Rapid target identification system with low storage complexity

By converting the flight time of the photon counting lidar into a complex index form and cumulatively, the problem of data storage and processing bottlenecks is solved, the recognition effect and storage efficiency of the lidar are improved, and the impact of ambient light is reduced.

CN118938247APending Publication Date: 2024-11-12ANHUI DAOJI QUANTUM TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411344644.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-25
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing photon counting lidars have severe time resolution losses and resolution compression caused by data storage and processing bottlenecks during the target recognition process, which affects the recognition effect.

Method used

By converting the flight time of the photon into a complex exponential form, and accumulating its corresponding complex exponential form parameters when the photon is received, only the accumulated value is stored, instead of storing the flight time one by one, the FPGA control module generates control signals and processes Data to generate target estimation parameters.

Benefits of technology

It greatly saves storage space, avoids large losses in time resolution and resolution compression, improves the recognition effect of lidar, and reduces inaccurate target recognition caused by ambient light.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118938247A_ABST
    Figure CN118938247A_ABST
Patent Text Reader

Abstract

The invention discloses a rapid target recognition system with low storage complexity, relates to the technical field of laser radars, and solves the technical problem of serious data storage and processing bottleneck caused by the fact that an existing photon counting laser radar recognizes a target in a photon counting mode. Comprising a photon emission module which emits photons after receiving a control signal generated by an FPGA control module; the photon receiving module is used for transmitting reflected photon data to the FPGA control module in an electric signal form after receiving the reflected photons; the FPGA control module is used for generating a control signal and processing photon data to obtain flight time corresponding to photons; the data processing module generates a time parameter average value according to the flight time; generating a target estimation parameter according to the time parameter average value; the calculation and storage complexity of the photon data is irrelevant to the resolution of the laser radar equipment and the number of detected photons, and the storage space is greatly saved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of laser radar technology and relates to laser radar technology, specifically a fast target recognition system with low storage complexity. Background Art

[0002] Photon counting lidar has become a valuable tool for 3D depth imaging and target identification because it can achieve high-precision detection over long distances. LiDAR emits light pulses and uses single-photon avalanche diodes (SPADs) to detect received photons. The time-to-digital converter (TDC) can accumulate time-of-flight (ToF) data over a series of light pulses, which can be represented by a histogram to provide surface depth and intensity information for each pixel in the detection scene, thereby achieving the purpose of target identification.

[0003] Existing photon counting lidars usually identify targets by counting photons; this traditional method requires a large amount of flight time data, which may lead to serious data storage and processing bottlenecks; as a result, practitioners often resort to using coarser time bin widths in TCSPC histograms, a technique usually called "coarse binning"; but this processing method usually results in a significant loss of temporal resolution and a limited trade-off in compression resolution, affecting the recognition effect of the lidar. Summary of the invention

[0004] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes a fast target recognition system with low storage complexity, which is used to solve the technical problem that the existing photon counting laser radar identifies the target by counting photons, resulting in serious data storage and processing bottlenecks; the present application solves the above problems by converting the flight time of the photon into a complex exponential form; when a photon is received, the corresponding complex exponential form parameters are accumulated; instead of storing the flight time of the photons one by one, the flight time of the corresponding photon is converted into a complex exponential form when the photon is received; then the accumulated value of the parameter is accumulated with the previously stored value to obtain the accumulated value of this time; only the accumulated value of this time is stored; and the target estimation parameters are generated according to the accumulated value to solve the above problems.

[0005] To achieve the above-mentioned purpose, the first aspect of the present application provides a fast target recognition system with low storage complexity, comprising: a photon transmitting module, a photon receiving module, an FPGA control module, and a data processing module;

[0006] The photon emission module emits photons after receiving a control signal generated by an FPGA control module connected thereto; the control signal includes a pulse signal generator control signal and an acousto-optic modulator control signal;

[0007] The photon receiving module: after receiving the reflected photons, sends the reflected photon data in the form of electrical signals to the FPGA control module connected thereto;

[0008] The FPGA control module is used to generate a control signal and process photon data to obtain the flight time corresponding to the photon; and send the flight time to the data processing module connected thereto;

[0009] The data processing module is used to obtain the flight time, generate a time parameter average value according to the flight time, and generate a target estimation parameter according to the time parameter average value.

