WO2022000333A1 - 一种雷达探测方法及相关装置 - Google Patents
一种雷达探测方法及相关装置 Download PDFInfo
- Publication number
- WO2022000333A1 WO2022000333A1 PCT/CN2020/099558 CN2020099558W WO2022000333A1 WO 2022000333 A1 WO2022000333 A1 WO 2022000333A1 CN 2020099558 W CN2020099558 W CN 2020099558W WO 2022000333 A1 WO2022000333 A1 WO 2022000333A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- signal
- time
- radar
- spectrum
- frequency
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
- G01S13/08—Systems for measuring distance only
- G01S13/32—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
- G01S13/34—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal
- G01S13/343—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal using sawtooth modulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
- G01S13/583—Velocity or trajectory determination systems; Sense-of-movement determination systems using transmission of continuous unmodulated waves, amplitude-, frequency-, or phase-modulated waves and based upon the Doppler effect resulting from movement of targets
- G01S13/584—Velocity or trajectory determination systems; Sense-of-movement determination systems using transmission of continuous unmodulated waves, amplitude-, frequency-, or phase-modulated waves and based upon the Doppler effect resulting from movement of targets adapted for simultaneous range and velocity measurements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
- G01S13/08—Systems for measuring distance only
- G01S13/32—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
- G01S13/34—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal
- G01S13/345—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal using triangular modulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
- G01S17/06—Systems determining position data of a target
- G01S17/08—Systems determining position data of a target for measuring distance only
- G01S17/32—Systems determining position data of a target for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
- G01S17/34—Systems determining position data of a target for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
- G01S17/06—Systems determining position data of a target
- G01S17/42—Simultaneous measurement of distance and other co-ordinates
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/023—Interference mitigation, e.g. reducing or avoiding non-intentional interference with other HF-transmitters, base station transmitters for mobile communication or other radar systems, e.g. using electro-magnetic interference [EMI] reduction techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/35—Details of non-pulse systems
- G01S7/352—Receivers
- G01S7/354—Extracting wanted echo-signals
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/35—Details of non-pulse systems
- G01S7/352—Receivers
- G01S7/356—Receivers involving particularities of FFT processing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/411—Identification of targets based on measurements of radar reflectivity
- G01S7/412—Identification of targets based on measurements of radar reflectivity based on a comparison between measured values and known or stored values
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
- G01S7/481—Constructional features, e.g. arrangements of optical elements
- G01S7/4817—Constructional features, e.g. arrangements of optical elements relating to scanning
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
- G01S13/42—Simultaneous measurement of distance and other co-ordinates
- G01S13/426—Scanning radar, e.g. 3D radar
- G01S13/428—Scanning radar, e.g. 3D radar within the pulse scanning systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/93—Radar or analogous systems specially adapted for specific applications for anti-collision purposes
- G01S13/931—Radar or analogous systems specially adapted for specific applications for anti-collision purposes of land vehicles
- G01S2013/9323—Alternative operation using light waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
- G01S7/4802—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Definitions
- the present application relates to the technical field of lidar, and in particular, to a radar detection method and a related device.
- Frequency modulated continuous wave (FMCW) radar is a ranging device.
- FMCW radar includes different subdivision types.
- the frequency modulated continuous wave radar using radio waves is called FMCW RADAR.
- Another example is the frequency modulation using laser.
- Continuous wave radar is called FMCW LIDAR.
- the radar generates a frequency-modulated radio frequency or laser signal, and the generated FMCW signal is divided into two channels, one of which is used as a local reference signal (also a local reference signal). It is called the local oscillator signal), and the other way is emitted to the measured target (also called the reflector) and reflected by the surface of the target to form an echo signal.
- Figure 2 illustrates the processing process of the reference signal and the echo signal by the FMCW radar, as shown in part (a) of Figure 2, the thick line indicates that the frequency of the FM signal of the transmitted signal and the reference signal changes with time, and the first half of the time , the signal frequency increases with time from low to high, and in the second half of the time, the signal frequency decreases from high to low with time.
- the thin line indicates the above-mentioned echo signal.
- the echo signal and the reference signal pass through the mixer to output a beat frequency signal, and the frequency of the beat frequency signal is the frequency difference between the reference signal and the echo signal, as shown in part (b) of FIG. 2 .
- the beat signal has a fixed frequency (the part inside the dashed line in the figure), as shown in part (c) of Figure 2, by performing frequency domain analysis (usually FFT) on the beat signal, it is possible to detect
- the frequency of the beat signal the frequency here has a one-to-one correspondence with the distance and speed of the target, so according to the frequency of the beat signal, the speed and distance information of the target can be calculated.
- FIG 3 there is usually a scanner in the lidar.
- the scanner emits laser light in different directions by rotating the laser, and then collects the measurement results in all directions to obtain a 3D point cloud image.
- Figure 4 shows a laser A 3D point cloud map of the radar.
- the point cloud resolution of lidar is an important indicator. High-resolution point clouds are very important for machine vision algorithms used in automatic driving, etc., and are crucial for the realization of point cloud segmentation, target recognition and other functions. At present, only one measurement can be completed within the time domain length of an upper and lower chirp pair (that is, one FM cycle of the radar). In this case, the resolution of the point cloud obtained is low.
- the commonly used method to improve the resolution of the point cloud is to More lasers are added to the lidar, and the 32-line LIDAR (that is, 32 lasers working simultaneously) is upgraded to a 128-line LIDAR (128 lasers, or 128 lasers are simulated with optics).
- the embodiment of the present application discloses a radar detection method and a related device, which can improve the radar output rate (ie, point cloud density) without losing the signal-to-noise ratio without significantly increasing the cost.
- the radar output rate ie, point cloud density
- an embodiment of the present application discloses a radar detection method, which includes:
- the radar beat signal into a two-dimensional time spectrum; intercept a plurality of measurement units MU whose time domain length is equal to the frequency modulation period of the radar from the two-dimensional time spectrum according to the time-domain sliding step; wherein, the time-domain sliding The length of the step is smaller than the frequency modulation period of the radar; the radar detection result is determined according to each MU in the plurality of MUs.
- a frequency domain signal with a smaller granularity in the time domain is obtained, and then a sliding window is used to intercept the measurement unit MU on the frequency domain signal with a smaller granularity. It is performed on frequency-domain signals with smaller granularity, so even if some MUs that are cut out have a common frequency band, because the granularity of frequency-domain signals is small, the differences between different MUs can still be distinguished.
- the length of the sliding step time domain in the time domain is smaller than the frequency modulation period of the radar, two adjacent MUs in any two time domains will share a part of the frequency information, so even if more MUs are intercepted, each MU can still be guaranteed. There is enough signal energy accumulation inside to ensure the signal-to-noise ratio. Therefore, in this way, the point-out rate can be improved without losing the signal-to-noise ratio without significantly increasing the cost.
- the converting the radar beat signal into a two-dimensional time spectrum includes: segmenting the beat signal in the time domain to obtain multiple sub-signals, wherein, The time domain length of each sub-signal is 1/N of the FM period of the radar, where N is greater than or equal to 2;
- a Fourier transform is performed on the sub-signals respectively to obtain frequency domain information of each sub-signal in the multi-segment sub-signals; a two-dimensional time spectrum is generated according to the frequency domain information of each sub-signal.
- the two-dimensional time spectrum may be obtained by synthesizing each sub-signal and the frequency domain information.
- the determining the radar detection result according to each MU in the multiple MUs includes: matching each MU with the highest similarity from the spectrum library.
- Reference spectrum wherein multiple reference spectrums are recorded in the spectrum library, each reference spectrum corresponds to detection information, and the detection information includes at least one item of distance information or speed information; the reference spectrum with the highest similarity of each MU is The corresponding detection information is used as the radar detection result of each MU.
- the spectrum library is used to match the reference spectrum that is closest to the MU and the detection information has been calculated, and then the detection information of the closest reference spectrum is used as the detection result of the MU, instead of being calculated in real time, which significantly reduces the detection rate. It improves the speed of determining the detection result of MU at a time; it effectively balances the calculation pressure caused by improving the point rate.
- the method before matching the reference spectrum with the highest similarity for each MU from the spectrum library, the method further includes: according to one item of Doppler frequency offset and one item of signal The time of flight generates a reference spectrum; a piece of speed information and a piece of distance information are used as the detection information corresponding to the one reference spectrum, and are stored in the spectrum library, wherein the item of speed information is the item of Doppler The ler frequency offset or the moving speed of the reflector corresponding to the item of Doppler frequency offset; the item of distance information is the time of flight of the signal or the distance of the reflector corresponding to the time of flight of the signal.
- the time-domain sliding step d is obtained by performing a rounding operation on d0, and d0 satisfies the following relationship:
- T chirp is the frequency modulation period of the radar
- R is the desired increase multiplier of the radar point cloud density
- the point cloud density can be adjusted to the desired density by configuring R, which is flexible and easy to implement.
- an embodiment of the present application provides a signal processing apparatus, and the apparatus includes:
- a conversion unit which is used to convert the beat frequency signal of the radar into a two-dimensional time spectrum
- An intercepting unit configured to intercept a plurality of measurement units MU whose time domain length is equal to the frequency modulation period of the radar from the two-dimensional time spectrum according to the time domain sliding step size; wherein the time domain sliding step length is less than the frequency modulation period of the radar cycle;
- An analysis unit configured to separately determine a radar detection result according to each of the multiple MUs.
- a frequency domain signal with a smaller granularity in the time domain is obtained, and then a sliding window is used to intercept the measurement unit MU on the frequency domain signal with a smaller granularity. It is performed on frequency-domain signals with smaller granularity, so even if some MUs that are cut out have a common frequency band, because the granularity of frequency-domain signals is small, the differences between different MUs can still be distinguished.
- the length of the sliding step time domain in the time domain is smaller than the frequency modulation period of the radar, two adjacent MUs in any two time domains will share a part of the frequency information, so even if more MUs are intercepted, each MU can still be guaranteed. There is enough signal energy accumulation inside to ensure the signal-to-noise ratio. Therefore, in this way, the point-out rate can be improved without losing the signal-to-noise ratio without significantly increasing the cost.
- the two-dimensional time-frequency transform is a short-time Fourier transform
- the conversion unit is specifically configured to: segment the radar beat signal in the time domain to obtain Multi-segment sub-signals, wherein the time domain length of each sub-signal is 1/N of the frequency modulation period of the radar, where N is greater than or equal to 2; Fourier transform is performed on the multi-segment sub-signals to obtain the frequency domain information of each sub-signal; and generating a two-dimensional time spectrum according to the frequency domain information of each sub-signal.
