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

Abstract

The invention discloses a forest canopy height estimation method, device and system, and a storage medium. The method comprises the following steps: S1, obtaining GEDI waveform data of a target area; s2, sequentially carrying out denoising processing, normalization processing and Gaussian fitting processing on the GEDI waveform data; s3, according to the preprocessed GEDI waveform data, determining a potential ground echo center position by comparing the amplitudes of the last two wave crests for an area with a large gradient, and calculating a mean value of elevations corresponding to waveform peak values as a ground elevation; s4, subtracting the spread width value corresponding to the fitting waveform of the central position of the potential ground echo from the top elevation of the forest canopy to obtain a corrected top elevation of the forest canopy; and S5, calculating a difference value between the corrected canopy top elevation and the ground elevation to obtain the forest canopy height. By adopting the technical scheme of the invention, the forest canopy height of a region with a relatively large terrain gradient can be accurately obtained.

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 an important part of the global ecosystem, forests not only provide shelter for biodiversity, but also play a vital role in climate regulation, soil and water conservation, air purification and many other aspects. As an important indicator of forest structure, forest canopy height is of great significance for evaluating the health of forest ecosystems, biodiversity, carbon storage and scientific management of forest resources. At present, in areas with large slopes, the increase in noise and broadening effects in the echoes brings certain difficulties to the accurate extraction of forest canopy height, resulting in the inability to accurately obtain the forest canopy height in areas with large terrain slopes. 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 solution:

[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: According to the pre-processed GEDI waveform data, for the area with a large slope, the potential ground echo center position is determined by comparing the amplitudes of the last two peaks, and the mean of the elevations corresponding to the waveform peaks is calculated as the ground elevation;

[0009] Step S4, subtracting the widening amplitude corresponding to the fitted waveform at the center position 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 optimizing the fitting.

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

[0014] An acquisition module, used to acquire GEDI waveform data of a 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 wave peaks based on the preprocessed GEDI waveform data for the area with a large slope, and calculate the average value of the elevation corresponding to the waveform peak as the ground elevation;

[0017] The second calculation module is used to subtract the widening 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 preprocessing 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, on which a computer program is stored, and 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; according to the pre-processed GEDI waveform data, for the area with a large slope, by comparing the amplitudes of the last two wave peaks, the potential ground echo center position is determined, and the mean of the elevation corresponding to the waveform peak is calculated as the ground elevation; the elevation of the top of the forest canopy minus the widening amplitude corresponding to the fitted waveform using the center of the potential ground echo is used as the elevation of the top of the forest canopy after correction; the height of the forest canopy is obtained by calculating the difference between the corrected canopy top elevation and the ground elevation. The technical solution of the present invention is used to accurately obtain the height of the forest canopy in the area with a large terrain slope. 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 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 embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying 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 Figure 1 is a schematic diagram of potential ground echo; GES is the vertical range of the ground and GEC is the center point of the vertical position of the ground. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work 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] Embodiment 1:

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

[0031] Step S1, obtaining GEDI (Global Ecosystem Dynamics Investigation, GEDI) 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: According to the pre-processed GEDI waveform data, for the area with a large slope, the potential ground echo center position is determined by comparing the amplitudes of the last two peaks, and the mean of the elevations corresponding to the waveform peaks is calculated as the ground elevation;

[0034] Step S4, subtracting the widening amplitude corresponding to the fitted waveform at the center position 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 of the first i high-frequency IMFs is removed 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 xth IMF. The improved wavelet threshold denoising (WTD-Improved) method is used to solve the oscillation phenomenon of 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 the embodiment of the present invention, in step S2, all GEDI waveforms after denoising are normalized to facilitate comparison of waveforms under different data acquisition conditions. Since the GEDI transmission pulse is approximately Gaussian distributed, the echo signal received 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 optimizing the fitting.

