Histogram-based signal detection with sub-regions corresponding to adaptive bin widths

Abstract

Various example embodiments relate to devices and methods, including devices having sensor circuitry and processing circuitry. In one example, the sensor circuitry generates and senses a detected signal corresponding to a physical object, the physical object being located in an operable region relative to the position of the sensor circuitry. The processing circuitry records and organizes information associated with the detected signal in a plurality of sub-histograms, the plurality of sub-histograms being respectively associated with different precision measures for corresponding sub-regions of the operable region, each of the plurality of sub-histograms comprising a set of histogram bins characterized by a bin width associated with a precision measure of the set of histogram bins, and the processing circuitry optimizes at least one of the precision measures by dynamically adapting one or more of the library widths in response to the detected signal.

Description

Technical Field

[0001] Aspects of various embodiments relate to histogram-based signal detection with sub-regions corresponding to adaptive bin widths. Background Art

[0002] For many applications, it may be beneficial to detect photon reception and generate data about the photon arrival time within a given timing window, for example, for time-of-flight applications. For example, sensor systems utilizing light detection and ranging (LIDAR) are increasingly deployed in vehicles to implement, for example, safety and / or autonomous driving features. A typical LIDAR system includes an illuminator (e.g., a light source) and a detector. For example, an illuminator may be a laser that emits light having a specific operating wavelength. The illuminator emits light toward a target, which then scatters the light. Some of the scattered light is received back at the detector. The system determines the distance to the target based on one or more characteristics associated with the returned light. For example, the system may determine the distance to the target based on the time of flight of the returned light pulse.

[0003] LIDAR plays an important role in autonomous vehicles due to its high range and angular resolution. Single-photon avalanche photodiode (SPAD), sometimes also referred to as Geiger-Mode Avalanche Photodiode, is one of the promising receiver technologies for autonomous LIDAR, thanks to its unique properties, including high sensitivity down to a single photon, high temporal resolution and low cost, and high array resolution enabled by a planar device structure (e.g., CMOS SPAD). By biasing the device above the collapse voltage, even a single released charge carrier (e.g., electron or "hole") can cause a self-sustaining avalanche. The release of charge carriers can be due to the absorption of incident photons (e.g., signal or noise), can be a thermally induced release (known as "dark count"), or the release of captured charge carriers from a previous avalanche (known as "residual pulse"). It should be noted that avalanche triggering can be considered as an event in this article. After the event, the SPAD device may be quenched shut down (eg, the bias voltage may be reduced below the breakdown voltage into a linear mode of operation) to avoid permanent damage to the device.

[0004] SPAD has a proven record of flight time in a time-correlated single photon counting (TCSPC) configuration for short-range and low-noise environmental applications, such as fluorescence lifetime microscopy. In such a configuration, the SPAD array may include a timing circuit consisting of, for example, a time-to-digital converter (TDC), or a time-to-analog converter (TAC), followed by an analog / digital converter, which together measure the time between a reference signal (e.g., a laser pulse) and the SPAD output signal caused by an avalanche. The corresponding flight time record can be stored in, for example, a register or latch for subsequent readout. In such a configuration, an earlier photon "blocks" a subsequent photon, and the probability that the flight time has not yet been recorded and the SPAD is activated can decrease exponentially over time.

[0005] To mitigate the effects of the single-photon mode of operation (e.g., blockage of the signal by prior noise, non-uniform photon detection probability, etc.), a single acquisition cycle can be composed of multiple measurements and the signal can be identified via statistical methods, for example, by construction of a histogram of the time-of-flight recordings and additional signal processing.

[0006] For many applications, these and other issues pose challenges to the efficiency of signal detection methods. Summary of the invention

[0007] Various example embodiments are directed to problems such as those set forth above, among others, which will become apparent from the following disclosure regarding recording and organizing information associated with detected signals in multiple sub-histograms with adaptive bin widths.

[0008] In certain example embodiments, aspects of the present invention relate to simultaneously identifying distances of two or more physical objects from a sensor circuit by generating and sensing detected signals corresponding to physical objects located in an operable area, and organizing information associated with the detected signals in multiple sub-histograms that correspond to sub-areas of the operable area and have adaptive bar widths.

[0009] In a more specific example embodiment, the device includes a sensor circuit and a processing circuit. The sensor circuit generates and senses detected signals corresponding to physical objects, which are located in an operable area relative to the position of the sensor circuit. The sensor circuit may include a single photon avalanche photodiode (SPAD) array circuit that senses light signals and, in response, generates detected signals. As further described herein, the sensor circuit can measure the time of flight by sensing (reflected) light signals. The processing circuit can record and organize information associated with the detected signals in a plurality of sub-histograms, which are respectively associated with different precision metrics for corresponding sub-areas of the operable area, and optimize at least one of the precision metrics by dynamically adapting one or more of the bar widths in response to the detected signal. Each of the plurality of sub-histograms includes a histogram bar set, which is characterized by a bar width associated with its precision metric. In various specific embodiments, a plurality of sub-histograms with one or more adapted bar widths can be used to simultaneously determine the distance of two or more physical objects from the sensor circuit.

[0010] The accuracy measure may indicate a level of detection reliability and / or resolution. For example, a plurality of sub-histograms correspond to different sub-areas of the operable area, wherein each of the different sub-areas at least temporarily covers a different range indicated by the (possible) distance between the sensor circuit and one or more of the physical objects. A plurality of sub-histograms may include at least two sub-histograms corresponding to two different sub-areas of the operable area and / or each of two or more sub-histograms has a different bar width compared to other sub-histograms, and wherein at least one of the different bar widths is dynamically adapted in response to at least a portion of the detected signal, however embodiments are not limited thereto. In a more specific embodiment, each of the plurality of histogram bars has a bar width that is less than half the width of the emitted light pulse.

[0011] The processing circuit can optimize at least one precision measure based on at least one of the indications of the power sensed in the optically received signal and the distance between the sensor circuit and one or more of the physical objects. As a specific example, the processing circuit optimizes at least one precision measure based on at least one of the indications of the power or distance sensed in the optically received signal, by changing one of the bar widths of one of the multiple different sub-histograms for coarser detection to optimize or improve the detection reliability of one or more far-end physical objects in the physical objects, and by changing another of the bar widths of another of the multiple sub-histograms for finer detection to optimize or improve the resolution in terms of detecting one or more near-end physical objects in the physical objects. According to various embodiments, the bar widths within a single sub-histogram are equal (e.g., the same) for all bars of the corresponding sub-histogram, and the bar widths of different sub-histograms are independent and unique in several embodiments. The bar widths for a single histogram can be changed and / or adjusted over time, for example, frame by frame. As another specific example, the processing circuit optimizes the accuracy measure in response to a detected signal indicating the presence of a proximal physical object in the physical objects and the presence of a distal physical object in the physical objects by adjusting a first bin width for a first one of a plurality of sub-histograms for optimizing or improving a resolution associated with the proximal physical object in the physical objects, and adjusting a second bin width corresponding to a second one of the plurality of sub-histograms for optimizing or improving a detection reliability associated with a distal physical object in the physical objects.

[0012] In a number of specific embodiments, a processing circuit provides an indication of the power of a detected signal in each of a plurality of histogram bins, and optimizes a precision metric by adjusting the bin width based on empirical data indicating a correspondence between the bin width and the detected signal power. The processing circuit may facilitate detection reliability based on a signal-to-noise ratio (SNR) of the detected signal so as to process peak detection of the detected signal, wherein at least one of the bin widths is adjusted as an iterative step to improve peak detection. Additionally, the processing circuit may at least temporarily generate a plurality of sub-histograms (e.g., a set of histogram bins) based on empirical data indicating a correspondence between the bin width and the detected signal power, the empirical data being provided from a storage circuit integrated with the processing circuit and / or received from a communication channel provided from an external circuit.

