WO2024119906A1 - 激光雷达、数据处理方法及光探测和数据采集处理装置 - Google Patents

激光雷达、数据处理方法及光探测和数据采集处理装置 Download PDF

Info

Publication number
WO2024119906A1
WO2024119906A1 PCT/CN2023/115383 CN2023115383W WO2024119906A1 WO 2024119906 A1 WO2024119906 A1 WO 2024119906A1 CN 2023115383 W CN2023115383 W CN 2023115383W WO 2024119906 A1 WO2024119906 A1 WO 2024119906A1
Authority
WO
WIPO (PCT)
Prior art keywords
detection unit
laser radar
array
detection
pixel
Prior art date
Application number
PCT/CN2023/115383
Other languages
English (en)
French (fr)
Inventor
陶俊
郑世伟
向少卿
Original Assignee
上海禾赛科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 上海禾赛科技有限公司 filed Critical 上海禾赛科技有限公司
Publication of WO2024119906A1 publication Critical patent/WO2024119906A1/zh

Links

Definitions

  • the present disclosure relates to the field of laser radar technology, and in particular to a laser radar, a laser radar data processing method, and an integrated light detection and data acquisition processing device.
  • LiDAR is a commonly used distance measurement sensor with the advantages of long detection distance, high resolution, strong anti-active interference ability, small size and light weight. It is widely used in intelligent robots, drones, unmanned driving and other fields.
  • LiDAR mainly includes mechanical rotating LiDAR, rotating mirror LiDAR and galvanometer mirror (such as micro-electromechanical system, MEMS) LiDAR, etc.
  • MEMS micro-electromechanical system
  • most LiDARs are based on discrete detection units such as avalanche photodiodes (APDs) and silicon photomultipliers (SiPMs) to achieve echo detection.
  • ADCs analog-to-digital converters
  • TDCs time-to-digital converters
  • FIG1a shows the composition of the transmitting unit and receiving unit of the existing laser radar based on discrete photosensitive devices
  • the transmitting device TX includes N transmitting units (refer to L1, L2, L3, ... LN shown in FIG1a)
  • the detecting device RX includes N detecting units (refer to D1, D2, D3, ... DN shown in FIG1a)
  • the detecting unit is, for example, an avalanche photodiode (APD), a silicon photomultiplier tube (SiPM), etc.
  • the N transmitting units and the N detecting units constitute N detection channels (i.e., N lines).
  • the transmitting unit emits detection light, which is reflected by an external object and detected by the corresponding detection unit. After subsequent circuit processing, a data point in the point cloud is generated.
  • the N transmitting units and the N detecting units are driven by a scanning device (such as a mechanical rotating radar), or the emitted light of the N transmitting units is deflected by a scanning device to form a detection of a certain vertical and horizontal field of view.
  • a scanning device such as a mechanical rotating radar
  • discrete devices in order to achieve higher beams and longer distance measurement, discrete devices usually adopt a method of overlapping light spots.
  • the point cloud resolution is 0.05° ⁇ 0.05°, taking the vertical direction as an example, due to technical limitations, the divergence angle of the light spot cannot usually be made smaller, and can usually only be 0.1° or even larger, resulting in overlapping areas of the echo spots of adjacent detection units (as shown in Figure 1b, the echo spots of detection units D1 and D2 overlap).
  • the overlapping area is repeatedly measured many times, resulting in low utilization of photon information, where the dark grid represents the emission/detection unit in operation, the solid circle represents the light spot on the detection unit after the detection light emitted by the emission unit in operation at the current moment is reflected, and the dotted circle represents the light spot on the detection unit after the detection light emitted by the emission unit at the previous moment is reflected.
  • the laser radar comprises:
  • a transmitting device wherein the transmitting device is configured to emit a detection light beam for detecting obstacles
  • a detection device comprising a plurality of detection units, each detection unit comprising a pixel array, wherein each pixel can respond to an echo reflected by the detection light beam on an obstacle and convert it into an electrical signal;
  • a data processing device is coupled to the detection device and configured to:
  • Each signal in the recombinant signal array is obtained according to the electrical signals output by a plurality of adjacent pixels.
  • each pixel comprises a plurality of single-photon avalanche diodes, each of which can be independently gated and addressed, and the data processing device is configured to At least one detection unit traverses the electrical signals output by the pixel array on the detection unit, and recombines the electrical signals output by the pixel array on the detection unit to obtain a recombined signal array.
  • the data processing device is configured to generate a laser radar point cloud based on the electrical signal output by the pixel array on a detection unit.
  • the at least one detection unit includes a first detection unit and a second detection unit, wherein the first detection unit and the second detection unit are two detection units with the same specifications, and the data processing device is configured as follows:
  • n points in the laser radar point cloud are generated, where m is greater than n.
  • the data processing device is configured to:
  • the electrical signal output by the pixel array on the detection unit is convolved by a convolution kernel according to a preset convolution step length to obtain the recombined signal array.
  • the convolution step size of the central area of the field of view of the laser radar is smaller than the convolution step size of the edge area of the field of view.
  • the dimension of the recombinant signal array is smaller than the dimension of the pixel array.
  • the present invention also relates to a data processing method for a laser radar, wherein the detection device of the laser radar comprises a plurality of detection units, each detection unit comprises a pixel array, wherein each pixel can respond to an echo reflected by the detection beam on an obstacle and convert it into an electrical signal, and the data processing method comprises:
  • each pixel includes a plurality of single-photon avalanche diodes arranged in a matrix, and each single-photon avalanche diode can be independently enabled and addressed.
  • Step S101 includes: for at least one detection unit, traversing the electrical signal output by the pixel array on the detection unit, and recombining the electrical signal output by the pixel array on the detection unit to obtain a recombined signal array.
  • the data processing method further comprises:
  • the data processing device is configured to generate a point in the laser radar point cloud based on the electrical signal output by the pixel array on a detection unit.
  • the step S102 comprises: generating a laser radar point cloud according to the first recombined signal array m points; according to the second reorganized signal array, n points in the lidar point cloud are generated, where m is greater than n.
  • the electrical signal output by the pixel array on the detection unit is convolved by a convolution kernel according to a preset convolution step length to obtain the recombined signal array.
  • the convolution step size can be adjusted according to the field of view corresponding to the detection unit.
  • FIG. 1a shows a schematic diagram of the composition of a transmitting unit and a receiving unit of an existing laser radar based on a discrete photosensitive device
  • FIG. 1b is a schematic diagram showing the overlap of echo spots detected by existing discrete photosensitive devices
  • FIG2 shows a schematic diagram of a laser radar according to an embodiment of the present invention
  • 3a and 3b are schematic diagrams of a transmitting device according to a preferred embodiment of the present invention.
  • FIG. 4a and 4b are schematic diagrams of a detection device according to a preferred embodiment of the present invention.
  • 5a and 5b are schematic diagrams respectively showing a method of performing convolution processing on the electrical signals output by the pixel array on the detection unit to obtain a reconstructed signal array according to a preferred embodiment of the present invention
  • FIG6 is a schematic diagram showing a recombined signal array obtained by pixel array recombining processing of a plurality of detection units in different field of view areas according to a preferred embodiment of the present invention
  • FIG7 is a schematic diagram showing a working mode of a detection unit according to a preferred embodiment of the present invention.
  • FIG8 shows a schematic diagram of a detection chip according to a preferred embodiment of the present invention.
  • FIG9 shows a schematic structural diagram of a laser radar according to a preferred embodiment of the present invention.
  • FIG10 is a schematic diagram showing an integrated light detection and data acquisition and processing device according to an embodiment of the present invention.
  • FIG11 shows a schematic diagram of an integrated light detection and data acquisition and processing device according to a preferred embodiment of the present invention.
  • FIG. 12 shows a flow chart of a laser radar data processing method according to a preferred embodiment of the present invention.
  • first and second are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated.
  • the features defined as “first” and “second” may explicitly or implicitly include one or more of the features.
  • the meaning of “multiple” is two or more, unless otherwise clearly and specifically defined.
  • the terms “installed”, “connected”, and “connected” should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, an electrical connection, or mutual communication; it can be directly connected, or indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements.
  • installed should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, an electrical connection, or mutual communication; it can be directly connected, or indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements.
  • it can be based on The specific meanings of the above terms in the present invention should be understood according to the specific circumstances.
  • a first feature being “above” or “below” a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them.
  • a first feature being “above”, “above” and “above” a second feature includes that the first feature is directly above and obliquely above the second feature, or simply means that the first feature is higher in level than the second feature.
  • a first feature being “below”, “below” and “below” a second feature includes that the first feature is directly above and obliquely above the second feature, or simply means that the first feature is lower in level than the second feature.
  • the present invention provides a laser radar.
  • a laser radar point cloud with a higher beam and flexibly adjustable angular resolution can be obtained, and the distance measurement capability is stronger, which is described in detail below.
  • FIG2 shows a schematic diagram of a laser radar 1 according to an embodiment of the present invention.
  • the laser radar 1 includes a transmitting device 10, a detecting device 20, and a data processing device 30.
  • the transmitting device 10 is configured to emit a detection light beam L for detecting obstacles (such as the cube exemplarily shown in FIG2 ).
  • the detecting device 20 includes a plurality of detection units (one detection unit is exemplarily shown in FIG2 ), each detection unit includes a pixel array, such as the 3 ⁇ 3 pixel array exemplarily shown in FIG2 , wherein each pixel can respond to the echo L' reflected by the detection light beam L on the obstacle and convert it into an electrical signal.
  • the data processing device 30 is coupled to the detection device 20 and is configured such that: for at least one of the detection units, the electrical signal output by the pixel array on the detection unit can be reorganized to obtain a reorganized signal array; a laser radar point cloud can be generated according to the reorganized signal array, wherein the Each signal in the recombined signal array can be obtained based on the electrical signals output by a plurality of adjacent pixels.
  • a point cloud is not generated based on the original electrical signal generated by the pixel array of the detection unit, but the original electrical signal is recombined to obtain a recombined signal array, and a laser radar point cloud is generated based on the recombined signal array. Since the signal strength of the recombined signal array is significantly increased, it is beneficial to improve the detection capability of the detection unit.
  • Fig. 3a shows a schematic diagram of a transmitting device 10 according to a preferred embodiment of the present invention.
  • the transmitting device 10 includes a plurality of transmitting units, such as N transmitting units L1, L2, L3, ... LN shown in Fig. 3a, where N is an integer greater than or equal to 1, and the plurality of transmitting units constitute a transmitting line array.
  • the transmitting device 10 is not limited to the case where it only includes a single column of transmitting units. According to another preferred embodiment of the present invention, the transmitting device 10 may also include multiple columns of transmitting units, which are coupled in parallel to form a transmitting array, such as the N ⁇ M transmitting unit array exemplified in Figure 3b, where N and M are both integers greater than 1.
  • the transmitting unit can be a vertical cavity surface emitting laser (VCSEL) or an edge emitting laser (EEL), etc., which can be selected according to actual needs.
  • VCSEL vertical cavity surface emitting laser
  • EEL edge emitting laser
  • the transmitting unit can poll the light in the vertical and/or horizontal direction according to a certain angular resolution (for example, 0.05°, 0.1°, 0.4°, etc.), thereby realizing the detection of the laser radar within a certain field of view.
  • Fig. 4a shows a schematic diagram of a detection device 20 according to a preferred embodiment of the present invention.
  • the detection device 20 includes a plurality of detection units, such as N detection units A1, A2, A3, ... AN shown in Fig. 4a, where N is an integer greater than or equal to 1, and the plurality of detection units constitute a detection line array.
  • the detection device 20 includes a column of detection units.
  • the transmitting device 20 may also include multiple columns of detection units, and the multiple columns of detection units are coupled in parallel to form a detection unit array, such as the N ⁇ M detection unit array shown in Figure 4b, where N and M are both integers greater than 1.
  • each detection unit includes a plurality of pixels, and the plurality of pixels constitute a pixel array. As exemplarily shown in FIG. 4a or FIG. 4b, each detection unit includes 5 ⁇ 5 pixels. Array. In some preferred embodiments, each pixel includes a plurality of single photon avalanche diodes (SPADs), as exemplarily shown in FIG. 4a or FIG.
  • SPADs single photon avalanche diodes
  • each pixel includes 3 ⁇ 3 single photon avalanche diodes (SPADs) totaling 9, wherein each single photon avalanche diode (SPAD) can be independently gated and addressed, that is, each single photon avalanche diode (SPAD) can independently respond to the echo L' reflected by the detection light beam L on the obstacle and convert it into an electrical signal.
  • the number of pixels included in each detection unit is not limited, nor is the number of single photon avalanche diodes included in each pixel.
  • the signal output of a pixel can be obtained according to the electrical signals output by multiple single photon avalanche diodes (SPADs) on a pixel, for example, the signal output of the pixel can be obtained by accumulating the electrical signals output by 9 single photon avalanche diodes (SPADs) on a pixel.
  • the signal output of a detection unit can also be obtained according to the electrical signals output by a pixel array on a detection unit, for example, the signal output of the detection unit can be obtained by accumulating the electrical signals output by a pixel array on a detection unit.
  • one emitting unit in the emitting device 10 corresponds to one detecting unit in the detecting device 20, and each detecting unit can be independently switched on and addressed. For example, one emitting unit emits a detection beam, and one corresponding detecting unit responds, while other detecting units are in a closed state.
  • the data processing device 30 does not reorganize the electrical signals of some pixels in the pixel array, but traverses the electrical signals output by the pixel array on the detection unit, and reorganizes the electrical signals output by the pixel array on the detection unit to obtain a reorganized signal array.
  • the intensity of the output signal obtained is significantly increased, which is conducive to improving the detection capability of the detection unit.
  • the data processing device 30 performs convolution processing on the electrical signal output by the pixel array to obtain a reconstructed signal array.
  • the electrical signal output by the pixel array on the detection unit is convolved according to a preset convolution step size through a convolution kernel to obtain a reconstructed signal array.
  • the convolution kernel is an accumulation window, and the window height H and width W of the convolution kernel can be configured, wherein at least one of H and W is a number greater than 1 pixel, and the two can be equal or unequal.
  • the window height H and width W of the convolution kernel are equal, for example,
  • the size of the convolution kernel is 3 pixels, where each pixel has a size of 30um ⁇ 30um, that is, the size of the convolution kernel can be a 3 ⁇ 3 matrix (for example, 90um ⁇ 90um). It should be understood that this embodiment is only an exemplary description and does not constitute a limitation of the present invention.
  • the size of the convolution kernel can also be other sizes, which can be determined according to actual conditions.
  • the convolution step is the translation step of the convolution kernel.
  • the convolution step is divided into the horizontal step Hs (Hstride) and the vertical step Vs (Vstride), which can be flexibly configured, wherein the horizontal step Hs and the vertical step Vs are both greater than or equal to 1, and the two can be equal or unequal.
  • the values of the horizontal step Hs and the vertical step Vs are equal, wherein the minimum horizontal step Hs/vertical step Vs is the size of one pixel.
  • the convolution processing of the electrical signal output by the pixel array on the detection unit is actually the data accumulation of the electrical signal (such as the echo waveform) output by the pixel array on the detection unit, thereby obtaining the recombined signal array (also called the recombined pixel array) on the detection unit, and subsequently calculating the distance and/or reflectivity of the obstacle according to each recombined signal.