- the two-dimensional time spectrum may be obtained by synthesizing each sub-signal and the frequency domain information.
- the intercepting unit is specifically configured to: match a reference spectrum with the highest similarity for each MU from a spectrum library, wherein the spectrum library records multiple Reference spectrum, each reference spectrum corresponds to detection information, and the detection information includes at least one item of distance information or speed information; the detection information corresponding to the reference spectrum with the highest similarity of each MU is used as the radar detection of each MU result.
- the spectrum library is used to match the reference spectrum that is closest to the MU and the detection information has been calculated, and then the detection information of the closest reference spectrum is used as the detection result of the MU, instead of being calculated in real time, which significantly reduces the detection rate. It improves the speed of determining the detection result of MU at a time; it effectively balances the calculation pressure caused by improving the point rate.
- a generating unit for generating a reference spectrum according to a Doppler frequency offset and a signal time-of-flight
- a storage unit configured to use an item of velocity information and an item of distance information as the detection information corresponding to the one reference spectrum, and save them in the spectrum library, wherein the item of velocity information is the item of Doppler frequency offset or the moving speed of the reflector corresponding to the item of Doppler frequency offset; the item of distance information is the time of flight of the signal or the distance of the reflector corresponding to the time of flight of the signal.
- the time-domain sliding step d is obtained by performing an integer operation on d0, and d0 satisfies the following relationship:
- T chirp is the frequency modulation period of the radar
- R is the desired increase multiplier of the radar point cloud density
- the point cloud density can be adjusted to the desired density by configuring R, which is flexible and easy to implement.
- an embodiment of the present application provides a processor, where the processor is configured to invoke a computer program stored in a computer-readable storage medium to implement the first aspect or any optional implementation manner described in the first aspect Methods.
- an embodiment of the present application provides a lidar system, where the lidar system includes a processor, a memory, and a laser, wherein the laser is used to emit a laser signal, the memory is used to store a computer program, and the The processor is configured to invoke the computer program to implement the method described in the first aspect or any optional implementation manner of the first aspect.
- an embodiment of the present application provides a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and when the computer program runs on a processor, the first aspect or the first aspect is implemented Any optional implementation of the method described.
- FIG. 1 is a schematic diagram of the principle of a laser radar provided by an embodiment of the present application
- FIG. 2 is a schematic diagram of a beat signal provided by an embodiment of the present application.
- FIG. 3 is a schematic diagram of deflection during scanning of a laser provided by an embodiment of the present application.
- FIG. 4 is a schematic diagram of a point cloud effect provided by an embodiment of the present application.
- FIG. 5 is a schematic diagram of the architecture of a lidar system provided by an embodiment of the present application.
- FIG. 6 is a schematic diagram of a beat signal generated by a triangular wave provided by an embodiment of the present application.
- FIG. 7 is a schematic diagram of a beat frequency signal generated by a sawtooth wave provided by an embodiment of the present application.
- FIG. 8 is a schematic diagram of the division of a beat frequency time provided by an embodiment of the present application.
- FIG. 9 is a schematic diagram of a triangular wave chirp coherent signal processing provided by an embodiment of the present application.
- FIG. 10 is a schematic diagram of the division of a beat frequency time provided by an embodiment of the present application.
- FIG. 11 is a schematic flowchart of a radar detection method provided by an embodiment of the present application.
- FIG. 12 is a schematic diagram of a beat signal provided by an embodiment of the present application.
- FIG. 13 is a schematic diagram of a time-frequency signal after a two-dimensional time-frequency change provided by an embodiment of the present application
- FIG. 14 is a schematic diagram of a two-dimensional time spectrum provided by an embodiment of the present application.
- 15 is a schematic diagram of a scene for intercepting a two-dimensional time spectrum provided by an embodiment of the present application.
- 16 is a schematic diagram of a scenario of a reference spectrum provided by an embodiment of the present application.
- FIG. 17 is a schematic structural diagram of a signal processing apparatus provided by an embodiment of the present application.
- the lidar in the embodiments of the present application can be applied to various fields such as intelligent transportation, automatic driving, atmospheric environment monitoring, geographic mapping, and unmanned aerial vehicles, and can perform functions such as distance measurement, speed measurement, target tracking, and imaging recognition.
- FIG. 5 is a schematic structural diagram of a lidar system provided by an embodiment of the present application.
- the lidar system is used to detect information of a target 505.
- the lidar system includes:
- the laser 501 such as a frequency-modulated laser (Tunable Laser, TL) is used to generate a laser signal
- the laser signal can be a linear frequency-modulated laser signal
- the modulation waveform of the frequency of the laser signal can be a sawtooth wave, or a triangle wave, or other forms waveform.
- the splitting device 502 is used for splitting the laser light generated by the laser 501 to obtain an emission signal and a local oscillator signal (Local Oscillator, LO), wherein the local oscillator signal is also called a reference signal.
- a collimating lens 500 may also be configured between the laser 501 and the branching device, and the lens 500 is used for beam shaping the laser signal sent to the branching device 502 .
- Collimator 503 for maximum efficient coupling of the transmitted signal into scanner 504.
- the scanner 504 also known as a 2D scanning mechanism, is used to transmit the transmission signal at a certain angle. After the transmission signal is transmitted, it is reflected by the target 505 to form an echo signal; at this time, the scanner 504 is also used for receiving The echo signal, after the echo signal passes through the corresponding optical devices (eg, the mirror 506 (optional), the receiving mirror 508 (optional)), is combined with the local oscillator signal in the mixer 510 .
- the corresponding optical devices eg, the mirror 506 (optional), the receiving mirror 508 (optional)
- the mixer 510 is used for performing mixing processing on the local oscillator signal and the echo signal to obtain a beat frequency signal.
- the detector 520 is used for extracting the beat signal from the mixer.
- the detector 520 may be Balanced Photo Detectors (BPD).
- An analog-to-digital converter (Analog digital converter, ADC) 511 is used to sample the beat frequency signal, and this sampling is essentially a process of converting an analog signal into a digital signal.
- the processor 512 which may include a digital signal processor (Digital signal processor, DSP), a central processing unit (CPU), an accelerated processing unit (APU), a graphics processing unit (GPU), a microprocessor or a microcontroller, etc.
- DSP digital signal processor
- CPU central processing unit
- APU accelerated processing unit
- GPU graphics processing unit
- microprocessor or a microcontroller etc.
- DSP digital signal processor
- the processor is used to process the sampled beat frequency signal, so as to obtain information such as the speed and distance of the target.
- the target 505 is also called a reflector, and the target 505 can be any object in the scanning direction of the scanner 504, for example, it can be a person, a mountain, a vehicle, a tree, a bridge, etc. illustrated as an example.
- the operation of processing the sampled beat frequency signal to obtain information such as the speed and distance of the target may be performed by one or more processors 512, for example, by one or more DSPs, Of course, it can also be completed by one or more processors 512 in combination with other devices, for example, a DSP combined with one or more central processing units (CPUs) to jointly complete.
- the processor 512 processes the beat frequency signal, it can be specifically realized by calling a computer program stored in a computer-readable storage medium, and the computer-readable storage medium includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read-only memory (compact disc read-only memory, CD-ROM), which can It is configured in the processor 512 or may be independent of the processor 512 .
- RAM random access memory
- ROM read-only memory
- EPROM erasable programmable read only memory
- CD-ROM compact disc read-only memory
- the laser 501 may be one or multiple.
- the one laser 501 may be In the time domain, a laser signal with a positive slope and a laser signal with a negative slope are alternately emitted; when there are two lasers 501, one of them emits a laser signal with a positive slope, and the other emits a laser signal with a negative slope, and the two lasers 501 can be synchronized emit a laser signal.
- the echo signal is mixed with the local oscillator signal LO after a period of flight time.
- the time from the start of emission to the return of the echo signal, the beat frequency signal generated by the echo signal and the local oscillator signal after the flight time is constant for a certain period of time, which can accurately reflect the distance and speed information of the target. This period of time is beat time.
- each measurement unit a measurement unit MU is used to obtain the detection result of one measurement point
- process the beat frequency In the process of signal accumulation, more energy is accumulated to achieve a higher signal-to-noise ratio.
- Practical applications often require lidar to have a higher point-out rate, so that lidar can obtain a higher field of view resolution and frame rate.
- the commonly used sawtooth chirp coherent signal processing method is shown in Figure 7.
- the time remaining after the modulation period T minus the flight time corresponding to the farthest detection distance is taken as the beat time TB, and the data in the beat time TB is collected.
- Fourier transform is performed to obtain the frequency of the real-time beat signal, and then the distance of the target is calculated accordingly.
- the measurement time of each measurement point TM ⁇ TB is obtained with other measurement channels with different linear frequency modulation slopes.
- multiple measurement units MU can be divided in a single frequency time. As shown in Figure 8, the beat frequency time is divided into three segments, and three MUs (corresponding to three measurement points) are obtained. Based on the signal of each MU, one detection can be completed, which can reach three times the original point rate, but at the same time, the beat frequency of each MU also becomes 1/3 of the original.
- the commonly used triangular wave chirp coherent signal processing method is shown in Figure 9.
- the respective flight time and beat frequency time are respectively divided, and the signal processing method in each beat frequency time is linear with the sawtooth wave.
- the FM coherent signal processing method is the same.
- the first half cycle of the measurement unit MU (which can also be described as the upper half cycle corresponding to the measurement point) is processed to obtain the beat frequency f 1
- the lower half cycle of MU (which can also be described as the lower half cycle corresponding to the measurement point) is obtained.
- the beat wave frequency f 2 is obtained by processing with half cycle), and thus, the distance and speed information of the target can be obtained through one measurement channel.
- the beat time in the positive slope period and the negative slope period can be divided into multiple measurement times.
- Figure 10 shows that the beat time T_ ⁇ in the negative slope period is divided into three measurement times.
- the above method of improving the point-out rate reduces the beat time of each measurement unit MU, and reduces the energy accumulation of useful signals during time-frequency transformation, thereby affecting the signal-to-noise ratio of the measurement and reducing the cost of coherent measurement. detection distance.
- the embodiment of the present application provides the method shown in FIG. 11 .
- FIG. 11 is a radar detection method provided by an embodiment of the present application.
- the method can be implemented based on the components in the lidar system shown in FIG. 5. Some operations in the subsequent description are performed by a signal processing device.
- the signal processing device can be the above-mentioned processor 512, or a device in which the above-mentioned processor 512 is deployed, for example, a lidar system or a certain module in the lidar system where the above-mentioned processor 5123 is deployed, and the method includes But not limited to the following steps:
- Step S1101 the signal processing device converts the radar beat signal into a two-dimensional time spectrum.