[0013] In various embodiments, the sensor circuit and the processing circuit operate simultaneously and in real time, wherein the sensor circuit generates a detected signal while the processing circuit constructs information for a plurality of sub-histograms. For example, the sensor circuit may operate in an automatic or semi-automatic driving (or flying) mode, wherein the detected signal is dynamically generated, and wherein the bars of the plurality of sub-histograms have dynamically changing information. In a specific embodiment, the apparatus further comprises an automobile, the sensor circuit and the processing circuit being fastened to the automobile, and wherein during movement of the automobile, the sensor circuit dynamically generates a detected signal, and in response the processing circuit dynamically optimizes a precision measure for at least one of the plurality of sub-histograms by adapting its bar width. However, embodiments are not limited to automotive applications and may relate to drones, robotic applications, and the like.

[0014] Other specific embodiments relate to methods of using the apparatus described above, for example, methods for sensing physical objects located in an operable area. The method includes using a sensor circuit to sense and generate detected signals corresponding to physical objects, which are located in the operable area relative to the position of the sensor circuit. The method further includes using a processing circuit that operates in response to the detected signal to record and organize information associated with the detected signal in a plurality of sub-histograms, each of which is associated with different precision measures for corresponding sub-areas of the operable area, each of which includes a set of histogram bars, characterized by a bar width associated with the precision measure of the histogram bar set, and optimizing at least one of the precision measures by dynamically adapting one or more of the bar widths in response to the detected signal. Adapting one or more of the bar widths may include dynamically adapting the bar width of at least one sub-histogram in response to the signal detected in one or more bars. In addition, in a specific embodiment, information is organized in two or more sub-histograms, wherein at least one of the two or more sub-histograms has a different bar width than another of the two or more sub-histograms.

[0015] The above discussion / summary is not intended to describe each embodiment or every implementation of the present invention. The figures and the following detailed description also illustrate various embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Various example embodiments may be more fully understood by considering the following detailed description taken in conjunction with the accompanying drawings, in which:

[0017] Figure 1 An example of a device according to the invention is shown;

[0018] Figure 2 An example of the operation of a device according to the invention is shown, for example, as with Figure 1consistent type of equipment;

[0019] Figures 3A to 3C shows an example of histogram bin widths according to the present invention;

[0020] Figures 4A to 4B An example pulse shape of a signal according to the present invention is shown;

[0021] Figure 5 shows examples of power adaptability histograms according to various embodiments;

[0022] Figures 6A to 6B An example method of using the device according to the present invention is shown;

[0023] Figure 7 shows a specific example of a sub-histogram with adaptive binning according to various embodiments; and

[0024] Figures 8A to 8B Example data processing of sub-histograms with range-adaptive binning in accordance with various embodiments is shown.

[0025] Although the various embodiments discussed herein are subject to modification and alternative forms, aspects of these embodiments have been shown by way of example in the drawings, and aspects of these embodiments will be described in detail. However, it should be understood that the invention is not intended to be limited to the specific embodiments described. On the contrary, it is intended to cover all modifications, equivalents, and alternatives of the various aspects defined in the claims that fall within the scope of the invention. In addition, the term "example" used throughout this application is only for illustration and is not intended to be limiting. DETAILED DESCRIPTION

[0026] Aspects of the present invention are considered to be applicable to a variety of different types of devices, systems and methods involving histogram-based signal detection with sub-areas corresponding to adaptive bar widths. In certain embodiments, aspects of the present invention have been shown to be beneficial when used to record and organize the context of information associated with signals detected in systems that involve light detection and ranging (LIDAR) and in which multiple sub-histograms are associated with or correspond to sub-areas of an operable area. In some embodiments, adaptive histograms are used together with adaptive bar widths to simultaneously detect the distance of two or more physical objects from a sensor circuit. Although various aspects can be understood through the following discussion of non-limiting examples using exemplary situations, they are not necessarily limited to this.

[0027] In the following description, various specific details are set forth to describe the specific examples presented herein. However, it should be apparent to those skilled in the art that one or more other examples and / or variations of these examples may be practiced without all of the specific details given below. In other cases, well-known features are not described in detail to avoid confusing the description of the examples herein. For ease of illustration, the same reference numerals may be used in different figures to refer to the same elements or additional examples of the same elements. Also, although various aspects and features may be described in a single figure in some cases, it should be understood that features from one figure or embodiment may be combined with features of another figure or embodiment, even if the combination is not explicitly shown or explicitly described as a combination.

[0028] Some of the specific examples discussed below relate to and / or include features related to detection of light detection and ranging (LIDAR), for example, for autonomous driving of cars or flying drones. Single-photon avalanche photodiodes (SPADs), avalanche diodes operating in single photon (or Geiger mode), are receiver technologies that can be used in LIDAR systems due to associated properties, including high sensitivity down to a single photon, high temporal resolution, and high array resolution enabled by planar device structures, such as complementary metal-oxide-semiconductor (CMOS) SPADs, which demonstrate sufficient flight time in a time-correlated single photon counting configuration. By biasing the device above the collapse voltage, even a single released charge carrier (e.g., electron or "hole") can cause a self-sustaining avalanche. The release of the charge carrier can be due to absorption of a photon (e.g., signal or noise), can be a thermally induced release (known as a "dark count"), or the release of trapped charge carriers from a previous avalanche (known as a "residual pulse"). It should be noted that avalanche triggering can be considered as an event in this article. After an event (e.g., detection of a reflected signal), the diode quenches and turns off (the bias voltage is reduced to below collapse) to a linear mode to avoid permanent damage, and therefore the diode remains in a linear mode for a certain time period to substantially reduce the probability of residual pulsation. This time period is generally referred to as the failure time, during which no signal can be detected. SPAD pixels can be followed by an array of time-to-digital converters for time-of-flight estimation. In order to mitigate the effects of the single-photon operating mode (e.g., blocking of signals by previous noise, non-uniform photon detection probability, etc.), the time-of-flight measurement is repeated many times and the true signal is statistically identified via the construction of a single-photon detection histogram.

[0029] Due to the effects of the single-photon mode of operation, a strong signal may be represented by a narrow full-width at half maximum (FWHM), while a weak signal may be represented by a wider FWHM given a longer integration time for SPAD saturation. Because the shape of the signal detected in the histogram may not represent the true shape of the incoming optical signal, a balance and / or compromise may occur between higher resolution for strong signals (e.g., high reflectivity and / or short range) and higher detection reliability for weak signals (e.g., low reflectivity and / or long range) when considering the bin width for the histogram. Various embodiments relate to techniques for overcoming this tradeoff using efficient / fast and inexpensive circuits as dictated by real-time automated applications (e.g., flying drones, among others) and by incorporating industry cost sensitivity for power-adaptive binning for SPAD-based LIDAR systems. The principle of the power adaptive histogram binning technique is to subdivide the complete operable area of ​​the sensor circuit into sub-areas, each of which has its own independent sub-histogram, each of which has a corresponding bin width. The complete operable area of ​​the sensor circuit can also be referred to as the "operating range" of the sensor circuit. The bin width is applicable to the signal power both spatially and temporally in a dynamic environment to improve or optimize the precision metric associated with one or more sub-histograms. All sub-histograms are used in combination to simulate a single frequency-based histogram over the complete operable area. This technology overcomes the resolution detection reliability trade-off between strong and weak signals caused by the probabilistic nature and inherent failure time of SPADs.

