Data refinement in optical imaging systems

Abstract

An imaging system is configured to output system output signals during a plurality of associated data periods. The frequency pattern of the system output signal is repeated as a function of time during each of the associated data periods. The lidar system includes an optical combiner that combines light from the system output signal returned to the lidar system with light from the reference signal to generate beat signals that are each beat at a beat frequency. Each of the beat frequencies is associated with a different one of the data periods. The system also includes electronics that calculate average frequencies, each being an average of the plurality of different beat frequencies, and each of the average frequencies being associated with a different one of the data periods. The electronics computes lidar data from the average frequency. The lidar data is indicative of a radial velocity and / or distance between the system and an object external to the system.

Description

[0001] Related Applications

[0002] This application is a continuation of U.S. patent application serial number 17 / 727,731, filed on April 23, 2022, entitled “Data Refinement in Optical Imaging Systems,” the entire contents of which are incorporated herein. Technical Field

[0003] The present invention relates to imaging systems. In particular, the present invention relates to data thinning in imaging systems. Background Art

[0004] The LIDAR system outputs a system output signal that is reflected by an object located outside the LIDAR system. The reflected light is returned to the LIDAR system as a system return signal. The LIDAR system includes electronics that use the system return signal to estimate LIDAR data (radial velocity and / or distance between the LIDAR system and the object) for a sample area illuminated by the system output signal. In some cases, the LIDAR data includes an estimate of the power of the system return signal.

[0005] An important measure of lidar system performance is the accuracy of lidar data measurements. However, the standard deviation of these estimates is affected by many factors, such as noise, speckle, interference, and practical limitations of the system. These effects can be a source of random variation in lidar data estimates and can reduce their accuracy.

[0006] Therefore, there is a need to increase the level of accuracy of the lidar data generated by the lidar system. Summary of the invention

[0007] An imaging system is configured to output a system output signal during a plurality of associated data periods. A frequency pattern of the system output signal as a function of time is repeated during each of the associated data periods. A laser radar system includes a light combiner that combines light from the system output signal returning to the laser radar system with light from a reference signal to generate a beat signal that beats at a beat frequency. Each of the beat frequencies is associated with a different one of the data periods. The system also includes an electronic device that calculates an average frequency, each of which is an average of a plurality of different beat frequencies, and each of the average frequencies is associated with a different one of the data periods. The electronic device calculates laser radar data from the average frequencies. The laser radar data indicates radial velocity and / or distance between the system and an object external to the system.

[0008] A method of operating an imaging system, comprising outputting a system output signal from the imaging system during a plurality of associated data periods. The frequency of the system output signal is repeated as a function of time during each of the associated data periods. The method further comprises combining light from the system output signal returned to a lidar system with light from a reference signal to generate a beat signal that is each beat at a beat frequency. Each of the beat frequencies is associated with a different one of the data periods. The method further comprises calculating an average frequency, each of which is an average of a plurality of different beat frequencies, and each of the average frequencies is associated with a different one of the beat frequencies. In addition, the method comprises calculating lidar data from the average frequencies. The lidar data indicates a radial velocity and / or distance between the lidar system and an object external to the lidar system.

[0009] In some cases, the number of beat frequencies averaged to calculate each of the average beat frequencies is different for different average beat frequencies. Additionally or alternatively, the value of the adjusted portion of the beat frequency is adjusted to a value different from the value of the beat frequency of the beat signal. At least a portion of the beat frequencies averaged to calculate one or more of the average beat frequencies may have an adjusted value other than the value of the beat frequency of the beat signal. The value of the adjusted portion of the beat frequency may be adjusted in response to a beat frequency in the adjusted portion being an outlier. One or more outlier identification criteria may be applied to the beat frequencies to identify the beat frequencies as outliers.

[0010] The beat frequency averaged to calculate the object frequency in the average frequency may be selected from the beat frequency values ​​appearing in the averaging window. The beat frequency is selected so that the size of the difference between each of the selected beat frequencies and the object beat frequency in the beat frequencies is less than a threshold value. Beat frequencies in the window but resulting in a size of the difference greater than a threshold value may be excluded from the beat frequencies averaged to calculate the object average frequency. The object beat frequency may be a beat frequency associated with the object average frequency.

[0011] A beat frequency that is averaged to calculate the object frequency in the average frequency is selected from the plurality of beat frequencies in the averaging window. A surface selection criterion may be applied to the beat frequencies in the averaging window to identify beat frequencies that are all a result of the system output signal impinging the same surface on the object. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1A is a top view of a schematic diagram of a lidar system, wherein the lidar system includes or consists of a lidar chip, which outputs a lidar output signal and receives a lidar input signal on a common waveguide.

[0013] Figure 1Bis a top view of a schematic diagram of a lidar system, wherein the lidar system includes or consists of a lidar chip, which outputs lidar output signals and receives lidar input signals on different waveguides.

[0014] Figure 1C is a top view of a schematic diagram of another embodiment of a lidar system, wherein the lidar system includes or is composed of a lidar chip, which outputs lidar output signals on different waveguides and receives multiple lidar input signals.

[0015] Figure 2 Is suitable for Figure 1B A top view of an example of a lidar adapter used together with a lidar chip.

[0016] Figure 3 Is suitable for Figure 1C A top view of an example of a lidar adapter used together with a lidar chip.

[0017] Figure 4 It is included on the common support Figure 1A LiDAR chips and Figure 2 A top view of an example of a lidar system with a lidar adapter.

[0018] Figure 5A An example of a processing component suitable for use with a lidar system is shown.

[0019] Figure 5B Provides suitable and Figure 5A Schematic diagram of the electronics used in conjunction with the constructed processing components.

[0020] Figure 5C It is a graph of the frequency of the system output signal versus time.

[0021] Figure 5D is another example of a frequency versus time graph of a system output signal.

[0022] Figure 5E Shows Figure 5C The relationship between the data time periods disclosed in and the field of view of the lidar system.

[0023] Fig. 5F is a graph of frequency versus period index (j).

[0024] Fig. 6A Two different objects are shown located in the field of view of the lidar system.

[0025] Figure 6BThe results of beat frequency versus time and average beat frequency versus time when the system output signal changes the surface are compared.

[0026] Figure 6C The results of beat frequency versus time when outliers were present and the average beat frequency versus time were compared.

[0027] Figure 7 The data used to refine the data to resolve the beat frequency value (f m,j ) in the presence of frequency outliers and / or surface variations.

[0028] Figure 8 shows a method suitable for performing Figure 7 An example of a processed buffer.

[0029] Fig. 9 Shown suitable for use as Figure 7 An example of a processing flow of the first algorithm and / or the second algorithm in the processing flow of .

[0030] Fig.10 is a flow chart illustrating one possible method of calculating the previous representative frequency.

[0031] Fig.11 is a flow chart illustrating one possible method of calculating subsequent representative frequencies.

[0032] Fig.12 is a cross-section of a portion of a lidar chip including a waveguide on a silicon-on-insulator platform. DETAILED DESCRIPTION

[0033] The imaging system is a frequency modulated continuous wave system that calculates a series of beat frequency values. The system includes electronics that calculate an average beat frequency value using a moving average of the beat frequencies, each of which is associated with one of the beat frequency values. The electronics estimates lidar data (radial velocity and / or distance between the system and an object external to the system) from the average beat frequency value rather than from the beat frequency value. The use of the average beat frequency value reduces the standard deviation of the lidar data and thus increases the accuracy of the lidar data.

[0034] The presence of outliers in the beat frequency values ​​introduces distortion into the average beat frequency value. The electronics can identify these outliers and adjust the beat frequency values ​​so as to reduce or eliminate the effect of the outliers on the average beat frequency value. The ability to correct for these outliers increases the probability of detection by the lidar system.

[0035] Edges of objects in the system's field of view may also introduce distortion into the average beat frequency value. However, the electronics may select the beat frequency values ​​that are averaged into the average frequency value. The selection of the beat frequency values ​​is performed so as to reduce or eliminate distortion. For example, the beat frequency values ​​that are averaged to calculate the average beat frequency value may be selected so that the average is generated by the same surface of the object. By averaging the beat frequency values ​​from the same surface, distortion caused by the presence of the edge is removed from the average beat frequency value. Thus, the lidar system may have the increased accuracy provided by averaging the beat frequency values ​​combined with reduced distortion from outliers and edges.

[0036] Figure 1A It is a top view of a schematic diagram of a laser radar chip that can be used as a laser radar system or can be included in a laser radar system, and the laser radar system includes components other than the laser radar chip. The laser radar chip may include a photonic integrated circuit (PIC) and may be a photonic integrated circuit chip. The laser radar chip includes a light source 4 that outputs a preliminary laser radar signal. Suitable light sources 4 include, but are not limited to, semiconductor lasers, such as external cavity lasers (ECL), distributed feedback lasers (DFB), discrete mode (DM) lasers, and distributed Bragg reflector lasers (DBR).

[0037] The lidar chip includes a utility waveguide 12 that receives an emitted lidar signal from a light source 4. The utility waveguide 12 terminates at a facet 14 and carries the emitted lidar signal to the facet 14. The facet 14 can be positioned so that the emitted lidar signal traveling through the facet 14 leaves the lidar chip and is used as a lidar output signal. For example, the facet 14 can be positioned at the edge of the chip so that the emitted lidar signal traveling through the facet 14 leaves the chip and is used as a lidar output signal. In some cases, a portion of the lidar output signal that has left the lidar chip can also be considered a system output signal. As an example, when the lidar output signal leaves the lidar chip is also the lidar output signal leaving the lidar system, the lidar output signal can also be considered a system output signal.

[0038] The lidar output signal travels away from the lidar system through free space in the atmosphere where the lidar system is located. The lidar output signal can be reflected by one or more objects in the path of the lidar output signal. When the lidar output signal is reflected, at least a portion of the reflected light travels back toward the lidar chip as the lidar input signal. In some cases, the lidar input signal can also be considered a system return signal. As an example, when the lidar output signal leaves the lidar chip, the lidar output signal also leaves the lidar system.

[0039] The lidar input signal may enter the utility waveguide 12 through the facet 14. The portion of the lidar input signal that enters the utility waveguide 12 is used as the incoming lidar signal. The utility waveguide 12 carries the incoming lidar signal to a splitter 16, which moves a portion of the outgoing lidar signal from the utility waveguide 12 to a comparison waveguide 18 as a comparison signal. The comparison waveguide 18 carries the comparison signal to a processing component 22 for further processing. Although Figure 1A A directional coupler is shown operating as the splitter 16, but other signal tapping components may also be used as the splitter 16. Suitable splitters 16 include, but are not limited to, directional couplers, optical couplers, Y-junctions, tapered couplers, and multi-mode interference (MMI) devices.

[0040] The utility waveguide 12 also carries the outgoing lidar signal to a splitter 16. The splitter 16 moves a portion of the outgoing lidar signal from the utility waveguide 12 as a reference signal onto a reference waveguide 20. The reference waveguide 20 carries the reference signal to a processing component 22 for further processing.

[0041] The percentage of light transmitted from the utility waveguide 12 by the splitter 16 can be fixed or substantially fixed. For example, the splitter 16 can be configured so that the power of the reference signal transmitted to the reference waveguide 20 is an emitted percentage of the power of the emitted lidar signal, or so that the power of the comparison signal transmitted to the comparison waveguide 18 is an incoming percentage of the power of the incoming lidar signal. In many splitters 16 such as directional couplers and multimode interferometers (MMIs), the emitted percentage is equal to or substantially equal to the incoming percentage. In some cases, the emitted percentage is greater than 30%, 40% or 49% and / or less than 51%, 60% or 70%, and / or the incoming percentage is greater than 30%, 40% or 49% and / or less than 51%, 60% or 70%. Splitters 16 such as multimode interferometers (MMIs) typically provide an emitted percentage and an incoming percentage of 50% or about 50%. However, multimode interferometers (MMIs) can be more easily manufactured in platforms such as silicon-on-insulator platforms than some alternatives. In one example, the separator 16 is a multi-mode interferometer (MMI), and the outgoing percentage and the incoming percentage are 50% or approximately 50%. As will be described in more detail below, the processing component 22 combines the comparison signal with the reference signal to form a composite signal, which carries the lidar data of the sample area on the field of view. Thus, the composite signal can be processed to extract the lidar data of the sample area (radial velocity and / or distance between the lidar system and objects external to the lidar system).

