Forest canopy height estimation method, device, system, and storage medium

Abstract

The present invention discloses a forest canopy height estimation method, device, system, and storage medium, comprising: step S1, obtaining GEDI waveform data for a target area; step S2, sequentially performing denoising, normalization, and Gaussian fitting on the GEDI waveform data; step S3, based on the pre-processed GEDI waveform data, determining the potential ground echo center position for areas with steep slopes by comparing the amplitudes of the last two peaks, and calculating the mean of the elevations corresponding to the waveform peaks as the ground elevation; step S4, subtracting the widening amplitude corresponding to the fitted waveform using the potential ground echo center position from the forest canopy top elevation to obtain the corrected forest canopy top elevation; and step S5, calculating the difference between the corrected canopy top elevation and the ground elevation to obtain the forest canopy height. The technical solution of the present invention accurately obtains the forest canopy height in areas with steep terrain slopes.

Description

Technical Field

[0001] The present invention belongs to the technical field of ecological management, and in particular relates to a forest canopy height estimation method and device, system, and storage medium. Background Art

[0002] As a vital component of global ecosystems, forests not only provide a sanctuary for biodiversity but also play a crucial role in climate regulation, soil and water conservation, and air purification. Forest canopy height, a key indicator of forest structure, is crucial for assessing the health of forest ecosystems, biodiversity, carbon storage, and the scientific management of forest resources. Currently, in areas with steeper terrain, the accurate extraction of forest canopy height is difficult due to the increased noise and broadening effects in echoes. Consequently, it is impossible to accurately determine forest canopy height in areas with steeper terrain. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a forest canopy height estimation method and device, system, and storage medium to accurately obtain the forest canopy height in areas with large terrain slopes.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] A method for estimating forest canopy height, comprising:

[0006] Step S1, obtaining GEDI waveform data of the target area;

[0007] Step S2, preprocessing the GEDI waveform data; wherein the preprocessing includes: denoising, normalization and Gaussian fitting;

[0008] Step S3: Based on the pre-processed GEDI waveform data, for areas with large slopes, the potential ground echo center position is determined by comparing the amplitudes of the last two peaks, and the average of the elevations corresponding to the waveform peaks is calculated as the ground elevation;

[0009] Step S4: subtract the broadening amplitude corresponding to the fitted waveform at the center of the potential ground echo from the forest canopy top elevation to obtain the corrected forest canopy top elevation;

[0010] Step S5: Calculate the difference between the corrected canopy top elevation and the ground elevation to obtain the forest canopy height.

[0011] Preferably, in step S2, a denoising algorithm combining empirical mode decomposition with an improved wavelet threshold is used to denoise the GEDI waveform data.

[0012] Preferably, in step S2, Gaussian fitting is performed on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and performing fitting optimization.

[0013] The present invention also provides a forest canopy height estimation device, comprising:

[0014] An acquisition module is used to obtain GEDI waveform data of the target area;

[0015] A preprocessing module is used to preprocess the GEDI waveform data; wherein the preprocessing includes: denoising, normalization and Gaussian fitting;

[0016] The first calculation module is used to determine the potential ground echo center position by comparing the amplitudes of the last two peaks of the pre-processed GEDI waveform data in the area with a large slope, and calculate the average of the elevations corresponding to the waveform peaks as the ground elevation;

[0017] The second calculation module is used to subtract the broadening amplitude corresponding to the fitted waveform using the center position of the potential ground echo from the forest canopy top elevation to obtain the corrected forest canopy top elevation;

[0018] The third calculation module is used to obtain the forest canopy height by calculating the difference between the corrected canopy top elevation and the ground elevation.

[0019] Preferably, the pre-processing module uses a denoising algorithm combining empirical mode decomposition with an improved wavelet threshold to denoise the GEDI waveform data.

[0020] Preferably, the preprocessing module performs Gaussian fitting on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and optimizing the fitting.

[0021] An embodiment of the present invention further provides a forest canopy height estimation system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a forest canopy height estimation method when executed by the processor.

[0022] An embodiment of the present invention further provides a storage medium having a computer program stored thereon. The computer program executes a forest canopy height estimation method when running.