[0030] Multiple embodiments are directed to range adaptive histogram binning for LIDAR systems using SPAD array circuits in addition to the power adaptive histogram binning technique described above. In long-range applications, such as in automobiles, the range inverse square power drop dominates the dynamic range. For example, the difference between the reflection from an object at 1 meter (m) away to the reflection from an object at 100m away is 10,000 times. In this case, various embodiments are directed to range adaptive application of the power adaptive histogram binning technique that compensates for the range inverse square power drop. The efficient circuit according to the present invention retains the low computational cost of equal width histograms to achieve real-time histogram construction parallel to data acquisition. The circuit determines how the full operational area is subdivided and how the bin widths are assigned.

[0031] More specific embodiments relate to a device including a sensor circuit and a processing circuit, wherein the sensor circuit generates and senses detected signals corresponding to physical objects, which are located in an operable area relative to the position of the sensor circuit. In a more specific embodiment, the sensor circuit includes a SPAD array circuit that senses an optical signal and, in response, generates a detected signal. For example, the sensor circuit may be configured to measure the flight time by sensing a (reflected or returned) optical signal. The processing circuit records and organizes information associated with the detected signal in a plurality of sub-histograms, which are respectively associated with different precision measures for corresponding sub-areas of the operable area, and optimizes at least one of the precision measures by dynamically adapting one or more of the bar widths in response to the detected signal. Each of the plurality of sub-histograms includes a histogram bar set, which is characterized by a bar width associated with its precision measure. In various specific embodiments, a plurality of sub-histograms having one or more of the adapted bar widths can be used to simultaneously determine the distance of two or more physical objects from the sensor circuit.

[0032] The precision metric indicates the level of detection reliability and / or resolution. For example, the bar width can be adjusted to improve or optimize the level of detection reliability and / or resolution for one or more of the sub-histograms. In a specific embodiment, a plurality of sub-histograms correspond to different sub-areas of the operable area, wherein each of the different sub-areas at least temporarily covers a different range indicated by a possible distance between the sensor circuit and one or more of the physical objects. A plurality of sub-histograms include at least two sub-histograms corresponding to two different sub-areas of the operable area and / or each of two or more sub-histograms has a different bar width compared to other sub-histograms, and wherein at least one of the different bar widths is dynamically adapted in response to at least a portion of the detected signal, however the embodiment is not limited thereto. In a more specific embodiment, each of the plurality of histogram bars has a bar width less than half the width of the emitted light pulse.

[0033] The processing circuit can optimize at least one precision measure based on at least one of an indication of power sensed in the optically received signal and an indication of the distance between the sensor circuit and one or more of the physical objects. As a specific example, the processing circuit optimizes at least one precision measure based on at least one of the indications of power or distance sensed in the optically received signal, by changing one or more of the bar widths of a plurality of different sub-histograms for coarser detection to optimize or improve the detection reliability of one or more far-end physical objects in the physical objects, and by changing another of the bar widths of another of the plurality of sub-histograms for finer detection to optimize or improve the resolution in terms of detecting one or more near-end physical objects in the physical objects. In various embodiments, changing one (or more) of the bar widths of a plurality of different sub-histograms includes changing the bar width of each of the bars of the corresponding sub-histograms to the same (changed or modified) bar width. For example, assuming that the histogram has four sub-histograms and each sub-histogram includes ten bars. The bin width of each (eg, all) of the ten bins of the corresponding sub-histogram is changed to the same modified bin width.

[0034] In a number of specific embodiments, a processing circuit provides an indication of the power of a detected signal in each of a plurality of histogram bins, and optimizes at least one of the precision metrics by adjusting the bin width based on empirical data indicating the correspondence between the bin width and the detected signal power. The processing circuit may promote detection reliability based on a signal-to-noise ratio (SNR) of the detected signal so as to process peak detection of the detected signal. At least one of the bin widths may be adjusted as an iterative step to improve peak detection. In addition, the processing circuit may generate a plurality of sub-histograms at least temporarily based on empirical data indicating the correspondence between the bin width and the detected signal power. The empirical data may be provided from a storage circuit integrated with the processing circuit and / or received from a communication channel provided from an external circuit.

[0035] In various embodiments, the sensor circuit and the processing circuit operate simultaneously in real time, wherein the sensor circuit generates a detected signal while the processing circuit constructs information for a plurality of sub-histograms. For example, the sensor circuit may operate in an automatic or semi-automatic driving mode, wherein the detected signal is dynamically generated, and wherein the bars of the plurality of sub-histograms have dynamically changing information. In a specific embodiment, the apparatus further includes an automobile, the sensor circuit and the processing circuit being fastened to the automobile, and wherein during movement of the automobile, the sensor circuit dynamically generates a detected signal, and in response the processing circuit dynamically optimizes a precision measure for at least one of the plurality of sub-histograms by adapting its bar width. However, embodiments are not limited to automotive applications and may include drones, robots, and the like.

[0036] Other specific embodiments are directed to methods of using the apparatus described above, for example, methods for sensing a physical object located in an operable region, as otherwise described herein.

[0037] Turning now to the accompanying drawings, Figure 1 An example of a device according to the invention is shown. Device 100 comprises sensor circuitry 104 and processing circuitry 102 .

[0038] As an example of the present invention, the sensor circuit 104 can generate and sense detected signals corresponding to physical objects that are located in an operable region relative to the position of the sensor circuit 104. In a specific embodiment, the sensor circuit 104 may include a SPAD array circuit that senses an optical signal and, in response, generates a detected signal. The SPAD array circuit may include an array of SPADs. For example, the sensor circuit 104 may use a transmitter circuit (e.g., a transmitter (TX) front-end circuit 114) to transmit a signal that may be reflected from a physical object within the operable region of the transmitted signal. The signal is reflected from the target object, and the reflected signal is detected via a receiver circuit (e.g., a receiver (RX) front-end circuit 116) of the sensor circuit 104. The generated signal and the received reflected signal are used to determine the flight time of the signal as transmitted and reflected.

[0039] As may be appreciated, SPADs (including Geiger-mode avalanche photodiodes) are detectors capable of capturing individual photons with very high arrival time resolution on the order of tens of picoseconds. SPADs can be fabricated in dedicated semiconductor processes or in standard CMOS technology using arrays of known SPAD sensors, e.g., for use in three-dimensional (3D) imaging cameras.

[0040] In a particular embodiment, the sensor circuit 104 includes a TX front-end circuit 114 and an RX front-end circuit 116. As further described herein, the TX front-end circuit 114 and the RX front-end circuit 116 are used to transmit a reference signal and detect (reflected) signals therefrom (e.g., via a TX lens 115 and an RX lens 117) that correspond to a physical object and are used to estimate a time of flight of the reference signal and the reflected signal that indicates a distance of the physical object from the sensor circuit 104.