[0042] The lidar chip may include a control branch for controlling the operation of the light source 4. The control branch includes a splitter 26 that moves a portion of the outgoing lidar signal from the utility waveguide 12 to a control waveguide 28. The coupled portion of the outgoing lidar signal serves as a tap signal. Figure 1A A directional coupler is shown for use as splitter 26, but other signal tapping components may also be used as splitter 26. Suitable splitters 26 include, but are not limited to, directional couplers, optical couplers, Y-junctions, tapered couplers, and multi-mode interference (MMI) devices.

[0043] The control waveguide 28 carries the tap signal to the control component 30. The control component may be in electrical communication with the electronics 32. All or a portion of the control component may be included in the electronics 32. During operation, the electronics may employ an output of the control component 30 from a control loop configured to control one, two, or three process variables of the loop control optical signal selected from the group consisting of the tap signal, the system output signal, and the emitted laser radar signal. Examples of suitable process variables include the frequency of the loop control optical signal and / or the phase of the loop control optical signal.

[0044] The lidar system can be modified so that the incoming lidar signal and the outgoing lidar signal are carried on different waveguides. For example, Figure 1B yes Figure 1A A top view of a lidar chip modified so that the incoming lidar signal and the outgoing lidar signal are carried on different waveguides. The outgoing lidar signal leaves the lidar chip through facet 14 and serves as a lidar output signal. When light from the lidar output signal is reflected by an object outside the lidar system, at least a portion of the reflected light returns to the lidar chip as a first lidar input signal. The first lidar input signal enters comparison waveguide 18 through facet 35 and serves as a comparison signal. Comparison waveguide 18 carries the comparison signal to processing component 22 for further processing. As shown in Figure 1A As described in the context of FIG. 1 , reference waveguide 20 carries the reference signal to processing component 22 for further processing. As will be described in more detail below, processing component 22 combines the comparison signal with the reference signal to form a composite signal that carries lidar data for a sample area on the field of view.

[0045] The lidar chip can be modified to receive multiple lidar input signals. For example, Figure 1C shows a modification to receive two laser radar input signals Figure 1BThe laser radar chip of the present invention. The separator 40 is configured to place a portion of the reference signal carried on the reference waveguide 20 on the first reference waveguide 42 and place another portion of the reference signal on the second reference waveguide 44. Therefore, the first reference waveguide 42 carries the first reference signal and the second reference waveguide 44 carries the second reference signal. The first reference waveguide 42 carries the first reference signal to the first processing component 46, and the second reference waveguide 44 carries the second reference signal to the second processing component 48. Examples of suitable separators 40 include, but are not limited to, Y junctions, optical couplers, and multi-mode interference couplers (MMI).

[0046] The emitted LiDAR signal leaves the LiDAR chip through facet 14 and is used as a LiDAR output signal. When light from the LiDAR output signal is reflected by one or more objects outside the LiDAR system, at least a portion of the reflected light returns to the LiDAR chip as a first LiDAR input signal. The first LiDAR input signal enters the comparison waveguide 18 through facet 35 and is used as a first comparison signal. The comparison waveguide 18 carries the first comparison signal to the first processing component 46 for further processing.

[0047] In addition, when light from the lidar output signal is reflected by one or more objects located outside the lidar system, at least a portion of the reflected signal is returned to the lidar chip as a second lidar input signal. The second lidar input signal enters the second comparison waveguide 50 through the facet 52 and serves as a second comparison signal carried by the second comparison waveguide 50. The second comparison waveguide 50 carries the second comparison signal to the second processing component 48 for further processing.

[0048] Although the light source 4 is shown as being located on the LiDAR chip, the light source 4 may be located outside the LiDAR chip. For example, the utility waveguide 12 may terminate at a second facet through which an emitted LiDAR signal may enter the utility waveguide 12 from the light source 4 located outside the LiDAR chip.

[0049] In some cases, according to Figure 1B or Figure 1CThe constructed lidar chip is used in combination with a lidar adapter. In some cases, the lidar adapter can be physically and optically positioned between the lidar chip and one or more reflective objects and / or the field of view because the optical path of (one or more) first lidar input signals and / or lidar output signals from the lidar chip to the field of view passes through the lidar adapter. In addition, the lidar adapter can be configured to operate on the first lidar input signal and the lidar output signal so that the first lidar input signal and the lidar output signal travel on different optical paths between the lidar adapter and the lidar chip, but travel on the same optical path between the lidar adapter and the reflective objects in the field of view.

[0050] Applies to Figure 1B An example of a lidar adapter for a lidar chip is Figure 2 . The lidar adapter includes a plurality of components positioned on a base. For example, the lidar adapter includes a circulator 100 positioned on a base 102. The optical circulator 100 shown includes three ports and is configured such that light entering one port exits from the next port. For example, the optical circulator shown includes a first port 104, a second port 106, and a third port 108. The lidar output signal enters the first port 104 from the utility waveguide 12 of the lidar chip and exits from the second port 106.

[0051] The lidar adapter can be configured such that the output of the lidar output signal from the second port 106 can also be used as the output of the lidar output signal from the lidar adapter and therefore from the lidar system. Therefore, the lidar output signal can be output from the lidar adapter such that the lidar output signal travels toward the sample area in the field of view. Therefore, in some cases, the portion of the lidar output signal that has left the lidar adapter can also be considered to be the system output signal. As an example, when the lidar output signal leaving the lidar adapter is also the lidar output signal leaving the lidar system, the lidar output signal can also be considered to be the system output signal.

[0052] The lidar output signal output from the lidar adapter includes, consists of, or consists essentially of light from the lidar output signal received from the lidar chip. Therefore, the lidar output signal output from the lidar adapter can be the same or essentially the same as the lidar output signal received from the lidar chip. However, there may be differences between the lidar output signal output from the lidar adapter and the lidar output signal received from the lidar chip. For example, the lidar output signal may experience optical losses as it travels through the lidar adapter, and / or the lidar adapter may optionally include an amplifier configured to amplify the lidar output signal as the lidar output signal travels through the lidar adapter.

[0053] When one or more objects in the sample area reflect the LiDAR output signal, at least a portion of the reflected light is returned to the circulator 100 as a system return signal. The system return signal enters the circulator 100 through the second port 106. Figure 2 The lidar output signal and the system return signal are shown traveling along the same optical path between the lidar adapter and the sample area.

[0054] The system return signal exits the circulator 100 through the third port 108 and is directed to the comparison waveguide 18 on the lidar chip. Thus, all or a portion of the system return signal can be used as the first lidar input signal, and the first lidar input signal includes or consists of light from the system return signal. Thus, the lidar output signal and the first lidar input signal travel along different optical paths between the lidar adapter and the lidar chip.

[0055] from Figure 2 It is apparent that the laser radar adapter may include optical components in addition to the circulator 100. For example, the laser radar adapter may include components for directing and controlling the optical path of the laser radar output signal and the system return signal. As an example, Figure 2 The adapter includes an optional amplifier 110 positioned to receive and amplify the Lidar output signal before the Lidar output signal enters the circulator 100. The amplifier 110 may be operated by the electronics 32, thereby allowing the electronics 32 to control the power of the Lidar output signal.

[0056] Figure 2Also shown is a lidar adapter including an optional first lens 112 and an optional second lens 114. The first lens 112 can be configured to couple the lidar output signal to a desired location. In some cases, the first lens 112 is configured to focus or collimate the lidar output signal at a desired location. In one example, when the lidar adapter does not include an amplifier 110, the first lens 112 is configured to couple the lidar output signal to the first port 104. As another example, when the lidar adapter includes an amplifier 110, the first lens 112 can be configured to couple the lidar output signal on the inlet port to the amplifier 110. The second lens 114 can be configured to couple the lidar output signal to a desired location. In some cases, the second lens 114 is configured to focus or collimate the lidar output signal at a desired location. For example, the second lens 114 can be configured to couple the lidar output signal to the facet 35 of the comparison waveguide 18.

[0057] The lidar adapter may also include one or more direction changing components, such as mirrors. Figure 2 A LiDAR adapter is shown that includes a reflector as a direction changing component 116 that redirects the system return signal from the circulator 100 to the facet 20 of the comparison waveguide 18 .

[0058] The lidar chip includes one or more waveguides that constrain the optical path of one or more optical signals. Although the lidar adapter may include waveguides, the optical path that the system return signal and the lidar output signal travel between components on the lidar adapter and / or between the lidar chip and components on the lidar adapter can be free space. For example, when traveling between different components on the lidar adapter and / or between components on the lidar adapter and the lidar chip, the system return signal and / or the lidar output signal can travel through the atmosphere in which the lidar chip, the lidar adapter and / or the base 102 are located. Therefore, optical components such as lenses and direction changing components can be used to control the characteristics of the optical path that the system return signal and the lidar output signal travel on the lidar adapter and to and from the lidar adapter.

[0059] Suitable bases 102 for the lidar adapter include, but are not limited to, substrates, platforms, and plates. Suitable substrates include, but are not limited to, glass, silicon, and ceramics. The components may be discrete components attached to a substrate. Suitable techniques for attaching discrete components to the base 102 include, but are not limited to, epoxy, solder, and mechanical clamping. In one example, one or more of the components are integrated components and the remaining components are discrete components. In another example, the lidar adapter includes one or more integrated amplifiers and the remaining components are discrete components.

[0060] A lidar system can be configured to compensate for polarization. Light from a laser source is typically linearly polarized, and therefore the lidar output signal is typically also linearly polarized. Reflections from an object can change the polarization angle of the returned light. Therefore, the system return signal can include light of different linear polarization states. For example, a first portion of the system return signal may include light of a first linear polarization state, and a second portion of the system return signal may include light of a second linear polarization state. The intensity of the resulting composite signal is proportional to the square of the cosine of the angle between the comparison and reference signal polarization fields. If the angle is 90 degrees, the lidar data may be lost in the resulting composite signal. However, the lidar system can be modified to compensate for changes in the polarization state of the lidar output signal.

[0061] Figure 3 Shows Figure 3 A lidar system modified so that the lidar adapter is suitable for use with Figure 1C The laser radar adapter includes a beam splitter 120 that receives the system return signal from the circulator 100. The beam splitter 120 splits the system return signal into a first part of the system return signal and a second part of the system return signal. Suitable beam splitters include, but are not limited to, Wollaston prisms and MEMS-based beam splitters.

[0062] The first part of the system return signal is directed to the comparison waveguide 18 on the lidar chip and used as Figure 1C A second portion of the system return signal is directed to a polarization rotator 122. The polarization rotator 122 outputs a second lidar input signal, which is directed to a second input waveguide 76 on the lidar chip and used as a second lidar input signal.

[0063] The beam splitter 120 may be a polarization beam splitter. An example of a polarization beam splitter is configured such that a first portion of the system return signal has a first polarization state but does not have or substantially does not have a second polarization state, and a second portion of the system return signal has a second polarization state but does not have or substantially does not have the first polarization state. The first polarization state and the second polarization state may be linear polarization states, and the second polarization state is different from the first polarization state. For example, the first polarization state may be TE and the second polarization state may be TM, or the first polarization state may be TM and the second polarization state may be TE. In some cases, the laser source may be linearly polarized so that the laser radar output signal has a first polarization state. Suitable beam splitters include, but are not limited to, Wollaston prisms and MEMS-based polarization beam splitters.

[0064] The polarization rotator can be configured to change the polarization state of the first portion of the system return signal and / or the second portion of the system return signal. Figure 3 The polarization rotator 122 shown in can be configured to change the polarization state of the second part of the system return signal from the second polarization state to the first polarization state. Therefore, the second laser radar input signal has the first polarization state, but does not have or substantially does not have the second polarization state. Therefore, the first laser radar input signal and the second laser radar input signal both have the same polarization state (the first polarization state in this example). Although carrying light of the same polarization state, due to the use of a polarization beam splitter, the first laser radar input signal and the second laser radar input signal are associated with different polarization states. For example, the first laser radar input signal carries light reflected in the first polarization state, and the second laser radar input signal carries light reflected in the second polarization state. Therefore, the first laser radar input signal is associated with the first polarization state, and the second laser radar input signal is associated with the second polarization state.

[0065] Since the first lidar input signal and the second lidar input signal carry light of the same polarization state, the comparison signal generated by the first lidar input signal has the same polarization angle as the comparison signal generated by the second lidar input signal.

[0066] Suitable polarization rotators include, but are not limited to, Faraday rotators, half-wave plates, MEMS-based polarization rotators, and integrated optical polarization rotators using asymmetric Y-branches, Mach-Zehnder interferometers, multimode interference couplers, and rotation of polarization maintaining fibers.