- the present application not only obtains the frequency domain information in the beat signal, but also obtains the time domain information in the beat signal, so after transformation, frequency domain information with smaller granularity in the time domain can be obtained.
- the transformation can be specifically two-dimensional time-frequency chip transformation, and the time domain length of each signal processed by the two-dimensional time-frequency transformation is the frequency modulation period of the radar (generally also called the frequency modulation period of the laser signal, or called the frequency modulation period of the laser signal.
- the frequency modulation period, or 1/N of the time domain length of an upper and lower chirp (chirp signal) pair), N is greater than or equal to 2, that is, a frequency with a granularity of 1/N of the radar’s frequency modulation period in the time domain is obtained. domain information.
- the two-dimensional time-frequency transform may specifically be short-time Fourier transform.
- the two-dimensional time-frequency transform is performed on the radar beat signal to obtain a two-dimensional time-frequency spectrum, which may specifically include:
- the above-mentioned beat signal is divided in the time domain to obtain multiple sub-signals, wherein the time domain length of each sub-signal is 1/N of the frequency modulation period of the radar;
- Fig. 12 shows a beat signal, in which the horizontal The axis represents the time domain information of the beat signal, and the vertical axis represents the amplitude information of the beat signal.
- the time domain length of this beat signal in Figure 12 is 6 microseconds (us), and the vertical line in Figure 12 indicates the Segmentation of the beat frequency signal of 6 microseconds.
- the expression of the l-segment beat signal s l (n) can be expressed as:
- s l (n) s[(l-1) ⁇ 200+n], 0 ⁇ n ⁇ 200, 0 ⁇ l ⁇ 100
- l is the sequence number of the segment
- n is the sequence number of the sampling point.
- the expression of FFT can be:
- F[] represents the FFT operation
- k is the sequence number of the frequency domain sequence (representing different frequencies).
- a two-dimensional time spectrum is generated according to the frequency domain information of each sub-signal in the multiple sub-signals.
- each sub-signal since each sub-signal is obtained by division in the time domain, each sub-signal corresponds to a piece of time-domain information, and because the frequency domain information in each sub-signal is obtained through Fourier transform, therefore, to this step has been every piece signal time domain information and frequency domain information, it is possible to store this information for each piece signal by respective rules, e.g., through specification of a table N fft ⁇ N slot T of storage, then this table T can be regarded as the two-dimensional time spectrum of the radar beat signal.
- Figure 14 shows a two-dimensional time spectrum, in which the horizontal axis represents the time domain information, and the vertical axis represents the frequency domain information.
- the brightness is used to indicate the frequency amplitude, which cannot be seen from Fig. 14, but in the experimental simulation and practical application, the amplitude information can be represented by the brightness in the two-dimensional time spectrum. Therefore, the two-dimensional time spectrum may be obtained by synthesizing each sub-signal and the frequency domain information.
- the positive rate part of the beat signal is often above the horizontal axis shown in FIG. 14 after the two-dimensional time-frequency transform
- the negative rate part of the beat signal is often after the two-dimensional time-frequency transform.
- the bands below the horizontal axis are usually mirrored above the horizontal axis, so the final effect seen in Figure 14 is that all the bands are above the horizontal axis, in Figure 14
- the second segment of the waveform is mirrored above the horizontal axis.
- the bands illustrated in Figure 14 appear to have only two segments, but in reality there may be more segments, eg, 4 segments, 6 segments, and so on.
- Step S1102 The signal processing apparatus intercepts a plurality of measurement units MU whose time domain length is equal to the frequency modulation period of the radar from the two-dimensional time spectrum according to the time domain sliding step.
- the time-domain sliding step size is preset according to the needs. If you want the radar to have a high point cloud density, that is, a high point-out rate, you can set the time-domain sliding step size to be smaller. For example, you can set the The length of the time domain sliding step is set to be smaller than the FM period of the radar.
- R can be introduced to calculate the time domain. Sliding step, R can be interpreted as: R is the desired increase in the density of radar point cloud, or interpreted as: the number of measurements that need to be completed in one frequency modulation period of the radar. After introducing R, the time domain sliding step size d0 is calculated as follows:
- the measurement unit (Measurement Unit, MU) is intercepted from the previously obtained two-dimensional time spectrum.
- the time domain length of each measurement unit MU is equal to the frequency modulation period of the radar;
- the method is to intercept according to the time domain sliding step size.
- the time-frequency range intercepted by the kth Measurement Unit (MU) is the sliding of the k-1 MU interception range, as shown in Figure 15, from left to right.
- Intercept time increases to the right
- R is equal to 5
- 5 measurement units MU are intercepted, which are MU1, MU2, MU3, MU4, and MU5 in sequence.
- the difference between the time domain starting points of two adjacent measurement units MU in any two time domains is is the above time domain sliding step, and the time domain starting points of MU1, MU2, MU3, MU4, and MU5 are all in the same frequency modulation period of the radar.
- MU2 spans three bands, of which the second band is complete, corresponding to the negative-slope beat signal; however, the first and third bands are incomplete, then, the tail of the first band can be A complete band is obtained by splicing with the beginning of the third band, and this complete band corresponds to the beat frequency signal with a positive slope.
- the time-domain sliding step size may not be a non-fixed value.
- the step size of the first sliding is different from that of the second sliding, and the step size of the second sliding Unlike the step of the third swipe.
- the general sliding step size can also be described as a sliding step size, which is used to indicate the length of the sliding.
- Step S1103 The signal processing apparatus determines a radar detection result according to each MU of the multiple MUs.
- the radar detection result may include one or more items of distance information and speed information, and of course may also include other information.
- the radar detection result is calculated from the frequency information in the MU, because the MU includes a positive slope part in the beat signal and a negative slope part of the beat signal, and the positive slope part in the beat signal and the negative slope of the beat signal
- the slope part is strongly correlated with the distance (the distance from the target to the radar) and the speed (the moving speed of the target). Therefore, information such as distance information and/or speed information can be obtained according to the MU calculation.
- each intercepted MU is matched with a reference spectrum with the highest similarity from the spectrum library, and then the detection information corresponding to the reference spectrum with the highest similarity of each MU is used as the radar detection result of each MU.
- the spectrum library records multiple reference spectrums, and each reference spectrum corresponds to detection information, and the detection information includes one or more items of distance information or speed information.
- the detection information corresponding to each reference spectrum in the spectrum library is pre-calculated and stored.
- the radar can simulate or experimentally obtain the spectrum under the condition of a specific Doppler spectrum (corresponding to a specific moving speed) and a specific flight time of the laser signal (corresponding to a specific distance), so different Doppler spectrum and flight time can be obtained.
- a specific Doppler spectrum corresponding to a specific moving speed
- a specific flight time of the laser signal corresponding to a specific distance
- different Doppler spectrum and flight time can be obtained.
- the first half of the spectrum corresponds to the positive slope part of the beat signal
- the second half of the spectrum corresponds to the negative slope of the beat signal
- the second half of the spectrum corresponds to the negative slope of the beat signal. It is usually a negative value. Therefore, after mirroring it as a positive value, the first half and the second half are positive values.
- the mirrored spectrum is the reference spectrum mentioned in the embodiment of this application.
- the Doppler spectrum (or specific velocity) and the specific flight time (or specific distance) are the detection information corresponding to the reference spectrum.
- FIG. 16 is
- step 1 a reference spectrum is generated according to a Doppler frequency offset and a signal time-of-flight.
- the Doppler frequency offset here corresponds to a moving speed (that is, the specific speed mentioned above), and the signal flight time here corresponds to a distance (that is, the specific distance mentioned above).
- Different Doppler frequency offsets and different The two-dimensional time spectrum of the corresponding beat signal after the combination is different. It is assumed that the sweep frequency curve of the radar laser is obtained, which is denoted as f est (t).
- the so-called frequency sweep curve is the curve of the frequency of the FMCW laser changing with time, as shown in part (a) of Figure 2.
- the way of obtaining can be through offline measurement or online measurement, which is not limited in this application.
- the two-dimensional time spectrum of the beat signal under different [Doppler, time of flight] combinations can be estimated according to the frequency sweep curve.
- the expression of its two-dimensional time spectrum can be estimated as:
- ⁇ is the flight time
- f d is the Doppler frequency offset
- f best (t, ⁇ , f d ) is the curve of the beat frequency over time
- [f d , ⁇ ] is determined
- f best (t, ⁇ ,f d ) is a definite two-dimensional time spectrum. Then, the above-mentioned reference spectrum can be obtained by mirroring the beat frequency of the negative slope part of the two-dimensional time spectrum.
- Step 2 use the item of velocity information and the item of distance information as the detection information corresponding to the one reference spectrum, and save them in the spectrum library together, wherein: the item of velocity information is the item of Doppler frequency offset, Or the moving speed of the reflector corresponding to the item of Doppler frequency offset; the item of distance information is the flight time of the item, or the distance of the reflector (also called the target) corresponding to the flight time of the signal .
- a spectrum library containing multiple reference spectrums can be obtained.
- the spectrum library can be updated, such as deleting or adding reference spectrums.
- f best (t, ⁇ ,f d )
- the measurement unit MU intercepted in step S1102 needs to be compared later.
- N chirp equal to N slot
- MU is recorded as a two-dimensional matrix T(l,k), 0 ⁇ l ⁇ N slot
- l Time sequence number
- N slot is the number of time slots (slots) of the beat frequency signal targeted for division
- 0 ⁇ k ⁇ N fft the frequency sequence number
- N fft is the number of FFT points.
- the time resolution of the recording T(l,k) is ⁇ t and the frequency resolution is ⁇ f.
- the so-called time resolution is the duration of each slot
- [ ⁇ est , f est ] argmax M rf ( ⁇ , f d )
- the intercepted detection results of each MU can be calculated.
- a frequency domain signal with a smaller granularity in the time domain is obtained, and then a sliding window is used to intercept the measurement unit MU on the frequency domain signal with a smaller granularity.
- Sliding window interception is performed on frequency-domain signals with smaller granularity, so even if some MUs that are intercepted have a common frequency band, due to the smaller granularity of frequency-domain signals, it is still possible to distinguish between different MUs. difference.
- the length of the sliding step time domain in the time domain is smaller than the frequency modulation period of the radar, two adjacent MUs in any two time domains will share a part of the frequency information, so even if more MUs are intercepted, each MU can still be guaranteed. There is enough signal energy accumulation inside to ensure the signal-to-noise ratio. Therefore, in this way, the point-out rate can be improved without losing the signal-to-noise ratio without significantly increasing the cost.
- FIG. 17 is a schematic structural diagram of a signal processing apparatus 170 provided by an embodiment of the present application.