[0041] The device further includes a processing circuit 102 in communication with the sensor circuit 104. The processing circuit 102 is used to generate an adaptive histogram using the detected signal, and to determine or estimate the distance of the physical object from the sensor circuit 104 using the adaptive histogram. The processing circuit 102 may include an array of processing circuits coupled to the SPAD array circuit. The adaptive histogram includes or refers to a histogram having a plurality of sub-histograms, wherein each sub-histogram has an adaptive bar width and the combination of the plurality of sub-histograms captures an operating area. As further described herein, the bar width may be adaptive to optimize one or more precision metrics based on an indication of power sensed in the received signal and / or an indication of a range or distance between the sensor circuit 104 and one or more of the physical objects. For example, the processing circuit 102 records and organizes information associated with the detected signal in a plurality of sub-histograms, which are respectively associated with different precision metrics for corresponding sub-areas of the operational area, and each of the plurality of sub-histograms includes a histogram bar set, which is characterized by a bar width associated with its precision metric. The processing circuit 102 additionally optimizes at least one of the accuracy metrics by dynamically adapting one or more of the bin widths in response to the detected signal.

[0042] pass Figure 1The specific embodiment shown shows a processing circuit 102, for example, as part of an automotive LIDAR system. In such embodiments, the processing circuit 102 includes a controller circuit 106, an adaptive histogram circuit 112, a digital signal processor 110, and a storage circuit (e.g., a random access memory (RAM)) 108. The controller circuit 106 synchronizes and arranges the complete system. The controller circuit 106 sends a reference signal or a timing signal to the TX front-end circuit 114 and the RX front-end circuit 116 to synchronize the time source for flight time estimation. The controller circuit 106 can also control histogram parameters and / or settings, for example, the column width setting for the sub-histogram. The TX front-end circuit 114 and the RX front-end circuit 116 may include a variety of known front-end signal receiving circuits and related logic for processing of I / O signals involving the controller circuit 106. The output of the RX front-end circuit 116 is a single detected flight time record (hereinafter time or data record) in a digital format. After the detection of the signal, the histogram is updated by the new data record. The adaptive histogram circuit 112 may be used to generate and store an adaptive histogram having sub-histograms with adaptive bin widths.

[0043] In various embodiments, the apparatus 100 further includes a car to which the sensor circuit 104 and the processing circuit 102 are fastened. During movement of the car, the sensor circuit 104 dynamically generates a detected signal, and in response, the processing circuit 102 dynamically optimizes the accuracy metric for at least one of the plurality of sub-histograms by adapting its bin width. As may be appreciated, the embodiments are not limited to automotive applications.

[0044] According to a number of embodiments, this accuracy measure can indicate the level of detection reliability and / or resolution. For example, the sensor circuit 104 can measure the flight time by sensing the optical signal, and the processing circuit 102 uses a plurality of sub-histograms with one or more adapted bar widths to determine the distance of two or more physical objects from the sensor circuit 104. In a specific embodiment, the plurality of sub-histograms include at least two sub-histograms corresponding to two different sub-areas of the operable area. Each of the two or more sub-histograms can have a different bar width than the other sub-histograms, and at least one of the different bar widths is dynamically adapted in response to at least a portion of the detected signal. In various specific embodiments, each of the plurality of histogram bars has a bar width that is less than half the width of the emitted light pulse.

[0045] According to various embodiments, column width is based on the power adjustment of the detected signal. As described above, column width can be applicable to the precision measurement of optimization histogram. The precision measure can indicate the detection reliability and / or the level of resolution of the subhistogram. This adaptive column width technology may include a variable column width relative to a range, and the column width is used to optimize resolution and detection reliability. By adapting the column width based on the power of the detected signal, a weak signal can have a higher detection reliability and a strong signal can have a higher resolution for the complete operating area of ​​the histogram. For example, narrow columns are associated with strong signals, for example, close targets and / or high reflectivity, to achieve higher resolution, and wider columns are associated with weak signals, for example, long-distance targets and / or low reflectivity, to ensure higher detection reliability. Column width can also change in time, for example, with an acquisition cycle, adapted to dynamic environments.

[0046] As further shown herein, for example, by Figure 5 , processing circuit 102 (via adaptive histogram circuit 112) can divide the complete operational area into sub-areas, and these sub-areas have their corresponding sub-histograms respectively. Each sub-area can be suitable for signal power by changing the corresponding column width independently of other sub-areas. For example, the operational area is subdivided into four sub-areas, and these four sub-areas have their corresponding sub-histograms respectively, and have the column width of 1,1,2,8x least significant bit (leastsignificant bit, LSB) (x times), respectively, wherein LSB is the least significant bit of flight time record. Each sub-histogram can be considered as an independent equal width histogram, with the simplicity of its maintenance and the ability to update in real time without partial or complete reconstruction. However, the sub-histograms of all combinations simulate the effect of the histogram based on frequency, for example, equal depth, because the column width for each sub-histogram is adjusted to the signal power and / or can be changed in time, with acquisition cycle. The plurality of sub-histograms correspond to different sub-areas of the operational area, wherein each of the different sub-areas at least temporarily covers a different range indicated by a (possible) distance between the sensor circuit 104 and one or more of the physical objects.

[0047] More specifically, the processing circuit 102 may optimize one or more accuracy metrics based on at least one of the indication of power or distance sensed in the optically received signal by changing one of the bin widths of one of the plurality of different sub-histograms for coarser detection to optimize or improve detection reliability of one or more far-end physical objects in the physical objects, and by changing another of the bin widths of another of the plurality of sub-histograms for finer detection to optimize or improve resolution with respect to detecting one or more near-end physical objects in the physical objects. The processing circuit 102 may facilitate detection reliability based on the SNR of the generated signal so as to process peak detection of the generated signal.

[0048] In other related embodiments, such accuracy metrics may additionally and / or alternatively be optimized based on a range or distance between the sensor circuit 104 and one or more physical objects (e.g., a sub-region of the operable region). The processing circuit 102 may optimize the accuracy metric based on at least one of an indication of power sensed in the optically received signal and an indication of a distance between the sensor circuit and one or more of the physical objects.

[0049] The processing circuit 102 may generate a plurality of sub-histograms at least temporarily based on empirical data indicating the correspondence between the bar width and the detected signal power. The empirical data may be provided from a storage circuit 108 integrated with the processing circuit 102. In other embodiments and / or in addition, the empirical data may be received from a communication channel provided from an external circuit. For example, the empirical data may come from an external system (e.g., an external navigation system) or from another circuit associated with the device 100, such as an internal navigation system of a car having a pre-programmed table based on an analysis (e.g., an analysis of the driver's habits and / or environment).

[0050] The sensor circuit 104 and the processing circuit 102 may operate simultaneously in real time. For example, the sensor circuit 104 generates a detected signal while the processing circuit constructs information for the plurality of sub-histograms. In a specific embodiment, the sensor circuit 104 operates in an automatic or semi-automatic driving mode, wherein the detected signal is generated dynamically, and wherein the bins of the plurality of sub-histograms have dynamically changing information.

[0051] Figure 2 An example of the operation of a device according to the invention is shown, for example, in combination with Figure 1 Type of device. Figure 2 The specific embodiment shown may include Figure 1The device 100 shown is used in an automobile, a robot, a drone, or other application related to the detection of a physical object. As previously described, the physical object can be detected using a time-of-flight technique by transmitting a signal via a TX front-end circuit and receiving a reflected signal via an RX back-end circuit in response thereto. The range (e.g., distance) r is calculated based on the time of flight (t1-t0). Timestamping can be via edges or peaks, using thresholding and a time-to-digital converter (TDC) or an analog / digital converter (ADC) and processing (e.g., where the TDC and ADC are high-speed circuits).