  • the recombined signal array also called the recombined pixel array
  • Figure 5a shows a schematic diagram of performing convolution processing on the electrical signal output by the pixel array on the detection unit to obtain a recombined signal array according to a preferred embodiment of the present invention, wherein the left part shows the situation of the pixel array on the detection unit before the convolution processing, the middle part shows the process of performing convolution processing on the pixel array on the detection unit, and the right part shows the situation of the pixel array on the detection unit after the convolution processing.
  • the size of the detection unit is, for example, 300um ⁇ 300um, which includes a 10 ⁇ 10 pixel array, wherein the size of each pixel may be 30um ⁇ 30um.
  • Each pixel may include 9 single photon avalanche diodes (SPADs), and the 9 single photon avalanche diodes (SPADs) constitute a 3 ⁇ 3 single photon avalanche diode (SPAD) array, wherein the size of each single photon avalanche diode (SPAD) may be 10um ⁇ 10um.
  • SPADs 9 single photon avalanche diodes
  • SPADs 9 single photon avalanche diodes
  • the 9 single photon avalanche diodes constitute a 3 ⁇ 3 single photon avalanche diode (SPAD) array, wherein the size of each single photon avalanche diode (SPAD) may be 10um ⁇ 10um.
  • the horizontal and vertical dimensions of the convolution kernel may both be the size of 3 pixels, i.e., 90um ⁇ 90um, i.e., the size of a 9 ⁇ 9 single photon avalanche diode (SPAD) array.
  • the horizontal and vertical dimensions of the convolution kernel may also be other dimensions, depending on the actual situation.
  • the size of the convolution step in the horizontal and vertical directions may both be the side length of 1 pixel, i.e., 30um, i.e., the size of 3 single photon avalanche diodes (SPADs).
  • the convolution kernel e.g., 9
  • the electrical signals (e.g., echo waveforms) output by the pixels (e.g., 9 pixels) corresponding to the convolution kernel are directly accumulated or weightedly accumulated according to a preset convolution step size (e.g., the size of 1 pixel), and the recombined signals (also called recombined pixels, refer to the portion framed by the dotted line frame on the right side of FIG. 5a) of the pixels corresponding to the convolution kernel can be obtained.
  • a preset convolution step size e.g., the size of 1 pixel
  • the recombined signals also called recombined pixels, refer to the portion framed by the dotted line frame on the right side of FIG. 5a
  • the pixel array (e.g., 10 ⁇ 10) on the entire detection unit can be traversed to obtain the recombined signal array of the pixel array (e.g., 10 ⁇ 10) on the detection unit, such as the 8 ⁇ 8 pixel array exemplarily shown in the right side of FIG. 5a (the middle part of FIG. 5a is shown in FIG.
  • the convolution is performed with the pixel of the 2nd row and the 2nd column as the center to generate the recombined pixel of the 1st row and the 1st column of the right part of Figure 5a.
  • the horizontal direction is convolved with the pixel of the 2nd row and the 9th column as the center with a step length of one pixel, and the recombined pixel of the 1st row and the 8th column of the right part of Figure 5a is generated;
  • the vertical direction is convolved with the pixel of the 9th row and the 2nd column as the center with a step length of one pixel, and the recombined pixel of the 8th row and the 1st column of the right part of Figure 5a is generated; therefore, after completing the traversal of the entire 10 ⁇ 10 pixel array, an 8 ⁇ 8 recombined pixel array is generated), and the distance and/or reflectivity of the obstacle can be calculated based on each recombined signal in the recombined signal array.
  • the size of the recombined pixel array is 240um ⁇ 240um, which is smaller than the size of the pixel array before recombining, 300um ⁇
  • FIG. 5a also shows two echoes E1 and E2, where echo E1 is the echo of a pixel of the pixel array on the detection unit before the convolution process, and echo E2 is the echo of a recombined pixel of the pixel array on the detection unit after the convolution process, i.e., a recombined signal.
  • the signal intensity of the echo E2 of the recombined pixel is significantly stronger than the signal intensity of the echo E1 of the pixel before the recombination (the figure shows the accumulation result of a time window).
  • the peak part is accumulated in a time window. Therefore, after the convolution process, the signal intensity output by the recombined signal array on the detection unit is significantly increased, which is conducive to improving the detection capability of the detection unit.
  • each detection moment in the pixel array (10 ⁇ 10 matrix) on the detection unit, the output of each pixel corresponds to a value.
  • the value may be, for example, 0, 1 or other values, representing the intensity of the echo corresponding to the pixel position, such as the intensity of echo E1 in the figure.
  • a reconstructed signal array (reconstructed pixel array) is formed, in which each element also corresponds to a value, representing the intensity of the reconstructed signal, such as the intensity of echo E2.
  • FIG5 shows that the size of the detection unit is 300um ⁇ 300um, including a 10 ⁇ 10 pixel array, and the size of the convolution kernel in the horizontal and vertical directions is 3 pixels, that is, 90um ⁇ 90um.
  • the size of the convolution kernel can also be larger, for example, the size of the horizontal and vertical directions is 4 pixels, that is, 120um ⁇ 120um.
  • the echo intensity may be greater, but the FOV corresponding to each reconstructed pixel will also be larger, which is not conducive to the detection of small-sized objects.
  • the horizontal and vertical fields of view corresponding to each pixel are both 0.05°
  • the horizontal and vertical fields of view corresponding to the generated reconstructed pixels are both 0.15°
  • the horizontal and vertical fields of view corresponding to the generated reconstructed pixels are both 0.2°. Therefore, the size of the detection unit, pixel and convolution kernel will be selected according to the actual application.
  • the above embodiment takes the convolution step size of 1 pixel as an example, and introduces in detail the whole process of performing convolution processing on the electrical signal output by the pixel array on the detection unit to obtain the reconstructed signal array.
  • the following takes the convolution step size of 2 pixels as an example to introduce the process of performing convolution processing on the electrical signal output by the pixel array on the detection unit to obtain the reconstructed signal array.
  • Figure 5b shows a schematic diagram of performing convolution processing on the electrical signal output by the pixel array on the detection unit to obtain a recombined signal array according to another preferred embodiment of the present invention, wherein the left part shows the situation of the pixel array on the detection unit before the convolution processing, the middle part shows the process of performing convolution processing on the pixel array on the detection unit, and the right part shows the situation of the pixel array on the detection unit after the convolution processing.
  • the electrical signal (e.g., the size of 9 pixels) output by the pixels (e.g., the size of 9 pixels) corresponding to the convolution kernel is convolved according to a preset convolution step size (e.g., the size of 2 pixels).
  • the reconstructed pixel corresponding to the convolution kernel can be obtained by directly accumulating or weighted accumulating the pixel array (e.g., 10 ⁇ 10) on the entire detection unit (e.g., 4 ⁇ 4 pixel array shown in the right part of FIG5b), and the reconstructed signal array of the pixel array (e.g., 10 ⁇ 10) on the detection unit can be obtained, such as the 4 ⁇ 4 pixel array shown in the right part of FIG5b.
  • the distance and/or reflectivity of the obstacle can be calculated based on each reconstructed signal in the reconstructed signal array.
  • the size of the reconstructed pixel array is 240um ⁇ 240um, which is smaller than the size of the pixel array before reconstructing, 300um ⁇ 300um.
  • the vertical and horizontal field of view angles of the detection unit are 0.1° ⁇ 0.1° respectively.
  • the above embodiments take the convolution step size (horizontal/vertical) of 1 pixel and 2 pixels as examples, respectively, and introduce the process of convolution processing the electrical signal output by the pixel array on the detection unit to obtain the recombined signal array.
  • the convolution step size is 1 pixel
  • an 8 ⁇ 8 recombined signal array can be obtained during the convolution processing of the 10 ⁇ 10 pixel array on the detection unit
  • the distance between adjacent recombined signals is 1 pixel
  • the resolution of the laser radar point cloud is 0.05° ⁇ 0.05°.
  • the convolution step size is 2 pixels
  • a 4 ⁇ 4 recombined signal array can be obtained
  • the distance between adjacent recombined signals is 2 pixels
  • the resolution of the laser radar point cloud is 0.1° ⁇ 0.1°.
  • the angular resolution of the LiDAR point cloud after convolution processing is also negatively correlated with the convolution step size, that is, the smaller the convolution step size, the higher the angular resolution of the LiDAR point cloud after convolution processing, the better the quality of the LiDAR point cloud, and the easier it is to present the details of the detected object. Therefore, the size of the convolution step size can be flexibly adjusted according to actual conditions, thereby achieving flexible configuration of the resolution of the LiDAR point cloud.
  • the convolution step size can be adjusted according to the field of view corresponding to the detection unit. In other words, different convolution step sizes can be used for different fields of view corresponding to the detection unit, as described in detail below.
  • the above-mentioned reorganization processing can be performed on the output signals of all detection units, which may place higher requirements on the data processing volume and processing speed of the laser radar. In some preferred embodiments, the above-mentioned reorganization processing can be performed on the output signals of the detection units corresponding to the specific field of view range.
  • the specific field of view range can be, for example, the central field of view area of the laser radar, which is usually the area that needs the most attention.
  • the data processing device 30 is configured to perform reorganization processing on the electrical signals output by the pixel array of at least one detection unit corresponding to the specific field of view range, obtain a reorganized signal array, and generate a laser radar point cloud based on the reorganized signal array. Since the reorganized signal array is equivalent to neighborhood accumulation, the signal strength of the laser radar point cloud generated based on the reorganized signal array also increases, which is conducive to improving the distance measurement performance of the laser radar.
  • the specific field of view area of the laser radar can be determined by an image recognition method.
  • the laser radar can work with a camera. According to the image captured by the camera, some objects that need to be focused on, such as pedestrians, vehicles, etc., can be identified. For the identified objects, the corresponding laser radar field of view area and the corresponding detection unit can be determined, and the output signals of these detection units are subjected to the above-mentioned reorganization processing for data processing.
  • the data processing device 30 is configured to generate a laser radar point cloud based on the electrical signal output by the pixel array on a detection unit corresponding to the field of view outside the specific field of view, without reorganizing the electrical signal output by the one detection unit, thereby effectively reducing the power consumption and processing burden of the laser radar.
  • the specific field of view range is not necessarily the central field of view area of the laser radar, but can also be other field of view areas of the laser radar, which can be set according to actual needs and is not limited by the present invention.
  • the first detection unit and the second detection unit in the detection device are preferably two detection units with the same specifications, that is, the size, the number of pixels included, the size of the pixels, the number and size of single-photon avalanche diodes (SPADs) of the two are the same, and the two correspond to different field of view areas, respectively, and different convolution step sizes can be used as needed.
  • the first detection unit and the second detection unit in the detection device are preferably two detection units with the same specifications, that is, the size, the number of pixels included, the size of the pixels, the number and size of single-photon avalanche diodes (SPADs) of the two are the same, and the two correspond to different field of view areas, respectively, and different convolution step sizes can be used as needed.
  • SSPADs single-photon avalanche diodes
  • the size of the first detection unit and the second detection unit are both 300um ⁇ 300um, and both include a 10 ⁇ 10 pixel array, wherein the size of each pixel is 30um ⁇ 30um, and each pixel includes 9 single-photon avalanche diodes (SPADs), and the 9 single-photon avalanche diodes (SPADs) constitute a 3 ⁇ 3 single-photon avalanche diode (SPAD) array, wherein the size of each single-photon avalanche diode (SPAD) is 10um ⁇ 10um.
  • the specifications of the first detection unit and the second detection unit are only exemplary and do not constitute a limitation to the present invention.
  • the data processing device 30 can reorganize the electrical signals output by the pixel array on the first detection unit to obtain a first reorganized signal array; generate m points in the laser radar point cloud according to the first reorganized signal array; can reorganize the electrical signals output by the pixel array on the second detection unit to obtain a second reorganized signal array; generate n points in the laser radar point cloud according to the second reorganized signal array, where m is greater than n.
  • FIG6 shows a reorganized signal array obtained by reorganizing the pixel arrays of two first detection units and two second detection units in this way, wherein the first detection unit may correspond to the central area of the field of view of the laser radar, and the second detection unit may correspond to the outside of the central area of the field of view of the laser radar.
  • the first detection unit is detected corresponding to the central area of the field of view, which can generate a denser laser radar point cloud (as shown in Figure 6, the signal array reorganized in the middle area is an 8 ⁇ 8 array), so as to present the morphological details of the detection target more finely, which is conducive to obtaining more effective and reliable detection results, especially for long-distance target detection; and for the second detection unit corresponding to the field of view area close to the edge, a larger convolution step size is used to generate a sparser laser radar point cloud (as shown in Figure 6, the signal array reorganized in the edge area is a 4 ⁇ 4 array).
  • non-key detection targets can be set outside the central area of the laser radar field of view, and the second detection unit is detected corresponding to the outside of the central area of the field of view, thereby achieving a certain degree of balance between effective and reliable detection results and processing efficiency.
  • the laser radar point cloud generated according to the first reorganized signal array is denser
  • the laser radar point cloud generated according to the second reorganized signal array is sparser, where the dark grid represents the point in the laser radar point cloud generated at the corresponding spatial position.
  • FIG6 shows a recombined signal array obtained by recombining the pixel arrays of two first detection units and two second detection units, which is only used for illustrative purposes and does not constitute a limitation to the present invention.
  • the pixel arrays of other numbers of first detection units and second detection units may also be recombined, and these are all within the scope of protection of the present invention.
  • the convolution step size of the central area of the laser radar's field of view can be adjusted to be smaller than the convolution step size of the edge area of the field of view.
  • a smaller convolution step size such as 1 pixel
  • the distribution of the laser radar point cloud can be expanded from uniform resolution to the ability to encrypt the central area of the field of view and sparse the edge area of the field of view.
  • the resolution of the central area of the field of view is 0.05° ⁇ 0.05°
  • the resolution of the edge area of the field of view is 0.1° ⁇ 0.1°.
  • the data processing device 30 is further configured to: obtain the ROI area around the laser radar; set the detection unit corresponding to the ROI area to at least one detection unit, that is, the detection unit that needs to perform signal reorganization.
  • the ROI area refers to the region of interest (ROI), which can be set according to user needs and/or interests.
  • ROI region of interest
  • the ROI area can be determined by image recognition.
  • at least one detection unit corresponding to the ROI area can be set according to the size and/or shape of the ROI area.
  • multiple detection units can be set to obtain a larger detection area, which is conducive to outputting a signal with stronger signal strength, thereby obtaining a higher quality laser radar point cloud, which is conducive to improving the effectiveness and reliability of the detection results of the ROI area.
  • the ROI area can also be set in the central area of the field of view of the laser radar, which will further improve the effectiveness and reliability of the detection results.
  • the convolution step size may be set according to the ROI area around the laser radar. For example, a smaller convolution step size (e.g., 1 pixel) may be used for the ROI area around the laser radar, while a larger convolution step size (e.g., 2 pixels) may be used for the non-ROI area. It should be understood that this embodiment is only an exemplary description and does not constitute a limitation of the present invention.
  • the convolution step size can also be adjusted according to the density of the lidar point cloud. For example, for the encrypted area in the lidar point cloud, a smaller convolution step size (for example, the size of 1 pixel) can be used, while for the sparse area (non-encrypted area), a larger convolution step size (for example, the size of 2 pixels) can be used.