- the apparatus may be the above-mentioned lidar system, or a processor in the lidar system, or the processor is deployed for Associated devices deployed in this lidar system.
- the signal processing apparatus 170 may include a converting unit 1701, an intercepting unit 1702 and an analyzing unit 1703, wherein the detailed description of each unit is as follows.
- a conversion unit 1701 for converting the radar beat signal into a two-dimensional time spectrum
- the intercepting unit 1702 is configured to intercept a plurality of measurement units MU whose time domain length is equal to the frequency modulation period of the radar from the two-dimensional time spectrum according to the time domain sliding step; FM cycle;
- the analyzing unit 1703 is configured to separately determine the radar detection result according to each MU in the multiple MUs.
- a frequency domain signal with a smaller granularity in the time domain is obtained, and then a sliding window is used to intercept the measurement unit MU on the frequency domain signal with a smaller granularity. It is performed on frequency-domain signals with smaller granularity, so even if some MUs that are cut out have a common frequency band, because the granularity of frequency-domain signals is small, the differences between different MUs can still be distinguished.
- the length of the sliding step time domain in the time domain is smaller than the frequency modulation period of the radar, two adjacent MUs in any two time domains will share a part of the frequency information, so even if more MUs are intercepted, each MU can still be guaranteed. There is enough signal energy accumulation inside to ensure the signal-to-noise ratio. Therefore, in this way, the point-out rate can be improved without losing the signal-to-noise ratio without significantly increasing the cost.
- the conversion unit 1701 is specifically configured to: segment the beat signal in the time domain to obtain multiple sub-signals, wherein the time domain length of each sub-signal is 1/N of the frequency modulation period of the radar, Wherein, N is greater than or equal to 2; Fourier transform is performed on the multi-segment sub-signals to obtain the frequency domain information of each sub-signal in the multi-segment sub-signals; two-dimensional information is generated according to the frequency domain information of each sub-signal. time spectrum.
- the intercepting unit 1702 is specifically configured to: match the reference spectrum with the highest similarity for each MU from the spectrum library, wherein the spectrum library records multiple reference spectrums, and each reference spectrum corresponds to Detection information, the detection information includes at least one item of distance information or speed information; the detection information corresponding to the reference spectrum with the highest similarity of each MU is used as the radar detection result of each MU.
- the spectrum library is used to match the reference spectrum that is closest to the MU and the detection information has been calculated, and then the detection information of the closest reference spectrum is used as the detection result of the MU, instead of being calculated in real time, which significantly reduces the detection rate. It improves the speed of determining the detection result of MU at a time; it effectively balances the calculation pressure caused by improving the point rate.
- a generating unit for generating a reference spectrum according to a Doppler frequency offset and a signal time-of-flight
- a storage unit configured to use an item of velocity information and an item of distance information as the detection information corresponding to the one reference spectrum, and save them in the spectrum library, wherein the item of velocity information is the item of Doppler frequency offset or the moving speed of the reflector corresponding to the item of Doppler frequency offset; the item of distance information is the time of flight of the signal or the distance of the reflector corresponding to the time of flight of the signal.
- the time-domain sliding step d is obtained by performing an integer operation on d0, and d0 satisfies the following relationship:
- T chirp is the frequency modulation period of the radar
- R is the desired increase multiplier of the radar point cloud density
- the point cloud density can be adjusted to the desired density by configuring R, which is flexible and easy to implement.
- each unit may also correspond to the corresponding description with reference to the method embodiment shown in FIG. 11 .
- the above units may be implemented in software, hardware, or a combination of the two.
- the hardware may be the aforementioned processor, and the software may include driver code running on the processor, which is not limited in this implementation.
- An embodiment of the present application further provides a chip system, where the chip system includes at least one processor, a memory, and an interface circuit.
- the memory, the interface circuit, and the at least one processor are interconnected through lines, and the at least one memory Instructions are stored in the ; when the instructions are executed by the processor, the method flow shown in FIG. 11 is implemented.
- Embodiments of the present application further provide a computer-readable storage medium, where instructions are stored in the computer-readable storage medium, and when the computer-readable storage medium runs on a processor, the method flow shown in FIG. 11 is implemented.
- the embodiment of the present application further provides a computer program product, which implements the method flow shown in FIG. 11 when the computer program product runs on the processor.