[0052] Figure 2 Another example of the present invention is also shown in accordance with the specific circuit-based manner of operation, which is combined with the above Figure 1 The aspects discussed are consistent. As shown, Figure 2 The approach shown in FIG. 2 relates to a vehicle 219 and its circuit, which includes an optical signal generating circuit 221 and an optical signal processing circuit 225, which form a sensor circuit, for example, composed of Figure 1 1. The optical signal generating circuit 221 generates optical radiation, which is detected by the optical signal processing circuit 225 when the optical radiation is reflected from a physical object (e.g., a tree, a building, a car). Additional circuits are configured to process the signals output from the optical signal generating circuit 221 and the optical signal processing circuit 225. In various embodiments, the optical signal generating circuit 221 generates optical radiation, which is detected by the optical signal processing circuit 225 when the optical radiation is reflected from the physical object (e.g., a tree, a building, a car). Figure 1 and 2 Shown (and by Fig. 8A The sensor circuit (also shown) forms a LiDAR system.

[0053] The optical signal generation circuit 221 includes an emitter driver 224, a timing engine 230, and an emitter illuminator 220 that form an emitter path (also referred to as an emission path). The emission of optical radiation by the emitter illuminator 220 is controlled by the emitter driver 224 and the timing engine 230. The emitter illuminator 220 constitutes an illuminator (e.g., a laser or otherwise). Example illuminators include light emitting diodes, edge emitting lasers, vertical cavity surface emitting lasers, and known arrays thereof. The optical radiation emitted by the emitter illuminator 220 travels until it is reflected by a physical object (e.g., a tree, a building, another motor vehicle). The reflected optical radiation is sensed / received by the optical signal processing circuit 225, at which time the flight time of the emitted optical radiation is determined.

[0054] The optical signal processing circuit 225 includes a single photon avalanche diode (SPAD) array 222 and a threshold (level detection) circuit 228 (and a quenching circuit 227). The SPAD array 222 is configured to detect the reflected optical radiation. Once detected by the SPAD array 222, the optical radiation is thresholded (e.g., compared to a static point through the threshold circuit 228), after which the optical radiation is passed to additional circuitry for additional processing. The sensor circuitry of the LIDAR system may additionally include a lens and a bandpass filter, which may be located in the LIDAR system relative to the SPAD array 222.

[0055] It should be noted that in some embodiments, the emitter path may include a micro-electro-mechanical system (MEMS) scanner and a MEMS driver for two-dimensional (2D) steering of a laser beam or one-dimensional (1D) steering of an array of laser beams or line lasers. In other embodiments, the emitter or emitter path may include an optical phase array and a driver for scanning the laser beam.

[0056] In yet other embodiments, the transmitter or transmitter path may include a VCSEL (vertical cavity surface emitting laser) array.

[0057] In still other embodiments, the emitter or emitter path may include a lens system for spreading the beam into a full field of view (e.g., a flash lamp) or for laser beam collimation. It will thus be appreciated that a number of different configurations may be implemented for the emitter or emitter path. The key is that the disclosed emitters may be configured from a number of different types of devices and components.

[0058] Additional circuitry configured to process at least one signal output from the optical signal generation circuit 221 and the optical signal processing circuit 225 includes a time-to-digital converter (TDC) 232, a histogramming circuit (e.g., block / IC) 234, and a signal processing circuit 236. The TDC 232 is configured to receive detected reflected optical radiation (which indicates the time of incidence of a single photon) from the optical signal processing circuit 225, as detected by the constituent SPAD array 222. The time of incidence is used to increment a count in a memory of the corresponding incident photon of the optical signal processing circuit 225 via the histogramming circuit 234 and for signal processing via the signal processing circuit 236.

[0059] After an event (e.g., detection of a reflected signal), a timing circuit (e.g., Figure 2The TDC 232 in the embodiment shown in FIG. 2 outputs the time difference between the event and the reference / signal, which is called the measured flight time record, which can be written in a flight time record register or latch. The record from the flight time register or latch can be read out and passed to the histogram block for subsequent histogram generation and storage. Although the timing circuit (e.g., TDC 232) is through Figure 2 Shown as separate from the optical signal processing circuit 225 , but in various embodiments the timing circuit may form part of the optical signal processing circuit 225 .

[0060] It should be noted that in alternative embodiments, the timing circuitry may be integrated in the pixel, or shared across columns, rows, or other sections via means of column / row decoding, multiplexing, etc. The timing circuitry may operate in both forward and reverse modes. The forward mode may imply the service of the reference signal as a start signal and the SPAD detection as an end signal, and vice versa in the reverse mode.

[0061] A digital signal processor (e.g., signal processing circuit 236) may be responsible for the execution of algorithms for signal detection (e.g., CFAR or "Constant False Alarm Rate" detection algorithms) and for writing detected signals to RAM. The final point cloud generated from the unique reflections may be retrieved from RAM. Processing by the digital signal processor may extend beyond signal detection, including, but not limited to, subsequent point cloud filtering, segmentation, object classification, and state estimation.

[0062] For example Figure 1 The storage circuit depicted in can be configured to store corresponding counts of photons that arrive at a plurality of different time bins within the operable region, the time bins having bin widths for different sub-regions, the bin widths being adapted by the histogramming circuit 234 to optimize a precision measure for one or more sub-histograms. When acquiring measurements, the memory associated with the optical signal processing circuit 225 stores corresponding counts of photons that arrive at the processing circuit in a plurality of different time bins having different widths that span the detection window set by the emitter driver 224 and the timing engine 230 for this optical signal processing circuit 225. The controller can process the histograms of corresponding counts on different bins for the sensing circuit to derive and output corresponding arrival time values ​​for the optical signal processing circuit 225.

[0063] As may be appreciated, embodiments are not limited to the use of a TDC and may include an apparatus having an analog-to-digital converter (ADC). For example, the detected reflected optical radiation is provided to the ADC via a correlated double sampling (CDS) circuit that receives the detected signal from the optical signal processing circuit 225 and outputs a sampled signal having a voltage proportional to the detected arrival time of the photon to the ADC. For general and specific information regarding estimating flight time and specific information regarding the use of a TDC or ADC, reference is made to U.S. Patent No. 9,502,458, filed on March 10, 2015, entitled “Circuit for generating direct timing histogram data in response to photon detection,” and U.S. Publication No. 14 / 830,760, filed on August 20, 2015, entitled “SPAD array with gated histogram construction.”

[0064] Figures 3A to 3C Examples of histogram bin widths according to the present invention are shown. Figures 3A to 3C Three equal width histograms with different bin widths for the same data are shown, where each histogram represents the full operational region. Assume a TDC resolution of 333 ps (eg, an LSB corresponds to 333 ps). As shown by Figure 3A As shown, in a histogram with a bin width of 1xLSB (e.g., bin width 1xLSB / 0.33ns and 2048 bins), a strong signal from a physical object at 17m has a distinct peak, as shown by 309, whereas a weaker signal from another physical object at 81m is indistinguishable, as shown by 311. As shown by Figure 3B As shown, in a histogram with a bin width of 8xLSB (e.g., bin width 8xLSB / 2.67ns and 256 bins), both peaks at 17m and at 81m are resolvable, as shown by 313 and 315, respectively. A possible disadvantage of the 8xLSB bin width (depending on the application and specific embodiment) is the loss of range resolution, 40cm vs. 5cm. Depending on the use case, the 40cm range resolution may be sufficient for other physical objects at 81m (e.g., as shown by 315), but not sufficient for the physical object at 17m (e.g., as shown by 313). Figure 3C(where the bin width is set to 64xLSB (e.g., bin width 64xLSB / 21.33ns and having 32 bins)) shows a different side effect at the other extreme: if the bin width is set too wide, then weak signals are eliminated along with the noise, as shown by 319 (e.g., peak at 81m) compared to 317 (e.g., peak at 17m).