  • the above embodiment introduces the situation of adjusting the convolution step size according to the field of view corresponding to the detection unit.
  • the resolution of the lidar point cloud can be flexibly configured, thereby not only obtaining the lidar point cloud of the required resolution but also reducing resource consumption.
  • FIG7 is a schematic diagram showing the working mode of the detection unit according to a preferred embodiment of the present invention.
  • the working mode of the detection unit can be: for scanning in the vertical direction, Scanning can be completed by scanning each pixel one by one.
  • each transmitting unit is polled and emitted light at a certain field of view angle (for example, 0.4°).
  • the spot of detection light emitted by a transmitting unit is basically the same size as that of a detection unit.
  • the corresponding detection unit responds, and the detected electrical signal is converted by an analog-to-digital conversion chip such as an analog-to-digital converter (ADC) and a time-to-digital converter (TDC) and then is used by a digital processing chip for echo recognition and time measurement.
  • ADC analog-to-digital converter
  • TDC time-to-digital converter
  • This can realize the detection of the vertical field of view range, which belongs to electronic scanning.
  • the transmitting unit can be driven to scan from one side of the laser radar's field of view to the other side by mechanical rotation such as the deflection of the scanning device or the rotation of the rotor, thereby realizing the detection of the horizontal field of view range, which belongs to mechanical scanning.
  • the horizontal angle corresponding to a column of detection units has a certain offset.
  • the detection device 20 can be implemented based on a detection chip that uses time-of-flight (TOF) measurement.
  • FIG8 shows a schematic diagram of a detection chip according to a preferred embodiment of the present invention. As shown in the left part of FIG8 , a plurality of independent detection units are integrated on the detection chip ( FIG8 exemplarily shows one of the detection units, refer to the part shown by the small white square), wherein each detection unit includes a pixel array.
  • the right part of FIG8 is an enlarged view of one of the detection units, and the size of the detection unit can be 300um ⁇ 300um, which can include a 10 ⁇ 10 pixel array, wherein each pixel can include a 3 ⁇ 3 single photon avalanche diode (SPAD) array.
  • SBAD single photon avalanche diode
  • the transmitting device includes a plurality of transmitting units, and the number of rows of the reorganized signal array is greater than the number of transmitting units.
  • the pixel array reorganized according to the present invention has multiple rows, such as the above-mentioned 8 rows or 4 rows, that is, one transmitting unit finally forms 8 lines or 4 lines.
  • the number of detection units and the number of transmitting units, as well as the number of lines of the laser radar are no longer one-to-one corresponding to each other as described in the background technology, but the number of lines is much greater than the number of transmitting units and detection units, so the solution of the present invention greatly expands the number of lines of the laser radar.
  • the dimension of the recombinant signal array is smaller than the dimension of the pixel array.
  • the dimension of the recombinant signal array is 8, and the dimension of the pixel array is 10.
  • the dimension of the recombinant signal array is 4, and the dimension of the pixel array is 10. It can be seen that the dimension of the recombinant signal array is smaller than the dimension of the pixel array.
  • the laser radar point cloud generated by the technical solution of the present invention may not have the same number of transmitting units/detecting units, but may be much larger than the number of transmitting units/detecting units. Therefore, a higher beam LiDAR point cloud can be achieved.
  • a 256-line laser radar point cloud can be realized based on 32 transmitting units and 32 detecting units. Specifically, 32 transmitting units and 32 detecting units are distributed in the vertical direction.
  • the convolution step size is 1 pixel
  • the laser radar may be a scanning laser radar, as shown in FIG9 , in addition to a transmitting device 10, a detecting device 20 and a data processing device 30, the laser radar also includes a scanning device 40, a first reflecting unit 51 and a second reflecting unit 52, wherein the scanning device 40 may be at least one of a rotating mirror, a galvanic mirror and a swinging mirror, and the first reflecting unit 51 and the second reflecting unit 52 may be reflecting mirrors.
  • the scanning device is a multi-faceted rotating mirror.
  • the transmitting device 10 and the detecting device 20 remain fixed.
  • the transmitting device 10 emits a detection beam L, which is reflected by the first reflecting unit 51, and then emitted to the external space through one surface of the scanning device 40. After being reflected by obstacles in the external space, an echo L' is formed.
  • the echo L' is incident on another surface of the scanning device 40, and then reflected by the second reflecting unit 52 and received by the detecting unit on the detecting device.
  • the scanning device 40 can adopt a one-dimensional scanning device or a two-dimensional scanning device.
  • the light emitting direction of each transmitting unit corresponds to a field of view or an angle in the horizontal/vertical direction
  • different transmitting units correspond to different fields of view or angles, thereby realizing detection of a certain field of view range in the horizontal/vertical direction.
  • the one-dimensional scanning device rotates around the rotation axis.
  • the one-dimensional scanning device can deflect the detection light beam emitted by the transmitting unit in the horizontal direction to different angles by rotating, thereby realizing scanning detection of a certain field of view range in the horizontal direction, and then combined with the detection of the transmitting unit in the vertical direction, realize three-dimensional detection of the surrounding environment; conversely, when the rotation axis is in the horizontal direction, the one-dimensional scanning device can deflect the detection light beam emitted by the transmitting unit in the vertical direction to different angles by rotating, thereby realizing scanning detection of a certain field of view range in the vertical direction, and then combined with the detection of the transmitting unit in the vertical direction, realize three-dimensional detection of the surrounding environment; conversely, when the rotation axis is in the horizontal direction, the one-dimensional scanning device can deflect the detection light beam emitted by the transmitting unit in the vertical direction to different angles by rotating, thereby realizing scanning detection of a certain field of view range in the vertical direction, and then combined with the transmitting unit in the vertical direction, realize three-dimensional detection of the
  • the scanning device 40 For a laser radar using a two-dimensional scanning device, the scanning device 40 has two rotating shafts at a certain angle (e.g., 90°), so the scanning device 40 can reflect the detection beam emitted by the transmitting unit to different angles in two directions (e.g., horizontal and vertical directions), thereby realizing two-dimensional scanning.
  • the two-dimensional scanning device may include two one-dimensional scanning devices, and the rotating shafts of the two one-dimensional scanning devices are at a certain angle (e.g., 90°), thereby also realizing two-dimensional scanning.
  • the laser radar can also be a mechanical laser radar.
  • the mechanical laser radar in addition to the transmitting device 10, the detecting device 20 and the data processing device 30, it also includes an optomechanical rotor (not shown in the figure).
  • the transmitting device 10 and the detecting device 20 are arranged on the optomechanical rotor, and at least one transmitting unit of the transmitting device 10 corresponds to at least one detecting unit of the detecting device 20.
  • each transmitting unit corresponds to a field of view or an angle in the vertical direction, and different transmitting units have different corresponding fields of view or angles in the vertical direction, thereby realizing the scanning detection of a certain field of view range of the laser radar in the vertical direction.
  • the optomechanical rotor rotates around the vertical axis, thereby realizing the scanning detection of the laser radar in the horizontal direction of 360°.
  • a detection beam L can be emitted once at a certain rotation angle (i.e., the horizontal angular resolution of the laser radar, such as 0.05°, 0.1°, etc.).
  • the detection device 20 receives the echo L' after the detection beam L is diffusely reflected on the obstacle.
  • the data processing device 30 recombines the electrical signal output by the pixel array on the detection unit to obtain a recombined signal array, and a laser radar point cloud of the required resolution can be generated based on the recombined signal array.
  • the laser radar can also be a fully solid-state laser radar, preferably a planar array laser radar.
  • one transmitting unit illuminates one detection unit, the detection unit is powered on, and the other detection units are in an off state.
  • the transmitting units and the detection units are distributed in a planar array, as shown in Figures 3b and 4b.
  • the detection units are powered on for detection according to the light emission sequence of the corresponding transmitting units, and perform corresponding convolution processing. Appropriate convolution kernels and convolution steps can be selected as needed.
  • the data processing device 30 recombines the electrical signals output by the pixel array on the detection unit to obtain a recombined signal array, and a laser radar point cloud of the required resolution can be generated according to the recombined signal array.
  • a laser radar point cloud of the required resolution can be generated according to the recombined signal array.
  • the above is a detailed introduction to the laser radar of the present invention.
  • a limited field of view area is exposed each time (i.e., TOF measurement), and the full scene scan around the laser radar can be completed by moving the field of view.
  • TOF measurement i.e., TOF measurement
  • the full scene scan around the laser radar can be completed by moving the field of view.
  • a laser radar point cloud with a higher beam and flexibly adjustable angular resolution can be generated, and the distance measurement capability is stronger.
  • the present invention also relates to an integrated light detection and data acquisition processing device 200.
  • the light detection and data acquisition processing device 200 includes a plurality of detection units 210 and a data acquisition processing device 220.
  • each detection unit includes a pixel array, wherein each pixel can respond to a light signal and convert it into an electrical signal.
  • the data acquisition processing device 220 is coupled to the plurality of detection units 210 and is configured as follows:
  • an electrical signal output by a pixel array on the detection unit is acquired, and the electrical signal output by the pixel array on the detection unit is recombined to obtain a recombined signal array;
  • a laser radar point cloud is generated according to the recombined signal array, wherein each signal in the recombined signal array is obtained according to electrical signals output by a plurality of adjacent pixels.
  • Figure 11 shows a schematic diagram of an integrated optical detection and data acquisition and processing device 300 according to a preferred embodiment of the present invention, wherein the data acquisition and processing device 220 includes a digital signal acquisition unit 220-1 and a digital signal processing unit 220-2, and the digital signal acquisition unit 220-1 and the digital signal processing unit 220-2 are coupled to each other, wherein the digital signal acquisition unit 220-1 is coupled to multiple detection units 210 and is configured to acquire electrical signals output by pixel arrays on multiple detection units 210; the digital signal processing unit 220-2 is configured to reorganize the electrical signals acquired by the digital signal acquisition unit 220-1 to obtain a reorganized signal array, and generate a laser radar point cloud of a required resolution based on the reorganized signal array.
  • the data acquisition and processing device 220 includes a digital signal acquisition unit 220-1 and a digital signal processing unit 220-2, and the digital signal acquisition unit 220-1 and the digital signal processing unit 220-2 are coupled to each other, wherein the digital signal acquisition unit 220-1 is coupled to multiple detection units
  • optical detection and data acquisition and processing device 200/300 The above is a detailed introduction to the optical detection and data acquisition and processing device 200/300.
  • the present invention also relates to a laser radar data processing method 100, wherein the laser radar detection device includes a plurality of detection units, each detection unit includes a pixel array, wherein each pixel can respond to the echo of the detection beam reflected on the obstacle and convert it into an electrical signal.
  • the data processing method 100 includes steps S101 and S102.
  • step S101 for at least one detection unit, the electrical signal outputted by the pixel array on the detection unit is recombined to obtain a recombined signal array;
  • step S102 generating a laser radar point cloud according to the recombined signal array
  • Each signal in the recombinant signal array is obtained according to the electrical signals output by a plurality of adjacent pixels.
  • each pixel includes a plurality of single-photon avalanche diodes arranged in a matrix, and each single-photon avalanche diode can be independently gated and addressed.
  • step S101 includes: for at least one detection unit, traversing the electrical signal output by the pixel array on the detection unit, recombining the electrical signal output by the pixel array on the detection unit, and obtaining a recombined signal array.
  • the data processing method 100 further comprises: in the field of view outside the specific field of view range of the laser radar, the data processing device is configured to generate a point in the laser radar point cloud based on an electrical signal output by a pixel array on a detection unit.
  • the at least one detection unit comprises a first detection unit and a second detection unit.
  • the step S101 comprises: recombining the electrical signals outputted by the pixel array on the first detection unit to obtain a first recombined signal array; recombining the electrical signals outputted by the pixel array on the second detection unit to obtain a second recombined signal array;
  • the step S102 includes: generating m points in the laser radar point cloud according to the first recombined signal array; generating n points in the laser radar point cloud according to the second recombined signal array, where m is greater than n.
  • the first detection unit corresponds to the central area of the field of view of the laser radar
  • the second detection unit corresponds to the outer side of the central area of the field of view of the laser radar.
  • a detection unit corresponding to the ROI area is set as the at least one detection unit.
  • step S101 comprises:
  • the electrical signal output by the pixel array on the detection unit is convolved by a convolution kernel according to a preset convolution step length to obtain the recombined signal array.
  • the convolution step size can be adjusted according to the field of view corresponding to the detection unit.
  • the convolution step size of the central area of the field of view of the laser radar is smaller than the convolution step size of the edge area of the field of view.
  • the laser radar further comprises a transmitting device, the transmitting device comprises a plurality of transmitting units, and the number of rows of the recombined signal array is greater than the number of the transmitting units.
  • the dimension of the recombinant signal array is smaller than the dimension of the pixel array.
  • the above introduces the data processing method 100 of the laser radar.
  • a laser radar point cloud with a higher beam and adjustable resolution can be generated, which is beneficial to improving the accuracy of the detection results.
  • the technical solution of the present invention is introduced above. Compared with the prior art, the technical solution of the present invention can be used to obtain a recombined signal array with a stronger signal by recombining the electrical signal output by the pixel array on the detection unit.
  • a laser radar point cloud with a higher beam can be generated according to the recombined signal array, which greatly improves the laser radar's ranging performance and is conducive to obtaining more effective and reliable detection results.
  • the angular resolution of the laser radar point cloud can be flexibly adjusted by configuring the convolution step size, and global point cloud encryption, partial area point cloud encryption and other configurations can be flexibly realized.
  • the technical solution of the present invention has a small spot overlap area, which can improve photon utilization and reduce overall machine power consumption.
  • the present invention further provides a computer-readable storage medium, comprising computer-executable instructions stored thereon, and the executable instructions implement the data processing method 100 as described above when executed by a processor.
  • the computer-readable storage medium may be any combination of one or more computer-readable media.
  • the computer-readable storage medium may be, for example, but not limited to, electrical, magnetic, optical, or semiconductor forms or devices, and more specific examples (a non-exhaustive list) include: An electrical connection having one or more conductors, a portable computer hard disk, a hard disk, a random access memory (RAM), a non-volatile random access memory (NVRAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
  • a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by an instruction execution system, apparatus or device or used in combination therewith.
  • the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.
  • CPU central processing unit
  • DSP digital signal processors
  • ASIC application-specific integrated circuits
  • FPGA field-programmable gate arrays
  • the present invention is not limited and depends on the specific circumstances.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Electromagnetism (AREA)
  • Optical Radar Systems And Details Thereof (AREA)

