A method for correcting distance effect radiation of hyperspectral lidar based on geometric factor

Abstract

The application discloses a hyperspectral laser radar distance effect radiation correction method based on a geometric factor, which comprises the following steps: step one, constructing a geometric factor model based on an optical system principle or an S-shaped curve fitting function; step two, constructing a distance effect model; and step three, constructing a distance effect radiation correction model. The geometric factor model is constructed by analyzing the coaxial system structure of a receiving and transmitting system of the hyperspectral laser radar. Through a ranging experiment, stable solution of model parameters is realized, and the method is verified and precision analysis is carried out. The distance effect of different targets at different wave bands shares the same geometric factor model parameters, and only a few parameters are used to carry out distance effect correction of various targets, and the universality of the parameters provides a simple and effective technical means for the distance effect problem of the hyperspectral laser radar.

Description

Technical Field

[0001] This invention belongs to the field of hyperspectral lidar, and particularly relates to a hyperspectral lidar range effect radiation correction method based on geometric factors. Background Technology

[0002] Hyperspectral lidar is a novel remote sensing method that combines the advantages of active lidar detection and hyperspectral imaging. It employs an active method of emitting broadband laser pulses and detecting the backscattered signals. This allows for the simultaneous acquisition of both spectral and range information of the target, generating point cloud data with "color" information. This provides richer target attribute information, including geometric characteristics, distance, material composition, state, and color. Hyperspectral lidar acquires spectral information primarily by measuring the power of the returned laser signal. The backscattered light power is converted into voltage within the receiver, amplified in the system, and finally converted into a DN (Digital Number) value, recorded as the laser backscattering intensity. This value characterizes the scattering ability and radiation characteristics of ground targets to the emitted laser pulse signal. Therefore, hyperspectral lidar possesses the ability to simultaneously acquire high-precision three-dimensional spatial information and high-resolution spectral information, exhibiting both high spatial detection capabilities and ground object physical attribute detection capabilities.

[0003] The hyperspectral lidar system generates point cloud data structures that combine hyperspectral backscattering intensity.

[0004] [x,y,z,I(λ)]

[0005] I(λ) represents the backscattering intensity as a continuous function of wavelength λ. The backscattering intensity of lasers at different wavelengths corresponds precisely to the three-dimensional coordinate information, exhibiting pixel-level fusion characteristics. During laser scanning, the laser backscattering intensity is affected by various factors such as instrument system characteristics, target surface characteristics, laser incident angle, and detection distance. To obtain accurate backscattering intensity information and correctly evaluate the response of ground objects to laser pulses, radiometric correction needs to be performed on the echo backscattering intensity of hyperspectral lidar in different bands to remove the influence of various factors and provide a good data foundation for subsequent data analysis and applications.

[0006] This invention primarily addresses the range effect of detection distance on backscattering intensity and conducts research on radiometric correction methods. To ensure that the received backscattering intensity data accurately reflects the target's radiation characteristics, it is essential to systematically analyze the radiation process of backscattering intensity, analyze the influence of detection distance, and perform radiometric correction of backscattering intensity. Summary of the Invention

[0007] This invention has the following beneficial technical effects. It proposes a method for correcting the radiation effect of hyperspectral lidar range based on geometric factors, specifically including the following steps:

[0008] Step 1: Construct a geometric factor model based on the principles of optical systems or an S-curve fitting function;

[0009] Step 2: Construct a distance effect model;

[0010] Step 3: Construct a distance effect radiation correction model.

[0011] Furthermore, the specific method for constructing a geometric factor model based on the principles of optical systems is as follows:

[0012] The effective receiving radius of the telescope is r T The radius of the emitted light spot is r L The radius r of the laser beam spot at different detection distances L (R), telescope field of view radius r T (R) is represented by formulas (6) and (7) respectively:

[0013] r L (R)=r L +R·θ L (6)

[0014] r T (R)=r T +R·θ T (7)

[0015] Where R is the detection distance; θ T θ is the telescope's half field of view. L The laser beam half-divergence angle;

[0016] The angle between the optical axis of the laser emitted by the hyperspectral lidar and the optical axis of the telescope is... The distance d(R) between the two optical axes is expressed by formula (8):

[0017]

[0018] Let the radius of the telescope's secondary mirror be r. S The obstruction radius r of the telescope's secondary mirror S (R) is represented by formula (9):

