the cover and the targets whose reflections caused the later peaks.

[0165] If the distribution comprises only a single peak, in step 370, the peak is compared to a reference peak (or one or more parameters thereof) for the one or more photodetectors whose detections generated the distribution. The reference peak or parameters thereof preferably being stored in a memory of the time of flight sensor.

[0166] The single peak may be compared to the reference peak by comparing the total number of counted photon detections in the single peak to the total number of counted photon detections in the reference peak. The total number of photon detections in a peak being proportional to the energy of the peak. The total number of photons in a peak may be the sum of the counts of each of the bins determined to be comprised by the peak, or the sum of the counts of each of those bins after an average background or noise level is subtracted therefrom.

[0167] If the single peak comprises a greater total count than the reference peak plus a predetermined threshold margin, the single peak may be determined to comprise counts of photons reflected off a target close to the cover.

[0168] The reference peak may be the single peak of a distribution obtained for the photodetector or group of associated photodetectors whose detections generated the distribution when performing a calibration measurement with no targets present in the range and field of view of the photodetector or group of associated photodetectors. Alternatively, the reference peak, and/or parameters thereof may be averages of reference peaks or parameters thereof obtained from multiple such calibration measurements.

[0169] The distributions obtained from performing calibration measurements contain only counts of background or noise photons and photons reflected off the cover of the sensor, which will produce the single reference peak.

[0170] Such calibration measurements may be performed when, or shortly after, the time of flight sensor is manufactured or assembled and are preferably performed in controlled conditions. The calibration measurements may be performed in situations with a pre-set level of, a minimum level of, or no ambient light; this may control or minimise noise or background counts. The reference peak and/or parameters thereof obtained in these initial measurements may be stored in a memory of the sensor for use when performing the method 300. The calibration measurements are preferably performed at room temperature, or at some other predetermined temperature at which the time of flight sensor is intended to primarily operate.

[0171] Alternatively, calibration measurements may be performed at times after the sensor has been manufactured and/or used, and the reference peak and/or parameters thereof may be updated. For example, when a time of flight measurement is performed using the time of flight sensor and a distribution for a photodetector or group of associated photodetectors is obtained with two or more separate peaks, the earliest peak may be assumed to correspond to counts of photons reflected off the cover only, and the earliest peak and/or parameters thereof may be recorded as the reference peak or parameters thereof, updating and/or replacing earlier stored peaks and/or parameters.

[0172] The predetermined threshold margin that the total count of a single peak must exceed for the peak to be determined to include counts of reflections off a target may be determined by performing additional calibration measurements with targets present and comparing the distributions obtained to the reference peak described above. For example, a minimum threshold margin for a target to be present may be equal to or less than the difference in the total counts of the reference peak, and the total counts of a distribution obtained from performing a calibration measurement with a very low reflectivity target present at a short distance from the cover, such that only a single peak is obtained.

[0173] If the peak is determined in step 370 to include counts of photons reflected off a target, in step 380, the separation between a reference time and a point on the peak is determined. The reference time is preferably a time corresponding to the time for a photon reflected off the cover to be detected by the one or more photodetectors. In preferred embodiments, the reference time is the time at which the maximum of the reference peak occurs.

[0174] The difference in the time taken by a photon reflected off the cover and a photon reflected off the target is determined by measuring the difference in time between the reference time and a point of the peak. The point may be the maximum of the single peak (the peak including reflections off the cover and off the close target), or may be a later time.

[0175] FIGS. 4A, 4B, and 4C show example histograms 400, 440, 480 of the times at which a photodetector (or a group of associated photodetectors) detected photons over the course of a series of light pulse emissions and detection periods repeated therewith.

[0176] Each histogram 400, 440, 480 comprises three hundred and fifty bins, each corresponding to a one hundred picosecond time interval of the repeated detection period, and counts of the total number of photons detected by the photodetector (or group of associated photodetectors) during each of the time periods. The histograms 400, 440, 480 being obtained by performing a number of iterations of the detection period 315 of the method 300 described above.