Abstract

一种激光雷达，包括：发射装置（10），发射装置配置成发出探测光束，用于探测障碍物；探测装置（20），包括多个探测单元，每个探测单元包括像素阵列，其中每个像素可对探测光束在障碍物上反射的回波作出响应并转换为电信号；数据处理装置（30），与探测装置（20）耦接，并且配置成：对于其中至少一个探测单元，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；根据重组信号阵列，生成激光雷达点云，其中重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。该激光雷达能够生成线束更高、角度分辨率可灵活调节的激光雷达点云，测远能力更强、功耗更低。

Description

激光雷达、数据处理方法及光探测和数据采集处理装置 技术领域
本公开涉及激光雷达技术领域，尤其涉及一种激光雷达、一种激光雷达的数据处理方法以及一种集成的光探测和数据采集处理装置。
背景技术
激光雷达是一种常用的测距传感器，具有探测距离远、分辨率高、抗有源干扰能力强、体积小、质量轻等优点，被广泛应用于智能机器人、无人机、无人驾驶等领域。目前的激光雷达主要包括机械旋转式激光雷达、转镜式激光雷达以及振镜式(如微机电系统，MEMS)激光雷达等，但大部分激光雷达都是基于分立的探测单元例如雪崩光电二极管(APD)、硅光电倍增管(SiPM)实现回波探测，所探测到的电信号经过模数转换芯片例如模拟数字转化器(ADC)、时间数字转化器(TDC)转换后被数字处理芯片进行回波识别和时间测量。
图1a示出了现有的基于分立的感光器件的激光雷达的发射单元和接收单元的组成,其中发射装置TX包括N个发射单元(参考图1a示例性示出的L1、L2、L3、……LN)，探测装置RX包括N个探测单元(参考图1a示例性示出的D1、D2、D3、……DN)，探测单元例如为雪崩光电二极管(APD)、硅光电倍增管(SiPM)等，N个发射单元和N个探测单元构成N个探测通道(亦即N线)。现有的激光雷达多数采用点扫描的方式进行探测，即发射单元发射探测光，探测光经外界物体反射后，被对应的探测单元探测，经后续电路处理之后，生成点云中的一个数据点。N个发射单元和N个探测单元经由扫描器件带动(例如机械式旋转雷达)，或者N个发射单元的出射光经由扫描器件偏折，形成一定竖直和水平视场范围的探测。但现有的激光雷达如果要进一步增加发射单元和探测单元的数量则非常困难，难以实现更高的线束， 例如256及以上的线束，而且点云的分辨率不便调节。
另外，分立器件为了实现更高的线束和更远的测距，通常采用光斑之间存在交叠的方式，如图1b所示，当点云分辨率为0.05°×0.05°时，以垂直方向为例，由于技术的限制，光斑的发散角通常无法做到更小，通常只能做到0.1°甚至更大，导致相邻探测单元的回波光斑存在交叠区域(如图1b中示出的探测单元D1与D2的回波光斑存在交叠)，交叠区域被重复测量多次，造成光子信息利用率低，其中深色格子代表工作中的发射/探测单元，实线圆圈代表当前时刻正在工作中的发射单元发射的探测光反射后在探测单元上的光斑，虚线圆圈代表上一时刻发射单元发射的探测光反射后在探测单元上的光斑。
背景技术部分的内容仅仅是公开人所知晓的技术，并不当然代表本领域的现有技术。
发明内容
针对现有技术存在问题中的一个或多个，本发明提供一种激光雷达，能够生成线束更高、角度分辨率可灵活调节的激光雷达点云，测远能力更强。
所述激光雷达包括：
发射装置，所述发射装置配置成发出探测光束，用于探测障碍物；
探测装置，包括多个探测单元，每个探测单元包括像素阵列，其中每个像素可对所述探测光束在障碍物上反射的回波作出响应并转换为电信号；和
数据处理装置，与所述探测装置耦接，并且配置成：
对于其中至少一个探测单元，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；和
根据所述重组信号阵列，生成激光雷达点云，
其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。
根据本发明的一个方面，其中每个像素包括多个单光子雪崩二极管，每个单光子雪崩二极管可独立选通和寻址，所述数据处理装置配置成对于其中 至少一个探测单元，遍历该探测单元上的像素阵列输出的电信号，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列。
根据本发明的一个方面，其中所述至少一个探测单元对应于所述激光雷达的特定视场范围，在所述激光雷达的所述特定视场范围以外的视场，所述数据处理装置配置成根据一个探测单元上的像素阵列输出的电信号生成激光雷达点云。
根据本发明的一个方面，其中所述至少一个探测单元包括第一探测单元和第二探测单元，其中所述第一探测单元和所述第二探测单元为规格相同的两个探测单元，其中所述数据处理装置配置成：
根据该第一探测单元上的像素阵列输出的电信号进行重组，获得第一重组信号阵列；根据该第二探测单元上的像素阵列输出的电信号进行重组，获得第二重组信号阵列；和
根据所述第一重组信号阵列，生成激光雷达点云中的m个点；根据所述第二重组信号阵列，生成激光雷达点云中的n个点，其中m大于n。
根据本发明的一个方面，其中所述第一探测单元对应于所述激光雷达的视场中央区域，所述第二探测单元对应于所述激光雷达的视场中央区域的外侧。
根据本发明的一个方面，其中所述数据处理装置配置成：
获取所述激光雷达周围的ROI区域；
将与所述ROI区域对应的探测单元设置为所述至少一个探测单元。
根据本发明的一个方面，其中所述数据处理装置配置成：
通过卷积核，按照预设的卷积步长，对该探测单元上的像素阵列输出的电信号进行卷积处理，获得所述重组信号阵列。
根据本发明的一个方面，其中所述卷积步长可根据所述探测单元对应的视场而调节。
根据本发明的一个方面，其中所述激光雷达的视场中央区域的卷积步长小于视场边缘区域的卷积步长。
根据本发明的一个方面，其中所述发射装置包括多个发射单元，所述重 组信号阵列的行数大于所述发射单元的个数。
根据本发明的一个方面，其中所述重组信号阵列的维数小于所述像素阵列的维数。
本发明还涉及一种激光雷达的数据处理方法，其中所述激光雷达的探测装置包括多个探测单元，每个探测单元包括像素阵列，其中每个像素可对所述探测光束在障碍物上反射的回波作出响应并转换为电信号，所述数据处理方法包括：
S101：对于其中至少一个探测单元，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；和
S102：根据所述重组信号阵列，生成激光雷达点云，
其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。
根据本发明的一个方面，其中每个像素包括成矩阵排列的多个单光子雪崩二极管，每个单光子雪崩二极管可独立选通和寻址，所述步骤S101包括：对于其中至少一个探测单元，遍历该探测单元上的像素阵列输出的电信号，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列。
根据本发明的一个方面，其中所述至少一个探测单元对应于所述激光雷达的特定视场范围，所述数据处理方法还包括：
在所述激光雷达的所述特定视场范围以外的视场，所述数据处理装置配置成根据一个探测单元上的像素阵列输出的电信号生成激光雷达点云中的一个点。
根据本发明的一个方面，其中所述至少一个探测单元包括第一探测单元和第二探测单元，其中所述第一探测单元和所述第二探测单元为规格相同的两个探测单元，
其中所述步骤S101包括：根据该第一探测单元上的像素阵列输出的电信号进行重组，获得第一重组信号阵列；根据该第二探测单元上的像素阵列输出的电信号进行重组，获得第二重组信号阵列；和
所述步骤S102包括：根据所述第一重组信号阵列，生成激光雷达点云中 的m个点；根据所述第二重组信号阵列，生成激光雷达点云中的n个点，其中m大于n。
根据本发明的一个方面，其中所述第一探测单元对应于所述激光雷达的视场中央区域，所述第二探测单元对应于所述激光雷达的视场中央区域的外侧。
根据本发明的一个方面，还包括：
获取所述激光雷达周围的ROI区域；
将与所述ROI区域对应的探测单元设置为所述至少一个探测单元。
根据本发明的一个方面，其中所述步骤S101包括：
通过卷积核，按照预设的卷积步长，对该探测单元上的像素阵列输出的电信号进行卷积处理，获得所述重组信号阵列。
根据本发明的一个方面，其中所述卷积步长可根据所述探测单元对应的视场而调节。
根据本发明的一个方面，其中所述激光雷达的视场中央区域的卷积步长小于视场边缘区域的卷积步长。
根据本发明的一个方面，其中所述激光雷达还包括发射装置，所述发射装置包括多个发射单元，所述重组信号阵列的行数大于所述发射单元的个数。
根据本发明的一个方面，其中所述重组信号阵列的维数小于所述像素阵列的维数。
本发明还涉及一种集成的光探测和数据采集处理装置，包括：：
多个探测单元，每个探测单元包括像素阵列，其中每个像素可对光信号作出响应并转换为电信号；和
数据采集处理装置，与所述多个探测单元耦接，并且配置成：
对于其中至少一个探测单元，获取该探测单元上的像素阵列输出的电信号，对该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；和
根据所述重组信号阵列，生成激光雷达点云，
其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号 获得。
采用本发明的技术方案，相比于现有技术，通过对探测单元上的像素阵列输出的电信号进行重组，可以获得信号更强的重组信号阵列，根据重组信号阵列可以生成线束更高的激光雷达点云，大幅度提升了激光雷达的测远性能，有利于获得更加有效和可靠的探测结果；通过配置卷积步长可以灵活调节激光雷达点云的角度分辨率，可以灵活实现全局点云加密、部分区域点云加密等配置；此外，本发明的技术方案光斑重叠区域小，可以提高光子利用率，降低整机功耗。
附图说明
构成本公开的一部分的附图用来提供对本公开的进一步理解，本公开的示意性实施例及其说明用于解释本公开，并不构成对本公开的不当限定。在附图中：
图1a示出了现有的基于分立的感光器件的激光雷达的发射单元和接收单元的组成示意图；
图1b示出了现有的基于分立的感光器件探测的回波光斑存在交叠的示意图；
图2示出了根据本发明的一个实施例的激光雷达的示意图；
图3a和图3b分别示出了根据本发明一个优选实施例的发射装置的示意图；
图4a和图4b分别示出了根据本发明一个优选实施例的探测装置的示意图；
图5a和图5b分别示出了根据本发明一个优选实施例的对探测单元上的像素阵列输出的电信号进行卷积处理获得重组信号阵列的示意图；
图6示出了根据本发明的一个优选实施例的针对不同视场区域的多个探测单元的像素阵列重组处理获得的重组信号阵列的示意图；
图7示出了根据本发明一个优选实施例的探测单元的工作模式的示意图；
图8中示出了根据本发明一个优选实施例的探测芯片的示意图；
图9示出了根据本发明一个优选实施例的激光雷达的结构示意图；
图10示出了根据本发明一个实施例的集成的光探测和数据采集处理装置的示意图；
图11示出了根据本发明一个优选实施例的集成的光探测和数据采集处理装置的示意图；以及
图12示出了根据本发明一个优选实施例的激光雷达的数据处理方法的流程图。
具体实施方式
在下文中，仅简单地描述了某些示例性实施例。正如本领域技术人员可认识到的那样，在不脱离本发明的精神或范围的情况下，可通过各种不同方式修改所描述的实施例。因此，附图和描述被认为本质上是示例性的而非限制性的。
在本发明的描述中，需要理解的是，术语"中心"、"纵向"、"横向"、"长度"、"宽度"、"厚度"、"上"、"下"、"前"、"后"、"左"、"右"、"竖直"、"水平"、"顶"、"底"、"内"、"外"、"顺时针"、"逆时针"等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本发明和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本发明的限制。此外，术语"第一"、"第二"仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有"第一"、"第二"的特征可以明示或者隐含地包括一个或者更多个所述特征。在本发明的描述中，"多个"的含义是两个或两个以上，除非另有明确具体的限定。
在本发明的描述中，需要说明的是，除非另有明确的规定和限定，术语"安装"、"相连"、"连接"应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或一体地连接:可以是机械连接，也可以是电连接或可以相互通讯；可以是直接相连，也可以通过中间媒介间接相连，可以是两个元件内部的连通或两个元件的相互作用关系。对于本领域的普通技术人员而言，可以根据 具体情况理解上述术语在本发明中的具体含义。
在本发明中，除非另有明确的规定和限定，第一特征在第二特征之"上"或之"下"可以包括第一和第二特征直接接触，也可以包括第一和第二特征不是直接接触而是通过它们之间的另外的特征接触。而且，第一特征在第二特征"之上"、"上方"和"上面"包括第一特征在第二特征正上方和斜上方，或仅仅表示第一特征水平高度高于第二特征。第一特征在第二特征"之下"、"下方"和"下面"包括第一特征在第二特征正上方和斜上方，或仅仅表示第一特征水平高度小于第二特征。
下文的公开提供了许多不同的实施方式或例子用来实现本发明的不同结构。为了简化本发明的公开，下文中对特定例子的部件和设置进行描述。当然，它们仅仅为示例，并且目的不在于限制本发明。此外，本发明可以在不同例子中重复参考数字和/或参考字母，这种重复是为了简化和清楚的目的，其本身不指示所讨论各种实施方式和/或设置之间的关系。此外，本发明提供了的各种特定的工艺和材料的例子，但是本领域普通技术人员可以意识到其他工艺的应用和/或其他材料的使用。
以下结合附图对本发明的优选实施例进行说明，应当理解，此处所描述的优选实施例仅用于说明和解释本发明，并不用于限定本发明。
本发明提供一种激光雷达，采用本发明的激光雷达，能够获得线束更高、角度分辨率可灵活调节的激光雷达点云，测远能力更强，下面详细介绍。
图2示出了根据本发明一个实施例的激光雷达1的示意图。如图2所示，激光雷达1包括发射装置10、探测装置20以及数据处理装置30。其中发射装置10配置成发出探测光束L，用于探测障碍物(如图2中示例性示出的正方体)。探测装置20包括多个探测单元(图2示例性示出了一个探测单元)，每个探测单元包括像素阵列，如图2中示例性示出的3×3像素阵列，其中每个像素可对所述探测光束L在障碍物上反射的回波L’作出响应并转换为电信号。数据处理装置30与探测装置20耦接，并且配置成：对于其中至少一个探测单元，可根据该探测单元上的像素阵列输出的电信号进行重组，从而获得重组信号阵列；根据所述重组信号阵列可生成激光雷达点云，其中所述 重组信号阵列中的每个信号可根据相邻的多个像素输出的电信号获得。本发明中，对于其中至少一个探测单元，并非是根据该探测单元的像素阵列产生的原始电信号而生成点云，而是在其原始电信号的基础上进行重组，获得重组信号阵列，基于重组信号阵列生成激光雷达点云。由于重组信号阵列的信号强度明显增加，因此有利于提高探测单元的探测能力。
图3a示出了根据本发明的一个优选实施例的发射装置10的示意图。如图3a所示，发射装置10包括多个发射单元，如图3a示例性示出的N个发射单元L1、L2、L3、……LN，其中N为大于等于1的整数，多个发射单元构成发射线列。
需要说明的是，发射装置10不限于只包括单列发射单元的情形，根据本发明另一优选实施例，发射装置10也可包括多列发射单元，多列发射单元并行耦接，构成发射面阵，如图3b示例性示出的N×M发射单元面阵，其中N和M均为大于1的整数。
