Multispectral and hyperspectral laser radar weak multi-echo signal extraction method

Abstract

The invention discloses a multispectral and hyperspectral laser radar weak multi-echo signal extraction method. According to the method, a B-spline model is adopted to establish models for transmitted pulses and received echoes. Compared with a traditional Gaussian model and a B-spline model, the method has local control characteristics, is more suitable for modeling and analysis of complex waveforms, improves the adaptability to diversified signals, remarkably reduces the probability that noise is misjudged as echo signals by conducting preliminary screening and positioning on received echoes of all channels, and improves the accuracy of noise detection. Therefore, error accumulation is effectively reduced. The process not only improves the accuracy of signal identification, but also greatly reduces the complexity of subsequent data processing, reduces the calculation burden of the system, and lays a solid foundation for refined waveform analysis.

Description

Technical Field

[0001] The present invention belongs to the technical field of remote sensing mapping, and particularly relates to a method for extracting weak multi-echo signals of multi-spectral and hyperspectral lidar. Background Art

[0002] As a new type of remote sensing technology, multi-spectral and hyperspectral lidar has been widely used in terrain surveying, environmental monitoring, agricultural vegetation monitoring and other fields by virtue of its high spatial resolution and rich spectral information. These technologies use lidar systems to obtain detailed information on the surface of ground objects and targets through high-precision laser scanning, and can achieve accurate measurement and analysis in complex environments. The advantage of multi-spectral and hyperspectral lidar is that it can simultaneously obtain spectral data in different bands, greatly improving the accuracy of target recognition and classification.

[0003] Multi-spectral and hyperspectral lidar usually uses a supercontinuum laser as the light source and adopts a single-shot multi-receive detection method to achieve broadband detection ability. The laser interacts with the target surface during transmission, and can even penetrate the gaps between multiple targets to detect the bottom information of complex scenes. The spectral characteristics and distance information of all measured targets within the laser irradiation area are implicitly contained in the backscattered echo data. Through high-precision interpretation of this new type of data, various functions such as target three-dimensional imaging, biomass inversion, and ground object classification and recognition can be realized.

[0004] Although the continuous progress of hardware systems and software technologies has improved the recording accuracy of hyperspectral lidar for target backscattered echo signals, according to the lidar equation, the echo signal intensity is inversely proportional to the square of the propagation distance, resulting in a significant attenuation of the signal with increasing distance. In practical applications, hyperspectral lidar usually uses a supercontinuum laser, and its energy distribution is uneven. At the same time, due to factors such as environmental noise and the collaborative work of multiple detectors, the system performance is more vulnerable to weak echo signals. In addition, due to the possible occlusion of multiple targets in the laser path, the echo signal is not only weak, but also signal overlap may occur in the case of short distances, further increasing the difficulty of signal separation and interpretation. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for extracting weak multi-echo signals of multi-spectral and hyperspectral lidar in view of the above problems existing in the prior art.

[0006] The above object of the present invention is achieved by the following technical means:

[0007] A method for extracting weak multi-echo signals of multi-spectral and hyperspectral lidar includes the following steps:

[0008] Step 1: Establish an emission pulse model and a received echo model based on the B-spline model respectively;

[0009] Step 2: Preprocess the emission pulse and the received echo respectively;

[0010] Step 3: Initialize the characteristic parameters of the emission pulse model and the received echo model respectively;

[0011] Step 4: Extract single echoes and overlapping echoes;

[0012] Step 5: Iterate the single echoes and the overlapping echoes through residual search to obtain the iterated received echo, extract the characteristic parameters of the iterated received echo, then discriminate the iterated received echo, screen out the effective received echoes, and incorporate the characteristic parameters of each effective received echo into the multi-channel characteristic parameter set.

[0013] As described above, the emission pulse model is based on the following formula:

[0014] F T,i (t) = A T,i ×F B (α T,i (t - τ T,i - Δτ T,i ))

[0015] In the formula, F T,i (t) is the emission pulse model in the i-th receiving channel, A T,i , α T,i , τ T,i are respectively the peak value, the full width at half maximum, and the center position of the emission pulse in the i-th receiving channel, Δτ T,i is the position offset of the center position of the emission pulse in the i-th receiving channel, t is the sampling time, i is the serial number of the receiving channel, n is the number of receiving channels, i takes values from 1 to n, F B is the mathematical expression of the B-spline waveform curve;

[0016] As described above, the received echo model is based on the following formula:

[0017]

[0018] f E,i,j (t) = A E,i,j ×F B (α E,i,j (t - τ E,i,j - Δτ E,i,j )

[0019] In the formula, F E,i (t) is the received echo model in the i-th receiving channel, f E,i,j(t) is the basis function of the j-th received echo in the i-th receiving channel, and Δτ E,i,j is the position offset of the center position of the j-th received echo in the i-th receiving channel; A E,i,j , α E,i,j , τ E,i,j are respectively the peak value, the full width at half maximum, and the center position of the j-th received echo in the i-th receiving channel. Each receiving channel includes j received echoes, where j is the serial number of the received echoes in a single receiving channel, and j takes values from 1 to N, and N is the number of received echoes in a single receiving channel.

[0020] Step 2 as described above specifically includes the following steps:

[0021] Step 2.1: Calibrate the center positions of the transmitted pulses and the received echoes in each receiving channel respectively, which specifically includes the following steps:

[0022] Step 2.1.1: First, calculate the signal-to-noise ratios of each receiving channel, and select the channel with the highest signal-to-noise ratio as the reference channel. The signal-to-noise ratios of each receiving channel are calculated based on the following formula:

[0023]

[0024] In the formula, SNR i is the signal-to-noise ratio of the i-th receiving channel, and A' i is the peak value of the i-th receiving channel. The peak value of the i-th receiving channel is the sum of the peak values of the transmitted pulse and each received echo in the i-th receiving channel;

[0025] Step 2.1.2: Taking the reference channel as a reference and the whiteboard as a standard target, calculate the position offset of the center position of the transmitted pulse in each receiving channel relative to the center position of the transmitted pulse in the reference channel, and obtain the position offset Δτ T,i of the center position of the transmitted pulse in the i-th receiving channel, where i takes values from 1 to n;

[0026] Step 2.1.3: Taking the reference channel as a reference and the whiteboard as a standard target, calculate the position offset of the center position of the received echo in each receiving channel relative to the center position of the received echo in the reference channel, and obtain the position offset Δτ E,i,j of the center position of the j-th received echo in the i-th receiving channel, where i takes values from 1 to n and j takes values from 1 to N;

[0027] Step 2.2: Denoise the full waveform of each receiving channel to obtain high-quality waveforms of the full waveforms in each receiving channel. The full waveform of each receiving channel includes the transmitted pulse and each received echo corresponding to the receiving channel. The high-quality waveform of the full waveform of the i-th receiving channel is F τ,i (t).