[0010] The present application generates a control signal through an FPGA control module to control a photon emission model to emit photons. After the photons contact a target object, they are reflected. The reflected photons are received by a photon receiving module. The received photon data is processed by the FPGA control module to obtain the flight time of the corresponding photons; the flight time is converted into a complex exponential form; when a photon is received, the corresponding complex exponential form parameters are accumulated; there is no need to store the flight time of the photons one by one, but only to convert the flight time of the corresponding photons into a complex exponential form when the photons are received; and then accumulate the accumulated value of the parameters stored previously to obtain the accumulated value of this time; only the accumulated value of this time is stored, so that the calculation and storage complexity of the photon data is independent of the resolution of the laser radar device and the number of photons detected, but only related to the number of surfaces of the detected target; the storage space is greatly saved; at the same time, it will not cause a significant loss of time resolution, and the trade-off of compression resolution is limited; the recognition effect of the laser radar is effectively improved.

[0011] Preferably, the photon transmitting module is composed of a pulse signal generator, a laser, an acousto-optic modulator and a collimator; the photon receiving module is composed of a collimator, an acousto-optic modulator, a single photon detector and a signal processing circuit.

[0012] Preferably, generating a parameter average value according to the flight time includes:

[0013] The flight time of the currently received photon is recorded as tp; through the formula Convert the flight time into a complex exponential parameter Φ p ; Where p is the number of the corresponding received photon, and w is the sampling frequency;

[0014] Determine whether the number p is equal to the set number n;

[0015] If yes, get the last parameter accumulated value And through the formula Calculate the average parameter z p , according to the time parameter average z pGenerate target estimation parameters;

[0016] If not, get the last parameter accumulated value. By formula X p =X p-1 +Φ p Calculate the cumulative value of the parameter X p .

[0017] The present application converts the flight time of photons into a complex exponential form. Within a set number of times, each time a photon is received, the corresponding complex exponential form parameters are accumulated. There is no need to store the flight time of photons one by one. It is only necessary to convert the flight time of the corresponding photon into a complex exponential form when the photon is received; and then add it to the previously stored parameter accumulated value to obtain the accumulated value of this time. Only the accumulated value of this time is stored, which greatly saves storage space; and at the same time is not limited by time resolution.

[0018] Preferably, the sampling frequency is given by the formula Calculated; wherein T is the number of timing boxes into which the collected flight time t is divided, and l is an optional constant; l = 1, 2, ..., T-1.

[0019] The present application sets the sampling frequency to a specific frequency to facilitate elimination of the contribution of ambient light in the characteristic function, thereby reducing the impact of ambient light in subsequent data processing and avoiding inaccurate target recognition due to ambient light.

[0020] Preferably, generating the target estimation parameter according to the time parameter average value includes:

[0021] Get the average value of the parameter z this time; and the characteristic function Ψ obeyed by the flight time; at the same time, the complex exponential function Φ of the characteristic function Ψ and the flight time = e iwt The same principle applies to the mathematical expectation of Here, k is the number of the reflection surface used for accumulation, K is the total number of reflection surfaces, and α k is the probability that the detected photon comes from the kth surface, that is, the corresponding reflectivity; α 0 is the probability that the detected photon comes from the ambient light; is the discrete Fourier transform of the impulse response function; w is the sampling frequency; T is the number of time bins divided by the acquired flight time t; d k is the depth of the kth reflection surface;

[0022] By formula The target estimation parameter θ is calculated; where n is the number of photons corresponding to the parameter average value z; γ is the positive definite Hermitian weight matrix; the estimated parameter θ includes the reflectivity and depth of the target.

[0023] Compared with the prior art, the beneficial effects of this application are:

[0024] 1. The present application generates a control signal through an FPGA control module to control the photon emission model to emit photons. After the photons contact the target object, they are reflected. The reflected photons are received by the photon receiving module. The received photon data is processed by the FPGA control module to obtain the flight time of the corresponding photons; the flight time is converted into a complex exponential form; when a photon is received, the corresponding complex exponential form parameters are accumulated; there is no need to store the flight time of the photons one by one, only the flight time of the corresponding photons needs to be converted into a complex exponential form when the photons are received; and then the accumulated value of the parameters stored previously is accumulated to obtain the accumulated value of this time; only the accumulated value of this time is stored, so that the calculation and storage complexity of the photon data is independent of the resolution of the laser radar device and the number of photons detected, but only related to the number of surfaces of the detected target; the storage space is greatly saved; at the same time, it will not cause a significant loss of time resolution, and the trade-off of compression resolution is limited; the recognition effect of the laser radar is effectively improved.