[0067] Since the emitted laser radar signal is linearly polarized, the first reference signal can have the same linear polarization state as the second reference signal. In addition, the components on the laser radar adapter can be selected so that the first reference signal, the second reference signal, the comparison signal, and the second comparison signal all have the same polarization state. Figure 3 In the examples disclosed in the context of , the first comparison signal, the second comparison signal, the first reference signal, and the second reference signal may all have light of a first polarization state.

[0068] As a result of the above configuration, the first composite signal generated by the first processing component 46 and the second composite signal generated by the second processing component 48 are both generated by combining the reference signal and the comparison signal of the same polarization state, and will provide the expected beat between the reference signal and the comparison signal accordingly. For example, the composite signal is generated by combining the first reference signal and the first comparison signal of the first polarization state, and excludes or substantially excludes light of the second polarization state, or the composite signal is generated by combining the first reference signal and the first comparison signal of the second polarization state, and excludes or substantially excludes light of the first polarization state. Similarly, the second composite signal includes the second reference signal and the second comparison signal of the same polarization state, and will provide the expected beat between the reference signal and the comparison signal accordingly. For example, the second composite signal is generated by combining the second reference signal and the second comparison signal of the first polarization state, and excludes or substantially excludes light of the second polarization state, or the second composite signal is generated by combining the second reference signal and the second comparison signal of the second polarization state, and excludes or substantially excludes light of the first polarization state.

[0069] The above configuration results in lidar data for a single sample area in the field of view being generated from a plurality of different composite signals (i.e., a first composite signal and a second composite signal) from the sample area. In some cases, determining the lidar data for the sample area includes an electronic device combining lidar data from different composite signals (i.e., the composite signal and the second composite signal). The combined lidar data may include taking an average, median, or mode of the lidar data generated from the different composite signals. For example, the electronic device may use the distance determined from the second composite signal to average the distance between the lidar system and the reflecting object determined from the composite signal, and / or the electronic device may use the radial velocity determined from the second composite signal to average the radial velocity between the lidar system and the reflecting object determined from the composite signal.

[0070] In some cases, determining the laser radar data of the sample area includes the electronic device identifying one or more composite signals (i.e., the composite signal and / or the second composite signal) as the source of the laser radar data (representative laser radar data) that best represents reality. The electronic device can then use the laser radar data from the identified composite signal as the representative laser radar data to be used for additional processing. For example, the electronic device can identify a signal (composite signal or second composite signal) with a larger amplitude as representative laser radar data, and can use the laser radar data from the identified signal for further processing by the laser radar system. In some cases, the electronic device combines the composite signal that identifies the representative laser radar data with the laser radar data that is combined from different laser radar signals. For example, the electronic device can identify each composite signal with an amplitude above an amplitude threshold as representative laser radar data, and when more than two composite signals are identified as representative laser radar data, the electronic device can combine the laser radar data from each identified composite signal. When a composite signal is identified as representative laser radar data, the electronic device can use the laser radar data from the composite signal as representative laser radar data. When no composite signals are identified as representative lidar data, the electronics may discard lidar data for the sample area associated with those composite signals.

[0071] although Figure 3 Described in the context of components being arranged such that the first comparison signal, the second comparison signal, the first reference signal, and the second reference signal all have a first polarization state, but Figure 3 Other configurations of the components in can be arranged so that a composite signal is generated by combining the reference signal and the comparison signal of the same linear polarization state and a second composite signal is generated by combining the reference signal and the comparison signal of the same linear polarization state. For example, the beam splitter 120 can be configured so that the second portion of the system return signal has a first polarization state and the first portion of the system return signal has a second polarization state, the polarization rotator receives the first portion of the system return signal, and the emitted lidar signal can have the second polarization state. In this example, both the first lidar input signal and the second lidar input signal have the second polarization state.

[0072] The above system configuration causes the first portion of the system return signal and the second portion of the system return signal to be directed into different composite signals. Therefore, because the first portion of the system return signal and the second portion of the system return signal are each associated with a different polarization state, but the electronic device can process each composite signal, the lidar system compensates for changes in the polarization state of the lidar output signal in response to reflections of the lidar output signal.

[0073] Figure 3The lidar adapter may include additional optical components, including passive optical components. For example, the lidar adapter may include an optional third lens 126. The third lens 126 may be configured to couple the second lidar output signal at a desired location. In some cases, the third lens 126 focuses or collimates the second lidar output signal at a desired location. For example, the third lens 126 may be configured to focus or collimate the second lidar output signal on the facet 52 of the second comparison waveguide 50. The lidar adapter also includes one or more direction changing components 124, such as reflectors and prisms. Figure 3 A LiDAR adapter is shown that includes a reflector as a direction changing component 124 that redirects a second portion of the system return signal from the circulator 100 to the facet 52 of the second comparison waveguide 50 and / or a third lens 126 .

[0074] When the LiDAR system includes a LiDAR chip and a LiDAR adapter, the LiDAR chip, the electronic device, and the LiDAR adapter can be positioned on a common mounting member. Suitable common mounting members include, but are not limited to, glass plates, metal plates, silicon plates, and ceramic plates. As an example, Figure 4 The common support 140 includes Figure 1A LiDAR chips and electronic devices 32 and Figure 2 140. A top view of a lidar system of a lidar adapter. Although the electronics 32 are shown as being located on a common support, all or a portion of the electronics may be located outside the common support. When the light source 4 is located outside the lidar chip, the light source may be located on or outside the common support 140. Suitable methods for mounting the lidar chip, electronics, and / or lidar adapter on a common support include, but are not limited to, epoxy, solder, and mechanical clamping.

[0075] The lidar system may include components with additional passive and / or active optical components. For example, the lidar system may include one or more components that receive a lidar output signal from a lidar chip or from a lidar adapter. Portions of the lidar output signal that exit from the one or more components may be used as a system output signal. As an example, the lidar system may include one or more beam steering components that receive a lidar output signal from a lidar chip or from a lidar adapter and output all or a portion of the lidar output signal for use as a system output signal. For example, Figure 4 A beam steering component 142 is shown that receives the LiDAR output signal from the LiDAR adapter. Figure 4The beam steering components are shown positioned on the common support 140, but the beam steering components may be positioned on the lidar chip, on the lidar adapter, external to the lidar chip, or external to the common support 140. Suitable beam steering components include, but are not limited to, movable mirrors, MEMS mirrors, optical phased arrays (OPAs), and actuators that move the lidar chip, lidar adapter, and / or common support.

[0076] The electronics may operate one or more beam steering components 142 to steer the system output signal to different sample regions 144. The sample regions may extend away from the lidar system to a maximum distance at which the lidar system is configured to provide reliable lidar data. The sample regions may be stitched together to define a field of view. For example, the field of view of the lidar system includes or consists of the space occupied by a combination of sample regions.

[0077] FIG. 5A to FIG. 5C An example of a suitable processing component is shown for use as all or part of a processing component selected from the group consisting of processing component 22, first processing component 46, and second processing component 48. The processing component receives a comparison signal from comparison waveguide 196 and a reference signal from reference waveguide 198. Figure 1A and Figure 1B The comparison waveguide 18 and the reference waveguide 20 shown in FIG. 1 can be used as the comparison waveguide 196 and the reference waveguide 198, Figure 1C The comparison waveguide 18 and the first reference waveguide 42 shown in FIG. 4 can be used as the comparison waveguide 196 and the reference waveguide 198, or Figure 1C The second comparison waveguide 50 and the second reference waveguide 44 shown in FIG. 4 may be used as the comparison waveguide 196 and the reference waveguide 198 .

[0078] The processing means comprises a second splitter 200 which divides the comparison signal carried on the comparison waveguide 196 into a first comparison waveguide 204 and a second comparison waveguide 206. The first comparison waveguide 204 carries a first portion of the comparison signal to an optical combining means 211. The second comparison waveguide 208 carries a second portion of the comparison signal to a second optical combining means 212.

[0079] The processing component includes a first splitter 202 that divides the reference signal carried on the reference waveguide 198 into a first reference waveguide 204 and a second reference waveguide 206. The first reference waveguide 204 carries a first portion of the reference signal to an optical combining component 211. The second reference waveguide 208 carries a second portion of the reference signal to a second optical combining component 212.

[0080] The second optical combining component 212 combines the second part of the comparison signal and the second part of the reference signal into a second composite signal. Due to the frequency difference between the second part of the comparison signal and the second part of the reference signal, the second composite signal is beat between the second part of the comparison signal and the second part of the reference signal.

[0081] The second optical combining component 212 also splits the resulting second composite signal onto a first auxiliary detector waveguide 214 and a second auxiliary detector waveguide 216. The first auxiliary detector waveguide 214 carries a first portion of the second composite signal to a first auxiliary optical sensor 218, which converts the first portion of the second composite signal into a first auxiliary electrical signal. The second auxiliary detector waveguide 216 carries a second portion of the second composite signal to a second auxiliary optical sensor 220, which converts the second portion of the second composite signal into a second auxiliary electrical signal. Examples of suitable optical sensors include germanium photodiodes (PDs) and avalanche photodiodes (APDs).

[0082] In some cases, the second optical combination component 212 separates the second composite signal so that a portion of the comparison signal included in the first portion of the second composite signal (i.e., a portion of the second portion of the comparison signal) is phase-shifted by 180° relative to a portion of the comparison signal in the second portion of the second composite signal (i.e., a portion of the second portion of the comparison signal), but a portion of the reference signal in the second portion of the second composite signal (i.e., a portion of the second portion of the reference signal) is not phase-shifted relative to a portion of the reference signal in the first portion of the second composite signal (i.e., a portion of the second portion of the reference signal). Alternatively, the second optical combination component 212 separates the second composite signal so that a portion of the reference signal in the first portion of the second composite signal (i.e., a portion of the second portion of the reference signal) is phase-shifted by 180° relative to a portion of the reference signal in the second portion of the second composite signal (i.e., a portion of the second portion of the reference signal), but a portion of the comparison signal in the first portion of the second composite signal (i.e., a portion of the second portion of the comparison signal) is not phase-shifted relative to a portion of the comparison signal in the second portion of the second composite signal (i.e., a portion of the second portion of the comparison signal). Examples of suitable optical sensors include germanium photodiodes (PDs) and avalanche photodiodes (APDs).

[0083] The first optical combining component 211 combines the first part of the comparison signal and the first part of the reference signal into a first composite signal. Due to the frequency difference between the first part of the comparison signal and the first part of the reference signal, the first composite signal is beat between the first part of the comparison signal and the first part of the reference signal.

[0084] The first optical combination component 211 also separates the first composite signal onto a first detector waveguide 221 and a second detector waveguide 222. The first detector waveguide 221 carries a first portion of the first composite signal to a first light sensor 223, which converts the first portion of the second composite signal into a first electrical signal. The second detector waveguide 222 carries a second portion of the second composite signal to a second light sensor 224, which converts the second portion of the second composite signal into a second electrical signal. Examples of suitable light sensors include germanium photodiodes (PDs) and avalanche photodiodes (APDs).

[0085] In some cases, the optical combination component 211 separates the first composite signal so that a portion of the comparison signal included in the first portion of the composite signal (the portion of the first portion of the comparison signal) is phase-shifted by 180° relative to a portion of the comparison signal in the second portion of the composite signal (i.e., the portion of the first portion of the comparison signal), but a portion of the reference signal in the first portion of the composite signal (i.e., the portion of the first portion of the reference signal) is not phase-shifted relative to a portion of the reference signal in the second portion of the composite signal (i.e., the portion of the first portion of the reference signal). Alternatively, the optical combination component 211 separates the composite signal so that a portion of the reference signal in the first portion of the composite signal (i.e., the portion of the first portion of the reference signal) is phase-shifted by 180° relative to a portion of the reference signal in the second portion of the composite signal (i.e., the portion of the first portion of the reference signal), but a portion of the comparison signal in the first portion of the composite signal (i.e., the portion of the first portion of the comparison signal) is not phase-shifted relative to a portion of the comparison signal in the second portion of the composite signal (i.e., the portion of the first portion of the comparison signal).

[0086] When the second optical combination component 212 separates the second composite signal so that the portion of the comparison signal in the first portion of the second composite signal is phase-shifted by 180° relative to the portion of the comparison signal in the second portion of the second composite signal, the optical combination component 211 also separates the composite signal so that the portion of the comparison signal in the first portion of the composite signal is phase-shifted by 180° relative to the portion of the comparison signal in the second portion of the composite signal. When the second optical combination component 212 separates the second composite signal so that the portion of the reference signal in the first portion of the second composite signal is phase-shifted by 180° relative to the portion of the reference signal in the second portion of the second composite signal, the optical combination component 211 also separates the composite signal so that the portion of the reference signal in the first portion of the composite signal is phase-shifted by 180° relative to the portion of the reference signal in the second portion of the composite signal.