关于发射单元的具体类型，本发明不进行限制，在一些优选实施例中，发射单元可以为垂直腔面发射激光器(VCSEL)或者边发射激光器(EEL)等，具体可根据实际需要进行选择。在探测过程中，发射单元可按照一定角度分辨率(例如0.05°、0.1°、0.4°等)在垂直和/或水平方向上轮询发光，从而实现激光雷达在一定视场范围的探测。
图4a示出了根据本发明的一个优选实施例的探测装置20的示意图。如图4a所示，探测装置20包括多个探测单元,如图4a示例性示出的N个探测单元A1、A2、A3、……AN，其中N为大于等于1的整数，多个探测单元构成探测线列。
上述实施例介绍了探测装置20包括一列探测单元的情形，此外，根据本发明另一优选实施例，发射装置20还可包括多列探测单元，多列探测单元并行耦接，构成探测单元面阵，如图4b示例性示出的N×M探测单元面阵，其中N和M均为大于1的整数。
继续参考图4a或图4b，其中每个探测单元包括多个像素，多个像素构成像素阵列，如图4a或图4b示例性示出的，每个探测单元包括5×5的像素 阵列。在一些优选实施例中，每个像素包括多个单光子雪崩二极管(SPAD)，如图4a或图4b中示例性示出的，每个像素包括3×3共9个单光子雪崩二极管(SPAD)，其中每个单光子雪崩二极管(SPAD)可独立选通和寻址，也就是说，每个单光子雪崩二极管(SPAD)可单独对探测光束L在障碍物上反射的回波L’作出响应并转换为电信号。本发明中，不限制每个探测单元所包括的像素的数目，也不限制每个像素所包括的单光子雪崩二极管的数目。
在一些优选实施例中，根据一个像素上的多个单光子雪崩二极管(SPAD)输出的电信号可获得该像素的信号输出，例如通过对一个像素上的9个单光子雪崩二极管(SPAD)输出的电信号进行累加，可获得该像素的信号输出。同理，根据一个探测单元上的像素阵列输出的电信号亦可获得该探测单元的信号输出，例如通过对一个探测单元上的像素阵列输出的电信号进行累加，可获得该探测单元的信号输出。
在一些优选实施例中，发射装置10中的一个发射单元与探测装置20中的一个探测单元相对应，每个探测单元可以独立选通和寻址。例如一个发射单元发出探测光束，与之对应的一个探测单元作出响应，其他探测单元处于关闭状态。
根据本发明的一个实施例，对于其所述至少一个探测单元，数据处理装置30并不是对像素阵列中的部分像素的电信号进行重组，而是遍历该探测单元上的像素阵列输出的电信号，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列。通过遍历、重组获得重组信号阵列，获得的输出信号强度明显增加，有利于提高探测单元的探测能力。
下面描述根据本发明不同实施例的重组方式。
在一些优选实施例中，其中数据处理装置30通过对像素阵列输出的电信号进行卷积处理，获得重组信号阵列。具体的，通过卷积核，按照预设的卷积步长，对探测单元上的像素阵列输出的电信号进行卷积处理，获得重组信号阵列。其中卷积核即为累加窗口，卷积核的窗口高度H和宽度W可以被配置，其中H和W中至少有一个为大于1个像素的数，两者既可以相等也可以不相等。在一些优选实施例中，卷积核的窗口高度H和宽度W相等，例如均 为3个像素的大小，其中每个像素的尺寸为30um×30um，也就是说，卷积核的尺寸可以为3×3的矩阵(例如90um×90um)。应理解，本实施例只是示例性说明，并不构成对本发明的限制，卷积核的尺寸也可以为其他尺寸，具体可根据实际情况而定。
卷积步长即为卷积核的平移步长，卷积步长被分为水平方向的步长Hs(Hstride)和垂直方向的步长Vs(Vstride)，均可以被灵活配置，其中水平步长Hs和垂直步长Vs均为大于等于1的数，两者既可以相等也可以不相等。在一些优选实施例中，水平步长Hs和垂直步长Vs的值相等，其中最小的水平步长Hs/垂直步长Vs为一个像素的大小。
另外，对探测单元上的像素阵列输出的电信号进行卷积处理，实际上是对探测单元上的像素阵列输出的电信号(例如回波波形)进行数据累加，由此可获得该探测单元上的重组信号阵列(也称为重组像素阵列)，后续可根据每一个重组信号计算障碍物的距离和/或反射率等。下面具体描述。
图5a示出了根据本发明一个优选实施例的对探测单元上的像素阵列输出的电信号进行卷积处理获得重组信号阵列的示意图，其中左边部分示出的是探测单元上的像素阵列在卷积处理之前的情形，中间部分示出的是探测单元上的像素阵列进行卷积处理的过程，右边部分示出的是探测单元上的像素阵列经卷积处理之后的情形。
如图5a左边部分示出的，探测单元的尺寸例如为300um×300um，其包括10×10像素阵列，其中每个像素的尺寸可以为30um×30um。每个像素可以包括9个单光子雪崩二极管(SPAD)，9个单光子雪崩二极管(SPAD)构成3×3单光子雪崩二极管(SPAD)阵列，其中每个单光子雪崩二极管(SPAD)的尺寸可以为10um×10um。另外，卷积核的水平和垂直方向的尺寸可以均为3个像素的大小，即90um×90um，亦即9×9单光子雪崩二极管(SPAD)阵列的大小，当然，卷积核的水平和垂直方向的尺寸也可以是其他尺寸，视实际情况而定。水平和垂直方向的卷积步长的尺寸可以均为1个像素的边长，即30um，亦即3个单光子雪崩二极管(SPAD)的大小。
如图5a中间部分所示出的，在卷积处理过程中，通过卷积核(例如9个 像素的大小，即90um×90um)，按照预设的卷积步长(例如1个像素的大小)对与卷积核对应的像素(例如9个像素)输出的电信号(例如回波波形)进行直接累加或者加权累加，可得到与卷积核对应的像素的重组信号(也称为重组像素,参考图5a右边部分虚线框框出的部分)，遍历整个探测单元上的像素阵列(例如10×10)，可获得该探测单元上的像素阵列(例如10×10)的重组信号阵列，如图5a右边部分示例性示出的8×8像素阵列(图5a中间部分以第2行第2列的像素为中心进行卷积，生成图5a右边部分第1行第1列的重组像素，水平方向按照一个像素的步长，…以第2行第9列的像素为中心进行卷积，生成图5a右边部分第1行第8列的重组像素；竖直方向按照一个像素的步长，…以第9行第2列的像素为中心进行卷积，生成图5a右边部分第8行第1列的重组像素；因此，完成整个10×10像素阵列的遍历后生成8×8重组像素阵列)，后续可根据重组信号阵列中的每一个重组信号计算障碍物的距离和/或反射率等。另外，根据图5a右边部分也可以明显看出，重组像素阵列的尺寸为240um×240um，小于重组之前的像素阵列的尺寸300um×300um。
另外，图5a还示出了两个回波E1和E2，其中回波E1为探测单元上的像素阵列在卷积处理之前的一个像素的回波，回波E2为探测单元上的像素阵列在卷积处理之后的一个重组像素的回波，即一个重组信号，由于重组像素等效于卷积核所对应的像素(例如9个像素)的邻域直接/加权累加，因此重组像素(重组信号)的回波E2的信号强度明显强于重组之前的像素的回波E1的信号强度(图中示出为一个时间窗口的累加结果)。具体地，由于单个spad只能输出0/1的信号，易饱和，单个像素包含的spad数量仍不足以获得较强的测远信号，通过对卷积核覆盖的多个像素输出的信号进行累加，在一个时间窗口，峰值部分得以累加。因此，在经过卷积处理之后，探测单元上的重组信号阵列输出的信号强度明显增加，有利于提高探测单元的探测能力。
另外，本领域技术人员容易理解，在图5a中，在每个探测时刻，探测单元上的像素阵列(10×10的矩阵)中，每个像素的输出均对应一个数值，该 数值例如可以为0、1或者其他数值，代表该像素位置处对应的回波的强度，例如图中回波E1的强度。经过上述的卷积处理之后形成了重组信号阵列(重组像素阵列)，其中每一个元素同样对应有一个数值，代表重组后信号的强度，例如回波E2的强度。
图5示出了探测单元的尺寸为300um×300um，包括10×10像素阵列，卷积核的水平和垂直方向的尺寸均为3个像素的大小，即90um×90um的情况，卷积核的尺寸也可以更大，例如水平和垂直方向的尺寸均为4个像素的大小，即120um×120um的情况，此时回波强度可能更大，但相应地每一个重组像素对应的FOV也会更大，不利于小尺寸物体的探测。例如，当每个像素对应的水平和垂直视场均为0.05°时，采用水平和垂直方向的尺寸均为3个像素大小的卷积核时，生成的重组像素对应的水平和垂直视场均为0.15°；当采用水平和垂直方向的尺寸均为4个像素大小的卷积核时，生成的重组像素对应的水平和垂直视场均为0.2°。因此会根据实际应用选择探测单元、像素及卷积核的尺寸。
上述实施例以卷积步长为1个像素的大小为例，对探测单元上的像素阵列输出的电信号进行卷积处理获得重组信号阵列的整个过程进行了详细介绍，下面以卷积步长为2个像素的大小为例，介绍探测单元上的像素阵列输出的电信号进行卷积处理获得重组信号阵列的过程。
图5b示出了根据本发明另一个优选实施例的对探测单元上的像素阵列输出的电信号进行卷积处理获得重组信号阵列的示意图，其中左边部分示出的是探测单元上的像素阵列在卷积处理之前的情形，中间部分示出的是探测单元上的像素阵列进行卷积处理的过程，右边部分示出的是探测单元上的像素阵列经卷积处理之后的情形。
关于探测单元上的像素阵列在卷积处理之前的情形，与前述实施例类似，此处不再赘述。
关于探测单元上的像素阵列进行卷积处理的过程，如图5b中间部分示出的，通过卷积核(例如9个像素的大小)，按照预设的卷积步长(例如2个像素的大小)对与卷积核对应的像素(例如9个像素)输出的电信号(例如 回波波形)进行直接累加或者加权累加，可得到与卷积核对应的像素的重组像素(参考图5b右边部分的虚线框框出的部分)，遍历整个探测单元上的像素阵列(例如10×10)，可获得该探测单元上的像素阵列(例如10×10)的重组信号阵列，如图5b右边部分示例性示出的4×4像素阵列，后续可根据重组信号阵列中的每一个重组信号计算障碍物的距离和/或反射率等。另外，根据图5b右边部分也可以明显看出，重组像素阵列的尺寸为240um×240um，小于重组之前的像素阵列的尺寸300um×300um。本实施例中，探测单元的垂直和水平视场角度分别为0.1°×0.1°。虽然图5b未示出，但应理解，在经过卷积处理之后，探测单元上的重组信号阵列输出的信号强度亦明显增加，有利于提高激光雷达的测远性能。
以上实施例分别以卷积步长(水平/垂直)为1个像素、2个像素的大小为例，介绍了对探测单元上的像素阵列输出的电信号进行卷积处理获得重组信号阵列的过程，通过对比图5a和图5b可知，对探测单元上的10×10像素阵列进行卷积处理过程中，当卷积步长为1个像素的大小时，可获得8×8重组信号阵列，相邻重组信号之间的距离为1个像素的大小，激光雷达点云的分辨率为0.05°×0.05°。当卷积步长为2个像素的大小时，可获得4×4重组信号阵列,相邻重组信号的间距为2个像素的大小，激光雷达点云的分辨率为0.1°×0.1°。由此可见，卷积步长最终体现为卷积处理之后的重组信号阵列的大小，并且卷积处理之后的重组信号阵列的大小与卷积步长呈负相关，即卷积步长越大，卷积处理之后的重组信号阵列的尺寸越小。此外，卷积处理之后的激光雷达点云的角度分辨率与卷积步长亦呈负相关，即卷积步长越小，卷积处理之后的激光雷达点云的角度分辨率越高，激光雷达点云的质量越好，越容易呈现被探测物体的细节。因此，可根据实际情况灵活调节卷积步长的大小，从而实现激光雷达点云的分辨率的灵活配置。
在一些优选实施例中，卷积步长可根据探测单元对应的视场而调节，换言之，针对探测单元所对应的不同视场，可以采用不同的卷积步长，下面具体描述。在一些实施例中，可以对全部探测单元的输出信号进行上述的重组处理，这样可能对于激光雷达的数据处理量和处理速度提出了较高的要求。 在一些优选实施例中，可以对那些对应于特定视场范围的探测单元的输出信号进行上述重组处理，特定视场范围例如可以为激光雷达的中央视场区域，中央视场区域通常是最需要关注的区域。因此数据处理装置30配置成可根据该特定视场范围所对应的至少一个探测单元的像素阵列所输出的电信号进行重组处理，获得重组信号阵列，并根据该重组信号阵列生成激光雷达点云。由于重组信号阵列等效于邻域累加，因此基于重组信号阵列生成的激光雷达点云的信号强度也随之增加，有利于提高激光雷达的测远性能。在一些优选实施例中，可以采用图像识别的方法确定激光雷达的特定视场区域。例如激光雷达可配合摄像头一起工作。根据摄像头拍摄的图像，可以识别出一些需要重点关注的物体，例如行人、车辆等。对于识别出的物体，可以确定相对应的激光雷达视场区域以及与此对应的探测单元，对于这些探测单元的输出信号，采用上述重组处理，进行数据处理。
针对在激光雷达的所述特定视场范围以外的视场，例如激光雷达的边缘视场区域，数据处理装置30配置成：可根据所述特定视场范围以外的视场所对应的一个探测单元上的像素阵列输出的电信号生成激光雷达点云，而不将所述一个探测单元所输出的电信号进行重组处理，由此可以有效降低激光雷达的功耗和处理负担。需要说明的是，所述特定视场范围并非一定是激光雷达的中央视场区域，也可以为激光雷达的其他视场区域，可根据实际需求进行设定，本发明不进行限制。
另外，根据本发明的优选实施例，也可以针对不同的区域采用不同的卷积步长。其中对于探测装置中的第一探测单元和第二探测单元优选为规格相同的两个探测单元，也就是说，两者的尺寸、包括的像素数量、像素的尺寸、单光子雪崩二极管(SPAD)的数量及大小都相同，两者分别对应不同的视场区域，可根据需要采用不同的卷积步长。例如，第一探测单元和第二探测单元的尺寸均为300um×300um，均包括10×10像素阵列，其中每个像素的尺寸均为30um×30um，每个像素包括9个单光子雪崩二极管(SPAD)，9个单光子雪崩二极管(SPAD)构成3×3单光子雪崩二极管(SPAD)阵列，其中每个单光子雪崩二极管(SPAD)的尺寸为10um×10um，应理解，本实施例中的 第一探测单元和第二探测单元的规格只是示例性说明，并不构成对本发明的限制。数据处理装置30可根据第一探测单元上的像素阵列输出的电信号进行重组，获得第一重组信号阵列；根据所述第一重组信号阵列，生成激光雷达点云中的m个点；可根据该第二探测单元上的像素阵列输出的电信号进行重组，获得第二重组信号阵列；根据所述第二重组信号阵列，生成激光雷达点云中的n个点，其中m大于n。例如在图6中示出了通过这样的方式对两个第一探测单元和两个第二探测单元的像素阵列重组获得的重组信号阵列，其中第一探测单元可对应于激光雷达的视场中央区域，第二探测单元可对应于激光雷达的视场中央区域的外侧。由于激光雷达在进行探测时通常更加关注其视场中央区域，因此将第一探测单元对应于该视场中央区域进行探测，可以生成更加密集的激光雷达点云(如图6中所示，中间区域重组的信号阵列为8×8的阵列)，以便更加精细地呈现探测目标的形态细节，有利于获得更加有效和可靠的探测结果，尤其适用于远距离目标探测；而对于靠近边缘的视场区域对应的第二探测单元，采用较大的卷积步长，生成较为稀疏的激光雷达点云(如图6中所示，边缘区域重组的信号阵列为4×4的阵列)。另外，在一些优选实施例中，可将非重点探测目标设置在激光雷达的视场中央区域的外侧，将第二探测单元对应于该视场中央区域的外侧进行探测，由此可实现在一定程度上兼顾有效并可靠的探测结果以及处理效率。由图6可以明显看出，根据第一重组信号阵列生成的激光雷达点云较为密集，而根据第二重组信号阵列生成的激光雷达点云较为稀疏，其中深色格子表示对应空间位置生成激光雷达点云中的点。需要说明的是，图6示出的是对两个第一探测单元和两个第二探测单元的像素阵列进行重组处理获得的重组信号阵列，只用于示例性说明，并不构成对本发明的限制，在实际处理过程中，也可以对其他数量的第一探测单元、第二探测单元的像素阵列进行重组，这些都在本发明的保护范围之内。