- a frequency domain signal with a smaller granularity in the time domain is obtained, and then a sliding window is used to intercept the measurement unit MU on the frequency domain signal with a smaller granularity. It is performed on frequency-domain signals with a smaller granularity, so even if some MUs that are cut out have a common frequency band, because the granularity of frequency-domain signals is small, the differences between different MUs can still be distinguished.
- the length of the sliding step time domain in the time domain is smaller than the frequency modulation period of the radar, two adjacent MUs in any two time domains will share a part of the frequency information, so even if more MUs are intercepted, each MU can still be guaranteed. There is enough signal energy accumulation inside to ensure the signal-to-noise ratio. Therefore, in this way, the out-point rate can be improved without losing the signal-to-noise ratio without significantly increasing the cost.
- the process can be completed by instructing the relevant hardware by a computer program, and the program can be stored in a computer-readable storage medium.
- the program When the program is executed , which may include the processes of the foregoing method embodiments.
- the aforementioned storage medium includes: ROM or random storage memory RAM, magnetic disk or optical disk and other mediums that can store program codes.
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Signal Processing (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
本申请实施例提供一种雷达探测方法及相关装置,该方法包括:将雷达的拍频信号变换为二维时频谱;按照时域滑动步长从所述二维时频谱中截取时域长度等于雷达的调频周期的多个测量单元MU;其中,所述时域滑动步长的长度小于雷达的调频周期;根据所述多个MU中的每个MU分别确定雷达探测结果。采用本申请实施例,能够在不明显提高成本的情况下,提高雷达出点率(即点云密度)且不丢失信噪比。
Description
本申请涉及激光雷达技术领域,尤其涉及一种雷达探测方法及相关装置。
调频连续波(frequency modulated continuous wave,FMCW)雷达是一种测距设备,FMCW雷达包含不同的细分种类,例如,采用无线电波的调频连续波雷达称为FMCW RADAR,再如,采用激光的调频连续波雷达称为FMCW LIDAR。不管哪种类型的FMCW雷达,其均包含图1所示结构,图1中,雷达产生出经过频率调制的射频或激光信号,将产生的调频信号分成两路,其中一路作为本地参考信号(也称本振信号),另一路出射至被测目标物(也称反射物)并被目标物表面反射,形成回波信号。
图2示意了FMCW雷达对参考信号和回波信号的处理过程,如图2中的(a)部分,粗线示意了发射信号和参考信号的调频信号的频率随时间发生变化,其中前一半时间,信号频率由低向高随时间递增,后一半时间,信号频率由高向低随时间递减。细线示意了上述回波信号。回波信号和参考信号经过混频器可输出拍频信号,拍频信号的频率是参考信号与回波信号的频率差,如图2的(b)部分所示。理想条件下,拍频信号有固定的频率(如图中的虚线内部分),如图2的(c)部分所示,通过对拍频信号进行频域分析(一般是做FFT),可以检测出拍频信号的频率,这里的频率与目标物的距离和速度有一一对应关系,所以根据拍频信号的频率,可以计算出目标物的速度与距离信息。
如图3所示,激光雷达中通常有扫描器,扫描器通过转动激光器向不同方向发射激光,然后将各个方向上测量的结果集中起来,就得到了3D点云图,图4示意了一种激光雷达的3D点云图。激光雷达的点云分辨率是一项重要的指标,高分辨率的点云对于自动驾驶等采用的机器视觉算法非常重要,对于点云分割,目标识别等功能的实现至关重要。目前,一个上下chirp对的时域长度内(即雷达的一个调频周期)只能完成一次测量,这种情况下得到的点云分辨率较低,目前常用的提高点云分辨率的方法是在激光雷达内增加更多的激光器,将32线LIDAR(即有32个激光器同时工作)升级为128线LIDAR(有128个激光器,或者用光学器件模拟出128个激光器)。
然而通过堆积硬件提高点云分辨率的方式,会显著增加激光雷达的成本、体积和功耗。
发明内容
本申请实施例公开了一种雷达探测方法及相关装置,能够在不明显提高成本的情况下,提高雷达出点率(即点云密度)且不丢失信噪比。
第一方面,本申请实施例公开了一种雷达探测方法,该方法包括:
将雷达的拍频信号变换为二维时频谱;按照时域滑动步长从所述二维时频谱中截取时域长度等于雷达的调频周期的多个测量单元MU;其中,所述时域滑动步长的长度小于雷达的调频周期;根据所述多个MU中的每个MU分别确定雷达探测结果。
在上述方法中,通过对拍频信号进行变换,得到时域上更小粒度的频域信号,然后在该更小粒度的频域信号上采用滑窗的方式截取测量单元MU,由于滑窗截取是在较小粒度的频域信号上进行的,因此即便截取出的某些MU存在共用的频段,但是由于频域信号的粒度较小,因此依旧可以分辨出不同的MU之间的差异。另外,由于时域上的滑动步进时域长度小于雷达的调频周期,因此任意两个时域上相邻的两个MU会共用一部分频率信息,因此即便截取较多MU依旧可以保证每个MU内有足够的信号能量积累,保证了信噪比。因此,采用这种方式可以在不明显提高成本的情况下,提高出点率且不丢失信噪比。
在第一方面的一种可选的方案中,所述将雷达的拍频信号变换为二维时频谱,包括:对所述拍频信号在时域上进行切分得到多段子信号,其中,每段子信号的时域长度为雷达的调频周期的1/N,其中,N大于或者等于2;
对所述多段子信号分别进行傅里叶变换得到所述多段子信号中每段子信号的频域信息;根据所述每段子信号的所述频域信息生成二维时频谱。所述二维时频谱可以是综合所述每段子信号以及所述频域信息所得到。
在第一方面的一种可选的方案中,所述根据所述多个MU中的每个MU分别确定雷达探测结果,包括:从频谱库中为所述每个MU分别匹配相似度最高的参考频谱,其中所述频谱库中记录有多个参考频谱,每个参考频谱对应探测信息,所述探测信息包括距离信息或速度信息中至少一项;将每个MU的相似度最高的参考频谱对应的探测信息作为所述每个MU的雷达探测结果。
上述方法中,通过频谱库来匹配与MU最接近且已计算出了探测信息的参考频谱,然后将该最接近的参考频谱的探测信息作为该MU的探测结果,而不是实时去计算,显著降低了单次确定MU的探测结果的速度;有效地平衡了因提升出点率而带来的计算压力。
在第一方面的一种可选的方案中,所述从频谱库中为所述每个MU分别匹配相似度最高的参考频谱之前,还包括:根据一项多普勒频偏和一项信号飞行时间生成一个参考频谱;将一项速度信息和一项距离信息作为所述一个参考频谱对应的探测信息,并保存到频谱库中,其中,所述一项速度信息为所述一项多普勒频偏或者所述一项多普勒频偏对应的反射物的移动速度;所述一项距离信息为所述一项信号飞行时间或者所述一项信号飞行时间对应的反射物的距离。
在第一方面的一种可选的方案中,所述时域滑动步长d是对d0进行取整运算得到,d0满足如下关系:
d0=T
chirp/R
其中,T
chirp为雷达的调频周期,R为期望的雷达点云密度提升倍数。
可以看出,采用配置R的方式可以将点云密度调整为期望的密度,灵活且易实现。
第二方面,本申请实施例提供一种信号处理装置,该装置包括:
转换单元,用于将雷达的拍频信号变换为二维时频谱;
截取单元,用于按照时域滑动步长从所述二维时频谱中截取时域长度等于雷达的调频周期的多个测量单元MU;其中,所述时域滑动步长的长度小于雷达的调频周期;
分析单元,用于根据所述多个MU中的每个MU分别确定雷达探测结果。
在上述方法中,通过对拍频信号进行变换,得到时域上更小粒度的频域信号,然后在该更小粒度的频域信号上采用滑窗的方式截取测量单元MU,由于滑窗截取是在较小粒度的频域信号上进行的,因此即便截取出的某些MU存在共用的频段,但是由于频域信号的粒度较小,因此依旧可以分辨出不同的MU之间的差异。另外,由于时域上的滑动步进时域长度小于雷达的调频周期,因此任意两个时域上相邻的两个MU会共用一部分频率信息,因此即便截取较多MU依旧可以保证每个MU内有足够的信号能量积累,保证了信噪比。因此,采用这种方式可以在不明显提高成本的情况下,提高出点率且不丢失信噪比。
在第二方面的一种可选的方案中,所述二维时频变换为短时傅里叶变换,所述转换单元具体用于:对雷达的拍频信号在时域上进行切分得到多段子信号,其中,每段子信号的时域长度为雷达的调频周期的1/N,其中,N大于或者等于2;对所述多段子信号分别进行傅里叶变换得到所述多段子信号中每段子信号的频域信息;根据所述每段子信号的所述频域信息生成二维时频谱。所述二维时频谱可以是综合所述每段子信号以及所述频域信息所得到。
在第二方面的一种可选的方案中,所述截取单元具体用于:从频谱库中为所述每个MU分别匹配相似度最高的参考频谱,其中所述频谱库中记录有多个参考频谱,每个参考频谱对应探测信息,所述探测信息包括距离信息或速度信息中至少一项;将每个MU的相似度最高的参考频谱对应的探测信息作为所述每个MU的雷达探测结果。
上述方法中,通过频谱库来匹配与MU最接近且已计算出了探测信息的参考频谱,然后将该最接近的参考频谱的探测信息作为该MU的探测结果,而不是实时去计算,显著降低了单次确定MU的探测结果的速度;有效地平衡了因提升出点率而带来的计算压力。
在第二方面的一种可选的方案中,还包括:
生成单元,用于根据一项多普勒频偏和一项信号飞行时间生成一个参考频谱;
保存单元,用于将一项速度信息和一项距离信息作为所述一个参考频谱对应的探测信息,并保存到频谱库中,其中,所述一项速度信息为所述一项多普勒频偏或者所述一项多普勒频偏对应的反射物的移动速度;所述一项距离信息为所述一项信号飞行时间或者所述一项信号飞行时间对应的反射物的距离。
在第二方面的一种可选的方案中,所述时域滑动步长d是对d0进行取整运算得到,d0满足如下关系:
d0=T
chirp/R
其中,T
chirp为雷达的调频周期,R为期望的雷达点云密度提升倍数。
可以看出,采用配置R的方式可以将点云密度调整为期望的密度,灵活且易实现。
第三方面,本申请实施例提供一种处理器,所述处理器用于调用计算机可读存储介质中存储的计算机程序,来实现第一方面或者第一方面的任一可选的实现方式所描述的方法。
第四方面,本申请实施例提供一种激光雷达***,所述激光雷达***包括处理器、存储器、激光器,其中,所述激光器用于发射激光信号,所述存储器用于存储计算机程序,所述处理器用于调用所述计算机程序来实现第一方面或者第一方面的任一可选的实现方式所描述的方法。
第五方面,本申请实施例提供一种计算机可读存储介质,所述计算机可读存储介质中 存储有计算机程序,当所述计算机程序在处理器上运行时,实现第一方面或者第一方面的任一可选的实现方式所描述的方法。
图1是本申请实施例提供的一种激光雷达的原理示意图;
图2是本申请实施例提供的一种拍频信号的示意图;
图3是本申请实施例提供的一种激光器扫描时的偏转示意图;
图4是本申请实施例提供的一种点云效果示意图;
图5是本申请实施例提供的一种激光雷达***的架构示意图;
图6是本申请实施例提供的一种三角波产生的拍频信号的示意图;
图7是本申请实施例提供的一种锯齿波产生的拍频信号的示意图;
图8是本申请实施例提供的一种拍频时间的划分示意图;
图9是本申请实施例提供的一种三角波线性调频相干信号处理的示意图;
图10是本申请实施例提供的一种拍频时间的划分示意图;
图11是本申请实施例提供的一种雷达探测方法的流程示意图;
图12是本申请实施例提供的一种拍频信号的示意图;
图13是本申请实施例提供的一种二维时频变化后的时频信号的示意图;