[0019]

[0020] The overlap area S between the laser spot and the detector element T (R) is represented by formula (10):

[0021]

[0022] In the formula, the parameter φ L and φ T The calculation formula is as follows: Formula (11):

[0023]

[0024] Therefore, the geometric factor Z(R) as a function of distance is expressed as formula (12):

[0025]

[0026] Furthermore, in step one, the method for constructing the geometric factor model based on the S-curve fitting function is as follows:

[0027] By fitting the Logistic function, a piecewise geometric factor calculation model for the function Z(R) is established, as shown in formula (13):

[0028]

[0029] Where α, β, γ are the parameters to be solved, e is the natural constant, and R T The detection distance threshold is greater than R. T When the distance increases, the backscattering intensity decreases; less than R T At that time, the backscattering intensity increases with increasing distance.

[0030] Furthermore, in step two,

[0031] The laser emission power of the lidar is P t The receiver aperture is D. r η atm η is the influencing factor of signal transmission in the atmosphere. sys Here are the parameters of the lidar system: the laser incident angle is θ, the detection range is R, the target reflectivity is ρ, and the different wavelengths of the hyperspectral lidar are represented by λ. During a single laser scan, P... t (λ),D r ,η atm (λ) and η sys (λ) represents four constants, representing the received power P of the hyperspectral lidar at different wavelengths. r (λ) is represented as:

[0032]

[0033] Where ρ(λ) represents the target reflectivity at wavelength λ, and C(λ) is expressed by formula (16), where λ is a constant.

[0034]

[0035] During laser scanning, the incident angle remains constant. For the same target, considering the influence of geometric factors, the relationship between backscattering intensity I(R,λ) and detection distance is expressed as shown in Equation 17:

[0036]

[0037] In the formula: g0(λ) is constant, expressed as formula (18):

[0038] g0(λ)=C(λ)ρ(λ)cosθ (18).

[0039] Furthermore, in step three,

[0040] Define a standard detection range R s The backscattering intensity at different detection distances is normalized to the standard distance. Based on formula (16), the influence of geometric factors is considered in the process of normalizing the distance effect of backscattering intensity. The distance normalization model is expressed as formula (19) as follows:

[0041]

[0042] In the formula, R s This is the standard detection range, Z(R) s I(R,λ) is the geometric factor over the standard detection range. c I(R,λ) is the corrected backscattering intensity, and I(R,λ) is the original backscattering intensity.

[0043] This invention offers the following beneficial technical effects: It analyzes the coaxial system structure of the receiving and transmitting systems of a hyperspectral lidar and constructs a geometric factor model. Through ranging experiments, stable solutions to the model parameters are achieved, and the method is verified and its accuracy analyzed. Experimental results show that different targets in different bands share the same geometric factor model parameters for range effects. Range effect correction for various targets can be carried out using only a limited number of parameters. The universality of these parameters provides a simple and effective technical means for addressing the range effect problem in hyperspectral lidar. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of a hyperspectral lidar range effect measurement platform.

[0045] Figure 2 The relationship between backscattering intensity and transmission distance of hyperspectral lidar for different types of targets;

[0046] Figure 3 The relationship between laser emission and reception in hyperspectral lidar;

[0047] Figure 4 The relationship between the laser echo and the detector elements;

[0048] Figure 5 The calculation results of the distance effect correction model parameters for different types of targets. Detailed Implementation

[0049] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0050] (1) Mechanism of backscattering intensity of hyperspectral lidar based on radar equation

[0051] The basic principle of lidar echo signal can be explained by the radar equation, as shown in formula (1).

[0052]

[0053] Where: P r The received power, σ0, is the target backscattering cross section, which is related to the reflection characteristics of the ground object. The laser emission power of the lidar is P. t The laser beamwidth is β t The receiver aperture is D. r η atm η is the influencing factor of signal transmission in the atmosphere. sys Here are the parameters for the lidar system, and R is the detection distance.