[0025] 2. The present application sets the sampling frequency to a specific frequency to facilitate elimination of the contribution of ambient light in the characteristic function, thereby reducing the impact of ambient light in subsequent data processing and avoiding inaccurate target recognition due to ambient light. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0027] Figure 1 This is a schematic diagram of the principle of the rapid target recognition system in this application;

[0028] Figure 2 Schematic diagram of the operating principle of the rapid target recognition system in this application. DETAILED DESCRIPTION

[0029] The technical solution of the present application will be described clearly and completely below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0030] See also Figure 1-Figure 2, the first aspect of the present application provides a fast target recognition system with low storage complexity, including: a photon transmitting module, a photon receiving module, an FPGA control module, and a data processing module;

[0031] Photon emission module: after receiving the control signal generated by the FPGA control module connected to it, it emits photons; the photon emission module is composed of a pulse signal generator, a laser, an acousto-optic modulator and a collimator; the control signal includes a pulse signal generator control signal and an acousto-optic modulator control signal; when the pulse signal generator receives the pulse signal generator control signal, it sends a pulse signal for controlling the laser; when the acousto-optic modulator receives the acousto-optic modulator control signal, it modulates the frequency, amplitude, phase and other parameters of the light according to the optical modulator control signal;

[0032] Photon receiving module: after receiving the reflected photons, the reflected photon data is sent to the FPGA control module connected to it in the form of electrical signals; the photon receiving module is composed of a collimator, an acousto-optic modulator, a single-photon detector and a signal processing circuit; the echo signal of the light pulse after being reflected by the surface of the object is received by the single-photon detector through the reverse process of the emission, and the photon data is obtained; the photon data includes the time of photon reception, the number of photons and the number, etc.

[0033] FPGA control module: used to generate control signals and process photon data to obtain the flight time corresponding to the photons; and send the flight time to the data processing module connected to it;

[0034] Data processing module: obtain flight time, which is the time interval from when the photon is emitted from the photon transmitting module to when it is received by the photon receiving module; generate the average value of time parameters based on the flight time; generate target estimation parameters based on the average value of time parameters; target estimation parameters include target depth and reflectivity, etc.

[0035] In this embodiment, the FPGA control module generates a control signal to control the photon emission model to emit photons. After the photons contact the target object, they are reflected. The reflected photons are received by the photon receiving module. The FPGA control module is used to process the received photon data to obtain the flight time of the corresponding photons; the flight time is converted into a complex exponential form; when a photon is received, the corresponding complex exponential form parameters are accumulated; there is no need to store the flight time of the photons one by one, only when the photon is received, the flight time of the corresponding photon is converted into a complex exponential form; and then the accumulated value of the parameter stored previously is accumulated to obtain the accumulated value of this time; only the accumulated value of this time is stored, so that the calculation and storage complexity of the photon data is independent of the resolution of the laser radar device and the number of photons detected, but only related to the number of surfaces of the detected target; the storage space is greatly saved; at the same time, it will not cause a significant loss of time resolution, and the trade-off of compression resolution is limited; the recognition effect of the laser radar is effectively improved.

[0036] Generates parameter averages based on flight time, including:

[0037] The flight time of the currently received photon is recorded as tp; through the formula Convert the flight time into a complex exponential parameter Φ p ; Where p is the number of the corresponding received photon. When the first photon is received, its corresponding number is 1, and w is the sampling frequency;

[0038] Determine whether the number p is equal to the set number n;

[0039] If yes, get the last parameter accumulated value By formula Calculate the average parameter z p , and according to the time parameter average z p Generate target estimation parameters; k in the above formula is a number used for accumulation, k = 1, 2, ..., p-1;

[0040] If not, get the last parameter accumulated value. By formula X p =X p-1 +Φ p Calculate the cumulative value of the parameter X p .

[0041] The set number is set according to expert experience to ensure that there is a certain number of photons as the sample size when estimating the target parameters, that is, the target parameter estimation can be performed only after a certain number of photons are received; in this application, the number p of the received photons is used to determine whether the sample size of the photons meets the set requirements. When the number of the received photons is equal to the set number, it means that the number of photons at this time has reached the set sample size, and the target estimation parameters can be performed; at this time, the parameter average value is calculated, and the target estimation parameters are generated based on the parameter average value.

[0042] In this embodiment, the flight time of the photon is converted into a complex exponential form. Within a set number of times, each time a photon is received, the corresponding complex exponential form parameter is accumulated. There is no need to store the flight time of the photon one by one. It is only necessary to convert the flight time of the corresponding photon into a complex exponential form when the photon is received; then add it to the previously stored parameter accumulated value to obtain the accumulated value of this time; only the accumulated value of this time is stored, which greatly saves storage space.

[0043] The sampling frequency is given by the formula Calculated; wherein T is the number of time bins into which the collected flight time t is divided, and l is an optional constant; l=1, 2, ..., T-1; This embodiment facilitates the elimination of the contribution of ambient light in the characteristic function by setting the sampling frequency to a specific frequency, so as to reduce the influence of ambient light in subsequent data processing and avoid the inaccurate target recognition caused by ambient light.

[0044] Generate target estimated parameters based on the time parameter average, including:

[0045] Get the average value of the parameter z; get the characteristic function Ψ obeyed by the flight time; at the same time, the complex exponential function Φ of the characteristic function Ψ and the flight time = e iwt The same principle applies to the mathematical expectation of Here, k is the number of the reflection surface used for accumulation, K is the total number of reflection surfaces, and α k is the probability that the detected photon comes from the kth surface, that is, the corresponding reflectivity; α 0 is the probability that the detected photon comes from the ambient light; is the discrete Fourier transform of the impulse response function; w is the sampling frequency; T is the number of time bins divided by the acquired flight time t; d k is the depth of the kth reflection surface;

[0046] By formula The target estimation parameter θ is calculated; where n is the number of photons corresponding to the parameter average value z; γ is the positive definite Hermitian weight matrix; the estimated parameter θ includes the reflectivity and depth of the target.

[0047] The flight time of photons obeys a mixed distribution that includes the influence of ambient stray light. This distribution is related to the target estimation parameters of different reflective surfaces, including reflectivity and depth, etc.; the characteristic function of the probability distribution obeyed by the photon flight time is obtained, and the characteristic function has the same mathematical expectation principle as the complex exponential function of the flight time; the characteristic function and the probability density distribution function of the flight time provide the same information of the flight time distribution data; the mathematical model constructed in this way can obtain the above-mentioned target parameter information through limited data under low signal-to-noise ratio through parameter estimation method, thereby achieving better target reconstruction; the flight time of the photon is mapped using a complex exponential function, and then accumulated to obtain the parameter average value, which obeys the central limit theorem; therefore, the required surface depth and reflectivity and other parameters can be obtained through parameter estimation.

[0048] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0049] How this application works:

[0050] The present application generates a control signal through an FPGA control module to control a photon emission model to emit photons. After the photons contact a target object, they are reflected. The reflected photons are received by a photon receiving module. The received photon data is processed by the FPGA control module to obtain the flight time of the corresponding photons; the flight time is converted into a complex exponential form; when a photon is received, the corresponding complex exponential form parameters are accumulated; there is no need to store the flight time of the photons one by one, but only to convert the flight time of the corresponding photons into a complex exponential form when the photons are received; and then accumulate the accumulated value of the parameters stored previously to obtain the accumulated value of this time; only the accumulated value of this time is stored, so that the calculation and storage complexity of the photon data is independent of the resolution of the laser radar device and the number of photons detected, but only related to the number of surfaces of the detected target; the storage space is greatly saved; at the same time, it will not cause a significant loss of time resolution, and the trade-off of compression resolution is limited; the recognition effect of the laser radar is effectively improved.

[0051] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical method of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. A fast target recognition system with low storage complexity, comprising: Photon transmitting module, photon receiving module, FPGA control module, data processing module; characterized in that, The photon emission module emits photons after receiving a control signal generated by an FPGA control module connected thereto; the control signal includes a pulse signal generator control signal and an acousto-optic modulator control signal; The photon receiving module: after receiving the reflected photons, sends the reflected photon data in the form of electrical signals to the FPGA control module connected thereto; The FPGA control module is used to generate a control signal and process photon data to obtain the flight time corresponding to the photon; and send the flight time to the data processing module connected thereto; The data processing module is used to obtain the flight time, generate a time parameter average value according to the flight time, and generate a target estimation parameter according to the time parameter average value.

2. A fast target recognition system with low storage complexity according to claim 1, characterized in that: The photon emission module is composed of a pulse signal generator, a laser, an acousto-optic modulator and a collimator; the photon receiving module is composed of a collimator, an acousto-optic modulator, a single photon detector and a signal processing circuit.

3. A fast target recognition system with low storage complexity according to claim 1, characterized in that: The generating a parameter average value according to the flight time comprises: The flight time of the currently received photon is recorded as tp; through the formula Convert the flight time into a complex exponential parameter Φ p ; Where p is the number of the corresponding received photon, and w is the sampling frequency; Determine whether the number p is equal to the set number n; If yes, get the last parameter accumulated value And through the formula Calculate the average parameter z p , and according to the time parameter average z p Generate target estimation parameters; If not, get the last parameter accumulated value. By formula X p =X p-1 +Φ p Calculate the cumulative value of the parameter X p .

4. A fast target recognition system with low storage complexity according to claim 3, characterized in that: The sampling frequency is given by the formula Calculated; wherein T is the number of timing boxes into which the collected flight time t is divided, and l is an optional constant; l = 1, 2, ..., T-1.

5. A fast target recognition system with low storage complexity according to claim 1, characterized in that: The generating target estimation parameters according to the time parameter average value comprises: Get the average parameter z of this time; and the characteristic function Ψ obeyed by the flight time; By formula The target estimated parameter θ is calculated, where n is the number of photons corresponding to the parameter average value z, and γ is the positive definite Hermitian weight matrix.