[0028] The initialization of the characteristic parameters in step 3 as described above is based on the following formula:

[0029] A i,j = y p,i,j

[0030]

[0031] τ i,j = t p,i,j

[0032]

[0033] In the formula, Δt i,j is the difference in inflection point positions, y p,i,j is the peak value, t p,i,j is the peak position, t l,i,j and t r,i,j are the inflection point positions in the left monotonic increasing interval and the right monotonic decreasing interval respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonic increasing interval and the right monotonic decreasing interval respectively, are the means of all inflection point positions in the left monotonic interval and the right monotonic interval respectively;

[0034] When initializing the characteristic parameters of the transmitted pulses in each receiving channel, A i,j is the peak value A of the transmitted pulse in the i-th receiving channel T,i , α i,j is the full width at half maximum α of the transmitted pulse in the i-th receiving channel T,i , τ i,j is the center position τ of the transmitted pulse in the i-th receiving channel T,i , Δt i,j is the difference in inflection point positions of the transmitted pulse in the i-th receiving channel, y p,i,j is the peak value of the transmitted pulse in the i-th receiving channel, t p,i,j is the peak position of the transmitted pulse in the i-th receiving channel, t l,i,j and t r,i,j are the inflection point positions in the left monotonic increasing interval and the right monotonic decreasing interval of the transmitted pulse in the i-th receiving channel respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonic increasing interval and the right monotonic decreasing interval of the transmitted pulse in the i-th receiving channel respectively, are the means of all inflection point positions in the left monotonic interval and the right monotonic interval of the transmitted pulse in the i-th receiving channel respectively. i ranges from 1 to n, each receiving channel includes one transmitted pulse, and j does not take values; the characteristic parameters of the transmitted pulses in each receiving channel are initialized in sequence;

[0035] When initializing the characteristic parameters of each received echo in each receiving channel, A i,j is the peak value of the j-th received echo in the i-th receiving channel A E,i,j , α i,j is the full-width at half maximum of the j-th received echo in the i-th receiving channel α E,i,j , τ i,j is the center position of the j-th received echo in the i-th receiving channel τ E,i,j , Δt i,j is the difference in inflection point positions of the j-th received echo in the i-th receiving channel, y p,i,j is the peak value of the j-th received echo in the i-th receiving channel, t p,i,j is the peak position of the j-th received echo in the i-th receiving channel, t l,i,j and t r,i,j are the inflection point positions in the left monotonic increasing interval and the right monotonic decreasing interval of the j-th received echo in the i-th receiving channel respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonic increasing interval and the right monotonic decreasing interval of the j-th received echo in the i-th receiving channel respectively, are the means of all inflection point positions in the left monotonic interval and the right monotonic interval of the j-th received echo in the i-th receiving channel respectively. i ranges from 1 to n, and j ranges from 1 to N; the characteristic parameters of each received echo in each receiving channel are initialized in sequence.

[0036] Step 4 as described above includes the following steps:

[0037] Step 4.1. Accumulate the high-quality waveform F τ,i (t) of the full waveform in each receiving channel to obtain the accumulated waveform F γ (t), and calculate the characteristic parameters of the transmitted pulse and the received echo of the accumulated waveform respectively, including the following steps:

[0038] Step 4.1.1. The accumulated waveform F γ (t) is calculated based on the following formula:

[0039]

[0040]

[0041] In the formula, ω i is the accumulation weight of the i-th receiving channel;

[0042] Step 4.1.2. Calculate the characteristic parameters of the transmitted pulse of the accumulated waveform according to Step 3. The characteristic parameters of the transmitted pulse of the accumulated waveform include the peak value A of the transmitted pulse of the accumulated waveformγ The full width at half maximum α of the emission pulse of the cumulative waveform γ and the center position τ of the emission pulse of the cumulative waveform γ ;

[0043] Step 4.1.3: Calculate the characteristic parameters of the received echo of the cumulative waveform according to Step 3. The characteristic parameters of the received echo of the cumulative waveform include the peak value, full width at half maximum, and center position of the received echo of the cumulative waveform. The peak value of the j-th received echo of the cumulative waveform is A′ γ,j , the full width at half maximum of the j-th received echo of the cumulative waveform is α′ γ,j , the center position of the j-th received echo of the cumulative waveform is τ′ γ,j , where j takes values from 1 to N.

[0044] As described above, Step 4 further includes the following steps:

[0045] Step 4.2: Extract the single echo and overlapping echo, including the following steps:

[0046] Step 4.2.1: Screen the received echoes that simultaneously meet the following screening conditions:

[0047] Screening condition 1: Retain the received echoes whose intersection value of the full width at half maximum in the monotonically increasing interval and the monotonically decreasing interval in each channel is greater than the noise threshold T i =λ i *S i +M i , where λ i is the threshold coefficient, S i is the mean square deviation of the noise in the i-th receiving channel, and M i is the mean value of the noise in the i-th receiving channel;

[0048] Screening condition 2: Retain the received echoes whose cross-correlation coefficient with the emission pulse in each receiving channel is greater than the set threshold;

[0049] Screening condition 3: If the number of the same y values corresponding to the extreme points of a certain received echo exceeds 2, then the above received echo is a saturated echo, and all saturated echoes are removed;

[0050] Step 4.2.2: Screen out the single echo from all the received echoes screened out in Step 4.2.1. The screening condition for the single echo is that the full width at half maximum of the j-th received echo in the i-th receiving channel is less than or equal to the full width at half maximum of the emission pulse, and there is only a single local extreme point in the waveform of the j-th received echo; and the number of sub-echo components of the j-th received echo of the cumulative waveform is 1;

[0051] Step 4.2.3: Screen out overlapping echoes from all the received echoes screened out in Step 4.2.1. The screening conditions for overlapping echoes are as follows: the full width at half maximum of the j-th received echo in the i-th receiving channel is greater than four-thirds of the full width at half maximum of the transmitted pulse, or there are multiple inflection points in the monotonic interval; and the number of sub-echo components of the j-th received echo in the cumulative waveform is greater than 1.

