WO2022217406A1 - 信号处理方法、装置及可读存储介质 - Google Patents

信号处理方法、装置及可读存储介质 Download PDF

Info

Publication number
WO2022217406A1
WO2022217406A1 PCT/CN2021/086497 CN2021086497W WO2022217406A1 WO 2022217406 A1 WO2022217406 A1 WO 2022217406A1 CN 2021086497 W CN2021086497 W CN 2021086497W WO 2022217406 A1 WO2022217406 A1 WO 2022217406A1
Authority
WO
WIPO (PCT)
Prior art keywords
frequency band
signal
pass
filtered
frequency
Prior art date
Application number
PCT/CN2021/086497
Other languages
English (en)
French (fr)
Inventor
牛犇
汪敬
朱琳
Original Assignee
深圳市速腾聚创科技有限公司
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 深圳市速腾聚创科技有限公司 filed Critical 深圳市速腾聚创科技有限公司
Priority to CN202180095632.9A priority Critical patent/CN117480403A/zh
Priority to PCT/CN2021/086497 priority patent/WO2022217406A1/zh
Publication of WO2022217406A1 publication Critical patent/WO2022217406A1/zh

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
    • 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

Definitions

  • the present application relates to the technical field of laser detection, and in particular, to a signal processing method, device and readable storage medium.
  • Lidar is a radar system that emits a laser beam to detect the position, velocity and other characteristic quantities of the target. Its working principle is to transmit a detection signal to the target, and then compare the received signal reflected from the target with the transmitted signal. After proper processing, the relevant information of the target can be obtained, such as target distance, azimuth, altitude, speed, Attitude, even shape and other parameters, so as to detect, track and identify targets such as aircraft and missiles.
  • the principle of ranging is to transmit a continuous wave whose frequency changes linearly during the frequency sweep period as the outgoing signal, a part of the outgoing signal is used as the local oscillator signal, and the rest is outgoing for detection, and the returned echo after being reflected by the object There is a certain frequency difference between the wave signal and the local oscillator signal, and the distance information between the detected target and the radar can be obtained by measuring the frequency difference.
  • FMCW Frequency-Modulated Continuous Wave
  • Embodiments of the present application provide a signal processing method, device, and readable storage medium, which can improve distance detection accuracy in a full frequency range.
  • an embodiment of the present application provides a signal processing method, including:
  • N-level decomposition on the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • the frequency bands to be filtered are M frequency bands Any one of the frequency bands, the M frequency bands cover the entire frequency range, and M ⁇ 2;
  • the next frequency band to be filtered is determined as the frequency band to be filtered according to the preset order, and the determining the number of target component layers according to the frequency band to be filtered is performed, and the target component layers are determined according to the frequency band to be filtered.
  • an embodiment of the present application provides a signal processing method, including:
  • N-level decomposition on the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • the filtered signal is output.
  • an embodiment of the present application provides a signal processing apparatus, including:
  • the first decomposition module is used for N-level decomposition of the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • a first denoising module configured to determine the number of target component layers according to the frequency band to be filtered, and perform wavelet threshold denoising on the components in the target component layer number located in the frequency band to be filtered to obtain a processed filtered signal;
  • the frequency band to be filtered is any one of the M frequency bands, and the M frequency bands cover the entire frequency range, and M ⁇ 2;
  • a first output module configured to output the filtered signal when the filtered signal satisfies a preset condition
  • a first determining module configured to determine the next frequency band to be filtered as the frequency band to be filtered according to a preset order when the filtered signal does not meet the preset condition, and call the first denoising module Determine the number of target component layers according to the frequency band to be filtered, perform wavelet threshold denoising on the components located in the frequency band to be filtered in the target component layers, and obtain the processed filtered signal until the last frequency band in the M frequency bands If the corresponding filtered signal does not meet the preset condition, the first output module is called to output the filtered signal corresponding to the last frequency band.
  • an embodiment of the present application provides a signal processing apparatus, including:
  • the second decomposition module is used to perform N-level decomposition on the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • the second denoising module is configured to determine the number of target component layers according to the frequency band to be filtered, and perform wavelet threshold denoising on the components in the target component layer number located in the frequency band to be filtered to obtain a processed filtered signal;
  • the second output module is used for outputting the filtered signal.
  • an embodiment of the present application provides a signal processing apparatus, including: a processor and a memory; wherein, the memory stores a computer program, and the computer program is adapted to be loaded by the processor and execute the first embodiment of the present application.
  • an embodiment of the present application provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the method provided in the first aspect or the second aspect of the embodiment of the present application.
  • the embodiments of the present application provide several new filtering algorithms, which realize the denoising function in different frequency bands by increasing the precision of wavelet coefficient decomposition and performing threshold processing on components of different frequency bands.
  • FIG. 1 is a schematic diagram of the distribution of a vehicle-mounted lidar according to an embodiment of the present application
  • FIG. 2 is a schematic diagram of the architecture of a signal processing system provided by an embodiment of the present application.
  • FIG. 3 is a schematic diagram of the frequency of the transmission and echo signals of the triangular wave-modulated FMCW system provided by an embodiment of the present application changing with time;
  • Figure 4a is a schematic diagram of the decomposition method of the classical wavelet threshold denoising algorithm
  • Figure 4b is a schematic diagram of the simulation result of the classical wavelet threshold denoising algorithm when the distance is 100 meters;
  • Figure 4c is a schematic diagram of the simulation result of the classical wavelet threshold denoising algorithm when the distance is 200 meters;
  • Figure 4d is a schematic diagram of the simulation result of the classical wavelet threshold denoising algorithm when the distance is 300 meters;
  • 4e is a schematic diagram of a decomposition method of a wavelet threshold denoising algorithm provided by an embodiment of the application;
  • 4f is a schematic diagram of a simulation result of the low-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 100 meters;
  • 4g is a schematic diagram of a simulation result of the low-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 200 meters;
  • 4h is a schematic diagram of a simulation result of the low-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 300 meters;
  • 4i is a schematic diagram of a simulation result of the bandpass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 100 meters;
  • 4j is a schematic diagram of a simulation result of the bandpass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 200 meters;
  • 4k is a schematic diagram of a simulation result of the bandpass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 300 meters;
  • 41 is a schematic diagram of a simulation result of the high-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is measured by 100 meters;
  • 4m is a schematic diagram of a simulation result of the high-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 200 meters;
  • FIG. 4n is a schematic diagram of a simulation result of the high-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the application when the distance is 300 meters;
  • FIG. 5 is a schematic flowchart of a signal processing method provided by an embodiment of the present application.
  • 6a is a schematic diagram of a full frequency range division provided by an embodiment of the present application.
  • FIG. 6b is a schematic diagram of another full frequency range division provided by an embodiment of the present application.
  • FIG. 6c is a schematic diagram of the detection capability at different distances of a signal processing method according to an embodiment of the present application.
  • FIG. 7a is a schematic diagram of another full frequency range division provided by an embodiment of the present application.
  • FIG. 7b is a schematic diagram of another full frequency range division provided by an embodiment of the present application.
  • FIG. 7c is a schematic diagram of the detection capability at different distances of another signal processing method according to an embodiment of the present application.
  • FIG. 8a is a schematic diagram of another full frequency range division provided by an embodiment of the present application.
  • FIG. 8b is a schematic diagram of another full frequency range division provided by an embodiment of the present application.
  • FIG. 8c is a schematic diagram of the detection capability at different distances of another signal processing method according to an embodiment of the present application.
  • FIG. 9 is a schematic flowchart of another signal processing method provided by an embodiment of the present application.
  • FIG. 10 is a schematic diagram of detection capability at different distances of a signal processing method provided by an embodiment of the present application.
  • FIG. 11 is a schematic diagram of the detection capability at different distances of another signal processing method provided by an embodiment of the present application.
  • FIG. 12 is a schematic diagram of the detection capability at different distances of another signal processing method provided by an embodiment of the present application.
  • FIG. 13 is a schematic structural diagram of a signal processing apparatus provided by an embodiment of the present application.
  • FIG. 14 is a schematic structural diagram of another signal processing apparatus provided by an embodiment of the present application.
  • FIG. 15 is a schematic structural diagram of another signal processing apparatus provided by an embodiment of the present application.
  • FIG. 16 is a schematic structural diagram of another signal processing apparatus provided by an embodiment of the present application.
  • FIG. 1 exemplarily shows a schematic diagram of the distribution of a vehicle-mounted lidar.
  • vehicle-mounted lidars 110a, 110b, 110c, and 110d
  • the lidar can also be distributed in any position on the left, right, front, rear, top, etc. of the vehicle according to the detection requirements.
  • the laser radar in the embodiment of the present application is an FMCW laser radar.
  • one of the lidars 110a is used as an example for description.
  • the lidar 110a can collect echo signals within its radiation range, obtain a denoised beat frequency signal after calculation according to the transmit signal and the echo signal, and send the denoised beat frequency signal to the vehicle terminal 120 .
  • the signal processing device determines a beat frequency signal with noise according to the transmitted signal and the received signal; and performs signal processing on the beat frequency signal to obtain a beat frequency signal after denoising.
  • the in-vehicle terminal 120 can process the denoised difference frequency signal obtained by the lidar 110a to obtain point cloud data, identify the type and location of obstacles around the vehicle and the movement trajectory of the obstacle according to the point cloud data, and combine its own operation According to the situation, make path planning and generate driving strategy.
  • FIG. 2 exemplarily shows a schematic diagram of the architecture of a signal processing system provided by an embodiment of the present application.
  • the FMCW lidar can send out a transmit signal, and the transmit signal hits the target object at a distance in space and then reflects and returns an echo signal, and the echo signal is received by the FMCW radar system.
  • the difference frequency signal is generated after the echo signal is mixed with the transmitted signal.
  • the signal processing device of the FMCW lidar can use the wavelet threshold denoising algorithm to denoise the beat frequency signal, and output the denoised beat frequency signal to the vehicle terminal.
  • the signal processing device may be disposed in the FMCW laser radar, may be independent of the FMCW laser radar, or may be disposed in the vehicle-mounted terminal 120, which is not limited in this embodiment of the present application.
  • the signal processing device is set in the FMCW laser radar as an example for description, and the FMCW laser radar uses a wavelet threshold denoising algorithm to perform denoising processing on the beat frequency signal.
  • the FMCW lidar provided in the embodiments of the present application can also be applied to devices such as robots and drones. That is to say, the difference frequency signal after denoising of the FMCW lidar output is not limited to being sent to the vehicle terminal. In other application scenarios, it can also be sent to the CPU in the robot or the UAV, etc. Not limited.
  • FIG. 3 is a schematic diagram showing the frequency of the transmitted signal and the echo signal of the triangular wave modulated FMCW lidar with time. As shown in Figure 3:
  • the difference frequency signal is shown in the following formulas (1) and (2) respectively:
  • f 0 is the initial frequency of each cycle of the transmitted signal
  • T is the signal period
  • B is the signal bandwidth
  • A is the amplitude of the difference frequency signal
  • c is the speed of light
  • t is the time variable.
  • the absolute value of the frequency of the beat frequency signal has a linear positive correlation with the distance of the detected target, that is to say, the farther the target is, the higher the frequency of the beat frequency signal, and vice versa.
  • the decomposition method of the classical wavelet region denoising algorithm is introduced.
  • the core of the classical wavelet threshold denoising algorithm is unilateral wavelet decomposition.
  • the specific process is as follows:
  • each level of decomposition keeps the detail part unchanged, and only decomposes the approximate part until the Nth level decomposition.
  • the specific number of decomposition layers can be selected according to actual needs.
  • the wavelet threshold denoising algorithm After the wavelet decomposition is completed, the wavelet threshold denoising algorithm performs threshold processing on the obtained wavelet components according to the threshold rules.
  • the echo signal is in the middle and high frequency.
  • the low-pass filtering property of the classical wavelet threshold denoising causes the signal in the middle and high frequency to be weakened or filtered, and the algorithm fails. That is to say, if the echo signal is in the middle and high frequency band, the classical wavelet algorithm not only cannot achieve the effect of denoising, but will weaken or filter the echo signal as noise, resulting in detection failure. Therefore, the classical wavelet algorithm is effective for the echo signal detected at a short distance, but in the face of a long distance, the detection ability will deteriorate and the detection will fail, which seriously limits the application of the actual FMCW lidar.
  • the minimum filtering range of the low-pass filtering of the classical wavelet threshold denoising algorithm is determined when the number of decomposition layers is 1, and the filtering range cannot be further reduced. That is to say, when the sampling frequency of the lidar is constant, the classical wavelet threshold denoising algorithm has a maximum effective detection distance, which cannot be adjusted to make the maximum effective detection distance larger, which seriously affects the actual FMCW lidar application.