[0054] Let the wavelength be λ. Under the condition that the transmission and detection receiving system parameters of different spectra are consistent in the design of the hyperspectral lidar system, the radar equation for the hyperspectral lidar is as shown in the following formula (2):

[0055]

[0056] Using R(θ) t ,φ t ,θ r ,φ r λ) represents the BRF of the target, which is defined as the ratio of the reflected radiance of a ground object in the same irradiance direction to the reflected radiance of a diffuse surface in that same direction. Where θ t Let φ be the incident zenith angle. t θ is the incident azimuth angle; r The backscattering zenith angle, φ r The azimuth angle is the backscattering angle. The derived radar equation can be expressed as the following formula (3):

[0057]

[0058] It can be concluded that, for the same target, when the incident angle is constant, the backscattering intensity is inversely proportional to the square of the detection distance R, and proportional to 1 / R. 2 Linear correlation.

[0059] (2) Distance effect of backscattering intensity

[0060] Establish a ranging platform to measure the variation of backscattering intensity from different targets as a function of detection, such as... Figure 1 As shown. Keeping the hyperspectral lidar position and laser incident direction constant, the laser detection range is changed by altering the target's position. The detection range is 1m-20m, with a sampling interval of 0.5m. Sampling is performed with the target perpendicular to the incident light (incident angle of 0°).

[0061] This invention selected a 100% standard diffuse reflector plate, bluestone bricks, smooth white floor tiles, and pothos leaves as experimental targets. Target types included diffuse reflective targets, rough targets, smooth targets, and plant targets, to determine the impact of different roughness and specular reflection on the distance effect. The hyperspectral lidar was set to operate in the high signal-to-noise ratio band of 650-850 nm, with a step size of 10 nm. To reduce noise and measurement errors, 50 echo data acquisitions were performed under identical conditions, and the average values ​​were used to calculate the backscattering intensity of different targets in different bands. The relationship between backscattering intensity and distance in different bands was statistically analyzed and calculated. Figure 2 The relationship between backscattering intensity obtained from hyperspectral lidar measurements and detection distance is shown. Figure 2 (a) is a 100% standard diffuse reflector. Figure 2 (b) is made of bluestone bricks. Figure 2 (c) is a white, smooth floor tile. Figure 2 (d) is a leaf of a pothos plant.

[0062] Depend on Figure 2 It can be seen that when the detection distance is greater than a certain distance threshold, the backscattering intensity decreases with the change of detection distance, showing an inverse relationship with the square of the transmission distance; when the detection distance is less than the distance threshold, the backscattering intensity increases with the change of detection distance.

[0063] The above experimental results show that when the detection distance is less than a certain threshold, the backscattering intensity increases with the detection distance, which does not satisfy formula (3). This phenomenon is mainly related to the laser echo reception process. In the lidar equation, the backscattering intensity and the detection distance satisfy 1 / R 2The premise of linear correlation is that after backscattered light is incident on the hyperspectral lidar receiving and detection system, all the echo energy is received by the detector. However, at close range, only a portion of the backscattered laser echo energy is received by the hyperspectral lidar detector; the proportion received varies at different distances.

[0064] (3) Geometric factor model construction method

[0065] This invention addresses the case where the hyperspectral lidar is a coaxial system, where the receiving and transmitting modules of the receiving and detection system are coaxial, meaning that the optical axis of the laser beam emitted by the lidar coincides with the optical axis of the telescope lens group.

[0066] The laser emitted by the hyperspectral lidar is considered as a uniform beam, and the cross-section of the laser beam at different distances is G. z ,like Figure 3 As shown. θ T θ is the half-field angle of the telescope. L This is the half-divergence angle of the laser beam. The receiving optical system of a hyperspectral lidar forms a focused spot by focusing the backscattered echo signal from the target onto a detector element. When the detection range R of the hyperspectral lidar is large, the focused spot is small and can be completely located within the detector element; when the detection range R is small, the focused spot is large, and the detector element cannot completely cover the focused spot, only the echo energy overlapping the detector element is received.

[0067] Based on the above principle, the ratio of the energy E(R) received on the detector surface element to the energy E0(R) received by the telescope primary mirror is defined as the geometric factor, as shown in the following formula (4):

[0068]

[0069] In the formula: R is the detection distance, and Z(R) is the geometric factor corresponding to the detection distance. Assuming that the energy of the laser spot of the hyperspectral lidar is uniformly distributed, when the detection distance is R, the geometric factor can be simplified to the ratio of the overlap area S(R) of the focused spot of the laser beam and the detector element to the total area S0(R) of the laser beam spot, as shown in the following formula (5):

[0070]

[0071] 1) Method 1: Construction of a geometric factor model based on the principles of optical systems