图14是本申请实施例提供的一种二维时频谱的示意图;
图15是本申请实施例提供的一种截取二维时频谱的场景示意图;
图16是本申请实施例提供的一种参考频谱的场景示意图;
图17是本申请实施例提供的一种信号处理装置的结构示意图。
下面结合本申请实施例中的附图对本申请实施例进行描述。
本申请实施例中的激光雷达能够应用于智能交通、自动驾驶、大气环境监测、地理测绘、无人机等各种领域,能够完成距离测量、速度测量、目标跟踪、成像识别等功能。
请参见图5,图5是本申请实施例提供的一种激光雷达***的结构示意图,该激光雷达***用于探测目标物505的信息,该激光雷达***包括:
激光器501,例如可以为调频激光器(Tunable Laser,TL),用于产生激光信号,该激光信号可以为线性调频激光信号,该激光信号频率的调制波形可以是锯齿波、或者三角波,或其他形式的波形。
分路器件502,用于对激光器501产生的激光进行分束,得到发射信号和本振信号(Local Oscillator,LO),其中,本振信号也称为参考信号。可选的,该激光器501与该分路器件之间还可以配置准直镜片500,该镜片500用于对输向分路器件502的激光信号进行光束整形。
准直器503,用于使发射信号最大效率的耦合进入扫描器504。
扫描器504,也称2D扫描机构,用于将发射信号按照一定的角度发射出去,发射信号 发射出去之后,被目标物505反射回来形成回波信号;这时,该扫描器504还用于接收该回波信号,回波信号经过相应光学器件(如反射镜506(可选)、接收镜片508(可选))之后,在混频器510与本振信号汇合。
混频器510,用于本振信号和回波信号进行混频处理,得到拍频信号。
探测器520,用于从混频器中提取出拍频信号。例如探测器520可以为平衡探测器(BalancedPhoto Detectors,BPD)。
模拟数字转换器(Analog digital converter,ADC)511,用于对拍频信号进行采样,这个采样实质是将模拟信号转换为数字信号的过程。
处理器512,该处理器可以包括数字信号处理器(Digital signal processor、DSP)、中央处理器(CPU)、加速处理单元(APU)、图像处理单元(GPU)、微处理器或微控制器等具有计算能力的器件,附图以DSP为例未介绍,该处理器用于对采样得到的拍频信号进行处理,从而得到目标物的速度、距离等信息。
本申请实施例中,目标物505也称为反射物,目标物505可以是扫描器504扫描方向上的任何物体,例如,可以是人、山、车辆、树木、桥梁等等,图5以车辆为例进行了示意。
本申请实施例中,对采样得到的拍频信号进行处理,从而得到目标物的速度、距离等信息的操作,可以由一个或多个处理器512,例如,由一个或多个DSP来完成,当然也可以由一个或多个处理器512结合其他器件来完成,例如,一个DSP结合一个或多个中央处理器CPU来共同完成。处理器512对拍频信号进行处理时,可以具体通过调用计算机可读存储介质存储的计算机程序来实现,该计算机可读存储介质包括但不限于是随机存储记忆体(random access memory,RAM)、只读存储器(read-only memory,ROM)、可擦除可编程只读存储器(erasable programmable read only memory,EPROM)、或便携式只读存储器(compact disc read-only memory,CD-ROM),其可以配置在处理器512中,也可以独立于处理器512。
本申请实施例中,上述提及的某些器件可以是单份,也可以是多份,例如,激光器501可以是一个,也可以是多个,当为一个激光器501时,这一个激光器501可以在时域上交替发射正斜率的激光信号和负斜率的激光信号;当存在两个激光器501时,其中一个发射正斜率的激光信号,另一个发射负斜率的激光信号,两个激光器501可以同步发射激光信号。
如图6所示,以该激光信号频率的调制波形为三角波线性调频为例,回波信号经过一段飞行时间之后与本振信号LO混频,这段飞行时间就是激光信号分出的发射信号从出射开始至回波信号返回的时间,回波信号经过飞行时间后与本振信号生成的拍频信号在一定时间内是恒定的,能准确反映目标物的距离和速度信息,这段时间即为拍频时间。拍频信号需要包括对应于正斜率的拍频f
1和对应于负斜率的拍频f
2,与目标物的速度相关的频谱f
速度可以表示为f
速度=(f
1-f
2)/2,与目标物的距离相关的频率f
距离可以表示为f
距离=(f
1+f
2)/2。得到f
速度和f
距离后就可以计算得到目标物(与激光雷达)的距离和目标物的移动速度。
相干激光雷达要测量更远距离的目标物的信息需要增加每个测量单元(Measurement Unit,MU)的拍频时间(一个测量单元MU用于得到一个测量点的探测结果),从而在处理拍频信号的过程中积累更多的能量,以达到更高的信噪比。而实际应用往往要求激光雷达具有更高的出点率,从而使激光雷达获得更高的视场分辨率和帧率。
常用的锯齿波线性调频相干信号处理方法如图7所示,将调制周期T减去最远探测距离对应的飞行时间后剩下的时间作为拍频时间TB,采集该拍频时间TB内的数据进行傅里叶变换求取实时拍频信号的频率,进而据此计算目标物的距离,此时,每一个测量点的测量时间TM≤TB。进一步若配合其他不同线性调频斜率的测量通道还能获得目标物的速度。
为提高激光雷达的出点率,可在拍单个频时间内划分多个测量单元MU,如图8所示,将拍频时间分为三段,得到三个MU(对应三个测量点),基于每个MU的信号分别可以完成一次探测,即可达到原出点率的三倍,但同时每个MU的拍频时间也变为原来的1/3。
常用的三角波线性调频相干信号处理方法如图9所示,在正斜率调制周期与负斜率调制周期内分别划分各自的飞行时间和拍频时间,各自拍频时间内的信号处理方式与锯齿波线性调频相干信号处理方法相同,测量单元MU的上半周期(也可描述为测量点对应的上半周期)处理得到拍波频率f
1,MU的下半周期(也可以描述为测量点对应的下半周期)处理得到拍波频率f
2,由此,通过一个测量通道即可获得目标的距离与速度信息。
为提高激光雷达的出点率,可以将正斜率周期和负斜率周期内的拍频时间划分为多个测量时间,如图10示意了将负斜率周期内的拍频时间T_τ划分为三个测量单元MU分别对应的三个测量时间T
0,基于其中每个测量单元MU分别可以完成一次探测,即可达到原出点率的三倍,但同时每个测量单元MU的拍频时间也变为原来的1/3。
可以看出,以上提高出点率的方式,减少了每个测量单元MU的拍频时间,降低了时频变换时有用信号的能量累积,从而影响了测量的信噪比,缩减了相干测量的探测距离。
为了在保证较高出点率和较高的信噪比,本申请实施例提供了图11所示的方法。
请参见图11,图11是本申请实施例提供的一种雷达探测方法,该方法可以基于图5所示的激光雷达***中的各部件来实现,后续的描述中一些操作是由信号处理装置来完成的,该信号处理装置可以为上述处理器512,或者部署了上述处理器512的装置,例如,部署了上述处理器5123的激光雷达***或者激光雷达***中的某个模块,该方法包括但不限于如下步骤:
步骤S1101:信号处理装置将雷达的拍频信号变换为二维时频谱。
也即是说,本申请并非仅获得拍频信号中的频域信息,还要获得拍频信号中的时域信息,因此变换之后可以获得从时域上看粒度更小的频域信息。变换可以具体为二维时频片换,所述二维时频变换所处理的每段信号的时域长度为雷达的调频周期(一般也可以称为激光信号的调频周期,或者称为激光器的调频周期,或者称为一个上下chirp(啁啾信号)对的时域长度)的1/N,N大于或等于2,即得到从时域上看粒度为雷达的调频周期的1/N的频域信息。
可选的,该二维时频变换可以具体为短时傅里叶变换,这种情况下,对雷达的拍频信号进行二维时频变换得到二维时频谱,可以具体包括:
首先,对上述拍频信号在时域上进行切分得到多段子信号,其中,每段子信号的时域 长度为雷达的调频周期的1/N;图12示意了一段拍频信号,其中,横轴表示拍频信号的时域信息,纵轴表示拍频信号的幅度信息,图12中的这一段拍频信号的时域长度是6微秒(us),图12中的竖线就示意了对该6微秒的拍频信号的切分情况。
举例来说,假若经过ADC采样得到的拍频信号序列有20000个采样点,计为s(n),0≤n≤20000,那么可以对s(n)进行切分,每200个采样点切分为1组,那么可以切分出100组,即100段子信号,每一组所占的时间我们可以称为一个时隙(slot),因此时隙总数N
slot=100;切分得到的第l段拍频信号s
l(n)的表达式可以表示为:
s
l(n)=s[(l-1)·200+n],0≤n<200,0≤l<100
其中,l是分段的序号,n是采样点的序号。
然后,对所述多段子信号分别进行快速傅里叶变换(fast Fourier transform,FFT)得到所述多段子信号中每段子信号的频域信息;如图13所示,示意了其中一段子信号的信息,其他子信号的信息可以参照该子信号的示意。例如,假若按照前面得到了100段子信号的信息,那么就需要进行100次的FFT计算,从而得到100段频域信息。每段频域信息对应的采样点数是N
fft=200。
可选的,进行FFT计算时FFT的表达式可以为:
其中,F[]表示FFT运算,k是频域序列的序号(代表不同的频率)。
接着,根据所述多段子信号中每段子信号的所述频域信息生成二维时频谱。
具体地,由于对每段子信号是按照时域进行切分得到的,因此每段子信号对应了一个时域信息,又因为通过傅里叶变换得到了每段子信号中的频域信息,因此,到本步骤已经得到了每段子信号的时域信息和频域信息,因此可以通过相应规则来存储每段子信号的这些信息,例如,通过规格为N
fft·N
slot的表格T来存储,那么这个表格T就可以认为是雷达的拍频信号的二维时频谱,图14示意了一种二维时频谱,其中,其横轴表示时域信息,纵轴表示频域信息,由于附图无法示意亮度差异,因此在实验中通过亮度示意频率幅度的方式,从附图14无法看出,但是实验模拟及实际应用中该二维时频谱中可以通过亮度来表示幅度信息的。因此,所述二维时频谱可以是综合所述每段子信号以及所述频域信息所得到。
需要说明的是,拍频信号中的正协率部分经过二维时频变换之后往往在图14所述的横轴的上方,拍频信号中的负协率部分经过二维时频变换之后往往在图14所述的横轴的下,但是为了操作方便,通常会将横轴下方的波段镜像到横轴上方,所以图14最终看到的效果是所有波段都在横轴上方,图14中的第二段波形就是镜像到横轴上方的。另外,图14中示意的波段看上去只有两段,但是实际上可以有更多段,例如,4段、6段、等等。
步骤S1102:信号处理装置按照时域滑动步长从所述二维时频谱中截取时域长度等于雷达的调频周期的多个测量单元MU。
其中,该时域滑动步长是根据需要预先设置的,如果希望雷达的点云密度高,即出点率高,那么可以将该时域滑动步长设置得小一些,例如,可以将所述时域滑动步长的长度设置为小于雷达的调频周期。
现有技术中平均一个雷达的调频周期内出一个测量点,那么如果希望点云密度提升R倍,则需要平均在一个雷达的调频周期内出R个测量点,因此可以引入R来计算时域滑动步长,R可以解释为:R为期望的雷达点云密度提升倍数,或解释为:雷达的一个调频周期内需要完成测量的次数。引入R之后,时域滑动步长d0的计算如下:
案例:先通过公式d=T
chirp/R计算初始值d,再对d取整得到d0,其中,T
chirp的单位可以是秒、毫秒、微秒、时隙等等。可以理解,T
chirp的单位为时隙时,T
chirp与上述N
slot相等。另外,对d取整时,采用的取整算法可以是向下取整,或者向上取整,或者四舍五入取整,或者其他方式取整,此处不做限制。
确定时域滑动步长之后,就从前面的得到的二维时频谱中截取测量单元(Measurement Unit,MU),可选的,每个测量单元MU的时域长度等于雷达的调频周期;截取的方式就是按照时域滑动步长截取,第k个测量单元(Measurement Unit,MU)截取出的时频范围是第k-1个MU截取范围的滑动,如图15所示,从左往右依次截取(时间往右增大),截取第一个测量单元MU1之后,时域上往右滑动时域滑动步长,然后进行第二次截取,得到第二个测量单元MU2,后续依次类推,如图15,R等于5,最终截取到了5个测量单元MU,依次为MU1、MU2、MU3、MU4、MU5,任意两个时域上相邻的两个测量单元MU的时域起点的差值即为上述时域滑动步长,且MU1、MU2、MU3、MU4、MU5的时域起点都在雷达的同一个调频周期。
在图15所示的案例中,有些波段虽然不完整,但是经过拼接之后是可以获得一个完整对应于正斜率的拍频信号的波段和一个完整的对应于负斜率的拍频信号的波段,例如,MU2横跨了三个波段,其中第二个波段是完整的,对应负斜率的拍频信号;但是,第一个波段和第三个波段不完整,那么,可以将第一个波段的尾与第三个波段的首进行拼接得到一个完整的波段,这个完整波段就对应正斜率的拍频信号。
可选的,时域滑动步长除了可以为固定值外,还可以不为非固定值,例如,第一次滑动的步长与第二次滑动的步长不同,第二次滑动的步进与第三次滑动的步进不同。一般滑动步长也可以描述为滑动步长,用于表示滑动的长度。
可以理解,按照时域滑动步长进行截取的话,通常可以截取到多个测量单元MU。
步骤S1103:信号处理装置根据所述多个MU中每个MU分别确定雷达探测结果。
具体地,所述雷达探测结果可以包括距离信息和速度信息中的一项或多项,当然还可以包括其他信息。
本申请实施例中,根据截取到的测量单元MU确定雷达探测结果的方式有很多。