  • Figures 4b-4d exemplarily show the schematic diagrams of the simulation results of the classical wavelet threshold denoising algorithm.
  • Figures 4b-4d show the simulation results when the range is 100m, 200m and 300m, respectively.
  • Each picture has three sub-pictures: the upper sub-picture shows the spectrum of the noise-free signal, the middle sub-picture is used to display the spectrum of the noisy signal (signal-to-noise ratio is -15dB), and the lower sub-picture shows the classical decomposition level of 1.
  • the spectrum of the signal after wavelet denoising It can be seen that the classical wavelet threshold denoising algorithm has obvious low-pass filtering effects. At a distance of 100 meters, the signal frequency is outside the filtered range.
  • the signal strength is weakened by the wavelet threshold denoising algorithm.
  • the signal strength is filtered out by the wavelet threshold denoising algorithm.
  • the embodiments of the present application provide several novel wavelet threshold denoising algorithms, which can improve the detection performance at medium and long distances compared with the classical wavelet threshold denoising algorithms.
  • FIG. 4e exemplarily shows a schematic diagram of a decomposition manner of a wavelet threshold denoising algorithm provided by an embodiment of the present application.
  • the specific process of wavelet decomposition of the signal by the wavelet threshold denoising algorithm is as follows:
  • the first-order wavelet coefficient decomposition is performed on the signal with noise to obtain the cA low-frequency part and the cD high-frequency part.
  • the embodiment of the present application provides three wavelet threshold denoising algorithms, which respectively implement low-pass filtering, band-pass filtering, and high-pass filtering. Next, the three wavelet threshold denoising algorithms are introduced respectively.
  • Low-pass filtering wavelet threshold denoising algorithm Specifically, the high-frequency components in the multiple components obtained after wavelet coefficient decomposition can be subjected to threshold processing, and other components remain unchanged.
  • the second-level decomposition as an example, four components can be obtained after the second-level decomposition: cAA, cAD, cDA, and cDD, and only the obtained high-frequency part cDD is subjected to threshold processing. Other parts remain unchanged.
  • FIGS. 4f to 4h exemplarily show schematic diagrams of simulation results of the low-pass filtering wavelet threshold denoising algorithm provided by the embodiments of the present application.
  • Fig. 4f-Fig. 4h show the simulation results when the range is 100m, 200m and 300m, respectively.
  • Each picture has three sub-pictures, upper, middle, and lower: the upper sub-picture shows the spectrum of the noise-free signal, the middle sub-picture is used to display the spectrum of the noisy signal (the signal-to-noise ratio is -15dB), and the lower sub-picture shows the spectrum of the signal provided by the embodiment of the present application.
  • the spectrum of the signal after denoising by the low-pass filtering wavelet threshold denoising algorithm Compared with FIG. 4b to FIG.
  • the frequency range of the low-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the present application is significantly reduced by the low-pass filtering. Specifically, it can be seen from the fact that the signal strength is not weakened when the distance is 200 meters.
  • Band-pass filtering wavelet threshold denoising algorithm Specifically, the low-frequency components and high-frequency components of the multiple components obtained after the wavelet coefficients are decomposed can be subjected to threshold processing, and other components remain unchanged.
  • the second-level decomposition as an example, four components can be obtained after the second-level decomposition: cAA, cAD, cDA, and cDD, and only the low-frequency part cAA and the high-frequency part cDD are thresholded. Other parts remain unchanged.
  • FIGS. 4i-4k exemplarily show schematic diagrams of simulation results of the bandpass filtering wavelet threshold denoising algorithm provided by the embodiments of the present application.
  • Figures 4i-4k show the simulation results when the range is 100m, 200m and 300m, respectively.
  • Each picture has three sub-pictures, upper, middle, and lower: the upper sub-picture shows the spectrum of the noise-free signal, the middle sub-picture is used to display the spectrum of the noisy signal (the signal-to-noise ratio is -15dB), and the lower sub-picture shows the spectrum of the signal provided by the embodiment of the present application.
  • the spectrum of the signal after denoising by the bandpass filtering wavelet threshold denoising algorithm Comparing with FIGS.
  • the bandpass filtering wavelet threshold denoising algorithm provided by the embodiment of the present application has obvious bandpass filtering properties. Specifically, it can be seen that the signal strength is not weakened when the ranging is 200 meters, and the signal strength is weakened when the ranging is 100 meters and 300 meters.
  • High-pass filtering wavelet threshold denoising algorithm Specifically, the low-frequency components in the multiple components obtained after the wavelet coefficients are decomposed can be subjected to threshold processing, and other components remain unchanged.
  • the second-level decomposition As an example, four components can be obtained after the second-level decomposition: cAA, cAD, cDA, and cDD, and only the obtained low-frequency part cAA is subjected to threshold processing. Other parts remain unchanged.
  • FIGS. 41-4n exemplarily show schematic diagrams of simulation results of the high-pass filtering wavelet threshold denoising algorithm provided by the embodiments of the present application.
  • Fig. 4l-Fig. 4n show the simulation results when the range is 100m, 200m and 300m, respectively.
  • Each picture has three sub-pictures, upper, middle, and lower: the upper sub-picture shows the spectrum of the noise-free signal, the middle sub-picture is used to display the spectrum of the noisy signal (the signal-to-noise ratio is -15dB), and the lower sub-picture shows the spectrum of the signal provided by the embodiment of the present application.
  • the spectrum of the signal after denoising by the high-pass filtering wavelet threshold denoising algorithm Comparing with FIGS.
  • the high-pass filtering wavelet threshold denoising algorithm provided by the embodiment of the present application has obvious properties of high-pass filtering. Specifically, it can be seen that the signal strength is not weakened when the ranging is 300 meters, and the signal strength is weakened when the ranging is 100 meters and 200 meters.
  • the above-mentioned threshold processing is specifically: setting the wavelet coefficients of components whose amplitudes are smaller than the preset threshold to zero.
  • the preset threshold is a preset critical value. Specifically, if the amplitude of a certain component is smaller than the preset threshold, it is determined that the component is mainly caused by noise, and the wavelet coefficient of the component is set to zero, thereby removing the noise.
  • FIG. 5 exemplarily shows a schematic flowchart of a signal processing method provided by an embodiment of the present application.
  • the signal processing method may include the following steps:
  • S501 Perform N-level decomposition on the signal to be processed to obtain 2 N components.
  • the signal to be processed is a signal with noise.
  • the decomposition manner shown in FIG. 4e For the manner of performing N-level decomposition on the signal to be processed, reference may be made to the decomposition manner shown in FIG. 4e, and details are not described herein again. Each decomposed component corresponds to a different frequency range.
  • the method may further include: acquiring the signal to be processed.
  • the signal to be processed may be the beat frequency signal mentioned in the embodiment of FIG. 2 .
  • S502 Determine the number of target component layers according to the frequency band to be filtered, and perform wavelet threshold denoising on the components in the target component layer number whose components are located in the frequency band to be filtered, to obtain a processed filtered signal.
  • the to-be-filtered frequency band is any one of M frequency bands, and the M frequency bands cover the entire frequency band range, and M ⁇ 2.
  • the M frequency bands are consecutively connected. That is, adjacent frequency bands among the M frequency bands do not overlap.
  • adjacent frequency bands among the M frequency bands partially overlap.
  • the signal to be processed is the difference frequency signal mentioned in the embodiment of FIG. 2 .
  • the N-level decomposition of the difference frequency signal 2 N components can be obtained, and each component corresponds to a frequency range.
  • the full frequency range mentioned in the embodiment of the present application is the difference corresponding to the ranging range of the lidar. frequency range of the frequency signal.
  • the ranging distance of the lidar is 300 meters
  • the frequency of the beat frequency signal is about 400 MHz.
  • the ranging range of this lidar is defined as 0-300 meters, its full frequency range is 0-400MHz.
  • wavelet threshold denoising may include, but are not limited to, the following: modulus maximum denoising, correlation denoising, wavelet shrinkage threshold denoising and translation invariant wavelet denoising, etc.
  • S503 Determine whether the frequency band to be filtered is the last frequency band in the M frequency bands; if so, execute S506; if not, execute S504 or S505.
  • the M frequency bands may be divided into at least two frequency bands.
  • the signal processing method provided by the embodiment of the present application may perform wavelet threshold denoising processing on the above at least two frequency bands according to a preset sequence.
  • the last frequency band in the M frequency bands is the only frequency band in the M frequency bands that has not been subjected to wavelet threshold denoising, that is to say, other frequency bands in the M frequency bands have been subjected to wavelet threshold denoising. Noise processed.
  • the signal to be processed may include a noise signal and a beat frequency signal. If the noise signal is completely or partially filtered in S502, the processed filtered signal may include a beat frequency signal.
  • the method further includes: extracting the difference frequency signal in the filtered signal.
  • the preset condition is that the beat frequency signal is successfully extracted.
  • the signal to be processed may include a noise signal and a beat frequency signal. If the noise signal is completely or partially filtered in S502, the processed filtered signal may include a beat frequency signal.
  • the beat frequency signal will be filtered out after the corresponding component is subjected to the wavelet threshold denoising process, then the beat frequency signal extraction fails; if the frequency of the beat frequency signal is not If it is located in the above frequency band to be filtered, after the wavelet threshold denoising process, the noise is filtered or partially filtered, and the difference frequency signal will be easily extracted.
  • a specific manner of extracting the difference frequency signal from the filtered signal may be to perform a Fast Fourier Transform (Fast Fourier Transform, FFT) on the filtered signal.
  • FFT Fast Fourier Transform
  • the filtered signal does not meet the preset conditions, it means that both the noise signal and the beat frequency signal are filtered out, which means that filtering the frequency band to be filtered cannot achieve only filtering the noise signal without weakening the beat frequency signal.
  • the to-be-processed signal is filtered for the next frequency band.
  • the above-mentioned full frequency range into M frequency bands which may be 2 frequency bands, 3 frequency bands or even more, and the division is selected according to the hardware computing capability of the signal processing, the precision requirements of the signal processing, etc.
  • the number of frequency bands, and each division method can correspond to a variety of preset sequences.
  • the preset sequence is relatively simple; when the whole frequency range is divided into 3 or more frequency bands, the preset sequence is relatively more complicated.
  • the whole frequency range is divided into 3 frequency bands as an example for description; when the whole frequency range is divided into more than 3 frequency bands, the processing logic of dividing the whole frequency range into 3 frequency bands is the same.
  • the M frequency bands may be a low-pass frequency band (which may be referred to as a third low-pass frequency band in this embodiment of the present application), a band-pass frequency band, and a high-pass frequency band (which may be referred to as a third high-pass frequency band in this embodiment of the present application) ).
  • the third low-pass frequency band is a frequency band whose frequency is less than the fifth threshold
  • the band-pass frequency band is a frequency band whose frequency is greater than the sixth threshold and less than the seventh threshold
  • the third high-pass frequency band is a frequency band whose frequency is greater than the eighth threshold.
  • the fifth threshold is greater than or equal to the sixth threshold
  • the seventh threshold is greater than or equal to the eighth threshold.
  • Fig. 6a exemplarily shows a schematic diagram of decomposing the full frequency range into 3 frequency bands.
  • the full frequency range can be divided into a low-pass frequency band, a band-pass frequency band, and a high-pass frequency band. Adjacent frequency bands are connected consecutively. That is, the fifth threshold is equal to the sixth threshold, and the seventh threshold is equal to the eighth threshold.
  • the bandwidth of each band can be the same or different.
  • Fig. 6b exemplarily shows another schematic diagram of decomposing the full frequency range into 3 frequency bands.
  • the full frequency range can be divided into a low-pass frequency band, a band-pass frequency band and a high-pass frequency band. Adjacent frequency bands partially overlap. That is, the fifth threshold is greater than the sixth threshold, and the seventh threshold is greater than the eighth threshold.
  • the bandwidth of each band can be the same or different.
  • This decomposition method can correspond to 6 preset sequences:
  • the first type the third low-pass frequency band, the band-pass frequency band, and the third high-pass frequency band.
  • the second type the third high-pass frequency band, the band-pass frequency band, and the third low-pass frequency band.
  • the third type the band-pass frequency band, the third low-pass frequency band, and the third high-pass frequency band.
  • the fourth type the third low-pass frequency band, the third high-pass frequency band, and the band-pass frequency band.
  • the sixth type the band-pass frequency band, the third high-pass frequency band, and the third low-pass frequency band.
  • the embodiment of the present application may first determine the third low-pass frequency band as the frequency band to be filtered, determine the number of target component layers according to the third low-pass frequency band, and determine the number of target component layers located in the to-be-filtered frequency band.
  • the components are subjected to wavelet threshold denoising to obtain the processed filtered signal; if the difference frequency signal can be extracted from the filtered signal, the filtered signal is output.
  • the band-pass frequency band is determined as the frequency band to be filtered, the number of target component layers is determined according to the band-pass frequency band, and the wavelet threshold is performed on the components in the target component layer number located in the frequency band to be filtered. noise to obtain a processed filtered signal; if the difference frequency signal can be extracted from the filtered signal, the filtered signal is output.
  • the third high-pass frequency band is determined as the frequency band to be filtered, the number of target component layers is determined according to the third high-pass frequency band, and wavelet is performed on the component of the target component layer number located in the to-be-filtered frequency band Threshold denoising to obtain a processed filtered signal.