[0046] The Gaussian fitting of the denoised and normalized GEDIL1B data includes two steps: initializing the Gaussian fitting parameters and optimizing the fitting. 1) The odd-even inflection point method is used to obtain the number of waveforms in the denoised and normalized GEDIL1B data. In order to reduce the impact of noise on the number of waveforms, waveforms with amplitudes less than the noise threshold need to be 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 waveforms by area 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 ground Gaussian waveform peaks (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 start and end positions of the ground waveform, and 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, the end point or the 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 GES range contains the maximum peak of the entire waveform, the elevation corresponding to the peak is the ground elevation; otherwise, the average of the elevations corresponding to all waveform peaks in the GES is calculated as the ground elevation.

[0050] As an implementation of an embodiment of the present invention, in step S4, the elevation of the top of the forest canopy is determined based on the background noise calculation. Specifically, the noise threshold is first determined based on the average noise value plus 3 times the noise standard deviation, and then the first signal position in the waveform data that is greater than the noise threshold is used as the elevation of the top of the forest canopy. The noise threshold expression is:

[0051] T = noise_mean * 3 noise_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 top of the forest canopy 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 top of the forest canopy 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 top of the forest canopy before correction, σ ground represents the standard deviation of the ground echo, σ transmit The embodiment of the present invention selects the standard deviation of the fitted waveform at the center of the potential ground echo as σ ground, and changing the coefficient in the formula from 3 times to 1.5 times can obtain a more accurate result. 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 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 adopts 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, the canopy waveform widening is inferred based on the ground waveform widening amplitude and the canopy top elevation correction is performed, which effectively improves the extraction accuracy of the forest canopy top elevation in areas with larger slopes.

[0060] Embodiment 2:

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

[0062] An acquisition module, used to acquire GEDI waveform data of a 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 wave peaks based on the preprocessed GEDI waveform data for the area with a large slope, and calculate the average value of the elevation corresponding to the waveform peak as the ground elevation;

[0065] The second calculation module is used to subtract the widening 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 preprocessing module uses a denoising algorithm combining 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 preprocessing module performs Gaussian fitting on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and fitting optimization.

[0069] Embodiment 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] Embodiment 4:

[0072] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and the computer program executes a forest canopy height estimation method when running.

[0073] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined 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: According to the pre-processed GEDI waveform data, for the area with a large slope, the potential ground echo center position is determined by comparing the amplitudes of the last two peaks, and the mean of the elevations corresponding to the waveform peaks is calculated as the ground elevation; Step S4, subtracting the widening amplitude corresponding to the fitted waveform at the center position of the potential ground echo from the forest canopy top elevation to obtain the corrected forest canopy top elevation; Step S5: Calculate the difference between the corrected canopy top elevation and the ground elevation to obtain the forest canopy height.

2. The method for estimating forest canopy height according to claim 1, characterized in that: In step S2, a denoising algorithm combining empirical mode decomposition and improved wavelet threshold is used to denoise the GEDI waveform data.

3. The method for estimating forest canopy height according to claim 2, characterized in that: In step S2, Gaussian fitting is performed on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and optimizing the fitting.

4. A forest canopy height estimation device, characterized in that: include: An acquisition module, used to acquire GEDI waveform data of a 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 potential ground echo center position by comparing the amplitudes of the last two wave peaks based on the preprocessed GEDI waveform data for the area with a large slope, and calculate the average value of the elevation corresponding to the waveform peak 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 potential ground echo from the forest canopy top elevation to obtain the corrected forest canopy top elevation; 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.

5. The forest canopy height estimation device according to claim 4, characterized in that: The preprocessing module uses a denoising algorithm that combines empirical mode decomposition with an improved wavelet threshold to denoise the GEDI waveform data.

6. The forest canopy height estimation device according to claim 5, characterized in that: The preprocessing module performs Gaussian fitting on the denoised and normalized GEDI waveform data by initializing Gaussian fitting parameters and fitting optimization.

7. 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 any one of claims 1 to 3 is executed.

8. A storage medium, characterized in that: The storage medium stores a computer program, which executes the forest canopy height estimation method according to any one of claims 1 to 3 when running.

Images