[0065] The tradeoff between resolution for strong signals and detection reliability for weak signals is the signal power and the resulting corresponding cumulative photon distribution. Due to the blocking nature of the SPAD array circuit, for example, earlier photons block later photons due to device quenching and subsequent dead time, a different signal shape than expected may exist in the resulting histogram. Strong signals saturate the SPAD array circuit almost instantaneously, while weak signals require longer integration times for the SPAD array circuit to become saturated.

[0066] Figures 4A to 4B An example pulse shape of a signal according to the present invention is shown. More specifically, Figure 4A and 4B A fine resolution (10ps) histogram with a higher number of measurements for target reflections at 17m and 81m, respectively, is shown. The original signal is a rectangular pulse with a full width at half maximum (FWHM) of 0.5 nanoseconds (ns) and a rise time and fall time of 0.5ns. The FWHM of the target reflection at 17m is about 0.7ns and the FWHM of the target reflection at 81m is about 5.0ns, for example, the latter corresponding to the FWHM of the original rectangular pulse signal. In general, a bin width of half the FWHM can be optimal because it ensures that at least one bin is fully included in the FWHM of the detected signal, for example, maximizing the SNR of the bin. Therefore, the optimal bin width for the corresponding physical objects is also different, with 0.35ns or approximately 1xLSB for the physical object at 17m (recall that in the example described above, LSB is equal to 333ps), and 2.5ns or approximately 8xLSB for another physical object at 81m.

[0067] Figure 5 An example of a power adaptive histogram according to various embodiments is shown. As previously described, the full operational area is divided into sub-areas, each of which has its own independent sub-histogram with bin widths adapted based on signal power. Because the bin widths are adjusted relative to signal strength with an acquisition cycle, the technique emulates the quality of a frequency-based histogram for real-time applications while retaining the low maintenance of a constant-width histogram. Figure 5The specific histogram shown has four sub-histograms 501, 503, 505, 507, which have Figures 3A to 3C The bin widths are 1, 1, 2, and 8 LSB on the data. Therefore, subhistogram 0 (identified as 501, for example) contains 512 bins, and subhistogram Figure 1 (e.g., identified as 503) contains 512 bins, sub-histograms Figure 2 (e.g., identified as 505) contains 256 bins, and sub-histogram 3 (e.g., identified as 507) contains 64 bins. However, as may be appreciated, embodiments are not limited to histograms having four sub-histograms and may include more than or less than four sub-histograms. In this example, reflections from a physical object at 17 m (e.g., as shown by 509) belong to sub-histogram 0 having a bin width of 1 x LSB, and reflections from another physical object at 81 m (e.g., as shown by 511) belong to sub-histogram 3 having a bin width of 8 x LSB.

[0068] In this example, the reflected signal from the physical object at 17m is stronger than the reflected signal from the other physical object at 81m, proportional to the inverse square of the range. However, it is worth noting that the signal from the closer physical object is not necessarily stronger than the signal from the farther physical object. For example, an object with a reflectivity of 0.8 or 80% at a distance of 50m can cause a stronger reflected signal at the receiver than an object with a reflectivity of 0.1 or 10% at a distance of 30m. Therefore, the sub-area containing 50m can have a smaller bar width than the sub-area containing the 30m range. Also, it should be noted that although for hard targets (as in the example provided above), the FWHM of the reflected signal is limited by the FWHM of the original transmitted signal, this may not be the case for so-called soft targets such as water droplets or glass.

[0069] The adaptive histogram described and illustrated above is an example of a power adaptive histogram segmentation. However, the embodiments are not so limited, and various segmentations of the full operating area may be performed based on what is best given the number of reflected signals, their strengths, memory limitations, etc. Finer sub-areas (e.g., the full area is divided into a higher number of sub-areas) may lead to better results, but may be accompanied by additional processing and memory costs. In each acquisition cycle (see Fig. 6A ), for example, when the final histogram is completed, the histogram is analyzed and the optimal histogram settings for the next acquisition cycle are set, such as: the number of sub-regions, the boundaries of the sub-regions, and the bin width for each sub-region.

[0070] Figures 6A to 6BAn example method of using the device according to the invention is shown. Such a method can be used to simultaneously detect more than one physical object within the operable area, at different sub-areas of the operating area.

[0071] More specifically, Fig. 6A is an example of a power adaptive histogram column bar. Process 640 includes, at 642, for a first acquisition cycle (e.g., an acquisition cycle shown by 639), setting the histogram by default parameters: the number of sub-histograms, the column width; allocating memory for the sub-histograms and initializing the memory with zeros. At 644, the TX front-end circuit and the RX front-end circuit transmit a signal and measure the flight time of the detection (noise or return signal). At 646, the sub-histogram to which the flight time record 645 (hereinafter, the time or data record) belongs is identified, and at 648, the corresponding sub-histogram is updated by incrementing the corresponding column bar. At 641, the measurement is repeated multiple times, and at 650, the final histogram 649 is completed. At 647, the histogram is processed, and at 643, the histogram is analyzed for the best histogram setting for the next acquisition cycle.

[0072] Figure 6B An example of a signal flow process for a LIDAR system operating using adaptive barcodes according to various embodiments is shown. As previously described, process 651 may be used to sense a physical object located in an operable region.

[0073] Process 651 includes, at 653, transmitting a signal by a sensor circuit (e.g., a TX front-end circuit). More specifically, at 653, a laser signal may be transmitted. Thereafter, as depicted at 655, the signal may be reflected from a target located in the path of the transmitted laser signal. Subsequently, after detection of the reflected light, as indicated at 657, the time of flight may be recorded in a time of flight record, as shown at 659. The measurement may be repeated multiple times, as indicated by arrow 654, and statistical methods may be used to identify the actual time of flight and, therefore, the range to the target, as depicted at 661. That is, the distance to the target may be estimated via statistical methods, as indicated at 661.

[0074] Combining the unique target reflections within a single time frame from the full receiver array, steps or operations may be performed as shown thereafter at 663, where a 2D / 3D point cloud of the environment may be generated at 668. In various alternative embodiments, the processing of the LIDAR system may extend beyond 2D / 3D point cloud generation and may include additional filtering (e.g., Kalman filter), segmentation, object classification, and state estimation of the point cloud.

[0075] It should be noted that as used herein, the term point cloud may refer to a collection of data points in space. A point cloud may be used to measure a large number of points on a surface and a large number of points of objects around such a surface. A point cloud may also refer to an organized point cloud map obtained by indexing 3D data into a LIDAR-specific 2D coordinate system or vice versa.

[0076] The device and / or process can be used to perform a variety of additional methods. The example method includes using a sensor circuit to sense and generate a detected signal corresponding to a physical object, which is located in an operable area relative to the position of the sensor circuit. The method further includes using a processing circuit that operates in response to the detected signal to record and organize information associated with the detected signal in a plurality of sub-histograms, each of which is associated with different precision measures for corresponding sub-areas of the operable area, each of the plurality of sub-histograms includes a histogram bar set, characterized by a bar width associated with the precision measure of the histogram bar set, and to optimize at least one of the precision measures by dynamically adapting one or more of the bar widths in response to the detected signal. As previously described, adapting one or more of the bar widths may include dynamically adapting the bar width of at least one of the sub-histograms in response to the signal detected in the one or more bars. In addition, the information can be organized in two or more sub-histograms, wherein at least one of the two or more sub-histograms has a different bar width than another of the two or more sub-histograms. In some embodiments, the information may be organized in three or more sub-histograms, for example, in four sub-histograms.