  • the filtered signal is output.
  • the wavelet threshold denoising is performed on the components located in the low-pass frequency band in the target component layer number, that is, the low-pass frequency band part of the beat frequency signal remains unchanged, and the rest of the beat frequency signal is noise filtered, which can be achieved. Filter covering the low-pass frequency band; perform wavelet threshold denoising on the components located in the band-pass frequency band in the target component layers, that is, the band-pass frequency band part of the beat frequency signal remains unchanged, and the rest of the beat frequency signal is filtered.
  • the filtering of the band-pass frequency band can be realized; the wavelet threshold denoising is performed on the components located in the high-pass frequency band in the target component layers, that is, the high-pass frequency band part of the beat frequency signal remains unchanged, and the rest of the beat frequency signal is filtered. Filtering covering the high-pass frequency band can be achieved.
  • the wavelet threshold denoising is performed on the components corresponding to different frequency bands, which can improve the detection probability corresponding to different detection distances.
  • the detection probability in the entire detection range can be improved, thereby improving the effective detection range of FMCW lidar and promoting the application of FMCW lidar.
  • Fig. 6c exemplarily shows a schematic diagram of the detection capability of a signal processing method corresponding to a decomposition manner at different distances.
  • the signal processing method corresponding to the decomposition method 1 has the detection capability at short distances, medium distances and long distances.
  • wavelet threshold denoising processing on the components corresponding to the low-pass frequency range of a signal can realize low-pass filtering, thereby realizing short-range detection.
  • wavelet threshold denoising is performed on the components corresponding to the low-pass frequency range and the high-pass frequency range of a signal, which can realize band-pass filtering, and then realize medium-distance detection.
  • wavelet threshold denoising processing on the components corresponding to the high-pass frequency range of a signal can realize high-pass filtering, thereby realizing long-distance detection.
  • the M frequency bands may be a low-pass frequency band (which may be referred to as a first low-pass frequency band in this embodiment of the present application) and a high-pass frequency band (which may be referred to as a first high-pass frequency band in this embodiment of the present application).
  • the first low-pass frequency band is a frequency band with a frequency lower than the first threshold
  • the first high-pass frequency band is a frequency band with a frequency greater than the second threshold.
  • the first threshold is greater than or equal to the second threshold.
  • the bandwidth (ie, the range length) of the first low-pass frequency band and the bandwidth of the first high-pass frequency band can be set according to the needs of signal processing. For example, the bandwidth of the first low-pass frequency band is larger than the bandwidth of the first high-pass frequency band.
  • FIG. 7a exemplarily shows a schematic diagram of decomposing the full frequency range into 2 frequency bands.
  • the full frequency range can be divided into a low-pass frequency band and a high-pass frequency band. Adjacent frequency bands are connected consecutively. That is, the first threshold is equal to the second threshold.
  • FIG. 7b exemplarily shows another schematic diagram of decomposing the full frequency range into 2 frequency bands.
  • the full frequency range can be divided into a low-pass frequency band and a high-pass frequency band. Adjacent frequency bands partially overlap. That is, the first threshold is greater than the second threshold.
  • This decomposition method can correspond to 2 preset sequences:
  • the first type the first low-pass frequency band and the first high-pass frequency band.
  • the second type the first high-pass frequency band and the first low-pass frequency band.
  • Fig. 7c exemplarily shows a diagram of the detection capability of the signal processing method corresponding to the second decomposition method at different distances.
  • the signal processing method corresponding to the decomposition method 2 has the detection capability at short distances, medium distances and long distances. Since in the embodiment of the present application, the bandwidth of the low-pass frequency band is relatively large, and the bandwidth of the high-pass frequency band is relatively small, the wavelet threshold denoising processing of the low-pass frequency band range of a signal can realize high-pass filtering, thereby realizing long-distance detection.
  • the wavelet threshold denoising process is performed on the components in the high-pass frequency range of a signal, which can realize low-pass filtering, and then realize short-distance and medium-distance detection.
  • the M frequency bands may be a low-pass frequency band (which may be referred to as a second low-pass frequency band in this embodiment of the present application) and a high-pass frequency band (which may be referred to as a second high-pass frequency band in this embodiment of the present application).
  • the second low-pass frequency band is a frequency band whose frequency is less than the third threshold
  • the second high-pass frequency band is a frequency band whose frequency is greater than the fourth threshold.
  • the third threshold is greater than or equal to the fourth threshold.
  • the bandwidth (ie, the range length) of the second low-pass frequency band is greater than the bandwidth of the second high-pass frequency band.
  • Fig. 8a exemplarily shows a schematic diagram of decomposing a full frequency range into 2 frequency bands.
  • the full frequency range can be divided into a low-pass frequency band and a high-pass frequency band. Adjacent frequency bands are connected consecutively. That is, the third threshold is equal to the fourth threshold.
  • FIG. 8b exemplarily shows another schematic diagram of decomposing the full frequency range into 2 frequency bands.
  • the full frequency range can be divided into a low-pass frequency band and a high-pass frequency band. Adjacent frequency bands partially overlap. That is, the third threshold is greater than the fourth threshold.
  • This decomposition method can correspond to 2 preset sequences:
  • the first type the second low-pass frequency band and the second high-pass frequency band.
  • the second type the second high-pass frequency band and the second low-pass frequency band.
  • Fig. 8c exemplarily shows a diagram of the detection capability of the signal processing method corresponding to the third decomposition method at different distances.
  • the signal processing method corresponding to the decomposition method 3 has the detection capability at short distances, medium distances and long distances. Because in the embodiment of the present application, the bandwidth of the low-pass frequency band is relatively small, and the bandwidth of the high-pass frequency band is relatively large, and the wavelet threshold denoising processing of the low-pass frequency band of a signal can realize high-pass filtering, thereby realizing long-distance and medium-distance detection. . Performing wavelet threshold denoising on the components in the high-pass frequency range of a signal can realize low-pass filtering, thereby realizing short-range detection.
  • the frequency band bandwidth corresponding to each component becomes narrower.
  • the M frequency bands obtained by the decomposition of a certain feature may be located in a certain component in different component layers respectively.
  • the wide-band low-pass frequency band can correspond to the component corresponding to the low-pass frequency band in the second component layer number (as shown in Figure 4e after the third-level decomposition is obtained).
  • the narrow-band high-pass frequency point can correspond to the component corresponding to the high-pass frequency band in the second component layer number (as shown in FIG. 4e, the cAA component obtained after three-level decomposition).
  • FIG. 9 shows a schematic flowchart of another signal processing method provided by an embodiment of the present application. As shown in Figure 9, the signal processing method may include the following steps:
  • S901 Perform N-level decomposition on the signal to be processed to obtain 2 N components.
  • S901 is the same as S501, and details are not repeated here.
  • S902 Determine the number of target component layers according to the frequency band to be filtered, and perform wavelet threshold denoising on the components in the target component layer number whose components are located in the frequency band to be filtered, to obtain a processed filtered signal.
  • the frequency band to be filtered may be a low-pass frequency band, specifically a frequency band whose frequency is less than the ninth threshold, for implementing low-pass filtering.
  • FIG. 10 exemplarily shows a graph of the detection capability of the embodiment of the present application at different distances after threshold processing is performed on components in the low-pass frequency range.
  • wavelet threshold denoising is performed on the components in the low-pass frequency range, that is, the low-frequency part of the beat frequency signal remains unchanged, and the remaining high-frequency parts of the beat frequency signal are filtered out. Therefore, the embodiment of the present application has an excellent detection capability at a short distance.
  • the frequency band to be filtered may be a high-pass frequency band, specifically a frequency band whose frequency is greater than the tenth threshold, for implementing high-pass filtering.
  • FIG. 11 exemplarily shows a graph of the detection capability of the embodiment of the present application at different distances after threshold processing is performed on the components of the high-pass frequency band range.
  • the wavelet threshold denoising is performed on the components in the high-pass frequency range, that is, the high-frequency part of the beat-frequency signal remains unchanged, and the remaining low-frequency parts of the beat-frequency signal are filtered out. After processing, high-pass filtering can be realized.
  • the embodiments of the present application have excellent detection capability at long distances.
  • the frequency band to be filtered may be a low-pass frequency band and a high-pass frequency band, specifically a frequency band with a frequency less than the eleventh threshold and a frequency band with a frequency greater than the twelfth threshold, for implementing band-pass filtering.
  • the eleventh threshold is smaller than the twelfth threshold.
  • FIG. 12 exemplarily shows a diagram of the detection capability of the embodiment of the present application at different distances after threshold processing is performed on the components in the band-pass frequency range.
  • wavelet threshold denoising is performed on the components in the band-pass frequency range, that is, the middle frequency band of the beat frequency signal remains unchanged, and the remaining low frequency and high frequency parts of the beat frequency signal are filtered out.
  • Bandpass filtering is implemented, and the embodiments of the present application have excellent detection capability at medium distances.
  • the short distance, the medium distance and the long distance mentioned in the embodiments of the present application are the distances between the obstacle and the lidar.
  • the distance between the obstacle and the lidar determines the frequency of the beat signal.
  • the short distance, the middle distance and the long distance are relative concepts, and the specific numerical values of the three are not limited in the embodiments of the present application.
  • the signal to be processed may include a noise signal and a beat frequency signal. If the frequency of the beat frequency signal is not in the above-mentioned frequency band to be filtered, after the wavelet threshold denoising process, the noise is filtered or partially filtered, and the beat frequency signal can be extracted.
  • the embodiments of the present application provide several new filtering algorithms. By increasing the accuracy of wavelet coefficient decomposition and performing threshold processing on components of different frequency bands, denoising functions in different frequency bands can be realized, and the filtering algorithms provided by the embodiments of the present application can be expanded. applicable scenarios.
  • FIG. 13 shows a schematic structural diagram of a signal processing apparatus provided by an exemplary embodiment of the present application.
  • the signal processing device can be implemented by software, hardware or a combination of both.
  • the signal processing apparatus 130 includes: a first decomposition module 1310 , a first denoising module 1320 , a first output module 1330 and a first determination module 1340 . in:
  • the first decomposition module 1310 is configured to perform N-level decomposition on the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • the first denoising module 1320 is configured to determine the number of target component layers according to the frequency band to be filtered, and perform wavelet threshold denoising on the components located in the frequency band to be filtered in the target component layer number to obtain a processed filtered signal; wherein, The frequency band to be filtered is any one of the M frequency bands, and the M frequency bands cover the entire frequency range, and M ⁇ 2;
  • a first output module 1330 configured to output the filtered signal when the filtered signal satisfies a preset condition
  • the first determination module 1340 is configured to determine the next frequency band to be filtered as the frequency band to be filtered according to a preset order in the case that the filtered signal does not meet the preset condition, and call the first denoising Module 1320 determines the number of target component layers according to the frequency band to be filtered, performs wavelet threshold denoising on the components in the target component layer number located in the frequency band to be filtered, and obtains a processed filtered signal, until the last of the M frequency bands. If the filtered signal corresponding to one frequency band does not meet the preset condition, the first output module 1330 is called to output the filtered signal corresponding to the last frequency band.
  • the M frequency bands are contiguous.
  • adjacent frequency bands among the M frequency bands partially overlap.
  • the signal processing apparatus 130 further includes: an extraction module, configured to, after the first denoising module 1320 obtains the processed filtered signal, the first output module 1330 when the filtered signal satisfies a preset condition In the case of , before outputting the filtered signal, extract the beat frequency signal in the filtered signal; the preset condition is that the beat frequency signal is successfully extracted.
  • M 2;
  • the M frequency bands are a first low-pass frequency band and a first high-pass frequency band;
  • the first low-pass frequency band is a frequency band with a frequency less than a first threshold, and the first high-pass frequency band is a frequency band with a frequency greater than a second threshold;
  • the bandwidth of the first low-pass frequency band is greater than the bandwidth of the first high-pass frequency band; the first threshold is greater than or equal to the second threshold.
  • the preset sequence is:
  • the first low-pass frequency band the first high-pass frequency band
  • the first high-pass frequency band and the first low-pass frequency band are the first high-pass frequency band and the first low-pass frequency band.
  • M 2;
  • the M frequency bands are a second low-pass frequency band and a second high-pass frequency band;
  • the second low-pass frequency band is a frequency band with a frequency less than a third threshold, and the second high-pass frequency band is a frequency band with a frequency greater than a fourth threshold;
  • the bandwidth of the second low-pass frequency band is smaller than the bandwidth of the second high-pass frequency band; the third threshold is greater than or equal to the fourth threshold.