[0023] The present invention sequentially performs denoising, normalization, and Gaussian fitting on GEDI waveform data. Based on the pre-processed GEDI waveform data, the potential ground echo center is determined by comparing the amplitudes of the last two peaks in areas with steeper slopes. The average of the elevations corresponding to the waveform peaks is calculated as the ground elevation. The corrected forest canopy top elevation is calculated by subtracting the widening amplitude corresponding to the fitted waveform using the potential ground echo center from the forest canopy top elevation. The forest canopy height is then calculated by calculating the difference between the corrected canopy top elevation and the ground elevation. Using the technical solution of the present invention, the forest canopy height in areas with steeper terrain slopes can be accurately determined. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0025] Figure 1 This is a flow chart of a method for estimating forest canopy height according to an embodiment of the present invention;

[0026] Figure 2 Schematic diagram of potential ground echo; GES is the vertical range of the ground, and GEC is the center point of the ground vertical position. DETAILED DESCRIPTION

[0027] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0028] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] Example 1:

[0030] like Figure 1 As shown, an embodiment of the present invention provides a forest canopy height estimation method, comprising:

[0031] Step S1, obtaining GEDI (Global Ecosystem Dynamics Investigation) waveform data of the target area;

[0032] Step S2, preprocessing the GEDI waveform data; wherein the preprocessing includes: denoising, normalization and Gaussian fitting;

[0033] Step S3: Based on the pre-processed GEDI waveform data, for areas with large slopes, the potential ground echo center position is determined by comparing the amplitudes of the last two peaks, and the average of the elevations corresponding to the waveform peaks is calculated as the ground elevation;

[0034] Step S4: subtract the broadening amplitude corresponding to the fitted waveform at the center of the potential ground echo from the forest canopy top elevation to obtain the corrected forest canopy top elevation;

[0035] Step S5: Calculate the difference between the corrected canopy top elevation and the ground elevation to obtain the forest canopy height.

[0036] As an implementation method of an embodiment of the present invention, in step S2, a denoising algorithm combining empirical mode decomposition and improved wavelet threshold is used to denoise the GEDI waveform data; specifically, the GEDI waveform data is decomposed into multiple IMFs using EMD, and the high-frequency noise is removed from the first i high-frequency IMFs by improved wavelet threshold denoising.

[0037] Furthermore, empirical mode decomposition (EMD) decomposes the GEDI waveform data into multiple intrinsic mode functions (IMFs) and a residual, which is specifically expressed as:

[0038]

[0039] Where n is the number of decomposed IMFs, r n (t) is the residual term after decomposition, c x (t) is the x-th IMF. The improved wavelet threshold denoising (WTD-Improved) method is used to solve the oscillation phenomenon during signal reconstruction in wavelet hard threshold denoising and the problem that the wavelet soft threshold denoising result is too smooth. It can be expressed as:

[0040]

[0041] Among them, w I are the wavelet coefficients of each scale after quantization by the improved threshold function, a and b are adjustment factors and a∈[0,1], b>0.

[0042] As an implementation method of an embodiment of the present invention, in step S2, all denoised GEDI waveforms are normalized to facilitate waveform comparison under different data acquisition conditions. Since the GEDI transmit pulse is approximately Gaussian distributed, the received echo signal can be approximately regarded as the sum of multiple Gaussian functions. The specific expression is as follows:

[0043]

[0044] Among them, A i represents the amplitude of the i-th Gaussian component, μ i represents the center position of the i-th Gaussian component, σ i represents the standard deviation of the i-th Gaussian component, and n(t) is the noise term representing random noise.

[0045] As an implementation of the embodiment of the present invention, in step S2, Gaussian fitting is performed on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and performing fitting optimization.

[0046] Gaussian fitting of the denoised and normalized GEDIL1B data involves two steps: initializing the Gaussian fitting parameters and optimizing the fit. 1) The even-odd inflection point method is used to determine the number of waveforms in the denoised and normalized GEDIL1B data. To reduce the impact of noise on the number of waveforms, waveforms with amplitudes less than the noise threshold are removed. The noise threshold is calculated as the noise mean plus three times the noise standard deviation. 2) The merge position interval is less than the standard deviation of the transmitted pulse σ. transmit If the number of waveforms is still greater than 20, they are merged according to their area. The specific method is to sort the waveform areas and merge the waveform with the smallest area with the adjacent waveform with a larger area until the number of waveforms is less than 20. During the Gaussian fitting optimization process, the value range of each parameter is restricted: the value range of amplitude A is [0,A peakmax ], where A peakmax is the amplitude of the maximum peak. Peak position t m Located between the start and end of the waveform. The standard deviation σ ranges from [σ transmit ,L extent ], L extent is the waveform range. Finally, when the residual of the fitting result is less than three times the noise standard deviation, the fitting is considered successful.

[0047] As an implementation method of the embodiment of the present invention, in step S3, all corresponding ground Gaussian waveforms (such as Figure 2 The ground elevation is calculated based on the average of the elevations corresponding to the peaks of the ground Gaussian waveform (shown as the black solid line waveform in the middle). The determination of the ground Gaussian waveform depends on the ground vertical range (GES), which defines the starting and ending positions of the ground waveform. Its expression is:

[0048] GES=tan(slope)*GLASFootPrintSize(m)

[0049] In order to accurately determine the position of the GES, it is necessary to set a reference point, such as the starting point, end point or center position. The embodiment of the present invention selects the terrain echo center position (GEC) as the reference point. The identification of the GEC is achieved by comparing the peak amplitudes of the last two Gaussian waveforms (L1 and L2), and the waveform with the larger peak amplitude is selected as the GEC. Specifically, the GEC is determined by comparing the peak amplitudes of L1 and L2, and the corresponding GES is found. Considering that the maximum peak in the waveform usually comes from the ground or the canopy, if the maximum peak of the entire waveform is included in the GES range, the elevation corresponding to the peak is the ground elevation; otherwise, the average of the elevations corresponding to the peaks of all waveforms in the GES is calculated as the ground elevation.

[0050] As one implementation of an embodiment of the present invention, in step S4, the elevation of the forest canopy top is determined based on background noise. Specifically, a noise threshold is first determined based on the average noise value plus three times the noise standard deviation. Then, the first signal position in the waveform data that is greater than the noise threshold is used as the elevation of the forest canopy top. The noise threshold expression is:

[0051] T = noise_mean*3noise_stddev

[0052] Among them, noise_mean is the average noise value, and noise_stddev is the noise standard deviation.

[0053] Furthermore, as the terrain slope increases, although the peak position remains unchanged, both the ground and canopy waveforms will widen. This widening of the canopy waveform may lead to inaccurate elevation of the forest canopy top obtained by the noise threshold method. Calculating the widening amplitude of the ground waveform can be used to quantitatively estimate the widening of the canopy waveform. The widening amplitude of the ground waveform is determined by the difference between the standard deviation of the ground waveform and the standard deviation of the transmitted pulse. In summary, the calculation formula for correcting the elevation of the forest canopy top is as follows:

[0054] H corrected_top =H top -3*(σ ground -σ transmit )

[0055] Among them, H correctedtop represents the corrected forest canopy top elevation, H top represents the elevation of the forest canopy top before correction, σ ground represents the standard deviation of the ground echo, σ transmit Indicates the standard deviation of the transmitted pulse. In the embodiment of the present invention, the standard deviation of the fitted waveform at the center of the potential ground echo is selected as σ ground, and changing the coefficient in the formula from 3 times to 1.5 times can obtain more accurate results. Therefore, the calculation formula for forest canopy top elevation correction used in the embodiment of the present invention is as follows:

[0056] H corrected_top =H top -1.5*(σ ground -σ transmit )

[0057] As an implementation method of the embodiment of the present invention, in step S5, the forest canopy height is obtained by subtracting the corrected forest canopy top elevation from the corresponding ground elevation, that is,

[0058] H canopy =H corrected_top -H ground .

[0059] The embodiment of the present invention uses the average elevation corresponding to all waveform peaks within the potential waveform range as the ground elevation, which significantly improves the extraction accuracy of the ground elevation in areas with larger slopes. At the same time, it infers the canopy waveform widening based on the ground waveform widening amplitude and performs canopy top elevation correction, effectively improving the extraction accuracy of the forest canopy top elevation in areas with larger slopes.

[0060] Example 2:

[0061] An embodiment of the present invention further provides a forest canopy height estimation device, comprising:

[0062] An acquisition module is used to obtain GEDI waveform data of the target area;

[0063] A preprocessing module is used to preprocess the GEDI waveform data; wherein the preprocessing includes: denoising, normalization and Gaussian fitting;

[0064] The first calculation module is used to determine the potential ground echo center position by comparing the amplitudes of the last two peaks of the pre-processed GEDI waveform data in the area with a large slope, and calculate the average of the elevations corresponding to the waveform peaks as the ground elevation;

[0065] The second calculation module is used to subtract the broadening amplitude corresponding to the fitted waveform using the center position of the potential ground echo from the forest canopy top elevation to obtain the corrected forest canopy top elevation;

[0066] The third calculation module is used to obtain the forest canopy height by calculating the difference between the corrected canopy top elevation and the ground elevation.

[0067] As an implementation method of the embodiment of the present invention, the pre-processing module uses a denoising algorithm that combines empirical mode decomposition with an improved wavelet threshold to denoise the GEDI waveform data.

[0068] As an implementation method of the embodiment of the present invention, the pre-processing module performs Gaussian fitting on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and fitting optimization.

[0069] Example 3:

[0070] An embodiment of the present invention further provides a forest canopy height estimation system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a forest canopy height estimation method when executed by the processor.

[0071] Example 4:

[0072] An embodiment of the present invention further provides a storage medium having a computer program stored thereon. The computer program executes a forest canopy height estimation method when running.

[0073] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A method for estimating forest canopy height, characterized in that: include: Step S1, obtaining GEDI waveform data of the target area; Step S2, preprocessing the GEDI waveform data; wherein the preprocessing includes: denoising, normalization and Gaussian fitting; Step S3: Based on the pre-processed GEDI waveform data, for areas with large slopes, the center position of the terrain echo is determined by comparing the amplitudes of the last two peaks, and the average of the elevations corresponding to the waveform peaks is calculated as the ground elevation; Step S4: subtract the widening amplitude corresponding to the fitted waveform at the center of the terrain echo from the elevation of the top of the forest canopy to obtain the corrected elevation of the top of the forest canopy; Step S5: Calculate the difference between the corrected canopy top elevation and the ground elevation to obtain the forest canopy height; In step S2, the GEDI waveform data is denoised using a denoising algorithm combining empirical mode decomposition with an improved wavelet threshold; Gaussian fitting is performed on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and performing fitting optimization; In step S3, the ground elevation is calculated based on the mean of the elevations corresponding to the peaks of the ground Gaussian waveform. The determination of the ground Gaussian waveform depends on the ground vertical range GES, which defines the start and end positions of the ground waveform. The expression is: GES=tan(slope)*GLASFootPrintSize(m) In step S3, the center of the terrain echo, GEC, is selected as the reference point to determine the location of the GES. The GEC is identified by comparing the peak amplitudes of the last two Gaussian waveforms, L1 and L2, and the waveform with the larger peak amplitude is selected as the GEC. The GEC is determined by comparing the peak amplitudes of L1 and L2, and the corresponding GES is found. If the GES contains the maximum peak of the entire waveform, the elevation corresponding to this peak is the ground elevation. Otherwise, the average elevation corresponding to the peaks of all waveforms in the GES is calculated as the ground elevation. In step S4, the elevation of the forest canopy top is determined based on background noise. Specifically, the noise threshold is first determined by adding three times the noise standard deviation to the average noise value, and then the first signal position in the waveform data that is greater than the noise threshold is used as the elevation of the forest canopy top. In step S4, the broadening of the ground waveform is calculated to quantitatively estimate the broadening of the canopy waveform. The broadening of the ground waveform is determined by the difference between the standard deviation of the ground waveform and the standard deviation of the transmitted pulse. The calculation formula for correcting the elevation of the top of the forest canopy is as follows: H corrected_top =H top -3*(s ground -s transmit ) Among them, H correctedtop represents the corrected forest canopy top elevation, H top represents the elevation of the forest canopy top before correction, σ ground represents the standard deviation of the ground echo, σ transmit Indicates the standard deviation of the transmitted pulse.

2. A forest canopy height estimation device for implementing the forest canopy height estimation method according to claim 1, characterized in that: include: An acquisition module is used to obtain GEDI waveform data of the target area; A preprocessing module is used to preprocess the GEDI waveform data; wherein the preprocessing includes: denoising, normalization and Gaussian fitting; The first calculation module is used to determine the center position of the terrain echo by comparing the amplitudes of the last two peaks in the area with a large slope based on the preprocessed GEDI waveform data, and calculate the average of the elevations corresponding to the waveform peaks as the ground elevation; The second calculation module is used to subtract the widening amplitude corresponding to the fitted waveform using the center position of the terrain echo from the elevation of the top of the forest canopy to obtain the corrected elevation of the top of the forest canopy; The third calculation module is used to calculate the difference between the corrected canopy top elevation and the ground elevation to obtain the forest canopy height; The pre-processing module uses a denoising algorithm that combines empirical mode decomposition with an improved wavelet threshold to denoise the GEDI waveform data; The preprocessing module performs Gaussian fitting on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and fitting optimization.

3. A forest canopy height estimation system, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the forest canopy height estimation method according to claim 1 is executed.

4. A storage medium, characterized in that The storage medium stores a computer program, which executes the forest canopy height estimation method according to claim 1 when running.

Images