例如,根据MU中的频率信息计算雷达探测结果,因为MU包括拍频信号中的正斜率部分和该拍频信号的负斜率部分,而拍频信号中的正斜率部分和该拍频信号的负斜率部分与距离(目标物到雷达的距离)、速度(目标物的移动速度)之间强相关,因此,可以根据MU计算得到距离信息和/速度信息等信息。
再如,从频谱库中为截取到的每个MU分别匹配相似度最高的参考频谱,然后将每个MU的相似度最高的参考频谱对应的探测信息作为所述每个MU的雷达探测结果。其中所述频谱库中记录有多个参考频谱,每条参考频谱对应探测信息,所述探测信息包括距离信 息或速度信息中的一项或者多项。该频谱库中的每个参考频谱对应的探测信息是预先计算并存储好的。
雷达可以在特定多普勒频谱(对应特定移动速度),以及激光信号的特定飞行时间(对应特定距离)的情况下模拟或实验得到频谱,因此不同的多普勒频谱和飞行时间可以得到的多个频谱,以其中一个频谱为例,该频谱中的前半段为对应于拍频信号中的正斜率部分,该频谱中的后半段对应于拍频信号中的负斜率部分,其中后半段通常为负值,因此将其镜像为正值后,前半段和后半段都为正值,经过镜像处理后的频谱就是本申请实施例中所提及的参考频谱,而前面用到的特定多普勒频谱(或特地速度)和特定飞行时间(或特定距离)就是该参考频谱对应的探测信息。图16示意了9个频谱经过镜像操作后得到的参考频谱的示意图。
可选的,下面例举一种可选的生成频谱库的方法:
步骤1,根据一项多普勒频偏和一项信号飞行时间生成一个参考频谱。
这里的多普勒频偏对应了一个运动速度(也就是前面说的特定速度),这里的信号飞行时间对应了一个距离(也就是前面说的特定距离),不同的多普勒频偏和不同的信号飞行时间,其结合后对应的拍频信号的二维时频谱是不同的,假设得到了雷达激光器的扫频曲线,记为f
est(t)。所谓扫频曲线,就是FMCW激光器的频率随时间变换的曲线,如图2的(a)部分所示。得到的方式可以是通过离线测量或者是在线测量,本申请不做限制。
那么可以根据扫频曲线,估计出不同的[多普勒,飞行时间]组合下,拍频信号的二维时频谱。例如针对特定的多普勒频谱和飞行时间[f
d,τ],可以估计出其二维时频谱的表达式为:
f
best(t,τ,f
d)=f
est(t)–[f
est(t-τ)+f
d]
其中,这里τ是飞行时间,f
d是多普勒频偏,f
best(t,τ,f
d)是拍频随时间的曲线,在[f
d,τ]确定了的情况下,f
best(t,τ,f
d)就是一段确定的二维时频谱。然后对二维时频谱中负斜率部分的拍频做镜像处理,就可以得到上述参考频谱。
步骤2,将该一项速度信息和该一项距离信息作为该一个参考频谱对应的探测信息,并一起保存到频谱库中,其中:该一项速度信息为该一项多普勒频偏,或为该一项多普勒频偏对应的反射物的移动速度;该一项距离信息为该一项飞行时间,或为该一项信号飞行时间对应的反射物(也称目标物)的距离。
针对多对多普勒频偏和信号飞行时间完成上述步骤1、步骤2后就可以得到一个包含多个参考频谱的所述频谱库。可选的,该频谱库可以进行更新,例如删除或增加参考频谱。
可选的,下面举例一种可选的匹配相似度最高的参考频谱的方法。
对于频谱库中的多个参考频谱,即基于多普勒频谱和飞行时间预估的一些参考频谱,记录为f
best(t,τ,f
d),后续需要比较步骤S1102中截取的测量单元MU与频谱库中哪个参考频谱最为相似,其中目标测量单元MU可以表示为:以N
chirp等于N
slot为例,MU记为二维矩阵T(l,k),0≤l<N
slot,l是时间序号,N
slot是切分时所针对的拍频信号的时隙(slot)数量,0≤k<N
fft,k是频率序号,N
fft是FFT的点数。
记录T(l,k)的时间分辨率是Δt,频率分辨率是Δf。所谓的时间分辨率,就是每个slot的时长,所谓的频率分辨率,就是FFT的频率间隔,有如下的关系,Δf=1/Δt。
那么,对于T(l,k)的l,其对应时间长度为l·Δt。
接着,对于不同的拍频曲线f
best(t,τ,f
d),计算每个l对应的频率,即f
best(l·Δt,τ,f
d)。
然后,将f
best(l·Δt,τ,f
d)折算为频率分辨率Δf的整数倍,即:
k
l,τ,fd=round(f
best(l·Δt,τ,fd)/Δf)
此后,将频谱库中每段参考频谱代入到如下公式得到匹配度值(即相似度)M
rf(τ,f
d):
其中哪个参考频谱带入到公式后计算得到的M
rf(τ,f
d)最大,则认为该参考频谱与上述目标MU的相似度最高,因此将该参考频谱中的[f
d,τ]作为上述目标MU的雷达探测结果,公式表示如下:
[τ
est,f
est]=argmax M
rf(τ,f
d)
基于目标MU的雷达探测结果的计算原理,可以计算出截取到的每个MU的探测结果。
图11所示的方法中,通过对拍频信号进行变换,得到时域上更小粒度的频域信号,然后在该更小粒度的频域信号上采用滑窗的方式截取测量单元MU,由于滑窗截取是在较小粒度的频域信号上进行的,因此即便截取出的某些MU存在共用的频段,但是由于频域信号的粒度较小,因此依旧可以分辨出不同的MU之间的差异。另外,由于时域上的滑动步进时域长度小于雷达的调频周期,因此任意两个时域上相邻的两个MU会共用一部分频率信息,因此即便截取较多MU依旧可以保证每个MU内有足够的信号能量积累,保证了信噪比。因此,采用这种方式可以在不明显提高成本的情况下,提高出点率且不丢失信噪比。
上述详细阐述了本申请实施例的方法,下面提供了本申请实施例的装置。
请参见图17,图17是本申请实施例提供的一种信号处理装置170的结构示意图,该装置可以为上述激光雷达***,或者该激光雷达***中的处理器,或者部署了该处理器用于部署在该激光雷达***中的相关器件。该信号处理装置170可以包括转换单元1701、截取单元1702和分析单元1703,其中,各个单元的详细描述如下。
转换单元1701,用于将雷达的拍频信号变换为二维时频谱;
截取单元1702,用于按照时域滑动步长从所述二维时频谱中截取时域长度等于雷达的调频周期的多个测量单元MU;其中,所述时域滑动步长的长度小于雷达的调频周期;
分析单元1703,用于根据所述多个MU中每个MU分别确定雷达探测结果。
在上述方法中,通过对拍频信号进行变换,得到时域上更小粒度的频域信号,然后在该更小粒度的频域信号上采用滑窗的方式截取测量单元MU,由于滑窗截取是在较小粒度的频域信号上进行的,因此即便截取出的某些MU存在共用的频段,但是由于频域信号的粒度较小,因此依旧可以分辨出不同的MU之间的差异。另外,由于时域上的滑动步进时域长度小于雷达的调频周期,因此任意两个时域上相邻的两个MU会共用一部分频率信息,因此即便截取较多MU依旧可以保证每个MU内有足够的信号能量积累,保证了信噪比。因此,采用这种方式可以在不明显提高成本的情况下,提高出点率且不丢失信噪比。
可选的,所述转换单元1701具体用于:对所述拍频信号在时域上进行切分得到多段子 信号,其中,每段子信号的时域长度为雷达的调频周期的1/N,其中,N大于或者等于2;对所述多段子信号分别进行傅里叶变换得到所述多段子信号中每段子信号的频域信息;根据所述每段子信号的所述频域信息生成二维时频谱。
可选的,所述截取单元1702具体用于:从频谱库中为所述每个MU分别匹配相似度最高的参考频谱,其中所述频谱库中记录有多个参考频谱,每个参考频谱对应探测信息,所述探测信息包括距离信息或速度信息中至少一项;将每个MU的相似度最高的参考频谱对应的探测信息作为所述每个MU的雷达探测结果。
上述方法中,通过频谱库来匹配与MU最接近且已计算出了探测信息的参考频谱,然后将该最接近的参考频谱的探测信息作为该MU的探测结果,而不是实时去计算,显著降低了单次确定MU的探测结果的速度;有效地平衡了因提升出点率而带来的计算压力。
可选的,还包括:
生成单元,用于根据一项多普勒频偏和一项信号飞行时间生成一个参考频谱;
保存单元,用于将一项速度信息和一项距离信息作为所述一个参考频谱对应的探测信息,并保存到频谱库中,其中,所述一项速度信息为所述一项多普勒频偏或者所述一项多普勒频偏对应的反射物的移动速度;所述一项距离信息为所述一项信号飞行时间或者所述一项信号飞行时间对应的反射物的距离。
可选的,所述时域滑动步长d是对d0进行取整运算得到,d0满足如下关系:
d0=T
chirp/R
其中,T
chirp为雷达的调频周期,R为期望的雷达点云密度提升倍数。
可以看出,采用配置R的方式可以将点云密度调整为期望的密度,灵活且易实现。
需要说明的是,各个单元的实现还可以对应参照图11所示的方法实施例的相应描述。以上各个单元可以软件、硬件或二者结合方式实现,硬件可以是之前所述处理器,软件可以包括运行于处理器上的驱动程序代码,本实施对此不限定。
本申请实施例还提供一种芯片***,所述芯片***包括至少一个处理器,存储器和接口电路,所述存储器、所述接口电路和所述至少一个处理器通过线路互联,所述至少一个存储器中存储有指令;所述指令被所述处理器执行时,实现图11所示的方法流程。
本申请实施例还提供一种计算机可读存储介质,所述计算机可读存储介质中存储有指令,当其在处理器上运行时,实现图11所示的方法流程。
本申请实施例还提供一种计算机程序产品,当所述计算机程序产品在处理器上运行时,实现图11所示的方法流程。
综上所述,通过对拍频信号进行变换,得到时域上更小粒度的频域信号,然后在该更小粒度的频域信号上采用滑窗的方式截取测量单元MU,由于滑窗截取是在较小粒度的频域信号上进行的,因此即便截取出的某些MU存在共用的频段,但是由于频域信号的粒度较小,因此依旧可以分辨出不同的MU之间的差异。另外,由于时域上的滑动步进时域长度小于雷达的调频周期,因此任意两个时域上相邻的两个MU会共用一部分频率信息,因此即便截取较多MU依旧可以保证每个MU内有足够的信号能量积累,保证了信噪比。因此,采用这种方式可以在不明显提高成本的情况下,提高出点率且不丢失信噪比。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,该流程可以 由计算机程序来指令相关的硬件完成,该程序可存储于计算机可读取存储介质中,该程序在执行时,可包括如上述各方法实施例的流程。而前述的存储介质包括:ROM或随机存储记忆体RAM、磁碟或者光盘等各种可存储程序代码的介质。
Claims (13)
- 一种雷达探测方法,其特征在于,包括:将雷达的拍频信号变换为二维时频谱;按照时域滑动步长从所述二维时频谱中截取时域长度等于雷达的调频周期的多个测量单元MU;其中,所述时域滑动步长的长度小于雷达的调频周期;根据所述多个MU中的每个MU分别确定雷达探测结果。
- 根据权利要求1所述的方法,其特征在于,所述将雷达的拍频信号变换为二维时频谱,包括:对所述拍频信号在时域上进行切分得到多段子信号,其中,每段子信号的时域长度为雷达的调频周期的1/N,其中,N大于或者等于2;对所述多段子信号分别进行傅里叶变换得到所述多段子信号中每段子信号的频域信息;根据所述每段子信号的所述频域信息生成二维时频谱。
- 根据权利要求1或2所述的方法,其特征在于,所述根据所述多个MU中的每个MU分别确定雷达探测结果,包括:从频谱库中为所述每个MU分别匹配相似度最高的参考频谱,其中所述频谱库中记录有多个参考频谱,每个参考频谱对应探测信息,所述探测信息包括距离信息或速度信息中至少一项;将每个MU的相似度最高的参考频谱对应的探测信息作为所述每个MU的雷达探测结果。
- 根据权利要求3所述的方法,其特征在于,所述从频谱库中为所述每个MU分别匹配相似度最高的参考频谱之前,还包括:根据一项多普勒频偏和一项信号飞行时间生成一个参考频谱;将一项速度信息和一项距离信息作为所述一个参考频谱对应的探测信息,并保存到频谱库中,其中,所述一项速度信息为所述一项多普勒频偏或者所述一项多普勒频偏对应的反射物的移动速度;所述一项距离信息为所述一项信号飞行时间或者所述一项信号飞行时间对应的反射物的距离。
- 根据权利要求1-4任一项所述的方法,其特征在于,所述时域滑动步长d是对d0进行取整运算得到,d0满足如下关系:d0=T chirp/R其中,T chirp为雷达的调频周期,R为期望的雷达点云密度提升倍数。
- 一种信号处理装置,其特征在于,包括:转换单元,用于将雷达的拍频信号变换为二维时频谱;截取单元,用于按照时域滑动步长从所述二维时频谱中截取时域长度等于雷达的调频周期的多个测量单元MU;其中,所述时域滑动步长的长度小于雷达的调频周期;分析单元,用于根据所述多个MU中的每个MU分别确定雷达探测结果。
- 根据权利要求6所述的装置,其特征在于,所述转换单元具体用于:对所述拍频信号在时域上进行切分得到多段子信号,其中,每段子信号的时域长度为雷达的调频周期的1/N,其中,N大于或者等于2;对所述多段子信号分别进行傅里叶变换得到所述多段子信号中每段子信号的频域信息;根据所述每段子信号的所述频域信息生成二维时频谱。
- 根据权利要求6或7所述的装置,其特征在于,所述截取单元具体用于:从频谱库中为所述每个MU分别匹配相似度最高的参考频谱,其中所述频谱库中记录有多个参考频谱,每个参考频谱对应探测信息,所述探测信息包括距离信息或速度信息中至少一项;将每个MU的相似度最高的参考频谱对应的探测信息作为所述每个MU的雷达探测结果。
- 根据权利要求8所述的装置,其特征在于,还包括:生成单元,用于根据一项多普勒频偏和一项信号飞行时间生成一个参考频谱;保存单元,用于将一项速度信息和一项距离信息作为所述一个参考频谱对应的探测信息,并保存到频谱库中,其中,所述一项速度信息为所述一项多普勒频偏或者所述一项多普勒频偏对应的反射物的移动速度;所述一项距离信息为所述一项信号飞行时间或者所述一项信号飞行时间对应的反射物的距离。
- 根据权利要求6-9任一项所述的装置,其特征在于,所述时域滑动步长d是对d0进行取整运算得到,d0满足如下关系:d0=T chirp/R其中,T chirp为雷达的调频周期,R为期望的雷达点云密度提升倍数。