  • the preset sequence is:
  • M 3;
  • the M frequency bands are a third low-pass frequency band, a band-pass frequency band, and a third high-pass frequency band;
  • the third low-pass frequency band is a frequency band with a frequency less than the fifth threshold, and the band-pass frequency band is a frequency greater than the sixth threshold and less than the frequency band of the seventh threshold, the third high-pass frequency band is the frequency band whose frequency is greater than the eighth threshold;
  • the fifth threshold is greater than or equal to the sixth threshold, and the seventh threshold is greater than or equal to the eighth threshold.
  • the preset sequence is:
  • the third low-pass frequency band, the band-pass frequency band, the third high-pass frequency band or
  • the third high-pass frequency band, the band-pass frequency band, the third low-pass frequency band or
  • the band-pass frequency band, the third low-pass frequency band, the third high-pass frequency band or.
  • the third low-pass frequency band, the third high-pass frequency band, the band-pass frequency band or
  • the band-pass frequency band, the third high-pass frequency band, and the third low-pass frequency band are the band-pass frequency band, the third high-pass frequency band, and the third low-pass frequency band.
  • FIG. 20 shows a schematic structural diagram of a signal processing apparatus provided by another exemplary embodiment of the present application.
  • the signal processing device can be implemented by software, hardware or a combination of both.
  • the signal processing apparatus 140 includes: a second decomposition module 1410 , a second denoising module 1420 and a second output module 1430 . in:
  • the second decomposition module 1410 is configured to perform N-level decomposition on the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • the second denoising module 1420 is configured to determine the number of target component layers according to the frequency band to be filtered, and perform wavelet threshold denoising on the components in the target component layer number located in the frequency band to be filtered to obtain a processed filtered signal;
  • the second output module 1430 is configured to output the filtered signal.
  • the target frequency band is a frequency band whose frequency is less than the ninth threshold.
  • the target frequency band is a frequency band whose frequency is greater than the tenth threshold.
  • the target frequency band is a frequency band with a frequency smaller than an eleventh threshold and a frequency band with a frequency greater than the twelfth threshold; wherein the eleventh threshold is smaller than the twelfth threshold.
  • the embodiments of the present application provide several new filtering algorithms. By increasing the accuracy of wavelet coefficient decomposition and performing threshold processing on components of different frequency bands, denoising functions in different frequency bands can be realized, and the filtering algorithms provided by the embodiments of the present application can be expanded. applicable scenarios.
  • the signal processing apparatus 150 may include: at least one processor 1501 , at least one network interface 1504 , user interface 1503 , memory 1505 , and at least one communication bus 1502 .
  • the communication bus 1502 is used to realize the connection and communication between these components.
  • the user interface 1503 may include a display screen (Display) and a camera (Camera), and the optional user interface 1503 may also include a standard wired interface and a wireless interface.
  • Display display screen
  • Camera Camera
  • the optional user interface 1503 may also include a standard wired interface and a wireless interface.
  • the network interface 1504 may optionally include a standard wired interface and a wireless interface (eg, a WI-FI interface).
  • the processor 1501 may include one or more processing cores.
  • the processor 1501 uses various excuses and lines to connect various parts of the entire signal processing device 150, and by running or executing the instructions, programs, code sets or instruction sets stored in the memory 1505, and calling the data stored in the memory 1505, Various functions of the signal processing device 150 are performed and data is processed.
  • the processor 1501 may use at least one of digital signal processing (Digital Signal Processing, DSP), field-programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). A hardware form is implemented.
  • DSP Digital Signal Processing
  • FPGA Field-Programmable Gate Array
  • PLA programmable logic array
  • the processor 1501 may integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU), a modem, and the like.
  • CPU Central Processing Unit
  • GPU Graphics Processing Unit
  • the CPU mainly deals with the operating system, user interface and application programs
  • the GPU is used for rendering and drawing the content that needs to be displayed on the display screen
  • the modem is used for processing wireless communication. It can be understood that, the above-mentioned modem may not be integrated into the processor 1501, but is implemented by a single chip.
  • the memory 1505 may include random access memory (Random Access Memory, RAM), or may include read-only memory (Read-Only Memory).
  • the memory 1505 includes a non-transitory computer-readable storage medium.
  • Memory 1505 may be used to store instructions, programs, codes, sets of codes, or sets of instructions.
  • the memory 1505 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), Instructions and the like used to implement the above method embodiments; the storage data area may store the data and the like involved in the above method embodiments.
  • the memory 1505 can optionally also be at least one storage device located remote from the aforementioned processor 1501 .
  • the memory 1505 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a signal processing application program.
  • the user interface 1503 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 1501 can be used to call the signal processing application program stored in the memory 1505, and Specifically do the following:
  • N-level decomposition on the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • the frequency bands to be filtered are M frequency bands Any one of the frequency bands, the M frequency bands cover the entire frequency range, and M ⁇ 2;
  • the next frequency band to be filtered is determined as the frequency band to be filtered according to the preset order, and the determining the number of target component layers according to the frequency band to be filtered is performed, and the target component layers are determined according to the frequency band to be filtered.
  • the components located in the frequency band to be filtered in the component layers are subjected to wavelet threshold denoising to obtain a processed filtered signal, until the filtered signal corresponding to the last frequency band in the M frequency bands does not meet the preset conditions, and output all The filter signal corresponding to the last frequency band is described.
  • the M frequency bands are contiguous.
  • adjacent frequency bands among the M frequency bands partially overlap.
  • the processor 1501 is further configured to perform: extracting the filtered signal. the difference frequency signal in the filtered signal;
  • the preset condition is that the beat frequency signal is successfully extracted.
  • M 2;
  • the M frequency bands are a first low-pass frequency band and a first high-pass frequency band;
  • the first low-pass frequency band is a frequency band with a frequency less than a first threshold, and the first high-pass frequency band is a frequency band with a frequency greater than a second threshold;
  • the bandwidth of the first low-pass frequency band is greater than the bandwidth of the first high-pass frequency band; the first threshold is greater than or equal to the second threshold.
  • the preset sequence is:
  • the first low-pass frequency band the first high-pass frequency band
  • the first high-pass frequency band and the first low-pass frequency band are the first high-pass frequency band and the first low-pass frequency band.
  • M 2;
  • the M frequency bands are a second low-pass frequency band and a second high-pass frequency band;
  • the second low-pass frequency band is a frequency band with a frequency less than a third threshold, and the second high-pass frequency band is a frequency band with a frequency greater than a fourth threshold;
  • the bandwidth of the second low-pass frequency band is smaller than the bandwidth of the second high-pass frequency band; the third threshold is greater than or equal to the fourth threshold.
  • the preset sequence is:
  • M 3;
  • the M frequency bands are a third low-pass frequency band, a band-pass frequency band, and a third high-pass frequency band;
  • the third low-pass frequency band is a frequency band with a frequency less than the fifth threshold, and the band-pass frequency band is a frequency greater than the sixth threshold and less than the frequency band of the seventh threshold, the third high-pass frequency band is the frequency band whose frequency is greater than the eighth threshold;
  • the fifth threshold is greater than or equal to the sixth threshold, and the seventh threshold is greater than or equal to the eighth threshold.
  • the preset sequence is:
  • the third low-pass frequency band, the band-pass frequency band, the third high-pass frequency band or
  • the third high-pass frequency band, the band-pass frequency band, the third low-pass frequency band or
  • the band-pass frequency band, the third low-pass frequency band, the third high-pass frequency band or.
  • the third low-pass frequency band, the third high-pass frequency band, the band-pass frequency band or
  • the band-pass frequency band, the third high-pass frequency band, and the third low-pass frequency band are the band-pass frequency band, the third high-pass frequency band, and the third low-pass frequency band.
  • Embodiments of the present application further provide a computer-readable storage medium, where instructions are stored in the computer-readable storage medium, when the computer or processor is run on the computer or the processor, the computer or the processor is made to execute the above-mentioned embodiment shown in FIG. 5 . one or more of the steps in . If each component module of the above signal processing device is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in the computer-readable storage medium.
  • the signal processing apparatus 160 may include: at least one processor 1601 , at least one network interface 1604 , user interface 1603 , memory 1605 , and at least one communication bus 1602 .
  • the communication bus 1602 is used to realize the connection communication between these components.
  • the user interface 1603 may include a display screen (Display) and a camera (Camera), and the optional user interface 1603 may also include a standard wired interface and a wireless interface.
  • Display display screen
  • Camera Camera
  • the optional user interface 1603 may also include a standard wired interface and a wireless interface.
  • the network interface 1604 may optionally include a standard wired interface and a wireless interface (eg, a WI-FI interface).
  • the processor 1601 may include one or more processing cores.
  • the processor 1601 uses various excuses and lines to connect various parts in the entire signal processing device 160, and by running or executing the instructions, programs, code sets or instruction sets stored in the memory 1605, and calling the data stored in the memory 1605, Various functions of the signal processing device 160 are performed and data is processed.
  • the processor 1601 may use at least one of digital signal processing (Digital Signal Processing, DSP), field-programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). A hardware form is implemented.
  • DSP Digital Signal Processing
  • FPGA Field-Programmable Gate Array
  • PLA programmable logic array
  • the processor 1601 may integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU), a modem, and the like.
  • CPU Central Processing Unit
  • GPU Graphics Processing Unit
  • the CPU mainly handles the operating system, user interface, and application programs
  • the GPU is used to render and draw the content that needs to be displayed on the display screen
  • the modem is used to handle wireless communication. It can be understood that, the above-mentioned modem may not be integrated into the processor 1601, but is implemented by a single chip.
  • the memory 1605 may include random access memory (Random Access Memory, RAM), or may include read-only memory (Read-Only Memory).
  • the memory 1605 includes a non-transitory computer-readable storage medium.
  • Memory 1605 may be used to store instructions, programs, codes, sets of codes, or sets of instructions.
  • the memory 1605 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), Instructions and the like used to implement the above method embodiments; the storage data area may store the data and the like involved in the above method embodiments.
  • the memory 1605 can optionally also be at least one storage device located remote from the aforementioned processor 1601 .
  • the memory 1605 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a signal processing application program.
  • the user interface 1603 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 1601 can be used to call the signal processing application program stored in the memory 1605, and Specifically do the following:
  • N-level decomposition on the signal to be processed to obtain 2 N components; wherein, N ⁇ 2, the signal to be processed is a signal with noise;
  • the filtered signal is output.
  • the target frequency band is a frequency band whose frequency is less than the ninth threshold.
  • the target frequency band is a frequency band whose frequency is greater than the tenth threshold.
  • the target frequency band is a frequency band with a frequency smaller than an eleventh threshold and a frequency band with a frequency greater than the twelfth threshold; wherein the eleventh threshold is smaller than the twelfth threshold.
  • the embodiments of the present application provide several new filtering algorithms. By increasing the accuracy of wavelet coefficient decomposition and performing threshold processing on components of different frequency bands, denoising functions in different frequency bands can be realized, and the filtering algorithms provided by the embodiments of the present application can be expanded. applicable scenarios.
  • the embodiment of the present application further provides a computer-readable storage medium, where instructions are stored in the computer-readable storage medium, when the computer or the processor runs on the computer, the computer or the processor causes the computer or the processor to execute the above-mentioned embodiment shown in FIG. 9 . one or more of the steps in . If each component module of the above signal processing device is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in the computer-readable storage medium.
  • the above-mentioned embodiments it may be implemented in whole or in part by software, hardware, firmware or any combination thereof.
  • software it can be implemented in whole or in part in the form of a computer program product.
  • the computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated.
  • the computer may be a general purpose computer, special purpose computer, computer network, or other programmable device.
  • the computer instructions may be stored in or transmitted over a computer-readable storage medium.
  • the computer instructions can be sent from a website site, computer, server, or data center via wired (eg, coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (eg, infrared, wireless, microwave, etc.) another website site, computer, server or data center for transmission.
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that includes an integration of one or more available media.
  • the available media may be magnetic media (eg, floppy disks, hard disks, magnetic tapes), optical media (eg, Digital Versatile Disc (DVD)), or semiconductor media (eg, Solid State Disk, SSD)) etc.
  • the aforementioned storage medium includes: a system memory (Read Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk and other media that can store program codes.
  • ROM Read Only Memory
  • RAM Random Access Memory
  • magnetic disk or an optical disk and other media that can store program codes.