[0177] A first histogram 400, shown in FIG. 4A, is obtained by performing the repetitions of the detection period 315 of the method 300 when no target was present within the range and field of view of the photodetector (or group of associated photodetectors). The histogram comprises a background noise level 402 of approximately one hundred counts per bin and a single peak 404 with a maximum count of approximately five hundred and ninety detections in the time interval 2.8-2.9 nanoseconds into the repeated detection period. The peak 404 has a full width at half maximum of approximately 7 time intervals between approximately 2.5 and 3.2 nanoseconds into the repeated detection period. This peak 404 is caused by reflections of light emitted in the light pulses off the cover of the sensor, or material thereon.

[0178] Such a histogram 400, may be obtained for one or more photodetectors when performing reference calibration measurements for a time of flight sensor, with no target present and in the field of view of the photodetectors. In such situations the peak 404 may be the reference peak. If peaks are considered to groups of adjacent bins with counts greater than twice the average noise count. Bins with greater than 200 counts per time interval are considered to be comprised by a peak, the peak 404 therefore comprises 7 bins and has a total count of approximately 2950, or 2250 when background or noise counts are subtracted therefrom.

[0179] The reference peak total count is therefore 2250 and the reference time is the 2.8-2.9 nanosecond time interval. A pre-set threshold margin that the total count of a single peak must exceed for the peak to be determined to include counts of reflections off a target is also determined to be 500 counts in this example.

[0180] The third histogram 480, shown in FIG. 4C, is obtained by performing the repetitions of the detection period 315 with a target present at a distance from the cover. The histogram 480 comprises a background noise level 482 of approximately one hundred counts per bin and two peaks 484, 486.

[0181] The earliest peak 484 has a maximum count of approximately five hundred and ninety detections within the time interval of 2.9-3.0 nanoseconds into the detection period, and a full width at half maximum of approximately 6 time intervals between approximately 2.5 and 3.2 nanoseconds into the repeated detection period.

[0182] The later peak 486 is has a maximum count of approximately five hundred and thirty counted detections within the time interval of 18.0-18.1 nanoseconds into the detection period, with a full width at half maximum of approximately 7 time intervals between 17.7 and 18.4 nanoseconds into the repeated detection period.

[0183] The earliest peak 484 is located at a reference time of 2.8-2.9 nanoseconds into the detection period and may therefore be assumed to contain counts of photons reflected off the cover. The earliest peak can also be assumed not to contain reflections off a target close to the cover because the distribution includes at least one later peak 486. A target close the cover would block photons from a more distant target and prevent the appearance of the later peak 486.

[0184] When such a histogram 480, with multiple separate peaks 484, 486 is obtained for a photodetector (or group of associated photodetectors) when performing the method 300. The separation in time from the maximum of the earliest peak, to the maximum of each later peak may be measured to determine the distance in travel time between photons reflected off targets and those reflected off the cover. In the histogram 480 shown in FIG. 4C, the separation in time is 15.1 nanoseconds.

[0185] The distance between the cover and the target whose reflections produced the later peak may be determined by multiplying half the separation in time by the speed of light, obtaining a distance of 2.26 metres (to 2 decimal places).

[0186] The second histogram 440 obtained by performing the repetitions of the detection period 315 with a target present close to the cover is shown in FIG. 4B. This histogram 440 also comprises a background noise level 442 of approximately one hundred counts per bin and a single peak 444. The single peak 444 includes counts of photons reflected off the cover and photons reflected off the close target. The peak 444 has a maximum count of approximately eight hundred and forty detections within the time interval 3.2-3.3 nanoseconds into the detection period and has a full width at half maximum of approximately 9 time intervals between approximately 2.6 and 3.5 nanoseconds into the detection period.

[0187] When a histogram obtained by performing the repetitions of the detection period 315 of the method 300 contains a single peak, such as the peak 444 of the second histogram 440, the single peak is analysed to determine if it comprises reflections off a target close to the cover, or if it comprises reflections off the cover only. This is performed by comparing the single peak 444 to the reference peak 404, for example by comparing the total counts of the single peak 444 to the total counts of the reference peak.