[0077] Figure 7 Specific examples of sub-histograms with adaptive binning according to various embodiments are shown. More specifically, Figure 7 Shown in the Figures 3A to 3B Example range adaptive histogram 775 on data of , where the full operational region is divided into sub-regions, each of which has its own independent sub-histogram with bin widths appropriate for the range of the operational region. Figure 7 The particular histogram shown has four subhistograms 777, 779, 781, 783, which have bin widths of 1, 2, 4, and 8 LSB, respectively. Thus, subhistogram 0 (e.g., identified as 777) contains 512 bins, and subhistograms 777, 779, 781, and 783 have bin widths of 1, 2, 4, and 8 LSB, respectively. Figure 1 (For example, identified as 779) contains 256 bins, sub-histograms Figure 2(e.g., identified as 781) contains 128 bins, and sub-histogram 3 (e.g., identified as 783) contains 64 bins. In this example, there is high range resolution for close targets, as shown by 785, and higher sensitivity for far targets, as shown by 787. Such embodiments involve an efficient (fast and cheap) architecture that compensates for the inverse square power drop in range, such as described elsewhere herein and by Figures 8A to 8B The architecture can be used to effectively partition the full operational range into sub-histograms, adjust the bin width or bin resolution of each sub-histogram, etc.

[0078] Figures 8A to 8B Shown is an example of data processing of a sub-histogram with a range adaptability column bar according to various embodiments. As previously described, embodiments are not limited to the power adaptation column bar width based on the detected reflected signal. In various embodiments, in addition to adapting the column bar width based on signal power, the column bar width can be applicable to compensating for the inverse square decline of the range, which is sometimes also referred to as the range adaptability histogram or range adaptability application of the power adaptability histogram column bar technology of the compensation range inverse square power decline in this article. Such range adaptability applications can be relatively cheap in terms of calculation and memory and well compensate for the range inverse square power decline.

[0079] Fig. 8A An example range adaptive histogram circuit according to various embodiments is shown. In various embodiments, the range adaptive histogram circuit may include Figure 1 860, a storage circuit (e.g., a histogram RAM 868), an adder 863, and a memory 864. Fig. 8A Operation of other circuit components of those shown (e.g., hist.[indi]++866).

[0080] Because histogram construction is a non-vectorizable task and cannot be effectively processed by a digital signal processor, it may be beneficial to have a technology equipped with a fast and simultaneously inexpensive way that enables real-time histogram construction to be parallel to data acquisition, for example, updating the histogram immediately once a new time record arrives. Through such embodiments, several advantages may be obtained. First, since data acquisition and histogram construction are completed in parallel, the overall cycle time can be reduced. In addition, such embodiments may include lower memory requirements, because the histogram is a compressed representation of data, but individual records are not stored. The range adaptability embodiment may be implemented by several considerations, including how the complete operational area is subdivided, and how to select the column width. Specifically, the way to divide the operational area is to make only the first M most significant bits (most significant bit, MSB) of any data record determine the corresponding sub-histogram and the remaining (NM) bits determine the corresponding column index inside the sub-histogram, where N is the total number of bits. There may be no overlap between the bits that determine the sub-histogram and the corresponding local column index. Another simplification is that the column width will be in the form of 2^wx LSB. Therefore, the local bin index 865 of the time record is simply retrieved via the right shift register 860 by shifting the last (NM) bits right by w (w is referred to as the bin resolution hereinafter). This limits the number of possible bin widths and significantly reduces computational costs. Block 862 uses the M MSBs to retrieve the global index 867 of the first bin of the corresponding sub-histogram within the complete array. Via adder 863, the local bin index 865 of the time record and the global index 867 of the first bin of the corresponding sub-histogram are used to determine the global bin index 861 of the time record. The corresponding bin is then updated in the histogram memory 868.

[0081] pass Fig. 8A The circuit shown can have various conditions. First, if the histogram is divided into a large number of sub-histograms, then the bin widths become too wide very quickly (exponential growth) and the signal is washed out by the noise (see Figure 3C ). As previously described, a bin width of half the FWHM of the detected signal can ensure that at least one bin is completely included in the FWHM, which can maximize the SNR of the bin. For hard physical objects (as opposed to soft physical objects such as water droplets, glass, etc.), the FWHM of the detected signal does not exceed the FWHM of the transmitted signal. Since the FWHM of the transmitted signal is a known parameter, the maximum bin width can be limited by half of this amount. Another condition is slightly opposite to the first one: even for additional physical objects (e.g., belonging to sub-histogram >=1, the signal can be strong enough for the optimal bin width of 1xLSB. To solve this problem, the full histogram is divided into three so-called pages, as shown by Figure 8B Shown.

[0082] If through Figure 8B 870, wherein the histogram is divided into pages, as depicted at 870, wherein bin_res represents the bin resolution, and w_max is calculated as the closest optimal resolution for the FWHM of the transmitted signal. Additional versatility of the range adaptive histogram can be achieved by controlling t1 and t2 and by additional conditional checks. After each acquisition, the histogram can be analyzed for optimal changes to the histogram settings, such as: t1, t2, the number of sub-histograms in the range adaptive page, etc.

[0083] Additionally, system parameters such as TDC resolution and / or transmit signal FWHM may be set during the design phase to further optimize the range adaptive histogram binning technique. For example, by ensuring:

[0084] FWHM TX =2 w *Δt TDC / 2,

[0085] Where w is the maximum number of sub-histograms in the range adaptability zone. This helps ensure that the coarsest bin resolution w is optimal for the weakest / farthest possible signals.

[0086] Embodiment as described above relates to the adaptive histogram that produces with subhistogram, and these subhistograms have adaptive column width.The column width of subhistogram is selected to maximize the SNR of detected signal for reliable subsequent peak detection processing, for example, constant false alarm rate (CFAR).The CFAR power of each column is compared with the local noise power of the protection column that excludes the column concerned.According to the adaptive histogram technology of various embodiments, provide the compromise in more efficient and computationally cheap mode compared with the potential adaptation in the peak detection processing on equal width histogram.By optimizing the column width of subhistogram to half of the detected signal width, maximum SNR is guaranteed, because at least one column will be included in the FWHM of signal completely.If the protection zone is narrower than the detected signal width (for example, the protection column number is lower and / or the column width is narrower), the equal width histogram can show possible shortcomings so, in this case the part of signal is included in the local noise estimation and this therefore reduces the speed of signal detection. And, if a wider local noise zone is selected to reduce the impact of the detected signal, this can cause a delayed and limited response to local noise changes. In addition, if the protection zone is wider than the detected signal width, the range resolution and / or accuracy is coarser than what can be physically resolved in other ways; and if a more advanced integration basis is used (a multiplication factor of 1 is used for the current bar, 0.5 is used for the protection bar to the right of it, 0.35 is used for another bar, etc.) when estimating the current bar power, then higher computing power requirements can be caused. Accordingly, the aspects disclosed herein take into account each of these factors or conditions and provide a balance of the compromise between the resolution for strong signals and the detection reliability for weak signals.

[0087] In addition, the power adaptability histogram can be constructed in parallel with data acquisition, and reconstructed from an equal width histogram with the narrowest possible column width (1xLSB). In the latter case, an additional process of reconstructing the histogram with complexity O (N) is added to the processing pipeline. In addition, the latter can use a higher memory capacity to store the original histogram. Because an example relates to the above discussion and relates to Figure 7 , so with 16 bits per bin, the power adaptive histogram can use 15,360 bits (per single SPAD pixel), while a constant width histogram with a bin width of 1xLSB can use 32.768 bits. More than double memory savings are achieved when the power adaptive histogram is built in parallel to data acquisition.