Landscapes

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

Abstract

一种信号处理方法、装置及可读存储介质。该方法包括：对待处理信号进行N级分解，得到2 N个分量；根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；待滤波频段为M个频段中的任意一个频段，M个频段覆盖全频段范围，M≥2；若滤波信号满足预设条件，输出滤波信号；若滤波信号不满足预设条件，按照预设顺序将下一个待滤波频段确定为待滤波频段，并确定待滤波频段对应的滤波信号，直至M个频段中的最后一个频段对应的滤波信号不满足预设条件，输出其对应的滤波信号，可以提高全频段范围内的距离探测精度。

Description

信号处理方法、装置及可读存储介质 技术领域
本申请涉及激光探测技术领域，尤其涉及一种信号处理方法、装置及可读存储介质。
背景技术
激光雷达，是以发射激光束探测目标的位置、速度等特征量的雷达系统。其工作原理是向目标发射探测信号，然后将接收到的从目标反射回来的信号与发射信号进行比较，作适当处理后，就可获得目标的有关信息，如目标距离、方位、高度、速度、姿态、甚至形状等参数，从而对飞机、导弹等目标进行探测、跟踪和识别。
对于激光雷达系统，其测距原理是在扫频周期内发射频率线性变化的连续波作为出射信号，出射信号的一部分作为本振信号，其余部分向外出射进行探测，被物体反射后返回的回波信号与本振信号有一定的频率差，通过测量频率差可以获得被探测目标与雷达之间的距离信息。
在调频连续波(Frequency-Modulated Continuous Wave，FMCW)系统中，发射信号经过空间中的一段距离击中目标物后发生漫反射。由于目标物体本身反射率有限，且经过目标物体漫反射后的回波信号向空间中大范围反射，在某个方向上接收到的回波信号往往很弱，信噪比很差。因此，需要一种有效的信号去噪算法对接收到的回波信号进行优化去噪以提高信噪比，这样才能提高FMCW系统探测的灵敏度。
发明内容
本申请实施例提供了一种信号处理方法、装置及可读存储介质，可以提高全频段范围内的距离探测精度。
第一方面，本申请实施例提供了一种信号处理方法，包括：
对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；其中，所述待滤波频段为M个频段中的任意一个频段，所述M个频段覆盖全频段范围，M≥2；
在所述滤波信号满足预设条件的情况下，输出所述滤波信号；
在所述滤波信号不满足所述预设条件的情况下，按照预设顺序将下一个待滤波频段确定为所述待滤波频段，并执行所述根据待滤波频段确定目标分量层数，对目标分量层数中分量位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号，直至所述M个频段中的最后一个频段对应的滤波信号不满足所述预设条件，输出所述最后一个频段对应的滤波信号。
第二方面，本申请实施例提供了一种信号处理方法，包括：
对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；
输出所述滤波信号。
第三方面，本申请实施例提供了一种信号处理装置，包括：
第一分解模块，用于对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
第一去噪模块，用于根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；其中，所述待滤波频段为M个频段中的任意一个频段，所述M个频段覆盖全频段范围，M≥2；
第一输出模块，用于在所述滤波信号满足预设条件的情况下，输出所述滤波信号；
第一确定模块，用于在所述滤波信号不满足所述预设条件的情况下，按照预设顺序将下一个待滤波频段确定为所述待滤波频段，并调用所述第一去噪模块根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号，直至所述M个频段中的最后一个频段对应的滤波信号不满足所述预设条件，调用所述第一输出模块输出所述最后一个频段对应的滤波信号。
第四方面，本申请实施例提供了一种信号处理装置，包括：
第二分解模块，用于对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
第二去噪模块，用于根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；
第二输出模块，用于输出所述滤波信号。
第五方面，本申请实施例提供了一种信号处理装置，包括：处理器和存储器；其中，所述存储器存储有计算机程序，所述计算机程序适于由处理器加载并执行本申请实施例第一方面或第二方面提供的方法步骤。
第六方面，本申请实施例提供了一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现本申请实施例第一方面或第二方面提供的方法。
本申请一些实施例提供的技术方案带来的有益效果至少包括：
在本申请一个或多个实施例中，可以通过按照预设的频段顺序依次尝试对待处理信号进行去噪处理，以确定噪声信号具体处于哪个频段，从而实现对该待处理信号在全频段范围内的精准去噪，相比于现有技术中只能够对低频段的噪声进行处理，提升了去噪精度，且扩展了有效探测的范围，进而增大FMCW系统的探测距离。另外，本申请实施例提供了几种新的滤波算法，通过增加小波系数分解的精度以及对不同频段的分量进行阈值处理，实现了不同频段范围内的去噪功能。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例中所需使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的一种车载的激光雷达分布示意图；
图2为本申请实施例提供的信号处理系统架构示意图；
图3为本申请实施例提供的三角波调制的FMCW系统的发射与回波信号的频率随时间变化的示意图；
图4a为经典小波阈值去噪算法的分解方式示意图；
图4b为经典小波阈值去噪算法在测距100米时的模拟结果示意图；
图4c为经典小波阈值去噪算法在测距200米时的模拟结果示意图；
图4d为经典小波阈值去噪算法在测距300米时的模拟结果示意图；
图4e为本申请实施例提供的一种小波阈值去噪算法的分解方式示意图；
图4f为本申请实施例提供的低通滤波小波阈值去噪算法在测距100米时的模拟结果示意图；
图4g为本申请实施例提供的低通滤波小波阈值去噪算法在测距200米时的模拟结果示意图；
图4h为本申请实施例提供的低通滤波小波阈值去噪算法在测距300米时的模拟结果示意图；
图4i为本申请实施例提供的带通滤波小波阈值去噪算法在测距100米时的模拟结果示意图；
图4j为本申请实施例提供的带通滤波小波阈值去噪算法在测距200米时的模拟结果示意图；
图4k为本申请实施例提供的带通滤波小波阈值去噪算法在测距300米时的模拟结果示意图；
图4l为本申请实施例提供的高通滤波小波阈值去噪算法在测距100米时的模拟结果示意图；
图4m为本申请实施例提供的高通滤波小波阈值去噪算法在测距200米时的模拟结果示意图；
图4n为本申请实施例提供的高通滤波小波阈值去噪算法在测距300米时的模拟结果示意图；
图5为本申请实施例提供的一种信号处理方法的流程示意图；
图6a为本申请实施例提供的一种全频段范围划分示意图；
图6b为本申请实施例提供的另外一种全频段范围划分示意图；
图6c为本申请实施例提供的一种信号处理方法在不同距离上的探测能力示意图；
图7a为本申请实施例提供的另外一种全频段范围划分示意图；
图7b为本申请实施例提供的另外一种全频段范围划分示意图；
图7c为本申请实施例提供的另外一种信号处理方法在不同距离上的探测能力示意图；
图8a为本申请实施例提供的另外一种全频段范围划分示意图；
图8b为本申请实施例提供的另外一种全频段范围划分示意图；
图8c为本申请实施例提供的另外一种信号处理方法在不同距离上的探测能力示意图；
图9为本申请实施例提供的另外一种信号处理方法的流程示意图；
图10为本申请实施例提供的一种信号处理方法在不同距离上的探测能力示意图；
图11为本申请实施例提供的另外一种信号处理方法在不同距离上的探测能力示意图；
图12为本申请实施例提供的另外一种信号处理方法在不同距离上的探测能力示意图；
图13为本申请实施例提供的一种信号处理装置的结构示意图；
图14为本申请实施例提供的另一种信号处理装置的结构示意图；
图15为本申请实施例提供的另一种信号处理装置的结构示意图；
图16为本申请实施例提供的另一种信号处理装置的结构示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述。
本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”、“第三”等是用于区 别不同对象，而不是用于描述特定顺序。此外，术语“包括”和“具有”以及它们任何变形，意图在于覆盖不排他的包含。例如包含了一系列步骤或单元的过程、方法、系统、产品或设备没有限定于已列出的步骤或单元，而是可选地还包括没有列出的步骤或单元，或可选地还包括对于这些过程、方法、产品或设备固有的其他步骤或单元。
图1示例性示出了一种车载激光雷达的分布示意图。如图1所示，车载的激光雷达(110a、110b、110c及110d)可分布于车辆的四个角落，可用于采集其辐射范围内(车辆周围)的点云数据。当然，激光雷达也可以根据探测需求分布于车辆的左侧、右侧、前侧、后侧、顶部等任意位置。其中，本申请实施例中的激光雷达为FMCW激光雷达。接下来以其中一个激光雷达110a为例进行说明。激光雷达110a可采集其辐射范围内的回波信号，根据发射信号和回波信号解算后获得去噪后的差频信号，并将去噪后的差频信号发送给车载终端120。具体的，信号处理装置根据发射信号和接收信号确定带有噪声的差频信号；并将差频信号进行信号处理得到去噪后的差频信号。车载终端120可对激光雷达110a获得的去噪后的差频信号进行处理得到点云数据，根据点云数据识别出车辆周围的障碍物类型、位置以及该障碍物的运动轨迹，并结合自身运行情况做出路径规划，生成驾驶策略。
图2示例性示出本申请实施例提供的一种信号处理系统架构示意图。如图2所示，FMCW激光雷达可以发出发射信号，发射信号经过空间中的一段距离里击中目标物体后发生反射返回回波信号，回波信号被FMCW雷达系统接收。回波信号与发射信号混频后产生差频信号。FMCW激光雷达的信号处理装置可以采用小波阈值去噪算法对该差频信号进行去噪处理，输出去噪后的差频信号给车载终端。在具体实现中，信号处理装置可以设置于该FMCW激光雷达，可以独立于该FMCW激光雷达，也可以设置于车载终端120，本申请实施例对此不做限定。本申请实施例中以信号处理装置设置于FMCW激光雷达为例进行说明，通过FMCW激光雷达采用小波阈值去噪算法对差频信号进行去噪处理。
不限于应用于车辆，本申请实施例提供的FMCW激光雷达还可以应用于机器人及无人机等设备中。也即是说，FMCW激光雷达输出去噪后的差频信号不限于发送给车载终端，在其他的应用场景中，还可以发送给机器人或无人机中的CPU等，本申请实施例对此不作限定。
图3示出了三角波调制的FMCW激光雷达的发射信号与回波信号的频率随时间变化的示意图。如图3所示：
假设一个与激光雷达之间的距离为R的目标，以速度为v向着激光雷达匀速运动(以远离激光雷达方向为正方向)，发射信号和回波信号的回波延迟为τ＝2(R+vt)/c，则在上扫频段和下扫频段，差频信号分别如下列公式(1)和(2)所示：
Figure PCTCN2021086497-appb-000001
Figure PCTCN2021086497-appb-000002
式中，f 0为发射信号每个周期的初始频率，μ＝B/T为调频的斜率，T为信号周期，B为信号带宽，A为差频信号的振幅，
Figure PCTCN2021086497-appb-000003
为差频信号的相位，c为光速，t为时间变量。
可以看到，在FMCW激光雷达中，差频信号的频率绝对值与探测目标的距离呈线性正相关的关系，也就是说，目标越远，差频信号的频率越高，反之越低。
接下来介绍经典的小波区域去噪算法的分解方式。如图4a所示，经典的小波阈值去噪算法的核心是单边小波分解，具体过程如下：
(1)对带有噪声的信号进行一级小波分解得到cA和cD两个部分，cA与cD分别表示近似部分与细节部分。这是第1级分解。
(2)细节部分cD不作处理，对近似部分cA继续分解得到近似部分cAA和细节部分 cAD两个部分。这是第2级分解。
(3)细节部分cAD不作处理，进一步对近似部分cAA继续分解得到cAAA和cAAD两个部分。这是第3级分解。
(4)以此类推，每一级分解时都保持细节部分保持不变，只分解近似部分，直到第N级分解。具体的分解层数可以根据实际需要选取。
小波分解完成后，小波阈值去噪算法根据阈值规则对得到的各个小波分量部分进行阈值处理。
由于经典的小波阈值去噪算法有明显的低通滤波特点，结合图3示出的三角波调制的FMCW系统的发射信号与回波信号的频率随时间变化的示意图，可以看出，经典的小波阈值去噪算法存在以下问题：
(1)当目标距离较远时，回波信号处在中高频，这种情况下经典的小波阈值去噪的低通滤波性质导致在中高频的信号被削弱或滤除，算法失效。也就是说，如果回波信号处在中高频段，经典的小波算法不但不能达到去噪的效果，反而会将回波信号作为噪声削弱或滤除，造成探测失效。所以，经典的小波算法对近距离探测的回波信号有效，面对长距离的情况会恶化探测能力使探测失败，严重限制了实际FMCW激光雷达的应用。
(2)经典的小波阈值去噪的分解层数越多，低通滤波被滤掉的频率范围越大。回波信号必须要在一个更窄的低频范围内才能保证不被削弱或滤除。也就是说，分解层数越大，经典的小波阈值去噪算法对回波信号有效距离越近，更加严重影响了实际FMCW激光雷达应用。
(3)经典的小波阈值去噪算法的低通滤波的最小滤波范围由分解层数为1时决定，无法继续减小滤波范围。也就是说，在激光雷达的采样频率一定时，经典的小波阈值去噪算法有一个最大有效探测距离，无法调节使这个最大有效探测距离变得更大，严重影响了实际FMCW激光雷达应用。
图4b-图4d示例性示出了经典小波阈值去噪算法在的模拟结果示意图。图4b-图4d分别显示测距100米、200米和300米时的模拟结果。每张图中有上中下三张子图：上面子图显示无噪声信号的频谱，中间子图用于显示带噪声信号的频谱(信噪比为-15dB)，下面子图显示分解层数为1的经典小波去噪后信号的频谱。可以看出，经典小波阈值去噪算法具备明显的低通滤波特效。在测距100米，信号频率在被滤波的范围以外。而在测距200米时，信号强度被小波阈值去噪算法削弱。在测距300米时，信号强度为小波阈值去噪算法滤除。(4)如果要在不改变经典的小波阈值去噪算法的基础上增加最大有效探测距离，需要提高激光雷达的采样频率。也就是说，需要更快的硬件配合。这会大大增加FMCW信号处理部分的整体硬件要求和成本，限制FMCW激光雷达的应用。
因此，本申请实施例提供了几种新型的小波阈值去噪算法，与经典的小波阈值去噪算法相比，可以提高中远距离的探测性能。
图4e示例性示出了本申请实施例提供的一种小波阈值去噪算法的分解方式示意图。如图4e所示，小波阈值去噪算法对信号进行小波分解的具体过程如下：
(1)对带有噪声的信号进行一级小波系数分解得到cA低频部分和cD高频部分。
(2)使用(1)的分解方法，把cA继续分解得到cAA低频部分和cAD高频部分，把cD继续分解得到cDD高频部分和cDA低频部分。这是第2级分解。
(3)对上一级分解得到cAA,cAD,cDA,cDD四个部分分别继续分解得到cAAA,cAAD,cADA,cADD,cDAA,cDAD,cDDA,cDDD。这是第3级分解。
(4)以此类推......直到第N级分解。具体的分解层数可以根据实际需要选取。
可知，不限于图4e示出的三级小波系数分解，在具体实现中可以有更少或更多级的小波系数分解，本申请实施例对此不作限定。
基于图4e提供的小波阈值去噪算法的分解方式，本申请实施例提供了三种小波阈值去噪算法，分别实现低通滤波、带通滤波及高通滤波。接下来分别介绍这三种小波阈值去噪算法。
低通滤波小波阈值去噪算法：具体可以对小波系数分解后得到的多个分量中的高频分量进行阈值处理，其他分量保持不变。以二级分解为例，二级分解后可以得到四个分量：cAA,cAD,cDA,cDD后，只对得到的高频部分cDD进行阈值处理。其它部分保持不变。
图4f-图4h示例性示出了本申请实施例提供的低通滤波小波阈值去噪算法的模拟结果示意图。图4f-图4h分别显示测距100米、200米和300米时的模拟结果。每张图中有上中下三张子图：上面子图显示无噪声信号的频谱，中间子图用于显示带噪声信号的频谱(信噪比为-15dB)，下面子图显示采用本申请实施例提供的低通滤波小波阈值去噪算法去噪后信号的频谱。与图4b-图4d对比可以看出，本申请实施例提供的低通滤波小波阈值去噪算法被低通滤波的频率范围明显减小。具体从测距为200米时信号强度未被削弱可以看出。