[0072] Ideally, the optical axis of the laser emitted by a hyperspectral lidar is parallel to the optical axis of the telescope. However, in reality, errors caused by manufacturing and assembly are difficult to completely eliminate, resulting in a slight deviation between the two optical axes. Let's assume the included angle is... In addition, considering the obstruction effect of the telescope's secondary mirror, the backscattered echo signal from the target is focused onto the detector element, forming a focused spot with an outer diameter of r. T (R) Inner diameter is r S The annular light spot (R) has a detector element radius of r. L (R). During telescope imaging, the size of the light spot formed by the object point on the focal plane depends on the distance to the object point and the parameters of the optical system. When the object point is close, the light spot size may exceed the size of the detector element, leading to signal loss or aliasing. Only a portion of the light spot enters the detector element, such as... Figure 4 As shown in (a) and (b), when the object point is far away, the image spot on the focal plane becomes smaller than the detector element and falls within the telescope's full field of view, such as... Figure 4 As shown in (c), the dark gray ring represents the laser spot, and the light gray circle represents the detector element.

[0073] The radius r of the laser beam spot at different detection distances L (R), telescope field of view radius r T (R) is represented by formulas (6) and (7) respectively:

[0074] r L (R)=r L +R·θ L (6)

[0075] r T (R)=r T +R·θ T (7)

[0076] The distance d(R) between the two optical axes is expressed by formula (8):

[0077]

[0078] Let the radius of the telescope's secondary mirror be r. S The obstruction radius r of the telescope's secondary mirror S (R) is represented by formula (9):

[0079]

[0080] The overlap area S between the laser spot and the detector element T (R) is represented by formula (10):

[0081]

[0082] In the formula, the parameter φ L and φ T The calculation formula is as follows: Formula (11):

[0083]

[0084] Therefore, the geometric factor Z(R) as a function of distance can be expressed as formula (12):

[0085]

[0086] 2) Construction of a geometric factor model based on an "S"-shaped curve fitting function

[0087] Based on the above theoretical foundation, when some parameters of the hyperspectral lidar are unknown, directly calculating the geometric factor according to formula (12) will result in a large error. At the same time, in order to simplify the function model of the geometric factor, the commonly used "S" curve fitting function can be selected for fitting. This invention takes the Logistic function fitting as an example to establish a piecewise geometric factor calculation model, as shown in formula (13):

[0088]

[0089] Where α, β, and γ are the parameters to be solved, and e is a natural constant with a value of approximately 2.718281828. R T The detection distance threshold is greater than R. T When the distance increases, the backscattering intensity decreases; when it is greater than R... T At that time, the backscattering intensity increases with increasing distance.

[0090] The calculation results of the geometric factors parameters α, β, γ for the four targets are as follows: Figure 5 As shown. Among them. Figure 5 (a) is a 100% standard diffuse reflector. Figure 5 (b) is made of bluestone bricks. Figure 5 (c) is a white, smooth floor tile. Figure 5 (d) is a leaf of a pothos plant.

[0091] The results show that although the four targets have different roughnesses and reflection characteristics, their calculated geometric factors are similar, and the geometric factor parameters are also similar across different spectral bands. Experimental results indicate that the geometric factor parameters of the backscattering intensity of hyperspectral lidar are not affected by the target's roughness, specular reflection, or spectral band. The results also verify that the geometric factor is mainly related to the detection equipment and has little to do with the characteristics of the measured target. Experimental results show that different targets share the same geometric factor model parameters for range effects across different spectral bands. Range effect correction for various targets can be carried out using only a limited number of parameters, and the universality of these parameters provides a simple and effective technical means for solving the range effect problem of hyperspectral lidar.

[0092] The geometric factor parameters of the 100% standard diffuse reflector plate in different bands are calculated using the following formula and used as the geometric factor parameters of the hyperspectral lidar.

[0093] Therefore, define the uniform α of this surface. mean ,β mean γ mean Let α be the average value of the geometric factor parameters, and n be the number of bands. The calculation of the geometric factor parameters is expressed by the following formula (14). Wherein, α(λ) n ), β(λ n ), γ(λ n You can select the parameter calculation results of a relatively stable target, or you can select the parameter calculation results of multiple targets.