[0088] For the range adaptive histogram in the application of power adaptive histogram column bar technology, the advantage can be realized in conjunction with certain embodiments.By ensuring that the column width is in the form of (2^w)x LSB, where w is called the column bar resolution, the column bar index can be retrieved via a simple shift operation (w cycles).The sub-histogram is predefined to compensate for the range inverse square power drop.Similarly, for the column bar index, the sub-histogram of the data record and the pointer to the first column bar in the memory are retrieved via a collection of simple arithmetic operations on the M most significant bits.In some of these embodiments, the versatility of the range adaptive histogram is added to via the introduction of pages, where page 0 corresponds to an equal width histogram of column bar resolution 0, page 1 is a range adaptive page, and page 2 corresponds to an equal width histogram of maximum column bar resolution, for additional subtraction operations at low prices.

[0089] Terms illustrating orientations, such as upper / lower, left / right, top / bottom, and above / below, etc., may be used herein to refer to the relative positions of elements as shown in the accompanying drawings. It should be understood that the terms are used for convenience of notation only and that in actual use, the disclosed structures may be oriented differently from the orientations shown in the accompanying drawings. Therefore, the terms should not be interpreted in a limiting manner.

[0090] Unless otherwise indicated, those skilled in the art will recognize that various terms used in the specification (including claims) mean ordinary meanings in the art. For example, the specification describes and / or illustrates aspects that can be used to combine the claimed disclosure with the help of various circuits or circuit systems, which can be shown as terms or using terms such as blocks, modules, devices, systems, units, controllers and / or other circuit type descriptions (e.g., Figures 1 to 2 Reference numerals 106 and 234 depict blocks / modules as described herein). Such circuits or circuit systems are used with other elements to illustrate how certain embodiments may be performed in form or structure, steps, functions, operations, activities, etc. For example, in some of the embodiments discussed above, one or more modules are discrete logic circuits or programmable logic circuits that are configured and arranged to perform these operations / activities, such as may be shown in FIG. Figure 2 and 8A In some embodiments, such a programmable circuit is one or more computer circuits, including a memory circuit for storing and accessing a program as a set (or multiple sets) of instructions to be executed (and / or as configuration data defining how the programmable circuit is to be executed), and the programmable circuit uses a program as described in Figures 6A to 6BThe algorithms or processes described herein are used to perform the relevant steps, functions, operations, activities, etc. Depending on the application, the instructions (and / or configuration data) may be implemented using logic circuits, where the instructions (whether characterized as being in the form of object code, firmware, or software) are stored in and accessible from a memory (circuit).

[0091] Based on the above discussion and examples, those skilled in the art will readily recognize that various modifications and changes may be made to the various embodiments without strictly following the exemplary embodiments and applications shown and described herein. For example, the methods illustrated in the figures may involve steps performed in various orders, wherein one or more aspects of the embodiments herein are maintained, or may involve fewer or more steps. For example, by Fig. 6A The process shown can be compared with Figure 6B As another example, by Figure 2 The circuit components shown can be used with Fig. 8A Such modifications do not depart from the true spirit and scope of the various aspects of the present invention including those set forth in the claims.

Claims

1. A device, characterized in that: include: a sensor circuit configured and arranged to generate and sense a detected signal corresponding to a physical object located in an operable region relative to a position of the sensor circuit; as well as A processing circuit configured and arranged to: recording and organizing information associated with the detected signals in a plurality of sub-histograms, the plurality of sub-histograms being respectively associated with different measures of precision for corresponding sub-regions of the operational region, each of the plurality of sub-histograms comprising a set of histogram bins characterized by a bin width associated with the measure of precision of the set of histogram bins; and At least one of the different accuracy measures is optimized by dynamically adapting one or more of the bin widths in response to the signal power of the detected signal, wherein each of the one or more bin widths is optimized to half of the full width at half maximum (FWHM) of the corresponding detected signal.

2. The device according to claim 1, characterized in that The at least one accuracy metric is indicative of detection reliability, and wherein the sensor circuit is configured and arranged to measure flight time by sensing a light signal, and the processing circuit is configured and arranged to simultaneously determine distances of two or more physical objects from the sensor circuit using the plurality of sub-histograms having adapted one or more of the bin widths.

3. The device according to claim 1, characterized in that The at least one precision measure is indicative of a level of resolution, and wherein the plurality of sub-histograms include at least two sub-histograms corresponding to two different sub-areas of the operational area.

4. The device according to claim 1, characterized in that Also included is a car, to which the sensor circuit and the processing circuit are secured, and wherein during movement of the car, the sensor circuit is further configured and arranged to dynamically generate the detected signal, and in response, the processing circuit is further configured and arranged to dynamically optimize the at least one accuracy metric for at least one of the multiple sub-histograms by adapting the bin width of at least one of the multiple sub-histograms.

5. The device according to claim 1, characterized in that The plurality of sub-histograms correspond to different sub-areas of the operable area, wherein each of the different sub-areas at least temporarily covers a different range indicated by a distance between the sensor circuit and one or more of the physical objects.

6. The device according to claim 1, characterized in that The processing circuit is further configured and arranged to optimize the at least one accuracy measure in response to the detected signal indicative of the presence of a proximal one of the physical objects and the presence of a distal one of the physical objects by: adjusting a first bin width for a first one of the plurality of sub-histograms for optimizing or improving a resolution associated with the proximal ones of the physical objects; as well as A second bin width corresponding to a second one of the plurality of sub-histograms is adjusted for optimizing or improving detection reliability associated with the far-end ones of the physical objects.

7. The device according to claim 1, characterized in that The processing circuit is further configured and arranged to optimize the at least one accuracy measure in dependence on at least one of an indication of power sensed in the optically received signal and an indication of a distance between the sensor circuit and one or more of the physical objects.

8. The device according to claim 1, characterized in that The processing circuit is further configured and arranged to optimize the at least one accuracy measure based on at least one of an indication of power or distance sensed in the optically received signal by changing one of the bin widths of one of the plurality of different sub-histograms for coarse detection to optimize or improve detection reliability of one or more distal ones of the physical objects, and changing another of the bin widths of another of the plurality of sub-histograms for fine detection to optimize or improve resolution with respect to detecting one or more proximal ones of the physical objects.

9. The device according to claim 1, characterized in that The processing circuit is further configured and arranged to facilitate detection reliability based on a signal-to-noise ratio (SNR) of the detected signal so as to process peak detection of the detected signal, wherein at least one of the bin widths is adjusted as an iterative step to improve the peak detection.

10. A method for sensing a physical object located in an operable area, characterized in that: The method comprises: sensing and generating a detected signal corresponding to a physical object with a position relative to the sensor circuit that is located in the operable region using a sensor circuit; and using processing circuitry operating in response to the detected signal to: recording and organizing information associated with the detected signals in a plurality of sub-histograms, the plurality of sub-histograms being respectively associated with different measures of precision for corresponding sub-regions of the operational region, each of the plurality of sub-histograms comprising a set of histogram bins characterized by a bin width associated with the measure of precision of the sub-histogram; and At least one of the different accuracy metrics is optimized by dynamically adapting one or more of the bin widths in response to the signal power of the detected signal, wherein each of the one or more bin widths is optimized to half of the full width at half maximum (FWHM) of the corresponding detected signal.

Images