带通滤波小波阈值去噪算法：具体可以对小波系数分解后得到的多个分量中的低频分量和高频分量进行阈值处理，其他分量保持不变。以二级分解为例，二级分解后可以得到四个分量：cAA,cAD,cDA,cDD后，只对得到的低频部分cAA及高频部分cDD进行阈值处理。其它部分保持不变。
图4i-图4k示例性示出了本申请实施例提供的带通滤波小波阈值去噪算法的模拟结果示意图。图4i-图4k分别显示测距100米、200米和300米时的模拟结果。每张图中有上中下三张子图：上面子图显示无噪声信号的频谱，中间子图用于显示带噪声信号的频谱(信噪比为-15dB)，下面子图显示采用本申请实施例提供的带通滤波小波阈值去噪算法去噪后信号的频谱。与图4b-图4d对比可以看出，本申请实施例提供的带通滤波小波阈值去噪算法具有明显的带通滤波的性质。具体从测距为200米时信号强度未被削弱，而测距为100米及300米时信号强度被削弱可以看出。高通滤波小波阈值去噪算法：具体可以对小波系数分解后得到的多个分量中的低频分量进行阈值处理，其他分量保持不变。以二级分解为例，二级分解后可以得到四个分量：cAA,cAD,cDA,cDD后，只对得到的低频部分cAA进行阈值处理。其它部分保持不变。
图4l-图4n示例性示出了本申请实施例提供的高通滤波小波阈值去噪算法的模拟结果示意图。图4l-图4n分别显示测距100米、200米和300米时的模拟结果。每张图中有上中下三张子图：上面子图显示无噪声信号的频谱，中间子图用于显示带噪声信号的频谱(信噪比为-15dB)，下面子图显示采用本申请实施例提供的高通滤波小波阈值去噪算法去噪后信号的频谱。与图4b-图4d对比可以看出，本申请实施例提供的高通滤波小波阈值去噪算法具有明显的高通滤波的性质。具体从测距为300米时信号强度未被削弱，而测距为100米及200米时信号强度被削弱可以看出。
可知，以上提到的阈值处理具体为：将幅值小于预设阈值的分量的小波系数设置为零。其中，该预设阈值为预先设置的临界值。具体地，若某分量的幅值小于预设阈值，则确定将该分量主要由噪声引起，则将该分量的小波系数设置为零，从而去除噪声。
接下来结合图1-图4n介绍本申请实施例提供的信号处理方法。图5示例性示出了本申请实施例提供的一种信号处理方法的流程示意图。如图5所示，信号处理方法可以包括以下几个步骤：
S501：对待处理信号进行N级分解，得到2 N个分量。
其中，N≥2，所述待处理信号为带噪声的信号。对待处理信号进行N级分解的方式可参考图4e示出的分解方式，此处不再赘述。分解得到的每个分量对应不同的频段范围。
在S501之前，该方法还可以包括：获取待处理信号。该待处理信号可以是图2实施例中提到的差频信号。
S502：根据待滤波频段确定目标分量层数，对所述目标分量层数中分量位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号。
具体地，所述待滤波频段为M个频段中的任意一个频段，所述M个频段覆盖全频段范围，M≥2。
可选地，M个频段连续相接。即M个频段中相邻频段不重叠。
可选地，M个频段中相邻频段部分重叠。
具体地，待处理信号即为图2实施例中提到的差频信号。对该差频信号进行N级分解后，可得到2 N个分量，每个分量对应一频段范围。由前述图3实施例中关于差频信号的频率与目标距离之间呈线性正相关的相关描述分析中可知，本申请实施例中提到的全频段范围为激光雷达的测距范围对应的差频信号的频率范围。示例性地，激光雷达的测距距离为300米时，差频信号的频率大约在400MHz。该激光雷达的测距范围定义为0-300米时，其全频段范围为0-400MHz。
具体地，小波阈值去噪的方法可以包括但不限于以下几种：模极大值法去噪、相关性去噪、小波收缩阈值法去噪和平移不变量小波法去噪等。
S503：判断所述待滤波频段是否为所述M个频段中的最后一个频段；若是，执行S506；若否，执行S504或S505。
具体地，M个频段可以被划分为至少两个频段。本申请实施例提供的信号处理方法可以按照预设的顺序对上述至少两个频段进行小波阈值去噪处理。所述M个频段中的最后一个频段即为所述M个频段中唯一一个未被进行小波阈值去噪处理的频段，也即是说所述M个频段中的其他频段均已进行小波阈值去噪处理过。
可知，待处理信号可以包括噪声信号及差频信号。若在S502中将噪声信号完全滤除或部分滤除之后，处理后的滤波信号可包含差频信号。
S504：在所述滤波信号满足预设条件的情况下，输出所述滤波信号。
具体地，在S502和S504之间，该方法还包括：提取所述滤波信号中的差频信号。预设条件为所述差频信号提取成功。
可知，待处理信号可以包括噪声信号及差频信号。若在S502中将噪声信号完全滤除或部分滤除之后，处理后的滤波信号可包含差频信号。
具体地，若差频信号的频率位于上述待滤波频段中，则对应的分量经过小波阈值去噪处理后，差频信号会被滤除，那么差频信号提取失败；若差频信号的频率未位于上述待滤波频段中，则经过小波阈值去噪处理后，噪声被滤除或部分滤除，差频信号将容易被提取。
其中，从滤波信号中提取差频信号的具体方式可以是对滤波信号进行快速傅里叶变换(Fast Fourier Transform，FFT)。
S505：在所述滤波信号不满足所述预设条件的情况下，按照预设顺序将下一个待滤波频段确定为所述待滤波频段，并执行所述S502。
具体地，若滤波信号不满足预设条件，表明噪声信号和差频信号均被滤除，说明对该待滤波频段进行滤波无法实现仅滤除噪声信号而不削弱差频信号，则需要将原始的待处理信号进行下一个频段的滤波处理。
本申请实施例中将上述全频段范围分为M个频段的方式有多种，可以是2个频段、3个频段甚至更多，根据信号处理的硬件运算能力、信号处理的精度要求等选择划分的频段个数，且每一种划分方式可以对应多种预设顺序。全频段范围分为2个频段时，预设顺序较简单；全频段范围分为3个或者3个以上频段时，预设顺序相对更复杂。为了表述起来更方便，以全频段范围分为3个频段为例进行说明；全频段范围分为3个以上频段时，与分为3个频段的处理逻辑相同。
接下来分别介绍三种分解方式。
方式一：低通+带通+高通(M＝3)
具体地，M个频段可以为低通频段(本申请实施例中可以将其称为第三低通频段)、带通频段及高通频段(本申请实施例中可以将其称为第三高通频段)。其中，第三低通频段为频率小于第五阈值的频段，带通频段为频率大于第六阈值且小于第七阈值的频段，第三高通频段为频率大于第八阈值的频段。其中，第五阈值大于或等于第六阈值，第七阈值大于或等于第八阈值。
图6a示例性示出了一种将全频段范围分解为3个频段的示意图。如图6a所示，全频段范围可被分为低通频段、带通频段及高通频段。相邻的频段连续相接。即第五阈值等于第六阈值，第七阈值等于第八阈值。每个频段的带宽可以相同也可以不同。
图6b示例性示出了另外一种将全频段范围分解为3个频段的示意图。如图6b所示，全频段范围可被分为低通频段、带通频段及高通频段。相邻的频段部分重叠。即第五阈值大于第六阈值，第七阈值大于第八阈值。每个频段的带宽可以相同也可以不同。
这种分解方式可以对应6种预设顺序：
第一种：所述第三低通频段、所述带通频段、所述第三高通频段。
第二种：所述第三高通频段、所述带通频段、所述第三低通频段。
第三种：所述带通频段、所述第三低通频段、所述第三高通频段。
第四种：所述第三低通频段、所述第三高通频段、所述带通频段。
第五种：所述第三高通频段、所述第三低通频段、所述带通频段。
第六种：所述带通频段、所述第三高通频段、所述第三低通频段。
以上述第一种为例，本申请实施例可以首先对第三低通频段确定为待滤波频段，根据第三低通频段确定目标分量层数，对目标分量层数中位于该待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；若能够从该滤波信号中提取出差频信号，则输出该滤波信号。若无法从该滤波信号中提取出差频信号，则将带通频段确定为待滤波频段，根据带通频段确定目标分量层数，对目标分量层数中位于该待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；若能够从该滤波信号中提取出差频信号，则输出该滤波信号。若无法从该滤波信号中提取出差频信号，则将第三高通频段确定为待滤波频段，根据第三高通频段确定目标分量层数，对目标分量层数中位于该待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号，此时无论能否从该滤波信号中提取出差频信号，都输出该滤波信号。
可以知道，对目标分量层数中位于低通频段的分量进行小波阈值去噪，即为差频信号的低通频段部分保持不变，对差频信号的其余部分进行噪声滤除，即可实现覆盖低通频段的滤波；对于目标分量层数中位于带通频段的分量进行小波阈值去噪，即为差频信号的带通频段部分保持不变，对差频信号的其余部分进行滤除，即可实现带通频段的滤波；对于目标分量层数中位于高通频段的分量进行小波阈值去噪，即为差频信号的高通频段部分保持不变，对差频信号的其余部分进行滤除，即可实现覆盖高通频段的滤波。
对不同频段对应的分量进行小波阈值去噪，可以提高不同探测距离对应的探测概率。通过低通、带通及高通频段的组合滤波，可以提高整个探测范围内的探测概率，以此提高FMCW激光雷达的有效探测距离，促进FMCW激光雷达的应用。
图6c示例性示出了分解方式一对应的信号处理方法在不同距离上的探测能力示意图。如图6c所示，分解方式一对应的信号处理方法在近距离、中距离和远距离上均具备探测能力。对某信号低通频段范围对应的分量进行小波阈值去噪处理，可以实现低通滤波，进而实现近距离的探测。同时对某信号低通频段范围及高通频段范围对应的分量进行小波阈值去噪处理，可以实现带通滤波，进而实现中距离的探测。对某信号高通频段范围对应的分量进行小波阈值去噪处理，可以实现高通滤波，进而实现远距离的探测。
具体地，对低通频段范围的分量进行阈值处理的方式可以参考图4e实施例中提到的低通滤波小波阈值去噪算法，此处不再赘述。
具体地，同时对低通频段范围及高通频段范围的分量进行阈值处理的方式可以参考图4e实施例中提到的带通滤波小波阈值去噪算法，此处不再赘述。
具体地，对高通频段范围的分量进行阈值处理的方式可以参考图4e实施例中提到的高通滤波小波阈值去噪算法，此处不再赘述。
方式二：宽带低通+窄带高通(M＝2)
具体地，M个频段可以为低通频段(本申请实施例中可以将其称为第一低通频段)、及高通频段(本申请实施例中可以将其称为第一高通频段)。其中，第一低通频段为频率小于第一阈值的频段，第一高通频段为频率大于第二阈值的频段。其中，第一阈值大于或等于第二阈值。第一低通频段的带宽(即范围长度)和第一高通频段的带宽可以根据信号处理的需要进行设置。以第一低通频段的带宽大于第一高通频段的带宽为例进行说明。
图7a示例性示出了一种将全频段范围分解为2个频段的示意图。如图7a所示，全频段范围可被分为低通频段及高通频段。相邻的频段连续相接。即第一阈值等于第二阈值。
图7b示例性示出了另外一种将全频段范围分解为2个频段的示意图。如图7b所示，全频段范围可被分为低通频段及高通频段。相邻的频段部分重叠。即第一阈值大于第二阈值。
这种分解方式可以对应2种预设顺序：
第一种：所述第一低通频段、所述第一高通频段。
第二种：所述第一高通频段、所述第一低通频段。
图7c示例性示出了分解方式二对应的信号处理方法在不同距离上的探测能力示图。如图7c所示，分解方式二对应的信号处理方法在近距离、中距离和远距离上均具备探测能力。由于本申请实施例中，低通频段的带宽较大，高通频段的带宽较小，对某信号的低通频段范围小波阈值去噪处理，可以实现高通滤波，进而实现远距离的探测。对某信号高通频段范围的分量进行小波阈值去噪处理，可以实现低通滤波，进而实现近距离及中距离的探测。
方式三：窄带低通+宽带高通(M＝2)
具体地，M个频段可以为低通频段(本申请实施例中可以将其称为第二低通频段)、及高通频段(本申请实施例中可以将其称为第二高通频段)。其中，第二低通频段为频率小于第三阈值的频段，第二高通频段为频率大于第四阈值的频段。其中，第三阈值大于或等于第四阈值。其中，第二低通频段的带宽(即范围长度)大于第二高通频段的带宽。
图8a示例性示出了一种将全频段范围分解为2个频段的示意图。如图8a所示，全频段范围可被分为低通频段及高通频段。相邻的频段连续相接。即第三阈值等于第四阈值。
图8b示例性示出了另外一种将全频段范围分解为2个频段的示意图。如图8b所示，全频段范围可被分为低通频段及高通频段。相邻的频段部分重叠。即第三阈值大于第四阈值。
这种分解方式可以对应2种预设顺序：
第一种：所述第二低通频段、所述第二高通频段。
第二种：所述第二高通频段、所述第二低通频段。
图8c示例性示出了分解方式三对应的信号处理方法在不同距离上的探测能力示图。如图8c所示，分解方式三对应的信号处理方法在近距离、中距离和远距离上均具备探测能力。由于本申请实施例中，低通频段的带宽较小，高通频段的带宽较大，对某信号的低通频段范围小波阈值去噪处理，可以实现高通滤波，进而实现远距离及中距离的探测。对某信号高通频段范围的分量进行小波阈值去噪处理，可以实现低通滤波，进而实现近距离的探测。
需要说明的是，随着分解层数的增加，每一个分量对应的频段带宽越窄。本申请实施 例中，对于全频段范围的分解可以参考分量分解的结果，以使M个频段中的任意一个频段可以与某一分量层数中的某一个或几个分量对应的频段一致。而采用某一特征的分解方式得到的M个频段，可以分别位于不同的分量层数中的某一个分量。示例性地，若采用方式二分解得到一个宽带低通频段及窄带高通频段，该宽带低通频段可对应第二分量层数中的低通频段对应的分量(如图4e中三级分解后得到的cDDD分量)；该窄带高通频点可对应第二分量层数中的高通频段对应的分量(如图4e中三级分解后得到的cAA分量)。
在本申请实施例中，可以通过按照预设的频段顺序依次尝试对待处理信号进行去噪处理，以确定噪声信号具体处于哪个频段，从而实现对该待处理信号在全频段范围内的精准去噪，相比于现有技术中只对低频段的噪声进行处理，提升了去噪精度，且扩展了有效探测的范围，进而增大FMCW系统的探测距离。
图9示出了本申请实施例提供的另外一种信号处理方法的流程示意图。如图9所示，信号处理方法可以包括以下几个步骤：
S901：对待处理信号进行N级分解，得到2 N个分量。
具体地，S901与S501一致，此处不再赘述。
S902：根据待滤波频段确定目标分量层数，对所述目标分量层数中分量位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号。
可选地，待滤波频段可以是低通频段，具体为频率小于第九阈值的频段，用于实现低通滤波。
具体地，对低通频段范围对应的分量进行阈值处理的方式可以参考图4e实施例中提到的低通滤波小波阈值去噪算法，此处不再赘述。图10示例性示出了对低通频段范围的分量进行阈值处理后本申请实施例在不同距离上的探测能力示图。如图10所示，对低通频段范围的分量进行小波阈值去噪，即为差频信号的低频部分保持不变，对差频信号其余的高频部分进行滤除，处理后可实现低通滤波，从而本申请实施例在近距离上具备优秀的探测能力。
可选地，待滤波频段可以是高通频段，具体为频率大于第十阈值的频段，用于实现高通滤波。
具体地，对高通频段范围的分量进行阈值处理的方式可以参考图4e实施例中提到的高通滤波小波阈值去噪算法，此处不再赘述。图11示例性示出了对高通频段范围的分量进行阈值处理后本申请实施例在不同距离上的探测能力示图。如图11所示，对高通频段范围的分量进行小波阈值去噪，即为差频信号的高频部分保持不变，对差频信号其余的低频部分进行滤除，处理后可实现高通滤波，本申请实施例在远距离上具备优秀的探测能力。
可选地，待滤波频段可以是低通频段及高通频段，具体为频率小于第十一阈值的频段及频率大于第十二阈值的频段，用于实现带通滤波。其中，第十一阈值小于第十二阈值。
具体地，同时对低通频段范围及高通频段范围的分量进行阈值处理的方式可以参考图4e实施例中提到的带通滤波小波阈值去噪算法，此处不再赘述。图12示例性示出了对带通频段范围的分量进行阈值处理后本申请实施例在不同距离上的探测能力示图。如图12所示，对带通频段范围的分量进行小波阈值去噪，即为差频信号的中间频段保持不变，对差频信号其余的低频部分和高频部分进行滤除，处理后可实现带通滤波，本申请实施例在中距离上具备优秀的探测能力。
可知，本申请实施例中提到的近距离、中距离及远距离为障碍物与激光雷达之间的距离。障碍物与激光雷达之间的距离决定了差频信号的频率。近距离、中距离及远距离为相对的概念，本申请实施例中对于这三者的具体数值不作限定。
S903：输出所述滤波信号。
可知，待处理信号可以包括噪声信号及差频信号。若差频信号的频率未位于上述待滤 波频段中，则经过小波阈值去噪处理后，噪声被滤除或部分滤除，差频信号可被提取。
本申请实施例提供了几种新的滤波算法，通过增加小波系数分解的精度以及对不同频段的分量进行阈值处理，可以实现不同频段范围内的去噪功能，扩大本申请实施例提供的滤波算法的适用场景。