[0052] As described above, Step 5 includes the following steps:

[0053] Step 5.1: Use a step-by-step iterative method to search for maxima respectively in the single echoes of Step 4.2.2, the overlapping echoes of Step 4.2.3, and the residual signal R m,i of the transmitted pulse, determine the number of new sub-echoes, and add the obtained new sub-echoes to the corresponding received echoes after each iteration to obtain the iterated received echoes. The j-th iterated received echo in the i-th receiving channel is f′ E,i,j (t);

[0054] The residual signal R m,i is calculated based on the following formula:

[0055] R m,i (t) = F T,i (t) - f E,i,j (t)

[0056] In the formula, m is the number of iterations. In each iteration, the peak value, center position, and pulse width of the residual waveform are gradually searched. The peak value, center position, and pulse width of the residual waveform of the j-th received echo in the i-th receiving channel are respectively denoted as A′ E,i,j , τ′ E,i,j , α′ E,i,j , until the maximum value of the residual waveform is less than the noise threshold T i and α′ E,i,j is less than the full width at half maximum α T,i of the transmitted pulse, and the iteration terminates;

[0057] Step 5.2: Sort the iterated received echoes in descending order according to the product of the peak value and the full width at half maximum;

[0058] Step 5.3: Calculate the cross-correlation coefficient ρ i,j (τ) between the sorted iterated received echoes and the transmitted pulses in the corresponding receiving channels in sequence. If the cross-correlation coefficient between a certain iterated received echo and the transmitted pulse in the corresponding receiving channel is not less than the threshold ρ′, it is regarded as a pending valid received echo;

[0059] Step 5.4: Screen out the pending valid echoes that appear repeatedly in at least half of the receiving channels among the pending valid received echoes in Step 5.3 as valid received echoes, and incorporate the characteristic parameters of each valid received echo into the multi-channel characteristic parameter set.

[0060] As described above, each iteration in Step 5.1 includes the following search process:

[0061] Step 5.1.1: Calculate the residual signal between the current transmitted pulse and the received echo;

[0062] Step 5.1.2: Search for local maximum points in the residual signal to determine the center position of the new sub-echo;

[0063] Step 5.1.3: Based on the center position of the newly searched sub-echo, calculate the full width at half maximum of the new sub-echo, determine the left boundary by subtracting the full width at half maximum from the sub-echo center position, determine the right boundary by adding the full width at half maximum, and use the corresponding left and right boundaries as the new time domain range;

[0064] Step 5.1.4: Add the newly extracted sub-echo to the received echo after the corresponding previous iteration;

[0065] Repeat Steps 5.1.2 to 5.1.4 until no more maximum points that meet the following conditions can be found, and then this iteration terminates:

[0066] Condition 1: A′ E,i,j >T i ;

[0067] Condition 2: There is at least one inflection point in the left and right monotonic intervals.

[0068] As described above, ρ i,j (τ) is calculated based on the following formula:

[0069]

[0070] The threshold ρ′ is calculated based on the following formula:

[0071]

[0072] In the formula, is the mean value of ρ N , ω represents the frequency of the transmitted pulse, represents the phase difference between the received echo and the transmitted pulse.

[0073] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of any one of the multi-spectral and hyperspectral lidar weak multi-echo signal extraction methods described above.

[0074] The present invention has the following beneficial effects compared with the prior art:

[0075] (1) The method of the present invention uses a B-spline model to establish models for the transmitted pulse and the received echo respectively. Compared with the traditional Gaussian model, the B-spline model has local control characteristics and is more suitable for the modeling and analysis of complex waveforms, improving the adaptability to diverse signals.

[0076] (2) By preliminarily screening and positioning the received echoes of all channels, the method of the present invention significantly reduces the probability of noise being misjudged as an echo signal, thereby effectively reducing error accumulation. This process not only improves the accuracy of signal recognition but also greatly reduces the complexity of subsequent data processing, lightens the computational burden on the system, and lays a solid foundation for refined waveform analysis.

[0077] (3) During the waveform parameter initialization, the method of the present invention sets a discrimination mechanism, effectively suppressing the pseudo-echo signals and noise interference that may occur during the initial retrieval process, reducing the error points in the point cloud from the source, and ensuring the reliability of data quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is a flowchart of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0079] To facilitate understanding and implementation of the present invention by those of ordinary skill in the art, the present invention will be further described in detail below with reference to embodiments. The embodiments described herein are only for illustrative and explanatory purposes and are not intended to limit the present invention.

[0080] Embodiment 1:

[0081] A method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar, comprising the following steps:

[0082] The received echo signal of the lidar system is the result of the complex interaction between the emitted laser and the objects in the transmission path. The waveform shape of the received echo does not follow a strict Gaussian symmetry and has a high correlation with the transmitted pulse (multi-spectral or hyperspectral lidar uses a single-shot multi-receive mode, and both the transmitted pulse and the received echo can be recorded simultaneously within one receive channel). Therefore, a transmitted pulse model and a received echo model can be established respectively, and the waveform components of the received echo can be obtained more precisely by using the method of gradually decomposing and approximating the transmitted pulse model to the received echo. Due to the restriction of the waveform shape, the traditional Gaussian model is difficult to accurately describe the complex received waveform. Therefore, the present invention uses the B-spline method to model the received echo of multiple receive channels. The B-spline method has higher flexibility and local control characteristics, and can fit complex waveforms more precisely, thus effectively improving the accuracy and robustness of signal modeling.

[0083] Step 1: Perform multi-channel waveform modeling. The present invention respectively establishes a transmitted pulse model and a received echo model based on the B-spline model, which specifically includes the following steps:

[0084] Assume there are n receive channels, and the complete waveform (full waveform) in each channel includes one transmitted pulse and N received echoes. Then the full waveform F i (t) in the i-th receive channel can be expressed as:

[0085] F i (t) = F T,i (t) + F E,i (t) + n i (t) (1)

[0086] In the formula, n is the number of receive channels, i is the serial number of the receive channel, i takes values from 1 to n, F T,i (t) and F E,i (t) are the transmitted pulse model and the received echo model respectively, t represents the sampling time, and n i (t) is the noise term of the i-th receive channel.