- 一种处理器,其特征在于,所述处理器用于调用计算机可读存储介质中存储的计算机程序,来实现权利要求1-5任一项所述的方法。
- 一种激光雷达***,其特征在于,所述激光雷达***包括处理器、存储器、激光器,其中,所述激光器用于发射激光信号,所述存储器用于存储计算机程序,所述处理器用于调用所述计算机程序来执行权利要求1-5任一项所述的方法。
- 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质中存储有计算机程序,当所述计算机程序在处理器上运行时,实现权利要求1-5任一所述的方法。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP20943763.1A EP4166982A4 (en) | 2020-06-30 | 2020-06-30 | RADAR DETECTION METHOD AND ASSOCIATED DEVICE |
CN202080004802.3A CN112639523B (zh) | 2020-06-30 | 2020-06-30 | 一种雷达探测方法及相关装置 |
PCT/CN2020/099558 WO2022000333A1 (zh) | 2020-06-30 | 2020-06-30 | 一种雷达探测方法及相关装置 |
CN202210436709.0A CN114966661A (zh) | 2020-06-30 | 2020-06-30 | 一种雷达探测方法及相关装置 |
US18/148,625 US20230140139A1 (en) | 2020-06-30 | 2022-12-30 | Radar Detection Method and Related Apparatus |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2020/099558 WO2022000333A1 (zh) | 2020-06-30 | 2020-06-30 | 一种雷达探测方法及相关装置 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US18/148,625 Continuation US20230140139A1 (en) | 2020-06-30 | 2022-12-30 | Radar Detection Method and Related Apparatus |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022000333A1 true WO2022000333A1 (zh) | 2022-01-06 |
Family
ID=75291246
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2020/099558 WO2022000333A1 (zh) | 2020-06-30 | 2020-06-30 | 一种雷达探测方法及相关装置 |
Country Status (4)
Country | Link |
---|---|
US (1) | US20230140139A1 (zh) |
EP (1) | EP4166982A4 (zh) |
CN (2) | CN112639523B (zh) |
WO (1) | WO2022000333A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115061115A (zh) * | 2022-07-26 | 2022-09-16 | 深圳市速腾聚创科技有限公司 | 点云的加密方法、装置、存储介质及激光雷达 |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117872313A (zh) * | 2021-12-29 | 2024-04-12 | 华为技术有限公司 | 一种激光雷达的发送装置、探测***以及探测方法 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008152373A1 (en) * | 2007-06-11 | 2008-12-18 | Mitsubishi Electric Information Technology Centre Europe B.V. | Object detection |
CN106291524A (zh) * | 2016-08-05 | 2017-01-04 | 南京理工大学 | 基于人体模型的lfmcw雷达探测运动人体目标的方法 |
CN107783099A (zh) * | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | 基于组合波形的旋翼无人机短距离防撞***信号处理***及方法 |
CN108387892A (zh) * | 2018-01-16 | 2018-08-10 | 上海交通大学 | 基于fmcw雷达的结构健康监测***与方法 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4376390B2 (ja) * | 1999-12-16 | 2009-12-02 | 富士通株式会社 | Fm−cwレーダの周波数変調特性測定方法及び装置 |
CN107169435B (zh) * | 2017-05-10 | 2021-07-20 | 天津大学 | 一种基于雷达仿真图像的卷积神经网络人体动作分类方法 |
US11187807B2 (en) * | 2017-07-24 | 2021-11-30 | Intel Corporation | Precisely controlled chirped diode laser and coherent lidar system |
CN107270952B (zh) * | 2017-07-27 | 2020-03-31 | 天津求实飞博科技有限公司 | 基于光频域反射长距离光纤分布式扰动传感信号处理方法 |
-
2020
- 2020-06-30 CN CN202080004802.3A patent/CN112639523B/zh active Active
- 2020-06-30 CN CN202210436709.0A patent/CN114966661A/zh active Pending
- 2020-06-30 EP EP20943763.1A patent/EP4166982A4/en active Pending
- 2020-06-30 WO PCT/CN2020/099558 patent/WO2022000333A1/zh unknown
-
2022
- 2022-12-30 US US18/148,625 patent/US20230140139A1/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008152373A1 (en) * | 2007-06-11 | 2008-12-18 | Mitsubishi Electric Information Technology Centre Europe B.V. | Object detection |
CN106291524A (zh) * | 2016-08-05 | 2017-01-04 | 南京理工大学 | 基于人体模型的lfmcw雷达探测运动人体目标的方法 |
CN107783099A (zh) * | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | 基于组合波形的旋翼无人机短距离防撞***信号处理***及方法 |
CN108387892A (zh) * | 2018-01-16 | 2018-08-10 | 上海交通大学 | 基于fmcw雷达的结构健康监测***与方法 |
Non-Patent Citations (1)
Title |
---|
See also references of EP4166982A4 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115061115A (zh) * | 2022-07-26 | 2022-09-16 | 深圳市速腾聚创科技有限公司 | 点云的加密方法、装置、存储介质及激光雷达 |
CN115061115B (zh) * | 2022-07-26 | 2023-02-03 | 深圳市速腾聚创科技有限公司 | 点云的加密方法、装置、存储介质及激光雷达 |
Also Published As
Publication number | Publication date |
---|---|
CN112639523B (zh) | 2022-04-29 |
CN112639523A (zh) | 2021-04-09 |
EP4166982A4 (en) | 2023-08-02 |
CN114966661A (zh) | 2022-08-30 |
EP4166982A1 (en) | 2023-04-19 |
US20230140139A1 (en) | 2023-05-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106062583B (zh) | 单个激光lidar*** | |
CN108415010B (zh) | 一种基于梯形lfmcw调制的雷达多目标检测方法 | |
US20230140139A1 (en) | Radar Detection Method and Related Apparatus | |
JP2019537013A (ja) | 光学測距システムによる自動リアルタイム適応走査の方法およびシステム | |
JP2022130554A (ja) | コヒーレントlidarのピッチ-キャッチスキャニングのための方法およびシステム | |
US9465108B1 (en) | System and method for target doppler estimation and range bias compensation using high duty cycle linear frequency modulated signals | |
JP2016029369A (ja) | タイミング同期を備えた周波数変調連続波(fmcw)レーダー | |
CN105738889B (zh) | 一种调频连续波测速测距方法 | |
CN104898119B (zh) | 一种基于相关函数的动目标参数估计方法 | |
WO2022061828A1 (zh) | 一种雷达探测方法及相关装置 | |
JP6179940B2 (ja) | ドップラーイメージング信号送信装置、ドップラーイメージング信号受信装置、ドップラーイメージングシステム及び方法 | |
US20230168381A1 (en) | Radar Detection Method and Related Apparatus | |
EP3418770A1 (en) | Synthetic-aperture-radar signal processing device | |
KR102146156B1 (ko) | 오브젝트 검출 방법 및 장치 | |
KR20190135267A (ko) | Cw 레이더 및 cw 레이더를 이용한 거리 측정 방법 | |
TWI693419B (zh) | 訊號處理方法 | |
CN115267721B (zh) | 一种基于双频sar的地面动目标径向速度估计方法 | |
KR20200053222A (ko) | 거리, 각도 속도를 정밀 측정하기 위한 레이더 장치 및 레이더 신호 처리 방법 | |
CN112099037B (zh) | 一种基于fmcw激光雷达高点云获取量的目标检测方法及装置 | |
KR20230081654A (ko) | 레이더 측정 방법 | |
CN112285730B (zh) | 基于轨道角动量调制的多维信息探测*** | |
CN112099038B (zh) | 一种基于fmcw激光雷达的多物体识别方法及装置 | |
CN110726995A (zh) | 激光雷达高精度测距方法及*** | |
KR20150135734A (ko) | 선형 주파수 변조 신호와 잡음 신호를 이용한 레이더 및 이의 제어 방법 | |
RU2542347C1 (ru) | Способ адаптивной настройки каналов ускорения в многоканальном обнаружителе маневрирующей цели |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20943763 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2020943763 Country of ref document: EP Effective date: 20230110 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |