Patent Application
20210173088
The present invention relates to a phase angle correction value calculation apparatus of the type that, for example, is configured to emit light and process received reflected light in accordance with an indirect time of flight measurement technique. The present invention also relates to a method of calculating a phase angle correction value, the method being of the type that, for example, employs an indirect time of flight measurement technique.
In so-called time-of-flight sensing systems and other systems, for example gaming console vision systems, it is known to employ an illumination source to illuminate a surrounding environment within a field of view of the illumination source, sometimes known as a “scene”, and process light reflected by features of the scene. Such so-called LiDAR (Light Detection And Ranging) systems illuminate a scene with light using the illumination source, and detect light reflected from an object in the scene using a detection device, for example an array of photodiodes, some optical elements and a processing unit. Light reflected from the object in the scene is received by the detection device and converted to an electrical signal, which is then processed by the processing unit by application of a time-of-flight (ToF) calculation in order to determine the distance of the object from the detection device. Although different varieties of LiDAR system are known to be based upon different operating principles, such systems nevertheless essentially illuminate a scene and detect reflected light.
In this regard, the so-called “Flash LiDAR” technique, which is a direct ToF ranging technique, employs a light source that emits pulses of light that are subsequently reflected by features of the scene and detected by a detector device. In such a technique, the distance to a reflecting feature is calculated directly using a measured time for a pulse of light to make a round trip to the reflecting feature and back to the detector device. The pulses of light incident upon the detector devices are sampled in the time domain at a very high sampling rate. The signal path in the processing circuitry to implement such a technique therefore requires a high bandwidth for signals as well as a large silicon “real estate”, i.e. such an implementation requires a relatively large area on a silicon wafer, which in turn limits the number of channels that can be supported on an integrated circuit. The practical spatial number of channels that such Flash LiDAR sensors can support is therefore usually below 100. To overcome this limitation, mechanical scanning systems are implemented requiring moving components.
Another known LiDAR system employs a so-called “indirect Time of Flight” (iToF) ranging technique. iTOF systems emit a continuous wave light signal and reflections of the continuous wave light signal are received by a detector device and analysed. Multiple samples, for example four samples, of the light reflected from a feature of the scene are taken, each sample being phase stepped by, for example, 90°. Using this illumination and sampling approach, a phase angle between illumination and reflection can be determined, and the determined phase angle can be used to determine a distance to the reflecting feature of the scene.
In iToF systems, high frequency signal processing (demodulation) occurs at the pixel level, and so the signal bandwidth post-pixel required to integrate a large number of pixels on the same chip is low. Consequently, iToF systems can support a larger number of channels and hence higher spatial resolution measurement than direct ToF systems. However, iToF systems have limited distance measurement capabilities. In this regard, to achieve low stochastic distance measurement errors, iToF systems require high modulation frequencies, which in turn lowers the distance range that can be measured unambiguously. For example, a 100 MHz modulation frequency results in an approximate unambiguous measurement range of 1.5 m. Also, a conventional iToF system is susceptible to errors due to multiple reflections and multiple propagation paths.
As explained above, iToF systems sample in respect of different applied phases. A typical iToF system comprises a buffer that stores analogue signals generated by a so-called photonic mixer device in respect of m phases employed for subsequent signal processing. A discrete Fourier transformation unit calculates a fundamental frequency of a complex signal stored by the buffer in terms of in-phase and quadrature components of the signals. Using the values of the in-phase and quadrature components, a phase angle of the complex signal and the amplitude can be calculated and the distance to an object can be solved using the phase angle information.
However, such amplitude modulated continuous wave type systems suffer from an inherent depth measurement error due to aliasing of the emitted light signal versus a reference signal correlation function. This is due to higher harmonics present in both the emitted light signal and the reference signal, which are not accounted for in modelling or measurement. This kind of error is known as wiggling or circular errors (see “Modeling “wiggling” as a multi-path interference problem in AMCW ToF imaging”, Feigin et al., OSA Optics Express, 2015).
In order to obviate or at least mitigate for the errors, it is known to take a phenomenological approach to compensation for circular errors. In this regard, B-spline or lookup table-based correction methods are known, for example as described in “Time-of-Flight sensor calibration for accurate range sensing” (Lindner et al., Computer Vision and Image Understanding, 114 (2010), pages 1318-1328), for compensating an aggregation of error phenomena. However, such a technique is sensitive to changes in system properties, for example temperature and frequency used to modulate an optical illumination signal emitted by an optical source and demodulate an optical signal received by the photonic mixer device.
For some implementations, the application of a high number of phases reduces the circular error of a given phase angle measurement to a negligible levels, especially when the number of phase offset values employed is odd. However, a significant drawback of this approach is that amount of time required for the photonic mixer device to generate a large number of electrical output signals. In this regard, the more time that is taken, the greater the chance that the phase angle measurement will be influenced by so-called motion artefacts as a result of the scene being dynamic. Therefore, from a practical standpoint it is advantageous to employ a smaller number of phase offsets values when measuring the phase angle (than the relatively high number of phase offset values required to minimise the circular errors) and to compensate for the circular errors using a model predicting circular error magnitude as suggested above.
Typically, a common method to reduce circular errors is deployed on a one-time basis during a calibration stage at the end of the manufacturing process for the iToF system. During calibration, an object is placed in front of an iToF camera and the object is regularly relocated from a minimum to the maximum measurable distance to the iToF camera in several steps. At each relocation step, a distance to the object is measured using the relatively low number of phases employed during normal operation of the iToF system, for example four phase offset values, the measured distance being recorded together with the known ideal distance to the object. After making measurements in respect of each relocation step, a curve is constructed and the circular error can be estimated for each phase angle measurement. The estimated circular error can then be recorded in a look-up table and used for circular error compensation in real-time operation of the iToF system. However, such a calibration technique is time consuming since it requires physical object relocation.
U.S. Pat. No. 7,936,449 describes an improved calibration procedure for circular error compensation where the object is statically positioned at a known distance during calibration and instead of relocating the object to multiple different distances to the iToF camera, measurement is performed for a set of known phase delays that are introduced to each distance measurement. This approach reduces the time needed for calibration, but requires a highly accurate delay line providing known phase delays, which can increase the cost of the production of a silicon die of the iToF system and/or the cost and time of calibration. Also, since this calibration is also only performed once at the end of the production cycle, corrections are not made in respect to drift of the circular error occurring over time in the iToF system due to aging or temperature changes.
According to a first aspect of the present invention, there is provided a phase angle error calculation apparatus for an optical ranging system, the apparatus comprising: a light source configured to emit light in accordance with an indirect time of flight measurement technique; a photonic mixer cell configured to generate and store a plurality of electrical output signals respectively corresponding to a plurality of predetermined phase offset values applied in accordance with the indirect time of flight measurement technique over a measurement cycle; a signal processing circuit configured to process the plurality of electrical output signals in accordance with the indirect time of flight measurement technique in order to calculate a reference vector and a reference phase angle from the reference vector, the plurality of electrical output signals corresponding to measurement at a first level of precision; wherein the signal processing circuit is configured to process a subset of electrical output signals from the plurality of electrical output signals in accordance with the indirect time of flight measurement technique in order to calculate a measurement vector and a measurement phase angle from the measurement vector, the subset of the electrical output signals corresponding to measurement at a second level of precision lower than the first level of precision; and the signal processing circuit is configured to calculate a phase angle correction value using the reference phase angle and the measurement phase angle.
The apparatus may further comprise: a time-of-flight modification unit configured to introduce a phase delay into the light that is emitted and reflected back to the photonic mixer cell; wherein the signal processing circuit may be configured to perform a repeat calculation of the phase angle correction value over a plurality of measurement cycles subsequent to the measurement cycle in order to calculate a plurality of phase angle correction values.
The phase delay may be unknown.
The time-of-flight modification unit may be a delay network providing a plurality of different time delays and may be configured to delay the phase of the light emitted by the light source by differing amounts between the plurality of measurement cycles.
The photonic mixer cell may comprise a photodetector element. A substantially static object may be located in front of the photodetector element. The substantially static object may provide a substantially single distance to the photodetector element.
The delay network may comprise an analogue delay line.
Application of the plurality of different time delays of the delay network may provide a range of phase measurements comprising respective phase errors; the range of phase measurements may be at least within a circular error period.
The apparatus may further comprise: a plurality of photonic mixer cells comprising the photonic mixer cell, the plurality of photonic mixer cells comprising a plurality of photodetector elements, respectively; a time-of-flight modification unit configured to introduce a plurality of respective phase delays into the light reflected back to the plurality of photodetector elements; wherein the signal processing circuit may be configured to perform the calculation of the phase angle correction value over the measurement cycle and in respect of the plurality of photonic mixer cells in order to calculate a plurality of phase angle correction values.
The time-of-flight modification unit may be an object located in a scene and arranged relative to the plurality of photodetector elements so as to extend away from the plurality of photodetector elements sufficiently to enable a range of phase angle correction values to be calculated in respect of and at least within a circular error period.
The signal processing circuitry may be configured to support an error modelling unit configured to analyse the plurality of phase angle correction values and to generate a model of the plurality of phase angle correction values.
The error modelling unit may be configured to generate a lookup table comprising phase error correction values over a predetermined range of calculated phase angles in respect of measurements at the second level of precision.
The phase error correction values may correspond to a plurality of uncorrected measurement phase angles; a range of the plurality of uncorrected measurement phase angles may comprise the circular error period. The plurality of uncorrected measurement phase angles may comprise a phase angle interval therebetween; the phase angle interval may be consistent.
The plurality of electrical output signals in respect of the measurement at the first level of precision may comprise a first set of electrical output signals in respect of the measurement at the second level of precision and a second set of electrical output signals in respect of the measurement at the second level of precision.
The plurality of predetermined phase offset values may comprise a first set of predetermined phase offset values and a second set of predetermined phase offset values; and the first set of electrical output signals in respect of the measurement at the second level of precision may correspond to the first set of predetermined phase offset values and the second set of electrical output signals in respect of the measurement at the second level of precision may correspond to the second set of predetermined phase offset values.
The second set of phase offset values may be different to the first set of phase offset values. The first set of phase offset values may comprise a number of phase offset values in common with the second set of phase offset values. The first set of phase offset values may be the same as the second set of phase offset values.
The signal processing circuitry may be configured to support a motion detection unit configured to detect motion during generation of the plurality of electrical output signals and to omit calculation of the phase angle correction value in response to detection of motion.
The motion detection unit may be configured to detect the motion by comparing a first electrical output signal in respect of a first predetermined phase offset value of the plurality of the predetermined phase offset values and a second electrical output signal in respect of a second subsequent predetermined phase offset value of the plurality of the predetermined phase offset values; the second subsequent predetermined phase offset value may follow application of a number of the plurality of predetermined phase offset values and the second phase offset value may be substantially the same as the first phase offset value.
The signal processing circuit may be configured to detect motion by calculating the difference between the first electrical output signal and the second electrical output signal.
According to a second aspect of the present invention, there is provided an optical ranging system comprising the phase angle error calculation apparatus as set forth above in relation to the first aspect of the invention, the system further comprising: the signal processing circuit configured to apply the calculated phase angle correction value to the measurement phase angle in order to correct the measured phase angle; and the signal processing circuit is configured to calculate a range using the corrected measured phase angle.
According to a third aspect of the present invention, there is provided a method of calculating a phase angle correction value, the method comprising: a light source emitting light in accordance with an indirect time of flight measurement technique; a photonic mixer cell generating and storing a plurality of electrical output signals respectively corresponding to a plurality of predetermined phase offset values applied in accordance with the indirect time of flight measurement technique over a measurement cycle; a signal processing circuit processing the plurality of electrical output signals in accordance with the indirect time of flight measurement technique in order to calculate a reference vector and a reference phase angle from the reference vector, the plurality of electrical output signals corresponding to measurement at a first level of precision; wherein the signal processing circuit processes a subset of electrical output signals from the plurality of electrical output signals in accordance with the indirect time of flight measurement technique in order to calculate a measurement vector and a measurement phase angle from the measurement vector, the subset of the electrical output signals being a smaller set than the plurality of electrical output signals and corresponding to measurement at a second level of precision lower than the first level of precision; and the signal processing circuit calculates the phase angle correction value using the reference phase angle and the measurement phase angle.
According to a fourth aspect of the present invention, there is provided a method of reducing errors in an indirect time of flight measurement system, the method comprising: illuminating a scene; repeatedly calculating a first phase angle to a first precision and a second phase angle to a second precision using electrical output signals in respect of a single measurement cycle in accordance with an indirect time of flight measurement technique, the calculation of the first phase angle employing a plurality of phase offset values and the calculation of the second phase angle employing a subset of the plurality of phase offset values; delaying propagation of light reflected by the scene; calculating a plurality of phase errors using the first and second phase angles calculated; generating a model of the plurality of phase errors; using the model of the plurality of phase errors to correct a subsequent phase angle measurement.
It is thus possible to provide an apparatus and method that are each capable of calculating circular errors, or so called wiggling, without any knowledge of distances to objects in a calibration scene. Reliance upon a priori distance data can be avoided owing to the use of a set of electrical output signals in respect of a measurement cycle as a reservoir of measurements that can be used to calculate phase angles at two different levels of precision and compared. Furthermore, obviating the need to rely on a priori distance data enables an arbitrary scene to be used for calibration. This also serves to improve the accuracy of calculation of phase error values, which hitherto has relied upon mechanical movement of reference objects in a scene as part of the calibration process. Additionally, the apparatus and method provide greater flexibility as to timing of performance of the calibration, as calibration can be performed in the field without the need for any special or dedicated calibration objection or arrangement. The apparatus and method also provide a degree of immunity to thermal changes (drift) and ageing of components in the apparatus.
At least one embodiment of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which:
Throughout the following description, identical reference numerals will be used to identify like parts.
Referring to
A phase signal generator 112 is configured to generate a continuous wave electrical signal. The phase offset of the continuous wave signal is selectable via a control input 114, the phase of the continuous wave signal being selectable from a set of phase offsets: [θ0, θ1, . . . , θm-1]. A first output of the phase signal generator 112 is coupled to a second input of photonic mixer 108 and a second output of the phase signal generator 112 is operably coupled to the source of electromagnetic radiation 102.
An output of the integrator 110 is coupled to an input of a Digital Fourier Transform (DFT) unit 116. In this respect, phase angle measurements are transferred serially to the DFT unit 116, thereby reducing memory requirements for the detection and ranging module. The DFT unit 116 comprises internal buffers (not shown) to support serial transfer of measurements from the integrator 110. In order to support this arrangement, the DFT unit 116 is operably coupled to a timing control unit 118 to maintain synchronisation of data processing.
The timing control unit 118 has a synchronisation output 120 operably coupled to a timing input 122 of the DFT unit 116. A control output 124 of the timing control unit 118 is operably coupled to the control input 114 of the phase signal generator 112.
In this example, the DFT unit 116 has a plurality of digital in-phase (I)/quadrature (Q) outputs 126. In this example, the DFT unit 116 comprises b pairs of digital I/O outputs corresponding to different harmonics of measured signals. As the output of the integrator 110 is an accumulated charge and, in this example in the analogue domain, the output of the integrator 110 needs to be converted to the digital domain. This can be achieved, for example, by employing a photon counter as the integrator 110 or providing an analogue-to-digital converter before the DFT unit 116.
A first pair of I/O outputs of the plurality of digital I/O outputs 126, relating to the first harmonic of received reflected optical signals, is coupled to a phase angle calculation unit, for example an arctan unit 128. A first output of the arctan unit 128 is coupled to an input 130 of an angular error analysis unit 132 having, in this example, a data modelling functionality. The angular error analysis unit 132 is provided to calculate phase error values and, in this example, generate a lookup table held in a data store. Consequently, the angular error analysis unit 132 is operably coupled to a data store supported by, for example, a digital memory. The data store stores a lookup table 134. A second output of the arctan unit 128 is operably coupled to an input 136 of an angle correction unit 138, a port 140 of the angle correction unit 138 is capable of accessing the lookup table 134. An output of the angle correction unit 138 is coupled to further downstream processing units that perform various other functions, for example range calculation using corrected measured angles. However, as the structure and operation of the downstream hardware is not core to an understanding of the embodiments set forth herein, for the sake of clarity and conciseness of description, the downstream hardware will not be descried further herein. In this example, the DFT unit 116, the arctan unit 128, the angular error analysis unit 132, the data store, and/or the angle correction unit 138 constitute a signal processing circuit.