[0087] Step 1.1: Establish a transmitted pulse model based on B-spline, specifically:

[0088] To reduce the sensitivity of the modeling to shape changes, the present invention uses the B-spline model for fitting. The B-spline model is obtained by the weighted sum of each control point p k , where the weight is determined by the B-spline basis function B k,l , and can be expressed as:

[0089]

[0090] In the formula, F B (t) is the mathematical expression of the B-spline waveform curve (that is, FB ), p k represents the control point corresponding to the k-th B-spline basis function, B k,l represents the k-th B-spline basis function of order l (l is the order of the B-spline basis function, which determines the smoothness of the curve. In this embodiment, l takes the value of 3, that is, cubic spline), I is the number of B-spline basis functions, k is the serial number of the B-spline basis function, and k takes values from 0 to I.

[0091] For a single receiving channel, its transmitted pulse model can be represented by a B-spline basis function, and the parameter A that describes the peak value of the transmitted pulse T,i and the parameter that describes B k,l (t) (α T,i , τ T,i ) together constitute the knot, then the transmitted pulse model F T,i (t) in the i-th receiving channel can be expressed as:

[0092] F T,i (t) = A T,i ×F B (α T,i (t - τ T,i - Δτ T,i )) (3)

[0093] In the formula, A T,i , α T,i , τ T,i are all characteristic parameters of the transmitted pulse in the i-th receiving channel, and the characteristic parameters they represent are peak value, full width at half maximum, and center position respectively. Δτ T,i represents the position offset of the center position of the transmitted pulse in the i-th receiving channel.

[0094] Step 1.2: Establish a received echo model based on B-spline, specifically: The received echo model in a single receiving channel is the sum of N received echoes in the corresponding receiving channel. Assuming that the j-th received echo is linearly superimposed by m sub-echoes, then the received echo model F E,i (t) in the i-th receiving channel can be expressed as:

[0095]

[0096] In the formula, N is the number of received echoes in a single receiving channel, j is the serial number of the received echoes in a single receiving channel, j takes values from 1 to N, f E,i,j (t) represents the basis function of the j-th received echo in the i-th receiving channel, t represents the sampling time, and Δτ E,i,j represents the position offset of the center position of the j-th received echo in the i-th receiving channel; A E,i,j , α E,i,j , τ E,i,jThey are all characteristic parameters of the j-th received echo in the i-th receiving channel, and the characteristic parameters represented are peak value, full width at half maximum, and center position respectively; due to reasons such as occlusion, the received echoes of adjacent receiving channels may overlap, and thus look like a single received echo. Therefore, a single received echo may correspond to M sub-echo components; respectively represent the -th sub-echo component of the j-th received echo in the i-th receiving channel, and the characteristic parameters represented are peak value, full width at half maximum, and center position respectively.

[0097] Step 2: Preprocess the transmitted pulse and the received echo. The preprocessing includes center position correction (i.e., Figure 1 the timing correction in

[0098] ), and noise suppression, which specifically includes the following steps:

[0099] Step 2.1: In a multi-spectral and hyperspectral lidar system, since there are multiple spectral receiving channels, there may be position offsets between the receiving channels when recording the transmitted pulse and the echo signal. It is necessary to calibrate the center positions of the receiving channels, including calibrating the center positions of the transmitted pulses in each receiving channel and the center positions of the received echoes in each receiving channel. Specifically, it includes the following steps:

[0099] Step 2.1.1: First, calculate the signal-to-noise ratio of each receiving channel, and select the channel with the highest signal-to-noise ratio as the reference channel. The signal-to-noise ratio of each receiving channel is calculated by the following formula:

[0100]

[0101] In the formula, SNR i is the signal-to-noise ratio of the i-th receiving channel, and A′ i is the peak value of the i-th receiving channel (i.e., the sum of the peak values of the transmitted pulse and each received echo in the i-th receiving channel);

[0102] Step 2.1.2: Taking the reference channel as a reference and the whiteboard as a standard target, calculate the position offset of the center position of the transmitted pulse in each receiving channel relative to the center position of the transmitted pulse in the reference channel, that is, obtain the position offset Δτ T,i of the center position of the transmitted pulse in the i-th receiving channel, where i takes 1 to n, to ensure that the center positions of the transmitted pulses in all receiving channels are aligned;

[0103] Step 2.1.3: Taking the reference channel as a reference and the whiteboard as a standard target, calculate the position offset of the center position of the received echo in each receiving channel relative to the center position of the received echo in the reference channel, that is, obtain the position offset Δτ E,i,j, where \(i\) ranges from 1 to \(n\) and \(j\) ranges from 1 to \(N\);

[0104] Since the surface of the whiteboard is uniform, taking the whiteboard as the standard target, theoretically, the central positions of all channels should be the same. Therefore, with the reference channel as the reference and the whiteboard as the standard target, correct the central positions of the transmitted pulses and received pulses in each receiving channel except the reference channel to ensure that the central positions of the transmitted pulses in all receiving channels are aligned and the central positions of the received pulses are also aligned, and calculate the position offset \(\Delta\tau\) of the central position of the \(j\)-th received echo in the \(i\)-th receiving channel. E,i,j is a fixed constant, and the position offset \(\Delta\tau\) of the central position of the transmitted pulse in the \(i\)-th receiving channel T,i is also a fixed constant, and the position offsets of the central positions of the transmitted pulse and the received echo in the reference channel are both 0.

[0105] Step 2.2: Although position correction has been carried out, due to the inherent characteristics of each receiving channel and environmental factors, there are still differences in the signal and noise characteristics of different receiving channels. To address this issue, the wavelet transform method is used to perform smoothing and denoising processing on the full waveforms (including transmitted pulses and each received echo) of each receiving channel to obtain high-quality waveforms of the denoised full waveforms in each receiving channel. Denote the high-quality waveform of the full waveform of the \(i\)-th receiving channel as \(F\) τ,i (t).

[0106] Wavelet transform has significant advantages in removing high-frequency noise. It can not only effectively eliminate high-frequency noise but also better retain the key feature information of the original signal. Through multi-scale decomposition, it can balance the relationship between the denoising effect and the retention of signal features, achieving the best balance between noise suppression and signal retention.

[0107] Step 3: Initialize the characteristic parameters (including peak value, full width at half maximum, and central position) of the transmitted pulse model and the received echo model respectively. The initialization of the characteristic parameters of the transmitted pulses and received echoes in each receiving channel is based on the following formulas:

[0108] A i,j = y p,i,j (7)

[0109]

[0110] τ i,j = t p,i,j (9)

[0111]

[0112] In the formula, \(\Delta t\) i,j is the difference in inflection point positions, \(y\) p,i,j is the peak value, and \(t\) p,i,jis the peak position, t l,i,j and t r,i,j are the inflection point positions in the left monotonic increasing interval and the right monotonic decreasing interval respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonic increasing interval and the right monotonic decreasing interval respectively, are the means of all inflection point positions in the left monotonic interval and the right monotonic interval respectively;

[0113] When initializing the characteristic parameters of the transmitted pulses in each receiving channel, A i,j is the peak A of the transmitted pulse in the i-th receiving channel A T,i , α i,j is the full width at half maximum α of the transmitted pulse in the i-th receiving channel T,i , τ i,j is the center position τ of the transmitted pulse in the i-th receiving channel T,i , Δt i,j is the difference between the inflection point positions of the transmitted pulse in the i-th receiving channel, y p,i,j is the peak of the transmitted pulse in the i-th receiving channel, t p,i,j is the peak position of the transmitted pulse in the i-th receiving channel, t l,i,j and t r,i,j are the inflection point positions in the left monotonic increasing interval and the right monotonic decreasing interval of the transmitted pulse in the i-th receiving channel respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonic increasing interval and the right monotonic decreasing interval of the transmitted pulse in the i-th receiving channel respectively, are the means of all inflection point positions in the left monotonic interval and the right monotonic interval of the transmitted pulse in the i-th receiving channel respectively, i ranges from 1 to n, each receiving channel includes a transmitted pulse, j does not take values; initialize the characteristic parameters of the transmitted pulses in each receiving channel in sequence;

[0114] When initializing the characteristic parameters of each received echo in each receiving channel, A i,j is the peak A of the j-th received echo in the i-th receiving channel A E,i,j , α i,j is the full width at half maximum α of the j-th received echo in the i-th receiving channel E,i,j , τ i,j is the center position τ of the j-th received echo in the i-th receiving channel E,i,j , Δt i,j is the difference between the inflection point positions of the j-th received echo in the i-th receiving channel, y p,i,j is the peak of the j-th received echo in the i-th receiving channel, t p,i,j is the peak position of the j-th received echo in the i-th receiving channel, tl,i,j and t r,i,j are the inflection point positions in the left monotonically increasing interval and the right monotonically decreasing interval of the j-th received echo in the i-th receiving channel, respectively. N l,i,j and N r,i,j are the numbers of inflection points in the left monotonically increasing interval and the right monotonically decreasing interval of the j-th received echo in the i-th receiving channel, respectively. are the means of all inflection point positions in the left monotonic interval and the right monotonic interval of the j-th received echo in the i-th receiving channel, respectively. i ranges from 1 to n, and j ranges from 1 to N; the characteristic parameters of each received echo in each receiving channel are initialized in sequence.

[0115] Step 4: Perform multi-channel waveform accumulation and extract single echoes and overlapping echoes, including the following steps:

[0116] Step 4.1: Accumulate the high-quality waveforms F τ,i (t) of the full waveforms in each receiving channel to obtain the accumulated waveform F γ (t), and calculate the characteristic parameters of the transmitted pulse and the received echo of the accumulated waveform, including the following steps:

[0117] The accumulated waveform F γ (t) is calculated by the following formula:

[0118]

[0119] In the formula, ω i is the accumulation weight of the i-th receiving channel.

[0120] Step 4.1.2: In the same way as the method of initializing the characteristic parameters of the transmitted pulse and the received echo in Step 3, calculate the characteristic parameters of the transmitted pulse of the accumulated waveform according to Step 3. The characteristic parameters of the transmitted pulse of the accumulated waveform include the peak value A γ of the transmitted pulse of the accumulated waveform, the half-height width α γ of the transmitted pulse of the accumulated waveform, and the center position τ γ of the transmitted pulse of the accumulated waveform;

[0121] Step 4.1.3: Calculate the characteristic parameters of the received echo of the accumulated waveform according to Step 3. The characteristic parameters of the received echo of the accumulated waveform include the peak value, the half-height width, and the center position of the received echo of the accumulated waveform. The peak value of the j-th received echo of the accumulated waveform is A′ γ,j , the half-height width of the j-th received echo of the accumulated waveform is α′ γ,j , and the center position of the j-th received echo of the accumulated waveform is τ′ γ,j , where j takes 1 to N.

[0122] Step 4.2. The cumulative waveform can generally be divided into three types of echo: single echo, overlapping echo (including strong and weak overlapping echoes), and saturated echo. Although the spatial position of the received echo can be generally retrieved through the cumulative process, in the case of overlapping waveforms, some masked waveform components cannot fully exhibit their characteristics. Therefore, it is necessary to further retrieve and analyze the characteristic parameters of the transmitted pulse and the received echo in all channels in detail.

[0123] Specifically, after the initial transmitted pulse characteristic parameters obtained in Step 3 are adjusted by the LM (Levenberg-Marquardt) optimizer, they will be used as the initial input of the transmitted pulse polynomial function and incorporated into the subsequent multi-channel step-by-step iterative decomposition process to achieve the accurate estimation and optimization of the sub-echo characteristic parameters.

[0124] To ensure that the extracted received signal has physical significance and avoid the interference of pseudo-echo caused by noise, before entering the progressive waveform decomposition process, it is necessary to screen the initial characteristic parameters of the received echo in each receiving channel. Only the echo signals that meet the following conditions simultaneously can be determined as valid echoes and enter the subsequent processing stage, which specifically includes the following steps:

[0125] Step 4.2.1. Screen the received echoes that meet the following three screening conditions simultaneously. The screened received echoes enter Step 4.2.2 and 4.2.3:

[0126] Screening condition 1: Retain the received echoes in each channel whose intersection value of the full width at half maximum in the monotonic increasing interval and the monotonic decreasing interval is greater than the noise threshold T i =λ i *S i +M i of the received echo, where λ i is the threshold coefficient, which can be selected according to the actual waveform characteristics, S i represents the mean square deviation of the noise in the i-th receiving channel, and M i represents the mean value of the noise in the i-th receiving channel;

[0127] Screening condition 2: Retain the received echoes in each receiving channel whose cross-correlation coefficient with the transmitted pulse is greater than the set threshold (the cross-correlation calculation is performed between the received echo and the transmitted pulse in a single channel. In this embodiment, the threshold of the cross-correlation coefficient is set to 0.01, which can be adjusted according to the actual waveform characteristics).

[0128] Screening condition 3: If the number of the same y values corresponding to the extreme points of a certain received echo exceeds 2, then the above received echo is regarded as a saturated echo, and all saturated echoes are removed.

[0129] Step 4.2.2: Further screen out single echoes from all the received echoes screened out in Step 4.2.1. The screening conditions for single echoes are as follows: the full width at half maximum of the j-th received echo in the i-th receiving channel is less than or equal to the full width at half maximum of the transmitted pulse, and there is only a single local extreme point (i.e., only one obvious peak) in the waveform of the j-th received echo, and the number of sub-echo components of the j-th received echo in the cumulative waveform is 1, then it is a to-be-determined single echo.

[0130] Step 4.2.3: Further screen out overlapping echoes from all the received echoes screened out in Step 4.2.1. The screening conditions for to-be-determined overlapping echoes are as follows: the full width at half maximum of the j-th received echo in the i-th receiving channel is greater than four-thirds of the full width at half maximum of the transmitted pulse, or there are multiple inflection points in the monotonic interval;

[0131] and the number of sub-echo components of the j-th received echo in the cumulative waveform is greater than 1, then it is a to-be-determined overlapping echo.

[0132] For single echoes and overlapping echoes, the following steps will be entered.

[0133] Step 5: Iterate the single echoes and overlapping echoes through residual search to obtain the iterated received echoes, extract the characteristic parameters of the iterated received echoes, then discriminate the iterated received echoes, screen out the valid received echoes, and incorporate the characteristic parameters of each valid received echo into the multi-channel characteristic parameter set, which specifically includes the following steps:

[0134] Step 5.1: After the preliminary discrimination of the received echoes, in order to more accurately extract the waveform parameters of the sub-echo components in the to-be-determined single echoes and to-be-determined overlapping echoes, a step-by-step iteration method is adopted to search for the maximum value in the residual signal R m,i between the single echoes in Step 4.2.2, the overlapping echoes in Step 4.2.3 and the transmitted pulse respectively, determine the number of new sub-echoes, and add the new sub-echoes obtained after each iteration to the corresponding received echo to obtain the iterated received echo. The j-th iterated received echo in the i-th receiving channel is f′ E,i,j (t);

[0135] The residual signal R m,i is calculated based on the following formula:

[0136] R m,i (t) = F T,i (t) - f E,i,j (t) (13)

[0137] Among them, m represents the number of iterations. The transmission pulse characteristic parameters of the i-th receiving channel after being adjusted by the LM optimizer are used as the initial input of the transmission pulse model. In each iteration, the peak value, center position, and pulse width of the residual waveform are gradually searched. The peak value, center position, and pulse width of the residual waveform of the j-th received echo in the i-th receiving channel are respectively denoted as A′ E,i,j , τ′ E,i,j , α′ E,i,j . These parameters are then input into the LM optimizer for optimization until the maximum value of the residual waveform is less than the noise threshold T i and α′ E,i,j is less than the half-height width α of the transmitted pulse T,i , and the iteration process terminates.

[0138] Each iteration includes the following processes:

[0139] Step 5.1.1: Calculate the residual signal between the single echo or overlapping echo and the transmitted pulse of the corresponding receiving channel.

[0140] Step 5.1.2: Find the local maximum points in the residual signal to determine the center position of the new sub-echo.

[0141] Step 5.1.3: Based on the center position of the newly searched sub-echo, calculate the half-height width of the new sub-echo. Determine the left boundary by subtracting the half-height width from the center position of the new sub-echo, and determine the right boundary by adding the half-height width. And use the corresponding left and right boundaries of the new sub-echo as the new time domain range, so as to incorporate the sub-echo into the current received echo model.

[0142] Step 5.1.4: Add the newly extracted sub-echo to the current received echo model and update the residual waveform. Repeat steps 2 - 4 until no more maximum points meeting the following conditions can be found:

[0143] (1) A′ E,i,j > T i ;

[0144] (2) There is at least one inflection point (i.e., the waveform has an obvious peak feature) in the left and right monotonic intervals.

[0145] Step 5.2: Sort the received echoes after iteration in descending order according to the product of the peak value and the half-height width;

[0146] Step 5.3: Calculate the cross-correlation coefficient ρ i,j (τ) between the sorted received echoes after iteration and the transmitted pulses in the corresponding receiving channels in turn. If the cross-correlation coefficient between a certain received echo after iteration and the transmitted pulse in the corresponding receiving channel is not lower than the threshold ρ′, it is regarded as a pending valid received echo;

[0147] ρi (τ) and the threshold ρ′ are calculated respectively based on the following formulas:

[0148]

[0149] In the formulas, is the mean value of ρ N ω represents the frequency of the transmitted pulse, represents the phase difference between the received echo and the transmitted pulse.

[0150] Step 5.4: Screen out the pending valid received echoes that appear repeatedly in at least half of the receiving channels among the pending valid received echoes in Step 5.3 as valid received echoes, and incorporate the characteristic parameters of each valid received echo into the multi-channel characteristic parameter set, denoted as vector Ψ E , which is expressed as:

[0151]

[0152] In one embodiment, a computer device is further provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0153] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0154] In one embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0155] It should be noted that the embodiments described in the present invention are only examples to illustrate the spirit of the present invention. Those skilled in the art of the present invention can make various modifications or supplements to the described embodiments or use similar ways to replace them, but will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar, characterized in that, It includes the following steps: Step 1: Establish an emission pulse model and a received echo model respectively based on the B-spline model; Step 2: Preprocess the emission pulse and the received echo respectively; Step 3: Initialize the characteristic parameters of the emission pulse model and the received echo model respectively; Step 4: Extract single echoes and overlapping echoes; Step 5: Iterate the single echoes and the overlapping echoes through residual search to obtain the iterated received echo, extract the characteristic parameters of the iterated received echo, then discriminate the iterated received echo, screen out the valid received echo, and incorporate the characteristic parameters of each valid received echo into the multi-channel characteristic parameter set.

2. The method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar according to claim 1, wherein The emission pulse model is based on the following formula: F T,i F(t) = A T,i × F B (α T,i (t - τ T,i - Δτ T,i )) where F T,i (t) is the transmitted pulse model in the i-th receiving channel, A T,i , α T,i , τ T,i are respectively the peak value, full width at half maximum, and center position of the transmitted pulse in the i-th receiving channel, Δτ T,i is the position offset of the center position of the transmitted pulse in the i-th receiving channel, t is the sampling time, i is the serial number of the receiving channel, n is the number of receiving channels, i takes values from 1 to n, F B is the mathematical expression of the B-spline waveform curve; The received echo model is based on the following formula: f E,i,j (t) = A E,i,j × F B (α E,i,j (t - τ E,i,j - Δτ E,i,j ) Where, F E,i (t) is the received echo model in the i-th receiving channel, and f E,i,j (t) is the basis function of the j-th received echo in the i-th receiving channel, and Δτ E,i,j is the position offset of the center position of the j-th received echo in the i-th receiving channel; A E,i,j , α E,i,j , τ E,i,j are respectively the peak value, the full width at half maximum, and the center position of the j-th received echo in the i-th receiving channel. Each receiving channel includes j received echoes. j is the serial number of the received echoes in a single receiving channel, and j takes values from 1 to N, where N is the number of received echoes in a single receiving channel.

3. The method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar according to claim 1, wherein The specific steps of Step 2 include the following steps: Step 2.1: Calibrate the center positions of the emission pulses and the center positions of the received echoes in each receiving channel respectively, which specifically includes the following steps: Step 2.1.1: First, calculate the signal-to-noise ratio of each receiving channel, and select the channel with the highest signal-to-noise ratio as the reference channel. The signal-to-noise ratio of each receiving channel is calculated based on the following formula: where SNR i is the signal-to-noise ratio of the i-th receiving channel, and A i ' is the peak value of the i-th receiving channel. The peak value of the i-th receiving channel is the sum of the peak values of the transmitted pulse and each received echo within the i-th receiving channel; Step 2.1.2: Taking the reference channel as a reference and the whiteboard as the standard target, calculate the position offset of the center position of the transmitted pulse in each receiving channel relative to the center position of the transmitted pulse in the reference channel, and obtain the position offset Δτ of the center position of the transmitted pulse in the i-th receiving channel, where i ranges from 1 to n. T,i , where i ranges from 1 to n; Step 2.1.3: Taking the reference channel as a reference and the whiteboard as the standard target, calculate the position offset of the center position of the received echo in each receiving channel relative to the center position of the received echo in the reference channel, and obtain the position offset Δτ of the center position of the j-th received echo in the i-th receiving channel, where i ranges from 1 to n and j ranges from 1 to N; E,i,j , where i ranges from 1 to n and j ranges from 1 to N; Step 2.2, denoise the full waveforms of each receiving channel to obtain high-quality waveforms of the full waveforms in each receiving channel. The full waveform of each receiving channel includes the transmitted pulse and each received echo of the corresponding receiving channel. The high-quality waveform of the full waveform of the i-th receiving channel is F τ,i (t).

4. A method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar according to claim 1, characterized in that, The initialization of the characteristic parameters in Step 3 is based on the following formula: A i,j = y p,i,j τ i,j = t p,i,j where Δt i,j is the difference in inflection point positions, y p,i,j is the peak value, t p,i,j is the peak position, t l,i,j and t r,i,j are the inflection point positions in the left monotonically increasing interval and the right monotonically decreasing interval respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonically increasing interval and the right monotonically decreasing interval respectively, t l,i,j and t r,i,j are the means of all inflection point positions in the left monotonic interval and the right monotonic interval respectively; When initializing the characteristic parameters of the transmitted pulses in each receiving channel, A i,j is the peak value A of the transmitted pulse in the i-th receiving channel T,i , α i,j is the full width at half maximum α of the transmitted pulse in the i-th receiving channel T,i , τ i,j is the center position τ of the transmitted pulse in the i-th receiving channel T,i , Δt i,j is the difference between the inflection point positions of the transmitted pulse in the i-th receiving channel, y p,i,j is the peak value of the transmitted pulse in the i-th receiving channel, t p,i,j is the peak position of the transmitted pulse in the i-th receiving channel, t l,i,j and t r,i,j are the inflection point positions in the left monotonic increasing interval and the right monotonic decreasing interval of the transmitted pulse in the i-th receiving channel respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonic increasing interval and the right monotonic decreasing interval of the transmitted pulse in the i-th receiving channel respectively, t l,i,j , t r,i,j are the means of all inflection point positions in the left monotonic interval and the right monotonic interval of the transmitted pulse in the i-th receiving channel respectively. i ranges from 1 to n, each receiving channel includes one transmitted pulse, and j does not take values; the characteristic parameters of the transmitted pulses in each receiving channel are initialized in sequence; When initializing the characteristic parameters of each received echo in each receiving channel, A i,j is the peak value of the j-th received echo in the i-th receiving channel, A E,i,j , α i,j is the full width at half maximum of the j-th received echo in the i-th receiving channel, α E,i,j , τ i,j is the central position of the j-th received echo in the i-th receiving channel, τ E,i,j , Δt i,j is the difference between the inflection point positions of the j-th received echo in the i-th receiving channel, y p,i,j is the peak value of the j-th received echo in the i-th receiving channel, t p,i,j is the peak position of the j-th received echo in the i-th receiving channel, t l,i,j and t r,i,j are the inflection point positions in the left monotonically increasing interval and the right monotonically decreasing interval of the j-th received echo in the i-th receiving channel respectively, N l,i,j and N r,i,j are the numbers of inflection points in the left monotonically increasing interval and the right monotonically decreasing interval of the j-th received echo in the i-th receiving channel respectively, t l,i,j , t r,i,j are the means of all inflection point positions in the left monotonic interval and the right monotonic interval of the j-th received echo in the i-th receiving channel respectively. i ranges from 1 to n, and j ranges from 1 to N; the characteristic parameters of each received echo in each receiving channel are initialized in turn.

5. The method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar according to claim 4, wherein Step 4 includes the following steps: Step 4.

1. Cumulate the high-quality waveforms F τ,i (t) of all full waveforms in each receiving channel to obtain the cumulative waveform F γ (t), and calculate the characteristic parameters of the transmitted pulse and received echo of the cumulative waveform respectively, including the following steps: τ,i (t), and calculate the characteristic parameters of the transmitted pulse and received echo of the cumulative waveform respectively, including the following steps: γ (t), and calculate the characteristic parameters of the transmitted pulse and received echo of the cumulative waveform respectively, including the following steps: Step 4.1.1, Cumulative waveform F γ (t) is calculated based on the following formula: where ω i is the cumulative weight of the i-th receiving channel; Step 4.1.2: Calculate the characteristic parameters of the emission pulse of the cumulative waveform according to Step 3. The characteristic parameters of the emission pulse of the cumulative waveform include the peak value A of the emission pulse of the cumulative waveform γ , the full width at half maximum α of the emission pulse of the cumulative waveform γ , and the center position τ of the emission pulse of the cumulative waveform γ ; Step 4.1.3: Calculate the characteristic parameters of the received echo of the cumulative waveform according to Step 3. The characteristic parameters of the received echo of the cumulative waveform include the peak value, full width at half maximum, and center position of the received echo of the cumulative waveform. The peak value of the j-th received echo of the cumulative waveform is A γ ′ ,j , the full width at half maximum of the j-th received echo of the cumulative waveform is α γ ′ ,j , the center position of the j-th received echo of the cumulative waveform is τ γ ′ ,j , where j ranges from 1 to N.

6. The method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar according to claim 5, wherein Step 4 also includes the following steps: Step 4.2: Extract single echoes and overlapping echoes, including the following steps: Step 4.2.1: Screen the received echoes that simultaneously meet the following screening conditions: Screening condition 1: Retain the value of the intersection point of the full width at half maximum of the received echo in the monotonically increasing interval and the monotonically decreasing interval in each channel, which is greater than the noise threshold T i = λ i * S i + M i of the received echo, where λ i is the threshold coefficient, S i is the mean square deviation of the noise in the i-th receiving channel, and M i is the mean value of the noise in the i-th receiving channel; Screening condition 2: Retain the received echoes in each receiving channel whose cross-correlation coefficient with the emission pulse is greater than the set threshold; Screening condition 3: If the number of received echoes corresponding to the same y value at the extreme point of a certain received echo exceeds 2, then the above received echo is a saturated echo, and all saturated echoes are removed; Step 4.2.2: Screen out single echoes from all the received echoes screened in Step 4.2.

1. The screening condition for single echoes is: the full width at half maximum of the j-th received echo in the i-th receiving channel is less than or equal to the full width at half maximum of the emission pulse, and there is only a single local extreme point in the waveform of the j-th received echo; and the number of sub-echo components of the j-th received echo in the cumulative waveform is 1; Step 4.2.3: Screen out overlapping echoes from all the received echoes screened in Step 4.2.

1. The screening condition for overlapping echoes is: the full width at half maximum of the j-th received echo in the i-th receiving channel is greater than four-thirds of the full width at half maximum of the emission pulse, or there are multiple inflection points in the monotonic interval; and the number of sub-echo components of the j-th received echo in the cumulative waveform is greater than 1.

7. A method for extracting weak multi - echo signals of a multi - spectral and hyperspectral lidar according to claim 6, characterized in that, The specific steps of Step 5 include the following steps: Step 5.1: Using a step-by-step iterative method, search for the maximum value in the single echo of Step 4.2.2, the overlapping echo of Step 4.2.3, and the residual signal R of the transmitted pulse respectively, determine the number of new sub-echoes, and add the obtained new sub-echoes to the corresponding received echo after each iteration to obtain the received echo after iteration. The j-th received echo after iteration in the i-th receiving channel is f m,i In it, search for the maximum value, determine the number of new sub-echoes, and add the obtained new sub-echoes to the corresponding received echo after each iteration to obtain the received echo after iteration. The j-th received echo after iteration in the i-th receiving channel is f E ′ ,i,j (t); Residual signal R m,i Calculated based on the following formula: R m,i R(t) = F T,i (t) - f E,i,j (t) where m is the number of iterations. In each iteration, the peak value, center position, and pulse width of the residual waveform are gradually searched. The peak value, center position, and pulse width of the residual waveform of the j-th received echo in the i-th receiving channel are denoted as A′ E,i,j , τ′ E,i,j , α′ E,i,j respectively, until the maximum value of the residual waveform is less than the noise threshold T i and α′ E,i,j is less than the half-height width α T,i of the transmitted pulse, and the iteration terminates; Step 5.2: Arrange the iterated received echoes in descending order according to the product of the peak value and the full width at half maximum; Step 5.

3. Calculate the cross-correlation coefficient ρ i,j (τ) between the sorted and iterated received echoes and the transmitted pulses in the corresponding receiving channels in sequence. If the cross-correlation coefficient between a certain iterated received echo and the transmitted pulse in the corresponding receiving channel is not lower than the threshold ρ′, it is regarded as a pending valid received echo; i,j (τ), if the cross-correlation coefficient between an iterated received echo and the transmitted pulse in the corresponding receiving channel is not less than the threshold ρ′, it is regarded as a pending valid received echo; Step 5.4: Screen out the pending valid received echoes that appear repeatedly in at least half of the receiving channels among the pending valid received echoes in Step 5.3 as valid received echoes, and incorporate the characteristic parameters of each valid received echo into the multi-channel characteristic parameter set.

8. A method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar according to claim 7, characterized in that, Each iteration in Step 5.1 includes the following search process: Step 5.1.1: Calculate the residual signal between the current emission pulse and the received echo; Step 5.1.2: Search for local maximum points in the residual signal to determine the center position of the new sub-echo; Step 5.1.3: Based on the center position of the newly searched sub-echo, calculate the full width at half maximum (FWHM) of the new sub-echo. Determine the left boundary by subtracting the FWHM from the sub-echo center position and the right boundary by adding the FWHM, and use the corresponding left and right boundaries as the new time domain range; Step 5.1.4: Add the newly extracted sub-echo to the corresponding received echo after the previous iteration; Repeat steps 5.1.2 to 5.1.4 until no more maximum points meeting the following conditions are found, at which point this iteration terminates: Condition 1: A′ E,i,j >T i ; Condition 2: There is at least one inflection point in the left and right monotonic intervals.

9. The method for extracting weak multi-echo signals of a multi-spectral and hyperspectral lidar according to claim 8, wherein Said ρ i,j (τ) is calculated based on the following formula: The threshold ρ′ is calculated based on the following formula: In the formula, is the mean value of ρ N , ω represents the frequency of the transmitted pulse, represents the phase difference between the received echo and the transmitted pulse.

10. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the multi-spectral and hyperspectral lidar weak multi-echo signal extraction method described in any one of claims 1 to 9.

Images