[0094]

[0095] (4) Distance effect model construction method

[0096] Based on formula (3), it is assumed that atmospheric conditions are constant during a single measurement. During the laser scanning process, P... t (λ),D r ,η atm (λ) and η sys (λ) can be considered as four constants. The received power P of the hyperspectral lidar at different wavelengths. r (λ) is represented as:

[0097]

[0098] Wherein, C(λ) is represented by formula (16) and can be regarded as a constant.

[0099]

[0100] Backscattering intensity I(R,λ) and received power P r (λ) Positive correlation. Assuming the incident angle is constant during laser scanning, for the same target, considering the influence of geometric factors, the relationship between backscattering intensity and detection distance is expressed as shown in formula (17):

[0101]

[0102] In the formula: g0(λ) can be considered constant, expressed as formula (18):

[0103] g0(λ)=C(λ)ρ(λ)cosθ (18)

[0104] (5) Method for constructing distance effect radiation correction model

[0105] To eliminate the influence of distance on backscattering intensity, a standard detection distance R is defined. s The backscattering intensity at different detection distances is normalized to the standard distance. Based on formula (16), the influence of geometric factors is considered simultaneously during the normalization of the distance effect of backscattering intensity. The distance normalization model is expressed as the following formula (19):

[0106]

[0107] In the formula, R s This is the standard detection range, Z(R) s I(R,λ) is the geometric factor over the standard detection range. c I(R,λ) is the corrected backscattering intensity, and I(R,λ) is the original backscattering intensity.

[0108] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for range effect radiometric correction of hyperspectral lidar based on geometric factors, characterized in that, Includes the following steps: Step 1: Construct a geometric factor model based on the principles of optical systems or an S-curve fitting function; Step 2: Construct a distance effect model; Step 3: Construct a distance-effect radiation correction model; The specific method for constructing a geometric factor model based on the principles of optical systems is as follows: The telescope's effective receiving radius is The radius of the emitted light spot is The radius of the laser beam spot at different detection distances Telescope field of view radius These are expressed as formulas (6) and (7) respectively: (6) (7) in, For detection distance; This is the telescope's half field of view. The laser beam half-divergence angle; The angle between the optical axis of the laser emitted by the hyperspectral lidar and the optical axis of the telescope is... The distance between the two optical axes Represented as formula (8): (8) Let the radius of the telescope's secondary mirror be... Telescope secondary mirror obstruction radius Represented as formula (9): (9) Overlap area of ​​laser spot and detector element Represented as formula (10): (10) In the formula, the parameters and The calculation formula is as follows: Formula (11): (11) Therefore, the geometric factor is a function of distance. Represented as formula (12): (12)。 2. The method for range effect radiation correction of hyperspectral lidar based on geometric factors according to claim 1, characterized in that, In step one, the method for constructing the geometric factor model based on the S-curve fitting function is as follows: The Logistic function is used for fitting, and a function is established. The piecewise geometric factor calculation model is shown in formula (13): (13) in, The parameters to be solved for the requirements. It is a natural constant. The detection distance threshold is greater than When the distance increases, the backscattering intensity decreases; when the distance is less than... At that time, the backscattering intensity increases with increasing distance.

3. The method for range effect radiation correction of hyperspectral lidar based on geometric factors according to claim 2, characterized in that, In step two, The laser emission power of the lidar is The receiver aperture is , These are the factors affecting signal transmission in the atmosphere. Here are the parameters for the lidar system, with the laser incident angle being... Detection range is The reflectivity of the target The different wavelengths of hyperspectral lidar are represented as During a laser scan, and The received power of hyperspectral lidar at different wavelengths is represented by four constants. Represented as: (15) in, This indicates that the target is at a wavelength of Target reflectivity at that time Represented by formula (16), where is a constant. (16) During laser scanning, the incident angle is constant. For the same target, considering the influence of geometric factors, the backscattering intensity... The relationship with the detection distance is expressed as shown in formula (17): (17) In the formula: Constant, expressed as formula (18): (18)。 4. As described in claim 3, characterized in that, In step three, Define a standard detection range The backscattering intensity at different detection distances is normalized to the standard distance. Based on formula (16), the influence of geometric factors is considered in the process of normalizing the distance effect of backscattering intensity. The distance normalization model is expressed as formula (19) as follows: (19) In the formula, This is the standard detection range. It is a geometric factor over the standard detection range. This is the corrected backscattering intensity. It is the original backscattering intensity.

Images