In operation (
The optical signal reflected by the object is received (Step 202) by each photonic mixer pixel device 104 of the array of photonic mixer devices of the apparatus 100. The phase signal generator 112 generates a continuous wave electrical signal, the timing control unit 118 controlling cycling through the set of phase offsets in respect of the electrical signal relative to the continuous wave optical signal. In this example, for calibration purposes, the phase signal generator 112 cycles through a plurality of phase offset values corresponding to a high precision measurement, for example that results in measurements with a very low level of circular errors, for example eight phase offset values. However, a different number of phase offset values can be employed, for example a greater number of phase offset values. A synchronisation signal is also applied by the synchronisation output 120 to the DFT unit 116.
For the sake of brevity, calculation of a phase error value will now be described with respect to a single pixel of the array of pixels, but the skilled person should appreciate that the functionality described is also performed by other pixels in the array of pixels. Thus, in order to calculate a phase angle error, a reference phase angle is calculated by applying the electrical signal generated by the phase signal generator 112 to the photonic mixer 108. The phase offset of the electrical signal is cycled through the set of phase offset values mentioned above and digital representations of the charges generated and stored (Step 204) in the integrator 110, constituting measurements, in respect of each phase offset value of the set of phase offset values are received by the DFT unit 116 in series and converted (Step 206) to a reference pair of I/O outputs using all the electrical output signals generated by the photonic mixer 108 in respect of all the phase offset values applied by the phase signal generator 112. In this regard, the plurality of electrical output signals generated during a measurement cycle or frame cycle constitutes measurement at a first level of precision. The reference pair of I/O outputs constitute an I/O reference vector, VR, representing the complex valued analogue electrical measurements in respect of the fundamental frequency.
It should be appreciated that, the integrator 110 provides a plurality of phase-separated amplitude measurement outputs in series representing respective accumulated charge levels for applied phase offset values in respect of the photonic mixer pixel device 102. As such, the DFT unit 116 calculates, for each measurement cycle, intermediate I and Q values for phase-separated amplitude measurements respectively received in series, which are accumulated in respect of the measurement cycle to generate final I and Q value results. In this example, operation of such an arrangement comprises vectors being calculated iteratively using the DFT unit 116 in respect of each incoming phase angle measurement. Consequently, as part of the process for calculating (Step 206) the reference pair of I/O outputs, the DFT unit 116 generates (Step 208) an interim result constituting a measurement pair of I/O outputs in respect of a subset of the plurality of electrical output values and corresponding to a subset of the phase offset values applied by the phase signal generator 112. The subset of electrical output signals used constitutes measurement at a second level of precision. The measurement pair of I/O outputs constitute an I/O measurement vector, VM, representing the complex valued analogue electrical measurements in respect of the fundamental frequency. The DFT unit 116 can also generate other I/O vectors in respect of harmonics of the charges measured by the integrator 108.
It should also be appreciated that where a sufficiently large number of phase offset values are applied in respect of measurement at the first level of precision, the plurality of electrical output signals can comprise multiple sets of electrical output signals corresponding to measurements at the second level of precision. For example, where 8 phase offset values are employed to measure at the first level of precision, the plurality of electrical output signals (eight) comprises a first set of electrical output signals (four) and a second set of electrical output signals (four) both corresponding to measurement at the second level of precision. It therefore follows that the plurality of phase offset values employed to generate the plurality of electrical output signals can be notionally divided into multiple sets of predetermined phase offset values, for example a first set of phase offset values and a second set of phase offset values, respectively corresponding to the first set of electrical output signals and the second set of electrical output signals. In such an example, the different sets of predetermined phase offset values comprise different phase offset values, i.e. in terms of set theory, an intersection of the multiple sets yields an empty set. As such, the first set of predetermined phase offset values is different from the second set of predetermined phase offset values. However, in other examples, the first set of predetermined phase offset values can comprise a number of phase offset values in common with the second set of predetermined phase offset values, i.e. the intersection is not an empty set. In some examples, the first set of predetermined phase offset values can be the same as the second set of predetermined phase offset values.
After the electrical measurement signals are converted to the frequency domain, the I- and Q-values for the fundamental frequency in respect of the reference pair of I/O outputs and the measurement pair of I/O outputs are provided by the DFT unit 116 at the output thereof. In this example, the synchronisation signal ensures that the fundamental frequency I/O outputs of a current measurement frame of the DFT unit 116 are synchronously received by the arctan unit 128. The arctan unit 128 then, in accordance with the indirect time of flight measurement technique, calculates (Step 210) an angle of the reference vector, VR, constituting a reference phase angle, φref, in the complex plane from the fundamental frequency reference I and Q values. Similarly, and also in accordance with the indirect time of flight measurement technique, the arctan unit 128 calculates (Step 212) an angle of the measurement vector, VM, constituting a measurement phase angle, φmeas, in the complex plane from the fundamental frequency measurement I and Q values. The reference phase angle, φref, corresponds to the measurement at the first level of precision and the measurement phase angle, φmeas, corresponds to the measurement at the second level of precision, the first level of precision being greater than the second level of precision. The angular error analysis unit 132 receives the calculated reference phase angle, φref, and the calculated measurement phase angle, φmeas, and calculates (Step 214) an error or correction value, for example a difference between the calculated reference phase angle, φref, and the calculated measurement phase angle, φmeas. The above process is performed in relation to each of the photonic mixer pixel devices 104 of the array and the angular error analysis unit 132 receives the calculated reference phase angles, φref, and the calculated measurement phase angles, φmeas, in respect of the all of the photonic mixer pixel devices 104 of the array of photonic mixer devices (unless data in respect of one or more of the pixels of the array are disregarded) and calculates (Step 214) a plurality of correction/error values. In this regard, the object 300, in this example, effectively applies different phase delays to the light emitted by the source of electromagnetic radiation 102 and so each photonic mixer pixel device of the array is capable of providing data to calculate a pair of a phase angles (reference, φref, and measurement, φmeas) in respect of different phase delays. Thus, the object or objects 300 extend away from the array of photonic mixer pixel devices 104 sufficiently to enable a range of phase angle error values to be calculated in respect of and within the circular error period.
Once the angular error analysis unit 132 has a sufficient number of calculated error values (within the circular error period), the angular error analysis unit 132 analyses the plurality of calculated phase errors and models (Step 216) the distribution of the plurality of error values with the respective calculated measurement phase angles, φmeas. Typically, the distribution can be fitted to a sinusoid or more than one sinusoid. The modelling can be, for example, by way of any suitable sinusoidal fitting algorithm.
Turning to
It can be seen that when plotting the calculated phase angle error values, ϕerr, against the measured phase angle, φmeas, calculated using the measurement vector, VM, in respect of the second level of precision, the plot is cyclic. In this regard, as the circular error is periodic, over a period of 2π in this example, the number of calculated values stored in the lookup table 134 can optionally be reduced by a factor of four for a four phase offset value measurement, i.e. π/2. This is the cyclic error period.
Referring back to
During non-calibration operation (
The source of electromagnetic radiation 102 again emits a continuous wave optical signal that illuminates (Step 220) a scene. An object in the scene, for example, reflects the emitted optical signal. The phase signal generator 112 generates a continuous wave electrical signal, the timing control unit 118 controlling cycling through the set of phase offsets in respect of the electrical signal relative to the continuous wave optical signal. In this example, the second, lower level precision is used and so four phase offset values are generated by the phase signal generator. The synchronisation signal is also applied by the synchronisation output 120 to the DFT unit 116.
In order to calculate a corrected phase angle, a phase angle is calculated by applying the electrical signal generated by the phase signal generator 112 to the photonic mixer 108 and the phase offset of the electrical signal is cycled through the set of phase offsets mentioned above and digital representations of the charges generated and stored (Step 222) in the integrator 110, constituting measurements, in respect of each phase offset value of the set of phase offset values are received by the DFT unit 116 in series and converted to a pair of I/O outputs constituting an I/O vector (Step 224), V, representing the complex valued analogue electrical measurements in respect of the fundamental frequency.
As in the previous example relating to the generation of the lookup table, the DFT unit 116 can also generate other I/O vectors in respect of harmonics of the charges measured by the integrator 110. After the electrical measurement signals are converted to the frequency domain, the I- and Q-values for the fundamental frequency are provided by the DFT unit 116 at the outputs thereof. In this example, the synchronisation signal ensures that the fundamental frequency I/O outputs of a current measurement frame of the DFT unit 116 are synchronously received by the arctan unit 128. The arctan unit 128 then, in accordance with the indirect time of flight measurement technique, calculates (Step 226) an angle of the vector, V, constituting an extracted (measured) calculated phase angle, φmeas, in the complex plane from the fundamental frequency I and Q values.
The extracted phase angle, φmeas, is then passed to the angle correction unit 138, which then access the lookup table stored by the lookup table 134. In response to receipt of the extracted phase angle, φmeas, the angle correction unit 138 accesses (Step 228) the lookup table 134 and obtains a phase angle correction value, −φerr, corresponding to the value of the extracted phase angle, φmeas, received, which is applied (Step 230) by the angle correction unit 138, for example added, to the extracted phase angle, φmeas, obtained. The combination of the extracted phase angle, φmeas, with the phase angle correction value, −φerr, yields a corrected phase angle, φcor, which is provided at an output of the angle correction unit 138. The corrected phase angle, φcor, can then be used by the signal processing circuit to calculate a range to the source of the reflection of the emitted light.
The above steps (Steps 220 to 230) are repeated (Step 232) until correction of measured angles is no longer required.
In another example, the object or objects 300 need not be specifically set up for the purpose of calibration and one or more objects can be sought by a user from a naturally occurring scene. In this regard, the calibration is performed in the field immediately prior to, during or after use in respect of a measurement application for which the apparatus 100 is being used. As such, calibration can be performed “on the spot” or on site in real time. In order to collect a sufficient quantity of data to span a circular error period, a scene that changes can be selected. The degree of movement in the scene can vary from periodically and not very often to more frequently. Where changes in the scene are sufficiently rapid to risk corrupting the data collected during calibration, motion detection can be employed to disregard measurements made where movement has occurred within the measurement cycle. An example of such motion detection is described later herein.
Referring to
In operation (
In this example, the analogue delay line 142 is configured to provide a plurality of phase delays to the source of electromagnetic radiation 102 so as span one circular error period, one per measurement cycle. In the case of a normal operating mode employing four phase offset values, the circular error period is π/2. Consequently, the application of the plurality of phase delays enables a range of phase measurements to be made, the range of phase measurements comprising a plurality of phase errors, respectively. However, in this example, the range of phase measurements is within the circular error period.
The optical signal reflected by the object 300 is received (Step 242) by the photonic mixer pixel device 104. The phase signal generator 112 generates a continuous wave electrical signal, the timing control unit 118 controlling cycling through the set of phase offsets in respect of the electrical signal relative to the continuous wave optical signal. In this example, for calibration purposes, the phase signal generator 112 again cycles through the plurality of phase offset values corresponding to a high precision measurement, for example that results in measurements with a very low level of circular errors, such as eight phase offset values. A synchronisation signal is also applied by the synchronisation output 120 to the DFT unit 116.
In order to calculate a phase angle error, the reference phase angle, φref, is calculated by applying the electrical signal generated by the phase signal generator 112 to the photonic mixer 108 and the phase offset of the electrical signal is cycled through the set of phase offsets mentioned above and digital representations of the charges generated and stored (Step 244) in the integrator 110, constituting measurements, in respect of each phase offset value of the set of phase offset values are received by the DFT unit 116 in series and converted (Step 248) to a reference pair of I/O outputs using all the electrical output signals generated by the photonic mixer 108 in respect of all the phase offset values applied by the phase signal generator 112. The reference pair of I/O outputs constitute an I/O reference vector, VR, representing the complex valued analogue electrical measurements in respect of the fundamental frequency.
It should be appreciated that the integrator 110 provides a plurality of phase-separated amplitude measurement outputs in series representing respective accumulated charge levels for applied phase offset values in respect of the photonic mixer pixel device 102. As such, the DFT unit 116 calculates, for each frame cycle or measurement cycle, intermediate I and Q values for phase-separated amplitude measurements respectively received in series, which are accumulated over a frame cycle to generate final I and Q value results. Operation of such an arrangement comprises vectors being calculated iteratively using the DFT unit 116 in respect of each incoming phase angle measurement. Consequently, as part of the process for calculating the reference pair of I/O outputs, the DFT unit 116 generates (Step 246) an interim result constituting a measurement pair of I/O outputs in respect of a subset of the phase offset values applied by the phase signal generator 112. The measurement pair of I/O outputs constitute an I/O measurement vector, VM, representing the complex valued analogue electrical measurements in respect of the fundamental frequency. As in previous examples, the DFT unit 116 can also generate other I/O vectors in respect of harmonics of the charges measured by the integrator 108.