在上述实施例中，激光雷达的视场中央区域的卷积步长可调节为小于视场边缘区域的卷积步长。例如针对激光雷达的视场中央区域，可以设置较小的卷积步长(例如1个像素的大小)；而针对激光雷达的视场边缘区域，可 以设置较大的卷积步长(例如2个像素的大小)，由此可使激光光雷达点云的分布由均匀的分辨率可以扩展为视场中央区域加密而视场边缘区域稀疏的能力，如图6示例性示出的，激光雷达点云中(以一个水平角度示意)，视场中央区域的分辨率为0.05°×0.05°，而视场边缘区域的分辨率为0.1°×0.1°。
在一些优选实施例中，其中所述数据处理装置30还配置成：可获取激光雷达周围的ROI区域；可将与所述ROI区域对应的探测单元设置为至少一个探测单元，即需要进行信号重组的探测单元。其中所述ROI区域指的是感兴趣区域(Region of Interest,ROI)，具体可根据用户需求和/兴趣自行设定，可选的，例如可以采用图像识别的方法确定ROI区域。在设定好ROI区域之后，可根据ROI区域的大小和/或形状设置与其对应的至少一个探测单元，优选的，可设置多个探测单元，以获得更大的探测区域，有利于输出信号强度更强的信号，从而获得更高质量的激光雷达点云，有利于提高ROI区域探测结果的有效性和可靠性。另外，也可以将ROI区域设置在激光雷达的视场中央区域，也将进一步提高探测结果的有效性和可靠性。
此外，还可根据激光雷达周围的ROI区域设置卷积步长。例如对于激光雷达周围的ROI区域可使用较小的卷积步长(例如1个像素的大小)，而针对非ROI区域使用较大的卷积步长(例如2个像素的大小)，应理解，本实施例只是示例性说明，并不构成对本发明的限制。
在另一些优选实施例中，还可根据激光雷达点云的疏密程度调节卷积步长，例如针对激光雷达点云中的加密区域，可使用较小的卷积步长(例如1个像素的大小)，而针对稀疏区域(非加密区)，可使用较大的卷积步长(例如2个像素的大小)。
以上实施例介绍了根据探测单元对应的视场而调节卷积步长的情形，通过调节卷积步长，可实现激光雷达点云的分辨率的灵活配置，由此不仅可以获得所需要的分辨率的激光雷达点云，还可以减少资源消耗。
图7示出了根据本发明的一个优选实施例的探测单元的工作模式的示意图。在一些优选实施例中，探测单元的工作模式可以为：对于垂直方向的扫 描，可由各个像素逐一扫描完成遍历，例如通过每个发射单元间隔一定视场角(例如0.4°)后轮询发光，一个发射单元发出的探测光的光斑与一个探测单元尺寸基本相同，由对应的探测单元作出响应，所探测到的电信号经过模数转换芯片例如模拟数字转化器(ADC)、时间数字转化器(TDC)转换后被数字处理芯片进行回波识别和时间测量，由此可实现垂直方向视场范围的探测，属于电子扫描。对于水平方向的扫描，可由扫描装置偏转或者转子转动等机械旋转的方式带动发射单元从激光雷达的视场一侧扫向另外一侧，由此实现水平视场范围的探测，属于机械扫描。图中，由于垂直方向的扫描和水平方向的扫描同步进行，所以一列探测单元对应的水平角度有一定偏移。
在一些优选实施例中，探测装置20可基于一片采用飞行时间(TOF)测量的探测芯片实现。图8中示出了根据本发明一优选实施例的探测芯片的示意图，如图8左边部分所示，探测芯片上集成有若干个独立的探测单元(图8示例性示出了其中一个探测单元，参考白色小方块示出的部分)，其中每个探测单元包括像素阵列。图8右边部分为其中一个探测单元的放大图，探测单元的尺寸可以为300um×300um，其可以包括10×10像素阵列，其中每个像素可以包括3×3单光子雪崩二极管(SPAD)面阵。
在一些优选实施例中，其中所述发射装置包括多个发射单元，重组信号阵列的行数大于发射单元的个数。例如对于一个发射单元，根据本发明重组后的像素阵列具有多行，例如上述的8行或者4行，也即一个发射单元最终形成了8线或4线。激光雷达的探测单元数和发射单元、以及线数不再如背景技术所述的一一对应，而是线数远大于发射单元和探测单元的数量，因此本发明的方案极大地扩充了激光雷达的线数。
在一些优选实施例中，其中重组信号阵列的维数小于像素阵列的维数。继续参考图5a，例如重组信号阵列的维数为8，像素阵列的维数为10。又如图5b所示，重组信号阵列的维数为4，像素阵列的维数为10。由此可见，重组信号阵列的维数小于像素阵列的维数。
需要说明的是，通过本发明的技术方案生成的激光雷达点云，其线束可以不与发射单元/探测单元的数量相等，可以远大于发射单元/探测单元的数 量，因此可以实现线束更高的激光雷达点云。
根据本发明的一个优选实施例，基于32个发射单元和32个探测单元可以实现256线激光雷达点云。具体的，32个发射单元和32个探测单元沿垂直方向分布，卷积步长为1个像素的大小时，一次遍历，可以获得256行(32×8＝256)×8列点云，即可得到256×8的点云阵列，等效256线的激光雷达。对于256线以上的激光雷达，与之类似，不再赘述。
根据本发明的一个优选实施例，激光雷达可以是扫描式激光雷达，如图9所示，激光雷达除了包括发射装置10、探测装置20以及数据处理装置30之外，还包括扫描装置40、第一反射单元51以及第二反射单元52，其中扫描装置40可以为转镜、振镜、摆镜中的至少一种，第一反射单元51和第二反射单元52可以为反射镜。
根据本发明的一个优选实施例，所述扫描装置为多面转镜，在激光雷达探测过程中，发射装置10和探测装置20保持固定，发射装置10发射出探测光束L经第一反射单元51反射之后，再经过扫描装置40的一个面出射至外部空间，经外部空间中的障碍物反射形成回波L’，回波L’入射至扫描装置40的另一个面，再经由第二反射单元52反射之后被探测装置上探测单元所接收。通过扫描装置40绕转轴旋转，可实现在水平方向上一定视场范围的探测。
在一些优选实施例中，扫描装置40可以采用一维扫描装置或者二维扫描装置。对于采用一维扫描装置的激光雷达，每个发射单元的发光方向对应于水平/竖直方向的一个视场或一个角度，不同发射单元对应的视场或角度不同，由此可实现在水平/竖直方向上一定视场范围的探测。具体的，激光雷达在进行探测过程中，一维扫描装置绕转轴旋转，当该转轴为竖直方向时，一维扫描装置通过旋转，可将发射单元发出的探测光束在水平方向上偏转到不同的角度，从而实现水平方向一定视场范围的扫描探测，再结合发射单元在竖直方向的探测，实现对周围环境的三维探测；反之，当该转轴为水平方向时，一维扫描装置通过旋转，可将发射单元发出的探测光束在竖直方向上偏转到不同角度，从而实现竖直方向上一定视场范围的扫描探测，再结合发射单元 在水平方向的探测，实现对周围环境的三维探测。
对于采用二维扫描装置的激光雷达，扫描装置40具有两个成一定角度(例如90°)的转轴，因此扫描装置40可以在两个方向(例如水平、竖直方向)上将发射单元发出的探测光束反射到不同的角度，从而实现二维扫描。在一些实施例中，二维扫描装置可以包括两个一维扫描装置，两个一维扫描装置的转轴成一定角度(例如90°)，由此也能够实现二维扫描。
以上对扫描式激光雷达进行了介绍，根据本发明另一个优选实施例，激光雷达也可以是机械式激光雷达。对于机械式激光雷达，除了包括发射装置10、探测装置20以及数据处理装置30之外，还包括光机转子(图中未示出)。发射装置10和探测装置20设置在光机转子上，发射装置10的至少一个发射单元和探测装置20的至少一个探测单元相对应。在探测过程中，每个发射单元的发光方向对应于竖直方向的一个视场或一个角度，不同发射单元在竖直方向对应的视场或角度不同，由此可实现激光雷达在竖直方向上一定视场范围的扫描探测。光机转子围绕竖直轴线旋转，由此可实现激光雷达在水平方向360°的扫描探测。光机转子在旋转过程中，可以每隔一定的旋转角度(即激光雷达的水平角分辨率，例如0.05°、0.1°等)发射一次探测光束L，探测装置20接收该探测光束L在障碍物上漫反射后的回波L’，经数据处理装置30根据探测单元上的像素阵列输出的电信号进行重组，可获得重组信号阵列，根据重组信号阵列可生成需要的分辨率的激光雷达点云。
根据本发明又一优选实施例，激光雷达还可以是全固态式激光雷达，优选为面阵激光雷达。在探测过程中，一个发射单元光照亮一个探测单元，该探测单元上电工作，其他探测单元处于关闭状态。在一些优选实施例中，发射单元和探测单元均呈面阵分布，如图3b和图4b所示，探测单元根据相对应的发射单元的发光顺序上电探测，并进行相应的卷积处理，可根据需要选择合适的卷积核和卷积步长，经数据处理装置30根据探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列，根据重组信号阵列可生成需要的分辨率的激光雷达点云。相比于扫描式或者机械式激光雷达，由于全固态式激光雷达未设置扫描装置40或者光子转子，因此可以进一步缩小激光雷达 的体积，更有利于激光雷达的小型化。
以上对本发明的激光雷达进行了具体介绍，采用本发明的激光雷达，通过每次对有限视场区域进行曝光(即TOF测量)，通过视场移动可完成对激光雷达周围的全场景扫描，通过将每个区域的激光雷达点云进行拼接即可生成线束更高、角度分辨率可灵活调节的激光雷达点云，测远能力更强。
此外，本发明还涉及一种集成的光探测和数据采集处理装置200，如图10所示，光探测和数据采集处理装置200包括多个探测单元210和数据采集处理装置220。其中多个探测单元210中，每个探测单元包括像素阵列，其中每个像素可对光信号作出响应并转换为电信号。数据采集处理装置220与所述多个探测单元210耦接，并且配置成：
对于其中至少一个探测单元，获取该探测单元上的像素阵列输出的电信号，对该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；
根据所述重组信号阵列，生成激光雷达点云，其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。
图11示出了根据本发明的一个优选实施例的集成的光探测和数据采集处理装置300的示意图，其中数据采集处理装置220包括数字信号获取单元220-1和数字信号处理单元220-2，数字信号获取单元220-1和数字信号处理单元220-2相互耦接，其中数字信号获取单元220-1与多个探测单元210耦接，配置成获取由多个探测单元210上的像素阵列输出的电信号；数字信号处理单元220-2配置成基于数字信号获取单元220-1获取的电信号进行重组，获得重组信号阵列，以及根据所述重组信号阵列生成需要的分辨率的激光雷达点云。
以上对光探测和数据采集处理装置200/300进行了详细介绍，通过将多个探测单元210和数据采集处理装置220集成，有利于实现整个装置的集成化和小型化，提高了整个装置的使用便利性。
此外，本发明还涉及一种激光雷达的数据处理方法100，其中所述激光雷达的探测装置包括多个探测单元，每个探测单元包括像素阵列，其中每个像素可对所述探测光束在障碍物上反射的回波作出响应并转换为电信号，所 述数据处理方法100包括步骤S101和S102。
如图12所示，在步骤S101：对于其中至少一个探测单元，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；
在步骤S102：根据所述重组信号阵列，生成激光雷达点云，
其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。
根据本发明的一个优选实施例，其中每个像素包括成矩阵排列的多个单光子雪崩二极管，每个单光子雪崩二极管可独立选通和寻址。根据本发明的一个优选实施例，所述步骤S101包括：对于其中至少一个探测单元，遍历该探测单元上的像素阵列输出的电信号，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列。
根据本发明的一个优选实施例，其中所述至少一个探测单元对应于所述激光雷达的特定视场范围，所述数据处理方法100还包括：在所述激光雷达的所述特定视场范围以外的视场，所述数据处理装置配置成根据一个探测单元上的像素阵列输出的电信号生成激光雷达点云中的一个点。
根据本发明的一个优选实施例，其中所述至少一个探测单元包括第一探测单元和第二探测单元，
其中所述步骤S101包括：根据该第一探测单元上的像素阵列输出的电信号进行重组，获得第一重组信号阵列；根据该第二探测单元上的像素阵列输出的电信号进行重组，获得第二重组信号阵列；
所述步骤S102包括：根据所述第一重组信号阵列，生成激光雷达点云中的m个点；根据所述第二重组信号阵列，生成激光雷达点云中的n个点，其中m大于n。
根据本发明的一个优选实施例，其中所述第一探测单元对应于所述激光雷达的视场中央区域，所述第二探测单元对应于所述激光雷达的视场中央区域的外侧。
根据本发明的一个优选实施例，还包括：
获取所述激光雷达周围的ROI区域；
将与所述ROI区域对应的探测单元设置为所述至少一个探测单元。
根据本发明的一个优选实施例，其中所述步骤S101包括：
通过卷积核，按照预设的卷积步长，对该探测单元上的像素阵列输出的电信号进行卷积处理，获得所述重组信号阵列。
根据本发明的一个优选实施例，其中所述卷积步长可根据所述探测单元对应的视场而调节。
根据本发明的一个优选实施例，其中所述激光雷达的视场中央区域的卷积步长小于视场边缘区域的卷积步长。
根据本发明的一个优选实施例，其中所述激光雷达还包括发射装置，所述发射装置包括多个发射单元，所述重组信号阵列的行数大于所述发射单元的个数。
根据本发明的一个优选实施例，其中所述重组信号阵列的维数小于所述像素阵列的维数。
以上对激光雷达的数据处理方法100进行了介绍，采用本发明的数据处理方法100，能够生成线束更高、分辨率可调节的激光雷达点云，有利于提高探测结果的准确度。
以上对本发明的技术方案进行了介绍，采用本发明的技术方案，相比于现有技术，通过对探测单元上的像素阵列输出的电信号进行重组，可获得信号更强的重组信号阵列，根据重组信号阵列可生成线束更高的激光雷达点云，大幅度提升了激光雷达的测远性能，有利于获得更加有效和可靠的探测结果；通过配置卷积步长可以灵活调节激光雷达点云的角度分辨率，可以灵活实现全局点云加密、部分区域点云加密等配置；此外，本发明的技术方案光斑重叠区域小，可以提高光子利用率，降低整机功耗。
本发明还提供一种计算机可读存储介质，包括存储于其上的计算机可执行指令，所述可执行指令在被处理器执行时实施如上所述的数据处理方法100。
在一些优选实施例中，所述计算机可读存储介质可以采用一个或多个计算机可读的介质的任意组合。所述计算机可读存储介质例如可以是但不限于电、磁、光、或半导体的形态或装置，更具体的例子(非穷举的列表)包括： 具有一个或多个导线的电连接、便携式计算机硬盘、硬盘、随机存取存储器(RAM)、非易失性随机访问存储器(NVRAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。
在本文中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其组合使用。所述处理器可以是中央处理单元(Central Processing Unit，CPU)，还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(ApplicationSpecific Integrated Circuit，ASIC)、现成可编程门阵列(Field-Programmable GateArray，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。本发明不进行限制，视具体情况而定。
需要说明的是，本说明书提供了如实施例或示意图所述的方法操作步骤，但基于常规或者无创造性的劳动可以包括更多或者更少的操作步骤。实施例中列举的步骤顺序仅仅为众多步骤执行顺序中的一种方式，不代表唯一的执行顺序。在实际中的系统或设备产品执行时，可以按照实施例或者流程图所示的方法顺序执行或者并行执行。
最后应说明的是：以上所述仅为本发明的优选实施例而已，并不用于限制本发明，尽管参照前述实施例对本发明进行了详细的说明，对于本领域的技术人员来说，其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (23)