[0087] The first reference waveguide 210 and the second reference waveguide 208 are configured to provide a phase shift between a first portion of the reference signal and a second portion of the reference signal. For example, the first reference waveguide 210 and the second reference waveguide 208 can be configured to provide a 90 degree phase shift between the first portion of the reference signal and the second portion of the reference signal. As an example, one reference signal portion can be an in-phase component and the other can be an orthogonal component. Therefore, one of the reference signal portions can be a sine function and the other reference signal portion can be a cosine function. In one example, the first reference waveguide 210 and the second reference waveguide 208 are configured so that the first reference signal portion is a cosine function and the second reference signal portion is a sine function. Therefore, a portion of the reference signal in the second composite signal is phase shifted relative to a portion of the reference signal in the first composite signal, however, a portion of the comparison signal in the first composite signal is not phase shifted relative to a portion of the comparison signal in the second composite signal.

[0088] The first light sensor 223 and the second light sensor 224 may be connected as a balanced detector, and the first auxiliary light sensor 218 and the second auxiliary light sensor 220 may also be connected as a balanced detector. Figure 5B A schematic diagram of the relationship between the electronics, first light sensor 223, second light sensor 224, first auxiliary light sensor 218, and second auxiliary light sensor 220 is provided. The symbol of a photodiode is used to represent first light sensor 223, second light sensor 224, first auxiliary light sensor 218, and second auxiliary light sensor 220, but one or more of these sensors may have other configurations. In some cases, Figure 5B All components shown in the schematic are included on the lidar chip. In some cases, Figure 5B The components shown in the schematic are distributed between the lidar chip and the electronic devices located outside the lidar chip.

[0089] The electronic device connects the first light sensor 223 and the second light sensor 224 as a first balanced detector 225, and connects the first auxiliary light sensor 218 and the second auxiliary light sensor 220 as a second balanced detector 226. In particular, the first light sensor 223 and the second light sensor 224 are connected in series. In addition, the first auxiliary light sensor 218 and the second auxiliary light sensor 220 are connected in series. The series connection in the first balanced detector communicates with the first data line 228, which carries the output from the first balanced detector as a first data signal. The series connection in the second balanced detector communicates with the second data line 232, which carries the output from the second balanced detector as a second data signal. The first data signal is an electrical representation of the first composite signal, and the second data signal is an electrical representation of the second composite signal. Therefore, the first data signal includes contributions from the first waveform and the second waveform, and the second data signal is a composite of the first waveform and the second waveform. The portion of the first waveform in the first data signal is phase-shifted relative to the portion of the first waveform in the first data signal, but the portion of the second waveform in the first data signal is in phase relative to the portion of the second waveform in the first data signal. For example, the second data signal includes a portion of the reference signal that is phase-shifted relative to a different portion of the reference signal included in the first data signal. In addition, the second data signal includes a portion of the comparison signal that is in phase with a different portion of the comparison signal included in the first data signal. The first data signal and the second data signal beat due to a beat between the comparison signal and the reference signal (i.e., a beat in the first composite signal and the second composite signal).

[0090] The electronic device 32 includes a transformation mechanism 238, which is configured to perform a mathematical transformation on the first data signal and the second data signal. For example, the mathematical transformation can be a complex Fourier transform with the first data signal and the second data signal as input. Since the first data signal is an in-phase component and the second data signal is an orthogonal component, the first data signal and the second data signal together act as a complex data signal, where the first data signal is the real component of the input and the second data signal is the imaginary component of the input.

[0091] The conversion mechanism 238 includes a first analog-to-digital converter (ADC) 264 that receives a first data signal from the first data line 228. The first analog-to-digital converter (ADC) 264 converts the first data signal from an analog form to a digital form and outputs a first digital data signal. The conversion mechanism 238 includes a second analog-to-digital converter (ADC) 266 that receives a second data signal from the second data line 232. The second analog-to-digital converter (ADC) 266 converts the second data signal from an analog form to a digital form and outputs a second digital data signal. The first digital data signal is a digital representation of the first data signal, and the second digital data signal is a digital representation of the second data signal. Therefore, the first digital data signal and the second digital data signal together act as a complex signal, wherein the first digital data signal acts as a real component of the complex signal, and the second digital data signal acts as an imaginary component of the complex data signal.

[0092] The transformation mechanism 238 includes a transformation component 268 that receives a complex data signal. For example, the transformation component 268 receives a first digital data signal as an input from a first analog-to-digital converter (ADC) 264, and also receives a second digital data signal as an input from a second analog-to-digital converter (ADC) 266. The transformation component 268 can be configured to perform a mathematical transformation on the complex signal to convert from the time domain to the frequency domain. The mathematical transformation can be a complex transformation such as a complex fast Fourier transform (FFT). A complex transformation such as a complex fast Fourier transform (FFT) provides an explicit solution for the frequency shift of the laser radar input signal relative to the laser radar output signal caused by the radial velocity between the reflecting object and the laser radar chip. The electronic device uses one or more frequency peaks output from the transformation component 268 for further processing to generate laser radar data (the distance and / or radial velocity between the reflecting object and the laser radar chip or the laser radar system). The transformation component 268 can use firmware, hardware, or software, or a combination thereof, to perform the attribute function.

[0093] The electronic device 32 includes a peak finder 270 that receives the output from the transform component 268. The peak finder 270 is configured to find a peak in the output of the transform component 268 in order to identify the beat frequency of the composite optical signal. The mathematical transform can be a complex transform such as a complex fast Fourier transform (FFT). A complex transform such as a complex fast Fourier transform (FFT) provides an unambiguous solution for the beat frequency of the composite optical signal. The peak finder 270 can store the beat frequency in a memory 271 for later use by a data refinement component 272 and a lidar data generator 274. As will be described in more detail below, the beat frequency can be stored as f m,j , where m represents the time period index and j represents the cycle index. The laser radar data generator 274 uses the beat frequency to generate laser radar data (the distance and / or radial velocity between the reflecting object and the laser radar chip or laser radar system). Suitable memory 271 includes, but is not limited to, a buffer.

[0094] Peak finder 270 may also store a power indicator (power level indicator) of the system return signal in memory 271. One power level indicator is the magnitude of the peak in the output of transform component 268. Thus, the magnitude of the peak frequency may be used as the power level indicator. Alternatively, the power of the system return signal may be calculated from the peak frequency and may be used as the power level indicator. The power level indicator may be obtained by P m,j , where m represents the time period index and j represents the cycle index. Therefore, each power level indicator can be associated with the data period, beat frequency (f m,j ) and / or average beat frequency (af m,j The power level indicator may be stored in the memory 271 for access by the data refinement component 272 and / or the lidar data generator 274.

[0095] although Figure 5A An optical combining component is shown that combines a portion of the reference signal with a portion of the comparison signal, but the processing component may include a single optical combining component that combines the reference signal with the comparison signal to form a composite signal. Thus, at least a portion of the reference signal and at least a portion of the comparison signal may be combined to form the composite signal. The combined portion of the reference signal may be the entire reference signal or a portion of the reference signal, and the combined portion of the comparison signal may be the entire comparison signal or a portion of the comparison signal.

[0096] The electronics tune the frequency of the system output signal over time. The system output signal has a frequency versus time pattern with a repeating period. Figure 5C An example of a suitable frequency versus time pattern of the system output signal is shown. The fundamental frequency (f o ) can be the frequency of the system output signal at the beginning of the cycle.

[0097] Figure 5C The cycle marked j and Cycle j+1 The frequency versus time relationship of a sequence of two cycles of , where j represents the cycle index. In some cases, the frequency versus time pattern is repeated in each cycle, such as Figure 5C The cycles shown do not include repositioning periods and / or the repositioning periods are not located between cycles. Therefore, Figure 5C Results of consecutive scans are shown.

[0098] Each cycle includes M data periods, each of which is associated with a period index m and is labeled DP m .exist Figure 5C In the example, each cycle includes a DP mThe frequency-time pattern is repeated for the data periods corresponding to each other (associated data periods) in different cycles, such as Figure 5C As shown in . Corresponding data periods are data periods with the same cycle index. Therefore, each data period DP1 can be considered as a corresponding data period, and the associated frequency-to-time pattern is Figure 5C At the end of the cycle, the electronics returns the frequency to the same frequency level at which it started the previous cycle.

[0099] In the data period DP m During this time, the electronics operates the light source so that the frequency of the system output signal changes at a linear rate α. m (Chirp rate) changes. Figure 5C , α2=-α1.

[0100] Figure 5C Mark sample regions, each sample region is associated with a sample region index k and marked as SR k . Figure 5C The sample area SR is marked k-1 To SR k+1 .exist Figure 5C Each sample region is illuminated with the system output signal during the data period associated with the sample region. For example, during the data period labeled DP2 in cycle j+1 and the data period labeled DP1 in cycle j+1, the sample region SR is illuminated with the system output signal k+1 Therefore, it is marked as SR k+1 The sample regions of SR1 and SR2 are associated with the data periods labeled DP1 and DP2 in cycle j+1. The sample region index k can be assigned relative to time. For example, the sample region can be illuminated by the system output signal in the sequence indicated by the index k. Thus, the sample region can be illuminated after the sample region SR9 and before the sample region SR1. 11 Before irradiating the sample area SR 10 .

[0101] The frequency output from the complex Fourier transform represents the beat frequency of the composite signal, each composite signal comprising the beat difference of the comparison signal and the reference signal. The beat frequencies from two or more different data periods associated with the same sample area can be combined to generate the lidar data. For example, when illuminating the sample area SR k The beat frequency determined from DP1 during the illumination of the sample region SR k The beat frequency combination determined from DP2 during the period is used to determine the sample region SR k As an example, the following equation is applied during a data period where the electronics increases the frequency of emitting the lidar signal during the data period, such as during Figure 5C The data period DP1 occurs: fub =-f d +α u τ, where f ub is the average frequency value af described below 1,j , τ is the round trip delay (the time between the system output signal leaving the lidar system and the system return signal returning to the lidar system), f d represents the Doppler frequency shift (f d =2V k f c / c), where f c Indicates the optical frequency (f o ), c represents the speed of light, V k is the radial velocity between the reflecting object and the lidar system, where the direction from the reflecting object toward the chip is assumed to be the positive direction, c is the speed of light, and α u represents the chirp rate during the data period (α m ), where the frequency of the system output signal increases over time (in this case α1). The following equation applies during the data period when the electronics reduces the frequency of the emitted lidar signal, such as during Figure 5C The data period DP2 occurs: db =-f d -α d τ, where f db is the average frequency value af described below 2,j , and α d represents the chirp rate during the data period (α m ), where the frequency of the system output signal increases with time (in this case by α2). In both equations, f d and τ are unknowns. These equations can be solved for the two unknowns, and the electronics can then k =c*f d / (2f c )) Determine the radial velocity (V) of sample area k k ), and / or the separation distance (R k ). Therefore, the average frequency value is a variable in the calculation of the lidar data. The radial velocity and / or separation distance determined for the sample area can represent the lidar data for the sample area.

[0102] Figure 5D Another approach for generating lidar data is shown. For example, Figure 5D Shows Figure 5C The frequency versus time pattern of , but modified so that the sample region is labeled SR k-1 To SR k+3 .exist Figure 5CEach sample region is illuminated with the system output signal during the data period associated with the sample region. For example, during the data period labeled DP2 in cycle j and the data period labeled DP1 in cycle j+1, the sample region SR is illuminated with the system output signal k+1 Therefore, it is marked as SR k+1 The sample region is associated with the data period labeled DP2 in cycle j and the data period labeled DP1 in cycle j+1.

[0103] from Figure 5D It is obvious that the data period is associated with multiple sample regions. For example, the data period marked as DP2 in period j is associated with the data period marked as SR k+1 The sample area and the SR k The sample areas are associated with the same data period. Therefore, different data period groups can share a common data period. However, the groups sharing a common data period can each include one or more data periods that are not shared by the group. The sample areas are illuminated by the system output signal during the associated data period, and different sample areas can be associated with the same data period, so different sample areas can overlap each other.

[0104] The system output signal may be chirped during at least a portion of a data period in the same cycle. The chirp may be constant and last for the duration of the data period (T). For example, during a data period labeled DP1 and a data period labeled DP2, the electronics operates the light source so that the frequency of the system output signal changes at a linear rate α(B / T). The direction of the frequency change during data period DP1 is opposite to the direction of the frequency change during data period DP2. Thus, the chirp of the system output signal may be different for different data periods within the same cycle.

[0105] The frequency output from the complex Fourier transform represents the beat frequency of the composite signals, each of which includes the beats of the comparison signal and the reference signal. The beat frequency (f LDP ) generates lidar data for the sample area. Thus, the electronics generates a lidar data set for the sample area from light included in the system output signal during a set of multiple data periods. For example, the beat frequency determined from DP1 in period j can be combined with the beat frequency determined from DP2 in period j to determine a value labeled SR k Thus, the electronic device generates a sample region SR from a set of data periods including the data periods labeled DP1 and DP2 in cycle j. k LiDAR dataset.

[0106] As an example of how to determine lidar data for a sample area from a set of data periods, the following equation is applied during a data period where the electronics increases the frequency of emitting lidar signals during the data period, such as during Figure 5D In the data period DP1 (m=1) of cycle j, the following occurs: ub =-f d +ατ, where f ub is the average frequency value af described below 1,j , τ is the round trip delay (the time between the system output signal leaving the lidar system and the system return signal returning to the lidar system), f d represents the Doppler frequency shift (f d =2νf c / c), where f c Indicates the optical frequency (f o ), c represents the speed of light, ν is the radial velocity between the reflecting object and the lidar system, where the direction from the reflecting object toward the lidar system is assumed to be the positive direction, and c is the speed of light. The following equation applies during the data period when the electronics reduces the frequency of the emitted lidar signal, such as during Figure 5D In the data period DP2 (m=2) of cycle j, the following occurs: db =-f d -ατ, where f db is the average frequency value af described below 2,j In both equations, f d and τ are unknown. The electronic device is designed for two unknowns f d and τ to solve these two equations. Then according to the Doppler frequency shift (ν=c*f d / (2f c )) quantifies the radial velocity of the sample region, and / or the separation distance of the sample region can be quantified according to c*τ / 2. Therefore, the average frequency value is a variable in the calculation of the lidar data.

[0107] The above example discloses the j ) generates a sample region (SR k However, the lidar data of the sample area can be generated for a set of data periods including data periods from different cycles. For example, the lidar data labeled SR can be generated using the associated data period. k+1 For example, the laser radar data from the data period marked as DP2 in period j can be used as described above. db The value of f from the data period marked as DP1 in cycle j+1 ub The value of SR is used to generate k+1LiDAR data for the sample area.

[0108] Figure 5E Shows Figure 5C The relationship between the data period disclosed in and the field of view of the lidar system. The field of view is represented by SR k-1 To SR k+1 The laser radar system outputs a system output signal that scans in the direction of the solid line marked as "scan". The system output signal is defined by a set of sample regions (SR k-1 To SR k+1 ) scanned by the system output signal. The collection of sample areas scanned by the system output signal constitutes the field of view of the LiDAR system. The object(s) in the field of view may change over time. Therefore, the position of the sample area is determined relative to the LiDAR system rather than relative to the environment and / or atmosphere in which the LiDAR system is located. For example, the sample area may be defined as being within an angular range relative to the LiDAR system. Figure 5D The dashed lines labeled as scans in the ?gure show that the scanning of the sample area in the ?eld of view can be repeated in a plurality of scan cycles. Thus, when the object in the ?eld of view has moved and / or changed, each scan cycle can be output by the same sample area scanning system. The sample area in the ?eld of view can be scanned in the same sequence during different scan cycles, or can be scanned in different sequences in different scan cycles.

[0109] The portion of each sample region corresponding to one of the data periods is Figure 5E are marked as DP1 or DP2. Figure 5C It is apparent that the chirp rate during the data period DP1 is α1, and the chirp rate during the data period DP2 is α2.

[0110] Each sample area includes a k The dotted line is marked as L k The dotted line can be used as a position reference line for the sample area k. Figure 5E Including marked L k-1 To L k+1 The position reference line. Figure 5E In the example, each position reference line (L) is drawn along the longitudinal axis of the sample region with sample region index k. k ).

[0111] Figure 5E The figure is marked as θ k where k represents the sample region index k. Orientation angle θ k The sample area SR can be measured k Angular orientation relative to the LiDAR system. In some cases, relative to a position reference line L k Measuring the orientation angle θk ,like Figure 5E Therefore, the orientation angle θ k Can measure the position reference line L k angular orientation. Figure 5E shows a lidar system with a two-dimensional field of view, so a single angle (θ k ) can limit the sample area SR k Angular orientation relative to the LiDAR system; however, the field of view is typically three-dimensional. Therefore, the LiDAR system may use two or more angles and / or other variables to define the orientation of the sample region relative to the LiDAR system.

[0112] The data refinement component 272 is configured to perform a beat frequency (f m,j ). For example, Fig. 5F is a graph of frequency versus period index (j). Labeled as “f 1,j The curve " represents the beat frequency (f) of the data period m = 1 as a function of the cycle index (j) 1,j The curve labeled “Moving Average” shows the beat frequency (f m,j ) is a moving average of the average. Fig. 5F It is shown for data period m=1, but a similar "moving average" curve can be generated for data period m=2.

[0113] Each point in the moving average curve can be represented by the average frequency (af m,j ), where m represents the time period index and j represents the cycle index. m,j ) is averaged to calculate the average frequency (af m,j ). For example, when moving average is performed on Q points, the moving average can be calculated by 1,j-(Q-1) / 2 to f 1,j+(Q-1) / 2 The beat frequency values ​​are averaged to calculate the average frequency af 1,j where Q is an odd number. Fig. 5F The "moving average" curve in is averaged over Q = 7 points. A suitable value for Q may depend on the application of the LiDAR system. Examples of suitable values ​​for Q include, but are not limited to, values ​​greater than or equal to 3 and less than or equal to 19. Fig. 5F Obviously, the average frequency value (af 1,j ) has a beat frequency value (f m,j )lower standard deviation, and therefore higher precision.

[0114] Fig. 6ATwo different objects are shown in the field of view of the LiDAR system. The LiDAR system outputs a system output signal that scans in the direction of the solid line labeled "Scan". The system output signal is represented by a solid line labeled SR k-1 To SR k+1 The sample area is scanned in a series of sample areas. Figure 5C The scheme is marked and therefore does not include Figure 5D The overlapping sample area occurs in the scheme of . k During scanning of the sample area, the system output signal switches from being incident on one surface to being incident on the other surface.

[0115] The presence of edges between different surfaces is the average frequency value (af m,j ) possible sources of distortion. As an example, Figure 6B is a graph of frequency versus period index (j). Labeled as “f 1,j The curve " represents the beat frequency (f) of the data period m = 1 as a function of the cycle index (j) 1,j The curve marked as “moving average” represents the beat frequency (f m,j ). The line labeled "Surface Change" indicates when the system output signal switches from being incident on one surface to being incident on the other surface.

[0116] From the label "f 1,j A comparison of the beat frequency curve with the “moving average” curve clearly shows that switching from one surface to another results in an average frequency value (af m,j ) becomes distorted and inaccurate. Similar distortion occurs when the system output signal switches from being incident on a surface to a location where no reflective surface exists or vice versa.

[0117] Average frequency value (af m,j ) is another source of distortion in the m,j ) exists in the outlier. The outlier is the value that provides the beat frequency value (f m,j ) The beat frequency value (f m,j ). Outliers do not necessarily appear at the transition between surfaces, but can appear in the middle of a surface. Therefore, outliers can be caused by the system output signal being incident on a surface but differing from other beat frequency values ​​(f m,j ) is inconsistent with the beat frequency value (f m,j ).

[0118] Figure 6C The outliers are plotted against the average frequency (af m,j ). From the 1,jThe comparison between the beat frequency curve of " and the moving average curve shows that the presence of outliers will increase the average frequency value (af m,j ) away from the exact value and causes the average frequency value (af m,j ) becomes distorted and inaccurate.

[0119] The data refinement component 272 may be configured to m,j ) to calculate the average frequency value (af m,j ) before or as part of it, solving for the frequency value (f m,j ) in the presence of frequency outliers and / or surface variations. Figure 7 The value for solving the beat frequency (f m,j ) in the presence of frequency outliers and / or surface variations.

[0120] At processing block 300, data refinement component 272 converts the beat frequency value (f m,j ) is identified as one that can be identified by f m,j’ denoted by , where j' represents the period index of the object beat frequency.

[0121] From processing block 300, data refinement component 272 proceeds to determination block 302, where data refinement component 272 determines the object beat frequency (f m,j’ ) is the shift beat frequency. The shift beat frequency value (f m,j’ ) has a beat frequency that is shifted by some amount relative to another beat frequency value used as a reference beat frequency value. For example, the shifted beat frequency value may have a value that differs from another of the beat frequency values ​​by δ / 2, where δ is a constant. In some cases, the reference beat frequency value is a beat frequency value that precedes the shifted beat frequency value and has the same cycle index (m). As an example, where |f m,j -f m,j-1 |>δ / 2 beat frequency value (f m,j ) can be a shifted beat frequency value, and where |f m,j -f m,j-1 |≤δ / 2 beat frequency value (f m,j ) can be the unshifted beat frequency value.

[0122] The value δ / 2 may be selected to be greater than the noise level typically present in the beat frequency. Additionally, the value of δ / 2 may be selected such that a change in the beat frequency from one beat frequency value to the next by an amount greater than δ / 2 is caused by an outlier or a change in the surface upon which the system output signal is incident. Suitable values ​​for δ include, but are not limited to, values ​​greater than 0.01 MHz and / or less than or equal to 100 MHz. In one example, δ is greater than or equal to 0.5 MHz and less than or equal to 2 MHz.

[0123] When the object beat frequency (f m,j’) is the shifted beat frequency, the data refinement component 272 may proceed from the determination block 302 to the processing block 304, where the data refinement component 272 may calculate a previous representative frequency. The previous representative frequency may be a frequency calculated to represent a signal having the same value as the object beat frequency (f m,j’ ) with the same data period index m and at the object beat frequency (f m,j’ ) in the time window before. Exemplary previous representative frequencies are all within the identified object beat frequency (f m,j’ ) appears in the time window before (f m,j ). In some cases, the data refinement component 272 applies a standard deviation reduction algorithm to the candidate beat frequency values ​​(cf m,j ) in order to identify the beat frequency value (f m,j ) in the first choice of the beat frequency value (f m,j The standard deviation reduction algorithm can be selected to reduce the value of the candidate beat frequency (cf m,j ) in the beat frequency value (f m,j ) is reduced by the standard deviation of the beat frequency value (f m,j ) in the first choice of the beat frequency value (f m,j )’s standard deviation.

[0124] The data refinement component 272 may proceed from processing block 304 to processing block 306, where the data refinement component 272 may calculate a subsequent representative frequency. The subsequent representative frequency may be a frequency calculated to represent a frequency having a beat frequency (f m,j’ ) with the same data period index m and at the object beat frequency (f m,j’ ) in the time window after the object beat frequency value. Exemplary subsequent representative frequencies are all within the object beat frequency (f m,j’ ) appears in the time window after m,j ). In some cases, the data refinement component 272 applies a standard deviation reduction algorithm to the candidate beat frequency values ​​(cf m,j ) in order to identify the beat frequency value (f m,j ) in the second choice of the beat frequency value (f m,j The standard deviation reduction algorithm can be selected to reduce the value of the candidate beat frequency (cf m,j ) in the beat frequency value (f m,j ) is reduced by the standard deviation of the beat frequency value (f m,j ) in the first choice of the beat frequency value (f m,j )’s standard deviation.

[0125] Data refinement component 272 may proceed from processing block 306 to processing block 308, where data refinement component 272 may apply one or more outlier detection criteria to the data obtained from the object beat frequency (f m,j’ ), a previous representative frequency, a subsequent representative frequency, and a combination thereof. One or more outlier detection criteria are selected to indicate the object beat frequency (f m,j’ ) is an outlier. When one or more outlier detection criteria indicate that the object beat frequency (f m,j’ ) is an abnormal value, the data refinement component 272 can adjust the object beat frequency (f m,j’ ) to adjust for the presence of outliers.

[0126] From processing block 308, data refinement component 272 may proceed to processing block 310, where data refinement component 272 may optionally apply one or more surface variation criteria to one or more components selected from the group consisting of a shifted beat frequency value, a previous representative frequency, a subsequent representative frequency, and combinations thereof. One or more outlier detection criteria are selected to indicate an outlier that is inconsistent with the shifted beat frequency value (f m,j’ ) whether the associated shift is the result of the system output signal moving onto or away from the surface.

[0127] The data refinement component 272 may proceed from processing block 310 to processing block 312. In addition, when the object beat frequency (f m,j’ ) is not a shifted beat frequency, the data refinement component 272 may proceed from the determination block 302 to the processing block 312. At the processing block 312, the data refinement component 272 performs a plurality of different beat frequency values ​​(f m,j ) to form a frequency with the same beat frequency as the object (f m,j’ ) The average frequency value (af) of the same time period index m and cycle index j m,j’ The average frequency value can be weighted or unweighted. For example, a value closer in time to the object beat frequency (f m,j’ ) of the beat frequency value (f m,j ) can be farther away in time than the object beat frequency (f m,j’ ) of the beat frequency value (f m,j ) is given more weight. The data refinement component 272 may apply the surface selection criteria to the beat frequency value (f m,j ) in order to select a particular beat frequency value for the average. The surface selection criteria may be configured such that beat frequency values ​​produced by system output signals illuminating the same surface are selected for inclusion in the average.

[0128] The data refinement component 272 returns from processing block 312 to processing block 300. When the data refinement component 272 returns to processing block 300, the data refinement component 272 may select another beat frequency value to be used as the object beat frequency. For example, the data refinement component 272 may select a beat frequency value that is next in time after the current object beat frequency and has the same time period index as the current object beat frequency to be used as the next object beat frequency. For example, when the object beat frequency is represented as f 2,j’ When the data refinement component 272 can select f 2,j’+1 The indicated beat frequency is taken as the next target beat frequency.

[0129] Figure 7 The blocks 300 to 312 in the method are included in a first algorithm labeled 314. The method may also include a second algorithm labeled 316, which includes the same blocks. The first algorithm 314 and the second algorithm 316 may be performed on beat frequency values ​​from different data periods. For example, the beat frequency value (f 1,j ) executes the first algorithm 314, and the beat frequency value (f 2,j ) executes the second algorithm 316. Therefore, the data refinement component 272 can generate each different data period (ie, af 1,j and af 2,j ) average frequency value (af m,j ).although Figure 7 The data thinning component 272 is shown executing the first algorithm 314 in parallel with the second algorithm 316 , but the first algorithm 314 and the second algorithm 316 may be executed in series.

[0130] Figure 8 shows a method suitable for performing Figure 7 An example of a buffer for a processing flow of FIG. One of the buffers is a flag buffer and one of the buffers is a frequency buffer. Each buffer includes 2P+1 buffer elements, where P is an integer selected such that P≥Q, where Q is the average depth. For illustration purposes, Figure 7 Use P=7. Suitable values ​​for P include, but are not limited to, values ​​greater than 3 and / or less than 100. In one example, P is greater than 3 and less than 20.

[0131] The beat frequency values ​​(f m,j ) is stored in the frequency buffer. The stored beat frequency value (f m,j ) changes with time, as shown by the arrows marking time. For example, an example of the frequency buffer contents is labeled t2, and another example of the frequency buffer contents is labeled t1. Time t2 occurs after time t1. Comparison of the frequency buffer contents shows that the more recent beat frequency value (f m,j) enters from the right, while the earlier beat frequency value (f m,j )Exit from the left.

[0132] The flag associated with each beat frequency value in the frequency buffer is stored in the flag buffer. m,j is the beat frequency f m,j .

[0133] The beat frequency value (f m,j ) can be used as the target beat frequency. Therefore, the beat frequency value in the frequency buffer element numbered P+1 can be selected as the target beat frequency. Figure 7 In the frequency buffer example, at the time marked as t2, the beat frequency value (f 1,9 ) will be selected as the object beat frequency. The object beat frequency can be determined by f m,j’ In which j' represents the period index of the object beat frequency. Figure 7 In the frequency buffer example of , at the time marked as t2, j'=9.

[0134] The data refinement component 272 may mark the beat frequency value (f m,j ). Shift beat frequency value (f m,j ) is a candidate for an outlier or a surface change. In some cases, the data refinement component 272 separates the beat frequency values ​​(f m,j ) is compared with a threshold value and / or by adding the beat frequency value (f m,j ) is compared with a threshold to identify the shifted beat frequency value. As an example, the data refinement component 272 may compare the beat frequency values ​​(f) in the buffer elements 2P+1 and 2P m,j ) is compared with the threshold. Figure 7 In the frequency buffer example, at the time marked as t2, the beat frequency value (f 1,16 ) occupies frequency buffer element number 2P+1, and the beat frequency value (f 1,15 ) occupies frequency buffer element number 2P. In one example, the beat frequency value f m,j’+P In |f m,j’+P -f m,j’+P-1 |>δ / 2 is marked as a shift frequency, but at |f m,j’+P -f m,j’+P-1 |≤δ / 2 is marked as unshifted. In this example, the beat frequency value f m,j’+P-1 Used as the beat frequency value f m,j’+P The reference frequency for comparison.

[0135] Fig. 9An example of a process flow suitable for use as the first algorithm 314 and / or the second algorithm 316 is shown. At processing block 340, the data refinement component 272 may update the frequency buffer and the flag buffer to reflect the new beat frequency value (f m,j ) is added to the buffer element. In response to the update of the frequency buffer, the data refinement component 272 may update the beat frequency value (f m,j ) as unshifted. For example, the data refinement component 272 may set the flag associated with the beat frequency value that has been most recently added to the frequency buffer to the flag value associated with the unshifted frequency value. Figure 7 In the example, the beat frequency value f m,j’+P Indicates the beat frequency value that has been recently added to the frequency buffer. Therefore, the data refinement component 272 can mark ff m,j’+P Set to the flag value associated with the unshifted frequency value.

[0136] The data refinement component 272 may proceed from processing block 340 to processing block 341, where the data refinement component 272 marks the shift beat frequency. As described above, when |f m,j’+P -f m,j’+P-1 |>δ / 2, the beat frequency value f m,j’+P You can enable the associated flag (ff m,j’+P ) is changed to a flag value associated with the shifted frequency value. In this example, when the beat frequency value changes by more than a threshold value (δ / 2) from the previous beat frequency value with the same time period index, the beat frequency value is marked as a shifted beat frequency value.

[0137] The data refinement component 272 may proceed from processing block 341 to processing block 342, where the data refinement component 272 identifies the beat frequency value (f m,j ) is used as the object beat frequency (f m,j’ ), where j' represents the period index of the object beat frequency. As described above, the data refinement component 272 can identify the beat frequency value (f m,j ) is used as the object beat frequency (f m,j’ ). Therefore, the beat frequency value in the frequency buffer element numbered P+1 can be selected to be used as the target beat frequency. Figure 7 In the frequency buffer example, at the time marked as t2, the beat frequency value (f 1,9 ) will be selected as the object beat frequency (f m,j’ ).

[0138] From processing block 342, data refinement component 272 proceeds to determination block 344, where data refinement component 272 determines the object beat frequency (f m,j’ ) is the shift beat frequency. For example, when ff m,j’When the flag at the position is associated with the shift frequency value, the data refinement component 272 determines the target beat frequency (f m,j’ ) is the shift beat frequency. m,j’ When the flag at is associated with the unshifted frequency value, the data refinement component 272 determines the target beat frequency (f m,j’ ) is the unshifted beat frequency.

[0139] From the determination block 344, the data refinement component 272 proceeds to a processing block 346, where the data refinement component 272 generates a previous representative frequency (PRF). As described above, the previous representative frequency may be calculated to represent a signal having a similar beat frequency (f m,j’ ) with the same data period index m and at the object beat frequency (f m,j’ ) is the frequency value of the beat frequency value that appeared in the window before. As will be discussed below, in some cases, the previous representative frequency (PRF) is marked as available or unavailable.

[0140] From processing block 346, data refinement component 272 proceeds to processing block 348, where data refinement component 272 generates a subsequent representative frequency. As described above, the subsequent representative frequency (SRF) may be calculated to represent a frequency having a similar beat frequency (f m,j’ ) with the same data period index m and at the object beat frequency (f m,j’ ) is the frequency value of the beat frequency value that appears in the window after ). As will be discussed below, in some cases, the subsequent representative frequency (SRF) is marked as available or unavailable.

[0141] From processing block 348, the data refinement component 272 proceeds to determination block 350, where the data refinement component 272 determines whether both the subsequent representative frequency (SRF) and the previous representative frequency (PRF) are marked as available.

[0142] When the data refinement component 272 determines at decision block 350 that the SRF and PRF are available, the data refinement component 272 proceeds from decision block 350 to decision block 352. At process block 352, the data refinement component 272 determines the object beat frequency (f m,j’ ) is an outlier. The data refinement component 272 may apply one or more outlier detection criteria to the data obtained from the object beat frequency (f m,j’ ), a previous representative frequency, a subsequent representative frequency, and a combination thereof. One or more outlier detection criteria are selected to indicate the object beat frequency (f m,j’ ) is an outlier. For example, the data refinement component 272 may determine the object beat frequency (f ) in response to the difference between the previous representative frequency and the subsequent representative frequency falling below a threshold. m,j’) is an outlier. As an example, the data refinement component 272 may determine the object beat frequency (f m,j’ ) are outliers.

[0143] In response to the data refinement component 272 determining the object beat frequency (f m,j’ ) is an abnormal value, the data refinement component 272 proceeds to processing block 354, where the data refinement component 272 can adjust the object beat frequency (f m,j’ ) value. For example, the data refinement component 272 can m,j’ ) is set to the value of the previous representative frequency (PRF). Therefore, the object beat frequency (f m,j’ ) is adjusted so that (f m,j’ ) value = PRF.

[0144] From processing block 354, data refinement component 272 proceeds to decision block 356. At decision block 356, data refinement component 272 determines the object beat frequency (f m,j’ ) should be marked as a shifted beat frequency. It is apparent from processing block 340 that the next beat frequency value (i.e., the object beat frequency (f) with the same time period index m) is m,j’ ) value and then by f m,j’+1 The flag associated with the object beat frequency (f m,j’ ) changes with the value of ). Therefore, due to the object beat frequency (f m,j’ ) is adjusted according to the shift value of the next beat frequency value (i.e., f m,j’+1 )The associated flag may be unreliable.

[0145] The data refinement component 272 may determine that the next beat frequency value should be marked as a shifted frequency value using a process flow similar to that disclosed in the context of process block 341. For example, in response to |PRF-f m,j’+1 |>δ / 2, the data refinement component 272 may mark the next beat frequency value as a shifted frequency value, or may mark the next beat frequency value as a shifted frequency value at processing block 360. m,j’ ) is set equal to PRF, the equation |PRF-f m,j’+1 |>δ / 2 can also be written as |f m,j’ -f m,j’+1 |>δ / 2, where f m,j’ The value of represents the f adjusted in processing block 354. m,j’ The value of .

[0146] In response to the data refinement component 272 determining the shifted object beat frequency (f m,j’) is not an outlier, the data refinement component 272 proceeds to decision block 358. m,j’ ) is not an outlier can be the result of |SRF-PRF|>δ / 2. This condition indicates that the beat frequency (f m,j’ ) is located at the edge of the surface where the system output signal moves away from the surface or onto the surface. Therefore, applying the |SRF-PRF|>δ / 2 condition to the representative frequency and subsequent representative frequencies is equivalent to applying the edge identification criterion to the beat frequency. Fig. 9 In the method data thinning process flow, the shift frequencies for which the SRF and PRF are available are basically regarded as outliers or located at the edge of the surface. Therefore, the shift frequencies for which the SRF and PRF are available and are not identified as outliers are regarded as located at the edge. Therefore, the one or more outlier identification criteria applied at the determination block 352 are also used as edge identification criteria.

[0147] When the data refinement unit 272 determines the object beat frequency (f m,j’ ) is not an outlier and determination block 352 determines |SRF-f m,j’ |>δ / 2, the object beat frequency (f m,j’ ) is displaced relative to the surface (or in the absence of a surface) that is the source of the SRF value. Determine the object beat frequency (f m,j’ ) is not an outlier and can be the result of |SRF-PRF|>δ / 2. m,j’+1 The symbol associated with the object beat frequency (f m,j’ ) (see processing block 341). Therefore, since the object beat frequency (f m,j’ ) relative to the previous beat frequency value (i.e., the value at the same time segment index m and determined by f m,j’-1 The object beat frequency (f m,j’ ) value) and also relative to the surface (or lack thereof) that is the source of the SRF value, and thus is consistent with the next beat frequency value (i.e., f m,j’+1 ) may be unreliable. To address this unreliability, the data refinement component 272 may m,j’+1 ) is compared with the SRF value to determine whether the next beat frequency value should be marked as a shifted frequency value. For example, the data refinement component 272 may respond to |SRF-f m,j’+1 |>δ / 2 and determines that the next beat frequency value should be marked as the shifted frequency value. Therefore, in response to |SRF-f m,j’ |>δ / 2 and |SRF-f m,j’+1|>δ / 2, at processing block 360, the data refinement component 272 may mark the next beat frequency value as the shifted frequency value, or may mark the next beat frequency value as the shifted frequency value.