After the electrical measurement signals are converted to the frequency domain, the I- and Q-values for the fundamental frequency in respect of the reference pair of I/O outputs and the measurement pair of I/O outputs are provided by the DFT unit 116 at the outputs thereof. In this example, the synchronisation signal ensures that the fundamental frequency I/O outputs of a current measurement frame of the DFT unit 116 are synchronously received by the arctan unit 128. The arctan unit 128 then, in accordance with the indirect time of flight measurement technique, calculates (Step 250) an angle of the reference vector, VR, constituting a reference phase angle, φref, in the complex plane from the fundamental frequency reference I and Q values. Similarly, and in accordance with the indirect time of flight measurement technique, the arctan unit 128 calculates (Step 252) an angle of the measurement vector, VM, constituting a measurement phase angle, φmeas, in the complex plane from the fundamental frequency measurement I and Q values. The reference phase angle, φref, constitutes a measurement at the first level of precision and the measurement phase angle, φmeas, constitutes a measurement at the second level of precision, the first level of precision being greater than the second level of precision. The angular error analysis unit 132 receives the calculated reference phase angle, φref, and the calculated measurement phase angle, φmeas, and calculates (Step 254) an error value, for example a difference between the calculated reference phase angle, φref, and the calculated measurement phase angle, φmeas.
As the process of applying a phase delay to the source of electromagnetic radiation 102 and subsequent calculation of the error value is in respect of a single pixel, the process has to be repeated until the angular error analysis unit 132 has collected a sufficient amount of data, for example spanning the cyclic error period. The signal processing circuit therefore determines (Step 256) whether the angular error analysis unit 132 has calculated a sufficient quantity of data to be able to analyse the plurality of calculated phase errors and model the data calculated. In the event that the angular error analysis unit 132 requires further error value data points, the phase delay generated by the analogue delay line 142 being random is changed automatically, and the above process is repeated (Steps 240 to 254) over successive measurement cycles, i.e. a plurality of measurement cycles, until a sufficient quantity of error correction values within the cyclic error period have been generated. In this regard, it should be appreciated that differing quantities of phase delay are applied to the light emitted by the source of electromagnetic radiation 102 between successive measurement cycles.
Thereafter, the angular error analysis unit 132 models (Step 260) the distribution of the plurality of error values with the respect to the calculated measurement phase angles, φmeas. As in relation to the example described with reference to
Referring back to
As the process of using the data stored in the lookup table 134 to correct calculated measured phase angles measured at the second level of precision, it should be appreciated that the description of the method according to
In the above examples, it is desirable to detect motion with respect to the scene, because motion in the scene can frustrate attempts to measure phase angles accurately during calibration. Referring to
The skilled person should appreciate that the above-described implementations are merely examples of the various implementations that are conceivable within the scope of the appended claims. Indeed, it should be appreciated that other techniques to calculate the phase angles described above can be employed. For example, in relation to the second level of precision employing four phase offset values, instead of employing the arctan unit 128 described above, a triangular calculation unit employing a triangular calculation technique to calculate amplitude and/or phase can be employed. In this regard, the circular error of calculated phase angles can be corrected in real time using “triangular” phase measurements taken at the second level of precision using DFT based measurements taken at the first level of precision.
It should also be appreciated that although, in the above examples, an analogue delay line is employed, digital delay lines should not be considered precluded.
In the examples set forth above in relation to
It should be appreciated that references herein to “light”, other than where expressly stated otherwise, are intended as references relating to the optical range of the electromagnetic spectrum, for example, between about 350 nm and about 2000 nm, such as between about 550 nm and about 1400 nm or between about 600 nm and about 1000 nm.
| Number | Date | Country | Kind |
|---|---|---|---|
| 19214952.4 | Dec 2019 | EP | regional |