[0188] When considering time intervals with counts exceeding two hundred (twice the average background noise count) to be comprised by the peak, the single peak 444 comprises 12 bins and has a total count of approximately 6400, or approximately 5200 when background or noise counts are subtracted therefrom. The total count of the reference peak is 2250 and the pre-set threshold margin that the total count of a single peak must exceed for the peak to be determined to include counts of reflections off a target is 500. As the total count of 5200 exceeds 2750, the single peak 444 is determined to include counts of reflections off the target.

[0189] As the counts of reflections off the target are included in the same peak 444 as the counts of reflections off the cover, it is not possible to determine the distance to the target by measuring the separation between two peaks as was performed with histogram 484 in FIG. 4C. Instead, the separation between the time maximum of the single peak 444 and the reference time at which the peak of the reference peak due to cover reflections only was located is measured.

[0190] The maximum of the single peak is in the time interval 3.2-3.3 nanoseconds into the detection period and the reference time is the 2.8-2.9 nanosecond time interval. The separation in time is therefore 0.4 nanoseconds and the distance to the close target is therefore 0.06 metres (to 2.d.p).

[0191] FIG. 5 is a graph 500 showing a pair of reference distributions 510, 520 obtained for two different photodetectors or groups thereof which detect reflections off the covers thereof at different rates, and two threshold functions 515, 525 for detecting targets for the two sensors. The thresholds being derived from the reference distributions.

[0192] The reference distributions 510, 520 are each a distributions of the times at which a single photodetector, or a group of associated photodetectors detected photons during reference calibration measurements with little or no ambient light and no objects within the range and field of view of the at least one photodetectors. The reference distributions 510, 520 are obtained by performing the repeated detection period steps 310, 320, 330 and 340 of the method 300 with little or no ambient light and no objects within the range and field of view of the at least one photodetectors.

[0193] Both reference distributions 510, 520 comprise negligible noise or background counts due to the lack of ambient light, and comprise only a single peak due to reflections off the cover of the photodetector with which they were obtained. The second reference distribution 520 comprises a larger peak with a greater maximum detection frequency than the first reference distribution 510, perhaps as a consequence of a greater degree of contamination on the cover over the photodetectors.

[0194] Thresholds 515, 525 for detecting a peak comprising counts of photons reflected off of a target is derived from each of the reference distributions 510, 520. The first threshold function 515 is derived from the first reference distribution 510 and the second threshold function 525 is derived from the second reference distribution.

[0195] Each of the threshold functions 515, 525 is a function of the time of the distribution. Each of the threshold functions 515, 525 comprises a first earliest portion from 0 nanoseconds to time at which the reference peaks of the reference distributions begin at approximately 14 nanoseconds, a second intermediate portion from approximately 14 nanoseconds to approximately 34 nanoseconds, which includes that times at which the reference peaks are located and a portion of time thereafter, and a third latest portion from approximately 34 nanoseconds onwards.

[0196] The threshold level for detecting a peak comprising counts of target reflections is close to 0 counts per unit time interval in the first and third portions of the threshold functions 515, 525. For such a threshold to avoid incorrect detections of peaks, a background or noise level must be subtracted from a distribution obtained using the method 300 with the photodetectors before comparing it to the threshold 515, 525. In alternative thresholds, the threshold functions may comprise constant thresholds above an expected noise level in these portions.

[0197] The threshold level for detecting a peak comprising counts of target reflections in the intermediate portions of the threshold functions 515, 525 is an inverse linear function sloping downwards from above the maximum frequency of photon detections of the reference distribution 510, 520 from which they are derived, to the threshold level of the third later portion.

[0198] Therefore, the earlier a peak is located after the time at which a peak due to reflections off of the glass occurs, and the more likely the peak is to include counts of photons reflected off of the glass, the lower higher the threshold level for detecting a peak which comprises counts of reflections off of a target. Such threshold functions 515, 525 may therefore be used to detect both peaks comprising photons reflected off both a target and the cover and later separate peaks comprising only counts of reflections off of targets..

[0199] The embodiments have been described by way of example only, and it will be appreciated that variation may be made to the embodiments described above without departing from the scope of the invention as defined by the claims.