[0148] The data refinement component 272 may proceed to processing block 364 in response to a negative determination at determination block 344, a negative determination at determination block 350, a negative determination at determination block 356, or a negative determination at determination block 358. The data refinement component 272 may proceed from processing block 360 to processing block 364. At processing block 364, the data refinement component 272 calculates the object beat frequency value (f m,j’ The average frequency (af m,j’ ). By m,j’-(Q-1) / 2 Extend to f m,j’+(Q-1) / 2 The beat frequency value (f m,j ) to calculate the average, where Q is the average depth. As an example, Figure 7 An averaging window labeled Q is shown. Figure 7 In the example, the averaging window is set from the beat frequency value f 1,6 Extended to the beat frequency f 1,12 .

[0149] In some cases, the beat frequency value (f m,j ) is used to calculate the average frequency (af m,j’ ). The data refinement component extracts the beat frequency value (f m,j ) Select the frequency to be used for calculating the average frequency (af m,j’ ) of the beat frequency value (f m,j ). For example, you can select the surfaces that are in the window and satisfy the surface selection criteria |f m,j -f m,j’ |≤δ / 2 beat frequency value (f m,j ) is used to calculate the average frequency (af m,j’ ). As an example, satisfying |f m,j -f m,j’ |≤δ / 2 m,j’-(Q-1) / 2 to f m,j’+(Q-1) / 2 Each beat frequency value can be used to calculate the average frequency (af m,j’ ). Since the average frequency (af m,j’ ) for each beat frequency value (f m,j ) needs to satisfy |f m,j -f m,j’ |≤δ / 2, so in some cases, the beat frequency value (f m,j ) is used to calculate the average frequency (af m,j’ In some cases, the object beat frequency value (f m,j’) is used to calculate the average frequency (af m,j’ ) is the unique beat frequency value (f m,j ) and f m,j’ =af m,j’ In these cases, the object beat frequency value (f m,j’ ) relative to the object beat frequency value (f m,j’ ) adjacent beat frequency values ​​(f m,j ) is shifted. Therefore, the object beat frequency value (f m,j’ ) is used as the average frequency (af m,j’ ) to avoid reducing the object beat frequency value (f m,j’ ) relative to its original value.

[0150] exist Figure 7 In the case of calculating the average frequency (af m,j’ ) using f 1,6 to f 1,12 Each beat frequency value of 1,j -f 1,9 |≤δ / 2. In some cases, all or part of the beat frequency value identified for calculating the average frequency is the adjusted object beat frequency (f m,j’ ) makes f m,j’ =PRF result, as described in the context of processing block 354. Although the object beat frequency (f m,j’ ) falls within the averaging window and must satisfy |f m,j -f m,j’ |≤δ / 2, but it is worth noting that the object beat frequency (f m,j’ ) is included in the identified frequency used to calculate the average frequency (af m,j’ ) of the beat frequency value (f m,j )middle.

[0151] Average frequency (af m,j’ ) can be identified to calculate the average frequency (af m,j’ ) of the beat frequency value (f m,j ) is the straight-line average, arithmetic mean, or weighted average. When the average frequency (af m,j ) is a weighted average, which is closer in time to the object beat frequency (f m,j’ ) of the beat frequency value (f m,j ) can be farther away in time than the object beat frequency (f m,j’ ) of the beat frequency value (f m,j ) are given more weight.

[0152] The electronics can convert the average frequency (af m,j’ ) is stored in the memory 271 for subsequent use by the lidar data generator 274 in calculating the lidar data, as disclosed above.

[0153] Fig. 9 The data thinning algorithm solves the Figure 6C The outlier values ​​disclosed in the context of distort the average beat frequency value because the deviating beat frequency value has its value adjusted at processing block 354. Since its value is adjusted to the value of the previous representative frequency, the effect of the deviating beat frequency value on the average beat frequency value is reduced or eliminated.

[0154] When selecting the beat frequency value (f m,j ) using the surface selection criteria |f m,j -f m,j’ |≤δ / 2 solves the Figure 6B The value of δ / 2 can be selected to approximate a low or minimum level of frequency variation that is characteristic of the system output signal across the edge of an object. Thus, when |f m,j -f m,j’ |>δ / 2, f m,j The values ​​of are considered to be beat frequency values ​​that are not generated by the system output signal incident on the surface on which the system output signal is incident to generate the object frequency value f m,j’ In contrast, the beat frequency value (where |f m,j -f m,j’ |≤δ / 2) are generated by the system output signal on the same surface. Therefore, the surface selection criterion effectively makes the beat frequency values ​​averaged to generate the average beat frequency value be, or at least approximate, the result of the system output signal illuminating the same surface. Since the averaged beat frequency values ​​are from the same surface, Figure 6B The average beat frequency value (af m,j ) is removed or reduced.

[0155] From processing block 364 , data refinement component 272 may return to processing block 340 .

[0156] As mentioned above, the power level indicator (P m,j ) is an indicator of the power of the system return signal which may be stored in memory 271. As the average beat frequency value (af) is calculated at processing block 264 m,j ), the data refinement component 272 may calculate an average power level indicator (aP m,j ), each average power level indicator is associated with a data period, a beat frequency and / or an average beat frequency. The electronic device may associate an average power level indicator (aP m,j ) is considered as a power level indicator associated with the data period, beat frequency and / or average beat frequency. In some cases, once the electronic device calculates the average power level indicator (aP m,j ), the electronic device can use the power level indicator (P m,j) is set equal to the average power level indicator (aP m,j ).

[0157] Average power level indicator (aP m,j ) can be similar to the average beat frequency value (af m,j ) is calculated. For example, the average beat frequency value (af m,j ) of the beat frequency value (f m,j ) associated with each power level indicator (P m,j ) is used to average the average power level indicator (aP m,j ). Therefore, the average beat frequency value (af m,j ) of the beat frequency value (f m,j ) can also be used to identify the average power level indicator (aP m,j ) power level indicator (P m,j ). Use the average to generate a power level indicator (aP m,j ) can reduce the impact of noise and increase the accuracy of power estimation.

[0158] and the average beat frequency (af m,j ), the average power level indicator (aP m,j ) can be identified to calculate the average power level indicator (aP m,j ) power level indicator (P m,j ) is the straight-line average, arithmetic mean or weighted average of the data.

[0159] Average power level indicator (aP m,j ) can be used in a variety of different applications. For example, different object materials cause the system output signal to be reflected at different power levels. Therefore, the average power level indicator (aP m,j ) may indicate what material the system output signal is irradiating, or may be an indicator of what material the system output signal is irradiating. Thus, the electronics may apply the material identification criteria to the average power level indicator (aP m,j ) to classify, identify and / or characterize the material illuminated by the system output signal. The ability to distinguish between water, concrete or metal may be important when the lidar system is used to guide an autonomous vehicle. The average power level indicator (aP m,j ) can be used to make these distinctions in autonomous vehicles and other lidar applications.

[0160] Fig.10is a flow chart showing one possible method of calculating a previous representative frequency. At processing block 370, data refinement component 272 may start the method by setting a counter (IP) to zero. Additionally, data refinement component 272 may mark the previous representative value as available.

[0161] From processing block 370, data thinning component 272 may proceed to processing block 372, where data thinning component 272 increases the value of counter (IP) by one.

[0162] From processing block 372, data refinement component 272 proceeds to processing block 374, where data refinement component 272 identifies a previous reference frequency. For example, data refinement component 272 may identify a beat frequency having the same period index as the object frequency and the IP beat frequency occurring before the object beat frequency as the previous reference frequency. As an example, data refinement component 272 may identify the beat frequency (f m,j-IP ) is identified as the previous reference frequency (PF).

[0163] From processing block 374, data refinement component 272 may proceed to processing block 376, where data refinement component 272 identifies a frequency value (f m,j ), and for the frequency value (f m,j ),|f m,j -PF|≤δ / 2. The previous frequency window may include the object frequency (f m,j’ ) before the frequency value (f m,j In one example, the previous frequency window is from the frequency value (f m,j’-XP ) extends to the frequency value (f m,j’-1 ), where XP is a positive integer greater than 1 and less than or equal to P. Therefore, the data refinement component 272 can identify the data from (f m,j’-XP ) to (f m,j’-1 ) frequency value, for which |f m,j -PF|≤δ / 2. In one example, XP=P. The identified frequency value (f m,j ) is restricted to values ​​for which |f m,j -PF|≤δ / 2 is used as a standard deviation reduction algorithm, which reduces the frequency value (f m,j )’s standard deviation.

[0164] Data refinement component 272 may proceed from process block 376 to determination block 378, where the frequency value (f m,j ) is greater than or equal to NP, where NP is a positive integer greater than or equal to 1. NP can be used as the frequency value (f) that needs to be identified in order to continue the calculation of the previous representative frequency.m,j ) minimum number.

[0165] When the determination at determination block 378 is negative, the data refinement component 272 may proceed from determination block 378 to determination block 380, where it is determined whether IP = XP. When the determination at determination block 380 is negative, the data refinement component 272 may return to processing block 372 and repeat the method with a different beat frequency value used as the previous reference frequency.

[0166] When the determination at determination block 380 is affirmative, the data refinement component 272 may proceed to block 382 where the previous representative frequency is marked as unavailable.

[0167] When the determination at decision block 378 is positive, the data refinement component 272 may proceed from decision block 378 to process block 384 where a previous representative value may be calculated. The previous representative value may be the frequency value (f m,j ). The frequency value used as the previous reference frequency must satisfy |f m,j -PF|≤δ / 2, and is therefore the frequency value (f m,j ). The average frequency value may be a linear average, an arithmetic mean, or a weighted average. For example, the frequency that is closer in time to the object beat frequency (f) identified at processing block 376 m,j’ ) frequency value (f m,j ) can be farther away in time than the object beat frequency (f m,j’ ) of the beat frequency value (f m,j ) is given more weight. The previous reference frequency is still marked as available.

[0168] Fig.11 is a flow chart illustrating one possible method of calculating a subsequent representative frequency. At processing block 390, data refinement component 272 may begin the method by setting a counter (IP) to zero. Additionally, data refinement component 272 may mark a subsequent representative value as available.

[0169] From processing block 390, data thinning component 272 may proceed to processing block 392, where data thinning component 272 increases the value of counter (IP) by one.

[0170] From processing block 392, data refinement component 272 may proceed to processing block 394, where data refinement component 272 identifies a subsequent reference frequency. For example, data refinement component 272 may identify a frequency having a similar frequency to the object frequency (f m,j’ ) and the beat frequency of the IP beat frequency that occurs after the object beat frequency is used as the subsequent reference frequency. As an example, the data refinement component 272 may use the beat frequency (f m,j+IP) is identified as the subsequent reference frequency (SF).

[0171] Data refinement component 272 may proceed from process block 394 to process block 396, where data refinement component 272 identifies a frequency value (f m,j ), and for the frequency value (f m,j ),|f m,j -SF|≤δ / 2. Subsequent frequency windows may include the object frequency (f m,j’ ) appears after the frequency value (f m,j In one example, the subsequent frequency window is from the frequency value (f m,j’+1 ) extends to the frequency value (f m,j’+XP ), where XP is a positive integer greater than 1 and less than or equal to P. Therefore, the data refinement component 272 can identify the data from (f m,j’+1 ) to (f m,j’+XP ) frequency value, for which |f m,j -SF|≤δ / 2. In one example, XP=P. The identified frequency value (f m,j ) is restricted to values ​​for which |f m,j -SF|≤δ / 2 is used as a standard deviation reduction algorithm, which reduces the frequency value (f m,j )’s standard deviation.

[0172] Data refinement component 272 may proceed from process block 396 to determination block 398, where the frequency value (f m,j ) is greater than or equal to NP, where NP is a positive integer greater than or equal to 1. NP can be used as a frequency value (f) that needs to be identified in order to continue the calculation of the subsequent representative frequency. m,j ) minimum number.