下述为本申请装置实施例，可以用于执行本申请方法实施例。对于本申请装置实施例中未披露的细节，请参照本申请方法实施例。
请参见图13，其示出了本申请一个示例性实施例提供的信号处理装置的结构示意图。该信号处理装置可以通过软件、硬件或者两者的结合实现。信号处理装置130包括：第一分解模块1310、第一去噪模块1320、第一输出模块1330及第一确定模块1340。其中：
第一分解模块1310，用于对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
第一去噪模块1320，用于根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；其中，所述待滤波频段为M个频段中的任意一个频段，所述M个频段覆盖全频段范围，M≥2；
第一输出模块1330，用于在所述滤波信号满足预设条件的情况下，输出所述滤波信号；
第一确定模块1340，用于在所述滤波信号不满足所述预设条件的情况下，按照预设顺序将下一个待滤波频段确定为所述待滤波频段，并调用所述第一去噪模块1320根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号，直至所述M个频段中的最后一个频段对应的滤波信号不满足所述预设条件，调用所述第一输出模块1330输出所述最后一个频段对应的滤波信号。
在一些可能的实施例中，所述M个频段连续相接。
在一些可能的实施例中，所述M个频段中相邻频段部分重叠。
在一些可能的实施例中，信号处理装置130还包括：提取模块，用于在第一去噪模块1320得到处理后的滤波信号之后，在第一输出模块1330在所述滤波信号满足预设条件的情况下，输出所述滤波信号之前，提取所述滤波信号中的差频信号；所述预设条件为所述差频信号提取成功。
在一些可能的实施例中，M＝2；
所述M个频段为第一低通频段及第一高通频段；所述第一低通频段为频率小于第一阈值的频段，所述第一高通频段为频率大于第二阈值的频段；
所述第一低通频段的带宽大于所述第一高通频段的带宽；所述第一阈值大于或等于所述第二阈值。
在一些可能的实施例中，所述预设顺序为：
所述第一低通频段、所述第一高通频段；或者
所述第一高通频段、所述第一低通频段。
在一些可能的实施例中，M＝2；
所述M个频段为第二低通频段及第二高通频段；所述第二低通频段为频率小于第三阈值的频段，所述第二高通频段为频率大于第四阈值的频段；
所述第二低通频段的带宽小于所述第二高通频段的带宽；所述第三阈值大于或等于所述第四阈值。
在一些可能的实施例中，所述预设顺序为：
所述第二低通频段、所述第二高通频段；或者
所述第二高通频段、所述第二低通频段。
在一些可能的实施例中，M＝3；
所述M个频段为第三低通频段、带通频段及第三高通频段；所述第三低通频段为频率 小于第五阈值的频段，所述带通频段为频率大于第六阈值且小于第七阈值的频段，所述第三高通频段为频率大于第八阈值的频段；
所述第五阈值大于或等于所述第六阈值，所述第七阈值大于或等于所述第八阈值。
在一些可能的实施例中，所述预设顺序为：
所述第三低通频段、所述带通频段、所述第三高通频段；或者
所述第三高通频段、所述带通频段、所述第三低通频段；或者
所述带通频段、所述第三低通频段、所述第三高通频段；或者。
所述第三低通频段、所述第三高通频段、所述带通频段；或者
所述第三高通频段、所述第三低通频段、所述带通频段；或者
所述带通频段、所述第三高通频段、所述第三低通频段。
在本申请实施例中，可以通过按照预设的频段顺序依次尝试对待处理信号进行去噪处理，以确定噪声信号具体处于哪个频段，从而实现对该待处理信号在全频段范围内的精准去噪，相比于现有技术中只对低频段的噪声进行处理，提升了去噪精度，且扩展了有效探测的范围，进而增大FMCW系统的探测距离。
请参见图20，其示出了本申请另一个示例性实施例提供的信号处理装置的结构示意图。该信号处理装置可以通过软件、硬件或者两者的结合实现。信号处理装置140包括：第二分解模块1410、第二去噪模块1420及第二输出模块1430。其中：
第二分解模块1410，用于对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
第二去噪模块1420，用于根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；
第二输出模块1430，用于输出所述滤波信号。
在一些可能的实施例中，所述目标频段为频率小于第九阈值的频段。
在一些可能的实施例中，所述目标频段为频率大于第十阈值的频段。
在一些可能的实施例中，所述目标频段为频率小于第十一阈值的频段及频率大于第十二阈值的频段；其中，所述第十一阈值小于所述第十二阈值。
本申请实施例提供了几种新的滤波算法，通过增加小波系数分解的精度以及对不同频段的分量进行阈值处理，可以实现不同频段范围内的去噪功能，扩大本申请实施例提供的滤波算法的适用场景。
请参见图15，为本申请实施例提供了另外一种信号处理装置的结构示意图。如图15所示，信号处理装置150可以包括：至少一个处理器1501，至少一个网络接口1504，用户接口1503，存储器1505，至少一个通信总线1502。
其中，通信总线1502用于实现这些组件之间的连接通信。
其中，用户接口1503可以包括显示屏(Display)、摄像头(Camera)，可选用户接口1503还可以包括标准的有线接口、无线接口。
其中，网络接口1504可选的可以包括标准的有线接口、无线接口(如WI-FI接口)。
其中，处理器1501可以包括一个或者多个处理核心。处理器1501利用各种借口和线路连接整个信号处理装置150内的各个部分，通过运行或执行存储在存储器1505内的指令、程序、代码集或指令集，以及调用存储在存储器1505内的数据，执行信号处理装置150的各种功能和处理数据。可选的，处理器1501可以采用数字信号处理(Digital Signal Processing，DSP)、现场可编程门阵列(Field-Programmable Gate Array，FPGA)、可编程逻辑阵列(Programmable Logic Array，PLA)中的至少一种硬件形式来实现。处理器1501可集成中央处理器(Central Processing Unit，CPU)、图像处理器(Graphics Processing Unit，GPU)和调制解调器等中的一种或几种的组合。其中，CPU主要处理操作系统、用户界面 和应用程序等；GPU用于负责显示屏所需要显示的内容的渲染和绘制；调制解调器用于处理无线通信。可以理解的是，上述调制解调器也可以不集成到处理器1501中，单独通过一块芯片进行实现。
其中，存储器1505可以包括随机存储器(Random Access Memory，RAM)，也可以包括只读存储器(Read-Only Memory)。可选的，该存储器1505包括非瞬时性计算机可读介质(non-transitory computer-readable storage medium)。存储器1505可用于存储指令、程序、代码、代码集或指令集。存储器1505可包括存储程序区和存储数据区，其中，存储程序区可存储用于实现操作系统的指令、用于至少一个功能的指令(比如触控功能、声音播放功能、图像播放功能等)、用于实现上述各个方法实施例的指令等；存储数据区可存储上面各个方法实施例中涉及到的数据等。存储器1505可选的还可以是至少一个位于远离前述处理器1501的存储装置。如图15所示，作为一种计算机存储介质的存储器1505中可以包括操作系统、网络通信模块、用户接口模块以及信号处理应用程序。
在图15所示的信号处理装置150中，用户接口1503主要用于为用户提供输入的接口，获取用户输入的数据；而处理器1501可以用于调用存储器1505中存储的信号处理应用程序，并具体执行以下操作：
对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；其中，所述待滤波频段为M个频段中的任意一个频段，所述M个频段覆盖全频段范围，M≥2；
在所述滤波信号满足预设条件的情况下，输出所述滤波信号；
在所述滤波信号不满足所述预设条件的情况下，按照预设顺序将下一个待滤波频段确定为所述待滤波频段，并执行所述根据待滤波频段确定目标分量层数，对目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号，直至所述M个频段中的最后一个频段对应的滤波信号不满足所述预设条件，输出所述最后一个频段对应的滤波信号。
在一些可能的实施例中，所述M个频段连续相接。
在一些可能的实施例中，所述M个频段中相邻频段部分重叠。
在一些可能的实施例中，所述处理器1501得到处理后的滤波信号之后，所述在所述滤波信号满足预设条件的情况下，输出所述滤波信号之前，还用于执行：提取所述滤波信号中的差频信号；
所述预设条件为所述差频信号提取成功。
在一些可能的实施例中，M＝2；
所述M个频段为第一低通频段及第一高通频段；所述第一低通频段为频率小于第一阈值的频段，所述第一高通频段为频率大于第二阈值的频段；
所述第一低通频段的带宽大于所述第一高通频段的带宽；所述第一阈值大于或等于所述第二阈值。
在一些可能的实施例中，所述预设顺序为：
所述第一低通频段、所述第一高通频段；或者
所述第一高通频段、所述第一低通频段。
在一些可能的实施例中，M＝2；
所述M个频段为第二低通频段及第二高通频段；所述第二低通频段为频率小于第三阈值的频段，所述第二高通频段为频率大于第四阈值的频段；
所述第二低通频段的带宽小于所述第二高通频段的带宽；所述第三阈值大于或等于所 述第四阈值。
在一些可能的实施例中，所述预设顺序为：
所述第二低通频段、所述第二高通频段；或者
所述第二高通频段、所述第二低通频段。
在一些可能的实施例中，M＝3；
所述M个频段为第三低通频段、带通频段及第三高通频段；所述第三低通频段为频率小于第五阈值的频段，所述带通频段为频率大于第六阈值且小于第七阈值的频段，所述第三高通频段为频率大于第八阈值的频段；
所述第五阈值大于或等于所述第六阈值，所述第七阈值大于或等于所述第八阈值。
在一些可能的实施例中，所述预设顺序为：
所述第三低通频段、所述带通频段、所述第三高通频段；或者
所述第三高通频段、所述带通频段、所述第三低通频段；或者
所述带通频段、所述第三低通频段、所述第三高通频段；或者。
所述第三低通频段、所述第三高通频段、所述带通频段；或者
所述第三高通频段、所述第三低通频段、所述带通频段；或者
所述带通频段、所述第三高通频段、所述第三低通频段。
在本申请实施例中，可以通过按照预设的频段顺序依次尝试对待处理信号进行去噪处理，以确定噪声信号具体处于哪个频段，从而实现对该待处理信号在全频段范围内的精准去噪，相比于现有技术中只对低频段的噪声进行处理，提升了去噪精度，且扩展了有效探测的范围，进而增大FMCW系统的探测距离。
本申请实施例还提供了一种计算机可读存储介质，该计算机可读存储介质中存储有指令，当其在计算机或处理器上运行时，使得计算机或处理器执行上述图5所示实施例中的一个或多个步骤。上述信号处理装置的各组成模块如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在所述计算机可读取存储介质中。
请参见图16，为本申请实施例提供了另外一种信号处理装置的结构示意图。如图16所示，所述信号处理装置160可以包括：至少一个处理器1601，至少一个网络接口1604，用户接口1603，存储器1605，至少一个通信总线1602。
其中，通信总线1602用于实现这些组件之间的连接通信。
其中，用户接口1603可以包括显示屏(Display)、摄像头(Camera)，可选用户接口1603还可以包括标准的有线接口、无线接口。
其中，网络接口1604可选的可以包括标准的有线接口、无线接口(如WI-FI接口)。
其中，处理器1601可以包括一个或者多个处理核心。处理器1601利用各种借口和线路连接整个信号处理装置160内的各个部分，通过运行或执行存储在存储器1605内的指令、程序、代码集或指令集，以及调用存储在存储器1605内的数据，执行信号处理装置160的各种功能和处理数据。可选的，处理器1601可以采用数字信号处理(Digital Signal Processing，DSP)、现场可编程门阵列(Field-Programmable Gate Array，FPGA)、可编程逻辑阵列(Programmable Logic Array，PLA)中的至少一种硬件形式来实现。处理器1601可集成中央处理器(Central Processing Unit，CPU)、图像处理器(Graphics Processing Unit，GPU)和调制解调器等中的一种或几种的组合。其中，CPU主要处理操作系统、用户界面和应用程序等；GPU用于负责显示屏所需要显示的内容的渲染和绘制；调制解调器用于处理无线通信。可以理解的是，上述调制解调器也可以不集成到处理器1601中，单独通过一块芯片进行实现。
其中，存储器1605可以包括随机存储器(Random Access Memory，RAM)，也可以包括只读存储器(Read-Only Memory)。可选的，该存储器1605包括非瞬时性计算机可读介 质(non-transitory computer-readable storage medium)。存储器1605可用于存储指令、程序、代码、代码集或指令集。存储器1605可包括存储程序区和存储数据区，其中，存储程序区可存储用于实现操作系统的指令、用于至少一个功能的指令(比如触控功能、声音播放功能、图像播放功能等)、用于实现上述各个方法实施例的指令等；存储数据区可存储上面各个方法实施例中涉及到的数据等。存储器1605可选的还可以是至少一个位于远离前述处理器1601的存储装置。如图16所示，作为一种计算机存储介质的存储器1605中可以包括操作系统、网络通信模块、用户接口模块以及信号处理应用程序。
在图16所示的信号处理装置160中，用户接口1603主要用于为用户提供输入的接口，获取用户输入的数据；而处理器1601可以用于调用存储器1605中存储的信号处理应用程序，并具体执行以下操作：
对待处理信号进行N级分解，得到2 N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；
输出所述滤波信号。
在一些可能的实施例中，所述目标频段为频率小于第九阈值的频段。
在一些可能的实施例中，所述目标频段为频率大于第十阈值的频段。
在一些可能的实施例中，所述目标频段为频率小于第十一阈值的频段及频率大于第十二阈值的频段；其中，所述第十一阈值小于所述第十二阈值。
本申请实施例提供了几种新的滤波算法，通过增加小波系数分解的精度以及对不同频段的分量进行阈值处理，可以实现不同频段范围内的去噪功能，扩大本申请实施例提供的滤波算法的适用场景。
本申请实施例还提供了一种计算机可读存储介质，该计算机可读存储介质中存储有指令，当其在计算机或处理器上运行时，使得计算机或处理器执行上述图9所示实施例中的一个或多个步骤。上述信号处理装置的各组成模块如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在所述计算机可读取存储介质中。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时，全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者通过所述计算机可读存储介质进行传输。所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(Digital Subscriber Line，DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质，(例如，软盘、硬盘、磁带)、光介质(例如，数字多功能光盘(Digital Versatile Disc，DVD))、或者半导体介质(例如，固态硬盘(Solid State Disk，SSD))等。
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，可以通过计算机程序来指令相关的硬件来完成，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的实施例的流程。而前述的存储介质包括：制度存储器(Read Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等各种可存储程序代码的介质。在不冲突的情况下，本实施例和实施方案中的技术特征可以任意组合。
以上所述的实施例仅仅是本申请的优选实施例方式进行描述，并非对本申请的范围进行限定，在不脱离本申请的设计精神的前提下，本领域普通技术人员对本申请的技术方案作出的各种变形及改进，均应落入本申请的权利要求书确定的保护范围内。