  1. 一种激光雷达，包括：
    发射装置，所述发射装置配置成发出探测光束，用于探测障碍物；
    探测装置，包括多个探测单元，每个探测单元包括像素阵列，其中每个像素可对所述探测光束在障碍物上反射的回波作出响应并转换为电信号；和数据处理装置，与所述探测装置耦接，并且配置成：
    对于其中至少一个探测单元，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；和
    根据所述重组信号阵列，生成激光雷达点云，
    其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。
  2. 根据权利要求1所述的激光雷达，其中每个像素包括多个单光子雪崩二极管，每个单光子雪崩二极管可独立选通和寻址，所述数据处理装置配置成对于其中至少一个探测单元，遍历该探测单元上的像素阵列输出的电信号，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列。
  3. 根据权利要求1所述的激光雷达，其中所述至少一个探测单元对应于所述激光雷达的特定视场范围，在所述激光雷达的所述特定视场范围以外的视场，所述数据处理装置配置成根据一个探测单元上的像素阵列输出的电信号生成激光雷达点云。
  4. 根据权利要求1-3中任一项所述的激光雷达，其中所述至少一个探测单元包括第一探测单元和第二探测单元，其中所述第一探测单元和所述第二探测单元为规格相同的两个探测单元，其中所述数据处理装置配置成：
    根据该第一探测单元上的像素阵列输出的电信号进行重组，获得第一重组信号阵列；根据该第二探测单元上的像素阵列输出的电信号进行重组，获得第二重组信号阵列；和
    根据所述第一重组信号阵列，生成激光雷达点云中的m个点；根据所述第二重组信号阵列，生成激光雷达点云中的n个点，其中m大于n。
  5. 根据权利要求4所述的激光雷达，其中所述第一探测单元对应于所述激光雷达的视场中央区域，所述第二探测单元对应于所述激光雷达的视场中央区域的外侧。
  6. 根据权利要求1-3中任一项所述的激光雷达，其中所述数据处理装置配置成：
    获取所述激光雷达周围的ROI区域；
    将与所述ROI区域对应的探测单元设置为所述至少一个探测单元。
  7. 根据权利要求1-3中任一项所述的激光雷达，其中所述数据处理装置配置成：
    通过卷积核，按照预设的卷积步长，对该探测单元上的像素阵列输出的电信号进行卷积处理，获得所述重组信号阵列。
  8. 根据权利要求1-3中任一项所述的激光雷达，其中所述卷积步长可根据所述探测单元对应的视场而调节。
  9. 根据权利要求8所述的激光雷达，其中所述激光雷达的视场中央区域的卷积步长小于视场边缘区域的卷积步长。
  10. 根据权利要求1-3中任一项所述的激光雷达，其中所述发射装置包括多个发射单元，所述重组信号阵列的行数大于所述发射单元的个数。
  11. 根据权利要求1-3中任一项所述的激光雷达，其中所述重组信号阵列的维数小于所述像素阵列的维数。
  12. 一种激光雷达的数据处理方法，其中所述激光雷达的探测装置包括多个 探测单元，每个探测单元包括像素阵列，其中每个像素可对所述探测光束在障碍物上反射的回波作出响应并转换为电信号，所述数据处理方法包括：
    S101：对于其中至少一个探测单元，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；和
    S102：根据所述重组信号阵列，生成激光雷达点云，
    其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。
  13. 根据权利要求12所述的数据处理方法，其中每个像素包括成矩阵排列的多个单光子雪崩二极管，每个单光子雪崩二极管可独立选通和寻址，所述步骤S101包括：对于其中至少一个探测单元，遍历该探测单元上的像素阵列输出的电信号，根据该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列。
  14. 根据权利要求12所述的数据处理方法，其中所述至少一个探测单元对应于所述激光雷达的特定视场范围，所述数据处理方法还包括：
    在所述激光雷达的所述特定视场范围以外的视场，所述数据处理装置配置成根据一个探测单元上的像素阵列输出的电信号生成激光雷达点云中的一个点。
  15. 根据权利要求12-14中任一项所述的数据处理方法，其中所述至少一个探测单元包括第一探测单元和第二探测单元，其中所述第一探测单元和所述第二探测单元为规格相同的两个探测单元，
    其中所述步骤S101包括：根据该第一探测单元上的像素阵列输出的电信号进行重组，获得第一重组信号阵列；根据该第二探测单元上的像素阵列输出的电信号进行重组，获得第二重组信号阵列；和
    所述步骤S102包括：根据所述第一重组信号阵列，生成激光雷达点云中 的m个点；根据所述第二重组信号阵列，生成激光雷达点云中的n个点，其中m大于n。
  16. 根据权利要求15所述的数据处理方法，其中所述第一探测单元对应于所述激光雷达的视场中央区域，所述第二探测单元对应于所述激光雷达的视场中央区域的外侧。
  17. 根据权利要求12-14中任一项所述的数据处理方法，还包括：
    获取所述激光雷达周围的ROI区域；
    将与所述ROI区域对应的探测单元设置为所述至少一个探测单元。
  18. 根据权利要求12-14中任一项所述的数据处理方法，其中所述步骤S101包括：
    通过卷积核，按照预设的卷积步长，对该探测单元上的像素阵列输出的电信号进行卷积处理，获得所述重组信号阵列。
  19. 根据权利要求12-14中任一项所述的数据处理方法，其中所述卷积步长可根据所述探测单元对应的视场而调节。
  20. 根据权利要求19所述的激光雷达，其中所述激光雷达的视场中央区域的卷积步长小于视场边缘区域的卷积步长。
  21. 根据权利要求12-14中任一项所述的激光雷达，其中所述激光雷达还包括发射装置，所述发射装置包括多个发射单元，所述重组信号阵列的行数大于所述发射单元的个数。
  22. 根据权利要求12-14中任一项所述的激光雷达，其中所述重组信号阵列的维数小于所述像素阵列的维数。
  23. 一种集成的光探测和数据采集处理装置，包括：
    多个探测单元，每个探测单元包括像素阵列，其中每个像素可对光信号作出响应并转换为电信号；和
    数据采集处理装置，与所述多个探测单元耦接，并且配置成：
    对于其中至少一个探测单元，获取该探测单元上的像素阵列输出的电信号，对该探测单元上的像素阵列输出的电信号进行重组，获得重组信号阵列；和
    根据所述重组信号阵列，生成激光雷达点云，
    其中所述重组信号阵列中的每个信号根据相邻的多个像素输出的电信号获得。
PCT/CN2023/115383 2022-12-07 2023-08-29 激光雷达、数据处理方法及光探测和数据采集处理装置 WO2024119906A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211560601.9A CN118151126A (zh) 2022-12-07 2022-12-07 激光雷达、数据处理方法及光探测和数据采集处理装置
CN202211560601.9 2022-12-07

Publications (1)

Publication Number Publication Date
WO2024119906A1 true WO2024119906A1 (zh) 2024-06-13

Family

ID=91186296

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/115383 WO2024119906A1 (zh) 2022-12-07 2023-08-29 激光雷达、数据处理方法及光探测和数据采集处理装置

Country Status (3)

Country Link
CN (1) CN118151126A (zh)
DE (1) DE102023133937A1 (zh)
WO (1) WO2024119906A1 (zh)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107607931A (zh) * 2017-08-07 2018-01-19 哈尔滨工业大学 一种激光雷达回波图像处理方法
CN107607960A (zh) * 2017-10-19 2018-01-19 深圳市欢创科技有限公司 一种光学测距的方法及装置
CN112861653A (zh) * 2021-01-20 2021-05-28 上海西井信息科技有限公司 融合图像和点云信息的检测方法、系统、设备及存储介质
CN112912890A (zh) * 2018-09-14 2021-06-04 华为技术有限公司 用于使用生成模型生成合成点云数据的方法和系统
CN113359142A (zh) * 2020-03-06 2021-09-07 上海禾赛科技有限公司 激光雷达及其测距方法
CN114022830A (zh) * 2020-07-17 2022-02-08 华为技术有限公司 一种目标确定方法以及目标确定装置
CN114174854A (zh) * 2019-08-07 2022-03-11 华为技术有限公司 飞行时间深度增强
CN114616489A (zh) * 2019-09-06 2022-06-10 奥斯特公司 Lidar图像处理
US20220196798A1 (en) * 2020-12-21 2022-06-23 Intel Corporation High end imaging radar
CN115187812A (zh) * 2022-07-14 2022-10-14 中国科学院空天信息创新研究院 高光谱激光雷达点云数据的分类方法、训练方法及装置

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107607931A (zh) * 2017-08-07 2018-01-19 哈尔滨工业大学 一种激光雷达回波图像处理方法
CN107607960A (zh) * 2017-10-19 2018-01-19 深圳市欢创科技有限公司 一种光学测距的方法及装置
CN112912890A (zh) * 2018-09-14 2021-06-04 华为技术有限公司 用于使用生成模型生成合成点云数据的方法和系统
CN114174854A (zh) * 2019-08-07 2022-03-11 华为技术有限公司 飞行时间深度增强
CN114616489A (zh) * 2019-09-06 2022-06-10 奥斯特公司 Lidar图像处理
CN113359142A (zh) * 2020-03-06 2021-09-07 上海禾赛科技有限公司 激光雷达及其测距方法
CN114022830A (zh) * 2020-07-17 2022-02-08 华为技术有限公司 一种目标确定方法以及目标确定装置
US20220196798A1 (en) * 2020-12-21 2022-06-23 Intel Corporation High end imaging radar
CN112861653A (zh) * 2021-01-20 2021-05-28 上海西井信息科技有限公司 融合图像和点云信息的检测方法、系统、设备及存储介质
CN115187812A (zh) * 2022-07-14 2022-10-14 中国科学院空天信息创新研究院 高光谱激光雷达点云数据的分类方法、训练方法及装置

Also Published As

Publication number Publication date
CN118151126A (zh) 2024-06-07
DE102023133937A1 (de) 2024-06-13

Similar Documents

Publication Publication Date Title
CN110596721B (zh) 双重共享tdc电路的飞行时间距离测量系统及测量方法
US10908266B2 (en) Time of flight distance sensor
US20210181317A1 (en) Time-of-flight-based distance measurement system and method
EP3593166B1 (en) Integrated lidar illumination power control
US20240045038A1 (en) Noise Adaptive Solid-State LIDAR System
US20200191959A1 (en) Solid state light detection and ranging (lidar) system and system and method for improving solid state light detection and ranging (lidar) resolution
US5682229A (en) Laser range camera
US11706400B2 (en) Hybrid sensor system and method for providing 3D imaging
CN110596723B (zh) 动态直方图绘制飞行时间距离测量方法及测量系统
CN110596724B (zh) 动态直方图绘制飞行时间距离测量方法及测量系统
US11592530B2 (en) Detector designs for improved resolution in lidar systems
WO2018217909A1 (en) Receive signal beam steering and detector for an optical distance measurement system
CN113970757A (zh) 一种深度成像方法及深度成像系统
CN212694038U (zh) 一种tof深度测量装置及电子设备
US11668801B2 (en) LIDAR system
US20210181308A1 (en) Optical distance measuring device
WO2019041268A1 (en) SEMICONDUCTOR LIGHT DISTANCE LIDAR DETECTION AND TELEMETRY SYSTEM
WO2024119906A1 (zh) 激光雷达、数据处理方法及光探测和数据采集处理装置
CN110967681B (zh) 用于三维扫描的结构振镜及应用其的激光雷达
US20220413149A1 (en) Operating method and control unit for a lidar system, lidar system, and device
US20210208257A1 (en) Spad array with ambient light suppression for solid-state lidar
CN114371463A (zh) 一种光感应模组及采用其的激光雷达
WO2024113328A1 (zh) 探测方法、阵列探测器、阵列发射器、探测装置及终端
WO2023061386A1 (zh) 激光雷达、接收系统、发射系统以及控制方法
CN118191787A (zh) 光探测和数据采集处理装置、激光雷达及其探测方法