[0173] When the determination at determination block 398 is negative, the data refinement component 272 may proceed from determination block 398 to determination block 400, where it is determined whether IP = XP. When the determination at determination block 400 is negative, the data refinement component 272 may return to processing block 392 and repeat the method with a different beat frequency value used as a subsequent reference frequency (SF).

[0174] When the determination at determination block 400 is affirmative, the data refinement component 272 may proceed to block 402 where subsequent representative frequencies are marked as unavailable.

[0175] When the determination at determination block 398 is positive, the data refinement component 272 may proceed from determination block 398 to process block 404, where a subsequent representative value may be calculated. The subsequent representative value may be the frequency value (f m,j ) is the average value of the frequency used as the subsequent reference frequency (SF). The frequency value used as the subsequent reference frequency (SF) must satisfy |f m,j -SF|≤δ / 2, and is therefore the frequency value (f m,j ). The average frequency value may be a linear average, an arithmetic mean, or a weighted average. For example, the frequency that is closer in time to the object beat frequency (f) identified at processing block 396 m,j’ ) frequency value (f m,j ) can be farther away in time than the object beat frequency (f m,j’ ) of the beat frequency value (f m,j ) are given more weight. Subsequent reference frequencies are still marked as available.

[0176] Suitable platforms for lidar chips include, but are not limited to, silicon dioxide, indium phosphide, and silicon-on-insulator wafers. Fig.12 is a cross section of a portion of a chip made of a silicon-on-insulator wafer. The silicon-on-insulator (SOI) wafer includes a buried layer 410 between a substrate 412 and an optical transmission medium 414. In the silicon-on-insulator wafer, the buried layer 410 is silicon dioxide, while the substrate 412 and the optical transmission medium 414 are silicon. The substrate 412 of an optical platform such as an SOI wafer can be used as a base for the entire lidar chip. For example, Figures 1A to 1C The optical components shown on the lidar chip can be positioned on or above the top side and / or lateral sides of substrate 412.

[0177] Fig.12 The portion of the chip shown in FIG. 4 includes a waveguide structure suitable for use in a lidar chip constructed from a silicon-on-insulator wafer. A ridge 416 of the optical transmission medium 414 extends away from a slab region 418 of the optical transmission medium. The optical signal is confined between the top of the ridge 416 and the buried oxide layer 410.

[0178] The size of the ridge waveguide is Fig.12. For example, the ridge has a width labeled w and a height labeled h. The thickness of the slab region is labeled T. For lidar applications, these dimensions may be more important than the other dimensions because higher levels of optical power than used in other applications need to be used. The ridge width (labeled w) is greater than 1 μm and less than 4 μm, the ridge height (labeled h) is greater than 1 μm and less than 4 μm, and the slab region thickness is greater than 0.5 μm and less than 3 μm. These dimensions can apply to straight or substantially straight portions of the waveguide, curved portions of the waveguide, and tapered portions of (one or more) waveguides. Therefore, these portions of the waveguide will be single-mode. However, in some cases, these dimensions apply to straight or substantially straight portions of the waveguide. Additionally or alternatively, the curved portion of the waveguide can have a reduced slab thickness in order to reduce optical losses in the curved portion of the waveguide. For example, the curved portion of the waveguide can have a ridge extending away from the slab region, whose thickness is greater than or equal to 0.0 μm and less than 0.5 μm. Although the above dimensions will generally provide a straight or substantially straight portion of the waveguide with a single-mode configuration, they may result in tapered section(s) and / or curved section(s) that are multi-mode. The coupling between the multi-mode geometry and the single-mode geometry may be accomplished using a taper that does not substantially excite higher order modes. Thus, a waveguide may be constructed such that a signal carried in the waveguide is carried in single mode even when carried in a waveguide section having multi-mode dimensions. Fig.12 The waveguide structure disclosed in the context of Figures 1A to 1C All or part of a waveguide on a constructed lidar chip.

[0179] The optical sensor that interfaces with the waveguide on the lidar chip can be a component that is separated from the chip and then attached to the chip. For example, the optical sensor can be a photodiode or an avalanche photodiode. Examples of suitable optical sensor components include, but are not limited to, InGaAs PIN photodiodes manufactured by Hamamatsu, located in Hamamatsu, Japan, or InGaAs APDs (avalanche photodiodes) manufactured by Hamamatsu, located in Hamamatsu, Japan. These optical sensors can be centrally located on the lidar chip. Alternatively, all or part of the waveguide that terminates at the optical sensor can terminate at a facet located at the edge of the chip, and the optical sensor can be attached to the edge of the chip above the facet so that the optical sensor receives light passing through the facet. The use of an optical sensor as a component separated from the chip is suitable for all or part of the optical sensors selected from the group consisting of a first auxiliary optical sensor 218, a second auxiliary optical sensor 220, a first optical sensor 223, and a second optical sensor 224.

[0180] As an alternative to the light sensor being a separate component, all or a portion of the light sensor may be integrated with the chip. For example, examples of light sensors interfaced with ridge waveguides on a chip constructed from a silicon-on-insulator wafer may be found in: Optics Express, Vol. 15, No. 21, 13965-13971 (2007); U.S. Pat. No. 8,093,080, issued on Jan. 10, 2012; U.S. Pat. No. 8,242,432, issued on Aug. 14, 2012; and U.S. Pat. No. 6,108,8472, issued on Aug. 22, 2000, all of which are incorporated herein in their entirety. The use of a light sensor integrated with the chip is suitable for all or a portion of a light sensor selected from the group consisting of the auxiliary light sensor 218, the second auxiliary light sensor 220, the first light sensor 223, and the second light sensor 224.

[0181] The light source 4 that interfaces with the utility waveguide 12 may be a laser chip that is separated from the lidar chip and then attached to the lidar chip. For example, the light source 4 may be a laser chip that is attached to the chip using a flip-chip arrangement. When the light source 4 is to be bonded to a ridge waveguide on a chip constructed from a silicon-on-insulator wafer, it is appropriate to use a flip-chip arrangement. Alternatively, the utility waveguide 12 may include a grating (not shown), such as a Bragg grating, that serves as a reflector for an external cavity laser. In these cases, the light source 4 may include a gain element that is separated from the lidar chip and then attached to the lidar chip in a flip-chip arrangement. Examples of suitable interfaces between flip-chip gain elements and ridge waveguides on chips constructed from silicon-on-insulator wafers may be found in U.S. Pat. No. 9,705,278, issued on July 11, 2017, and U.S. Pat. No. 5,991,484, issued on November 23, 1999; each of the above patents is incorporated herein in its entirety. When the light source 4 is a gain element or a laser chip, the electronics 32 may change the frequency of the emitted laser radar signal by changing the level of current applied through the gain element or the laser cavity.

[0182] The above-mentioned laser radar system includes multiple optical components, such as a laser radar chip, a laser radar adapter, a light source, a light sensor, a waveguide and an amplifier. In some cases, in addition to the optical components shown or as a substitute for the optical components shown, the laser radar system includes one or more passive optical components. Passive optical components can be solid-state components that do not include moving parts. Suitable passive optical components include, but are not limited to, lenses, reflectors, gratings, reflective surfaces, separators, demultiplexers, multiplexers, polarizers, polarization separators and polarization rotators. In some cases, in addition to the optical components shown or as a substitute for the optical components shown, the laser radar system includes one or more active optical components. Suitable active optical components include, but are not limited to, optical switches, phase tuners, attenuators, steerable reflectors, steerable lenses, tunable demultiplexers, and tunable multiplexers.

[0183] Although the imaging system is disclosed in the context of a lidar system having a lidar chip, the disclosed data refinement can be used in conjunction with other lidar systems. For example, the disclosed data refinement can be used for.

[0184] In view of these teachings, other embodiments, combinations and modifications of the present invention will be readily apparent to those skilled in the art. Therefore, the present invention is limited only by the appended claims, which include all such embodiments and modifications when viewed in conjunction with the above description and accompanying drawings.

Claims

1. A system comprising: an imaging system configured to output a system output signal during a plurality of associated data periods, repeating a frequency of a system output signal as a function of time during each of said associated data periods; a laser radar system including a light combiner that combines light from the system output signal returning to the laser radar system with light from a reference signal to generate a beat signal that both beat at a beat frequency, Each of the beat frequencies is associated with a different one of the data periods; as well as electronics for calculating average frequencies, each of which is an average of a plurality of different beat frequencies, and each of which is associated with a different one of the beat frequencies, The electronics calculates lidar data from the average frequency, the lidar data indicating a radial velocity and / or distance between the lidar system and an object external to the lidar system.

2. The system according to claim 1, wherein: The number of beat frequencies averaged to calculate each of the average beat frequencies is different for different average beat frequencies.

3. The system according to claim 1, wherein: The electronic device adjusts a value of the adjustment portion of the beat frequency to an adjustment value different from a value of the beat frequency of the beat signal.

4. The system according to claim 3, wherein: At least a portion of the beat frequencies averaged to calculate one or more of the average beat frequencies has the adjusted value instead of the value of the beat frequency of the beat signal.

5. The system according to claim 3, wherein: The electronic device adjusts the value of the adjustment portion in response to a beat frequency in the adjustment portion of the beat frequency being an abnormal value.

6. The system according to claim 3, wherein: The electronics are configured to apply one or more outlier identification criteria to the beat frequencies in order to identify beat frequencies that are outliers.

7. The system according to claim 1, wherein: selecting, from the beat frequency values ​​occurring in the averaging window, the beat frequency of the object frequency that is averaged in order to calculate the average frequency, The beat frequencies are selected so that a magnitude of a difference between each of the selected beat frequencies and a target beat frequency among the beat frequencies is smaller than a threshold value, The beat frequencies in the window that result in the magnitude of the difference being greater than a threshold are excluded from the beat frequencies that are averaged to calculate the object mean frequency, and The object beat frequency is a beat frequency associated with the object average frequency.

8. The system according to claim 1, wherein: selecting a beat frequency that is averaged to calculate the object frequency in the average frequency from a plurality of beat frequencies in the average window, The electronics applies a surface selection criterion to the beat frequencies in the averaging window to identify beat frequencies that are all the result of the system output signal impinging the same surface on the object.

9. The system according to claim 1, wherein: The number of beat frequencies averaged to calculate at least a portion of the average frequency is between 3 and 19.

10. A method of operating an imaging system, the method comprising: outputting a system output signal from the imaging system during a plurality of associated data periods, repeating a frequency of a system output signal as a function of time during each of said associated data periods; combining light from the system output signal returning to the lidar system with light from the reference signal to generate a beat signal, both beating at a beat frequency, Each of the beat frequencies is associated with a different one of the data periods; calculating average frequencies, each of the average frequencies being an average of a plurality of different beat frequencies, and each of the average frequencies being associated with a different one of the beat frequencies; as well as Lidar data is calculated from the average frequency, the Lidar data indicating radial velocity and / or distance between the Lidar system and objects external to the Lidar system.

11. The method according to claim 10, wherein: The number of beat frequencies averaged to calculate each of the average beat frequencies is different for different average beat frequencies.

12. The method according to claim 10, further comprising: The value of the adjustment portion of the beat frequency is adjusted to an adjustment value different from the value of the beat frequency of the beat signal.

13. The method according to claim 12, wherein: At least a portion of the beat frequencies averaged to calculate one or more of the average beat frequencies has the adjusted value instead of the value of the beat frequency of the beat signal.

14. The method according to claim 12, wherein: The value of the adjustment portion is adjusted in response to the beat frequency in the adjustment portion of the beat frequency being an abnormal value.

15. The method according to claim 12, further comprising: One or more outlier identification criteria are applied to the beat frequencies to identify beat frequencies that are outliers.

16. The method according to claim 10, wherein: selecting, from the beat frequency values ​​occurring in the averaging window, the beat frequency of the object frequency that is averaged in order to calculate the average frequency, The beat frequencies are selected so that a magnitude of a difference between each of the selected beat frequencies and a target beat frequency among the beat frequencies is smaller than a threshold value, The beat frequencies in the window that result in the magnitude of the difference being greater than a threshold are excluded from the beat frequencies that are averaged to calculate the object mean frequency, and The object beat frequency is a beat frequency associated with the object average frequency.

17. The method according to claim 10, wherein: selecting a beat frequency that is averaged to calculate the object frequency in the average frequency from a plurality of beat frequencies in the average window, The electronics applies a surface selection criterion to the beat frequencies in the averaging window to identify beat frequencies that are all the result of the system output signal impinging the same surface on the object.

18. The method according to claim 10, wherein: The number of beat frequencies averaged to calculate at least a portion of the average frequency is between 3 and 19.

Images