Claims (13)

  1. 一种信号处理方法，其特征在于，包括：
    对待处理信号进行N级分解，得到2N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
    根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；其中，所述待滤波频段为M个频段中的任意一个频段，所述M个频段覆盖全频段范围，M≥2；
    在所述滤波信号满足预设条件的情况下，输出所述滤波信号；
    在所述滤波信号不满足所述预设条件的情况下，按照预设顺序将下一个待滤波频段确定为所述待滤波频段，并执行所述根据待滤波频段确定目标分量层数，对目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号，直至所述M个频段中的最后一个频段对应的滤波信号不满足所述预设条件，输出所述最后一个频段对应的滤波信号。
  2. 如权利要求1所述的方法，其特征在于，所述M个频段连续相接。
  3. 如权利要求1所述的方法，其特征在于，所述M个频段中相邻频段部分重叠。
  4. 如权利要求1-3任一项所述的方法，其特征在于，所述得到处理后的滤波信号之后，所述在所述滤波信号满足预设条件的情况下，输出所述滤波信号之前，所述方法还包括：提取所述滤波信号中的差频信号；
    所述预设条件为所述差频信号提取成功。
  5. 如权利要求1所述的方法，其特征在于，M＝2；
    所述M个频段为第一低通频段及第一高通频段；所述第一低通频段为频率小于第一阈值的频段，所述第一高通频段为频率大于第二阈值的频段；
    所述第一低通频段的带宽大于所述第一高通频段的带宽；所述第一阈值大于或等于所述第二阈值。
  6. 如权利要求5所述的方法，其特征在于，所述预设顺序为：
    所述第一低通频段、所述第一高通频段；或者
    所述第一高通频段、所述第一低通频段。
  7. 如权利要求1所述的方法，其特征在于，M＝2；
    所述M个频段为第二低通频段及第二高通频段；所述第二低通频段为频率小于第三阈值的频段，所述第二高通频段为频率大于第四阈值的频段；
    所述第二低通频段的带宽小于所述第二高通频段的带宽；所述第三阈值大于或等于所述第四阈值。
  8. 如权利要求7所述的方法，其特征在于，所述预设顺序为：
    所述第二低通频段、所述第二高通频段；或者
    所述第二高通频段、所述第二低通频段。
  9. 如权利要求1所述的方法，其特征在于，M＝3；
    所述M个频段为第三低通频段、带通频段及第三高通频段；所述第三低通频段为频率小于第五阈值的频段，所述带通频段为频率大于第六阈值且小于第七阈值的频段，所述第三高通频段为频率大于第八阈值的频段；
    所述第五阈值大于或等于所述第六阈值，所述第七阈值大于或等于所述第八阈值。
  10. 如权利要求9所述的方法，其特征在于，所述预设顺序为：
    所述第三低通频段、所述带通频段、所述第三高通频段；或者
    所述第三高通频段、所述带通频段、所述第三低通频段；或者
    所述带通频段、所述第三低通频段、所述第三高通频段；或者
    所述第三低通频段、所述第三高通频段、所述带通频段；或者
    所述第三高通频段、所述第三低通频段、所述带通频段；或者
    所述带通频段、所述第三高通频段、所述第三低通频段。
  11. 一种信号处理装置，其特征在于，包括：
    第一分解模块，用于对待处理信号进行N级分解，得到2N个分量；其中，N≥2，所述待处理信号为带噪声的信号；
    第一去噪模块，用于根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号；其中，所述待滤波频段为M个频段中的任意一个频段，所述M个频段覆盖全频段范围，M≥2；
    第一输出模块，用于在所述滤波信号满足预设条件的情况下，输出所述滤波信号；
    第一确定模块，用于在所述滤波信号不满足所述预设条件的情况下，按照预设顺序将下一个待滤波频段确定为所述待滤波频段，并调用所述第一去噪模块根据待滤波频段确定目标分量层数，对所述目标分量层数中位于所述待滤波频段的分量进行小波阈值去噪，得到处理后的滤波信号，直至所述M个频段中的最后一个频段对应的滤波信号不满足所述预设条件，调用所述第一输出模块输出所述最后一个频段对应的滤波信号。
  12. 一种信号处理装置，其特征在于，包括：处理器和存储器；其中，所述存储器存储有计算机程序，所述计算机程序适于由处理器加载并执行如权利要求1-10任一项的方法步骤。
  13. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如权利要求1-10任一项所述的方法。
PCT/CN2021/086497 2021-04-12 2021-04-12 信号处理方法、装置及可读存储介质 WO2022217406A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202180095632.9A CN117480403A (zh) 2021-04-12 2021-04-12 信号处理方法、装置及可读存储介质
PCT/CN2021/086497 WO2022217406A1 (zh) 2021-04-12 2021-04-12 信号处理方法、装置及可读存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/086497 WO2022217406A1 (zh) 2021-04-12 2021-04-12 信号处理方法、装置及可读存储介质

Publications (1)

Publication Number Publication Date
WO2022217406A1 true WO2022217406A1 (zh) 2022-10-20

Family

ID=83639362

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/086497 WO2022217406A1 (zh) 2021-04-12 2021-04-12 信号处理方法、装置及可读存储介质

Country Status (2)

Country Link
CN (1) CN117480403A (zh)
WO (1) WO2022217406A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110255710A1 (en) * 2010-04-14 2011-10-20 Keisuke Toyama Signal processing apparatus, signal processing method, and program
CN107144829A (zh) * 2017-06-29 2017-09-08 南京信息工程大学 一种高效的激光雷达回波信号去噪方法
CN107247933A (zh) * 2017-05-27 2017-10-13 北京理工大学 一种烟雾环境中fmcw激光定距系统差频信号提取方法
CN107315991A (zh) * 2017-05-05 2017-11-03 华南理工大学 一种基于小波阈值去噪的ifra频响曲线去噪方法
CN108712222A (zh) * 2018-05-17 2018-10-26 广东工业大学 一种协作频谱感知方法及相关装置
CN109001703A (zh) * 2018-08-10 2018-12-14 南京信息工程大学 一种基于小波包多阈值处理的海杂波去噪办法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110255710A1 (en) * 2010-04-14 2011-10-20 Keisuke Toyama Signal processing apparatus, signal processing method, and program
CN107315991A (zh) * 2017-05-05 2017-11-03 华南理工大学 一种基于小波阈值去噪的ifra频响曲线去噪方法
CN107247933A (zh) * 2017-05-27 2017-10-13 北京理工大学 一种烟雾环境中fmcw激光定距系统差频信号提取方法
CN107144829A (zh) * 2017-06-29 2017-09-08 南京信息工程大学 一种高效的激光雷达回波信号去噪方法
CN108712222A (zh) * 2018-05-17 2018-10-26 广东工业大学 一种协作频谱感知方法及相关装置
CN109001703A (zh) * 2018-08-10 2018-12-14 南京信息工程大学 一种基于小波包多阈值处理的海杂波去噪办法

Also Published As

Publication number Publication date
CN117480403A (zh) 2024-01-30

Similar Documents

Publication Publication Date Title
CN109407071B (zh) 雷达测距方法、雷达测距装置、无人机和存储介质
CN110376559B (zh) 单通道雷达主瓣多源干扰分离方法、装置和设备
WO2022061597A1 (zh) 一种信号噪声滤除方法、装置、存储介质及激光雷达
CN114002658B (zh) 基于点迹曲线关联曲线分离的雷达目标微动特征提取方法
WO2022061596A1 (zh) 一种信号噪声滤除方法、装置、存储介质及激光雷达
WO2022217407A1 (zh) 信号处理方法、装置及可读存储介质
CN111045002A (zh) 基于TRT和SNuFFT的机动目标相参积累方法
Dubey et al. Region based single-stage interference mitigation and target detection
KR102146156B1 (ko) 오브젝트 검출 방법 및 장치
CN112444814B (zh) 基于pcie光纤采集卡的数字阵列天气雷达信号处理器
WO2022217406A1 (zh) 信号处理方法、装置及可读存储介质
WO2022000333A1 (zh) 一种雷达探测方法及相关装置
JP2019194583A (ja) レーダ信号の処理
WO2022000332A1 (zh) 一种雷达探测方法及相关装置
JP3773779B2 (ja) レーダ信号処理装置
JP5933245B2 (ja) クラッタ抑圧装置
JP5564244B2 (ja) 観測信号処理装置
WO2022061598A1 (zh) 一种信号噪声滤除方法、装置、存储介质及激光雷达
KR102049402B1 (ko) 레이더 시스템의 씨파를 기반으로 하는 신호 처리 방법 및 그 장치
CN115561746A (zh) Fmcw雷达系统中的干扰抑制
CN114305354A (zh) 一种生命特征检测方法及装置
CN110231603B (zh) 一种基于gmti的快速解算目标速度的方法
CN115047419A (zh) 一种毫米波雷达恒虚警率检测方法及装置
CN113109796A (zh) 一种目标检测方法及装置
JP2020112564A (ja) 信号処理装置及び信号処理方法

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: 21936302

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 202180095632.9

Country of ref document: CN

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21936302

Country of ref document: EP

Kind code of ref document: A1