CN-121981926-A - Method and system for correcting forest structure data of space laser radar

CN121981926ACN 121981926 ACN121981926 ACN 121981926ACN-121981926-A

Abstract

The invention relates to a method and a system for correcting forest structure data of a space laser radar, and belongs to the technical field of space laser remote sensing and forest resource monitoring. Aiming at GEDI data with four errors of geometric positioning offset, gradual distortion, terrain stretching effect and canopy structure influence, the invention adopts a hierarchical layout strategy to lay UAV control points, establishes homonymous point relation through waveform registration and calculates weight values, carries out overall geometric correction based on regular grid control points, carries out local distortion correction based on encryption control points, establishes a physical compensation model to compensate the terrain stretching effect, establishes a statistical regression model based on canopy structure parameters and GEDI signal quality indexes to correct canopy height residual errors, finally obtains high-precision canopy height convergence results through spatial smoothing treatment, and outputs corrected data. The invention realizes the step-by-step precision improvement from the whole domain to the local domain and from the geometry to the height, and has low cost, high precision and wide application range.

Inventors

BAO GUANGDAO
DING MINGMING
LIU TING
YU DONGBIN
LIN CAN
BAO YING
JIANG XUEFEI

Assignees

吉林省林业科学研究院(吉林省林业生物防治中心站)

Dates

Publication Date: 20260505
Application Date: 20260408

Claims (10)

1. The method for correcting the forest structure data of the space laser radar is characterized by comprising the following steps of: Firstly, arranging UAV control points of a regular grid in a target area according to a hierarchical arrangement strategy, arranging the UAV control points in a complex terrain area in an encrypted manner, acquiring UAV data of all the UAV control points, carrying out waveform registration on the UAV data and GEDI footprint points, establishing a homonymy point relation, and calculating corresponding weight values; performing overall geometric correction on plane coordinates of GEDI footprint according to a track section or a period of time based on UAV control points of a regular grid, wherein the overall geometric correction comprises overall translation, rotation and scale adjustment, so as to obtain primary correction coordinates; Thirdly, carrying out local bending and distortion correction on the GEDI footprint after the overall geometric correction based on the UAV control points distributed in an encrypted manner to obtain secondary correction coordinates; step four, based on the terrain gradient and the laser beam direction, a physical compensation model is established, and terrain broadening effect compensation is carried out on the canopy height in GEDI data, so that the once corrected canopy height is obtained; establishing a statistical regression model based on the canopy structure parameters and GEDI signal quality indexes, and correcting the canopy height residual error to obtain a secondary corrected canopy height; And step six, eliminating local scattered residual errors of the secondary correction canopy height through a space smoothing method to obtain a canopy height convergence result, and finally outputting corrected GEDI forest structure data, wherein the corrected GEDI forest structure data comprise the secondary correction coordinates and the canopy height convergence result.
2. A method for spatially lidar forest structure data correction according to claim 1, wherein the step of registering the waveforms in step one comprises: taking GEDI footprint diameter as a window, and extracting corresponding analog waveforms from UAV data; and determining the translation quantity which is most similar to the GEDI waveform of the analog waveform by maximizing the normalized cross-correlation in a preset micro-shift search window, taking the translation quantity as a homonymy point matching result, and recording the maximum normalized cross-correlation coefficient.
3. The method for correcting forest structure data of spatial lidar according to claim 2, wherein the weight calculation formula in the step one is: Wherein, the Is the first Weights corresponding to the GEDI footprints; Respectively represent the first Sensitivity, received energy, and number of samples of the GEDI footprints; are all experience indexes; Is the first Maximum normalized cross-correlation coefficients corresponding to the GEDI footprints; In order to provide a truncated function, The lower threshold and the upper threshold of the weight cutoff are respectively.
4. The method for correcting forest structure data of spatial lidar according to claim 1, wherein the integral geometric correction in the second step adopts an affine adjustment model, and an observation equation is as follows: Wherein, the For GEDI the original footprint coordinates, Matching coordinate references for UAV true coordinates or homonymies; Method for solving affine transformation parameter vector by weighted least square method And according to the obtained affine transformation parameter vector Correcting GEDI footprint coordinates to obtain corresponding first-order corrected coordinates 。
5. The method for correcting forest structure data of space laser radar according to claim 1, wherein in the third step, the local bending and distortion correction is performed by fitting the residual error of the primary correction coordinates by using a quadric residual error model or a thin plate spline model.
6. The method for correcting forest structure data of spatial lidar according to claim 1, wherein the physical compensation model is: Wherein, the Correcting the height of the canopy for one time; A diameter of GEDI footprints; is an index of the roughness of the ground; Is the included angle between the laser beam direction and the maximum slope direction; Is a slope angle; The coefficients calibrated by the combination with the UAV truth values, respectively.
7. The method for correcting forest structure data of spatial lidar according to claim 1, wherein the expression of the statistical regression model is: Wherein, the Correcting the height residual error of the canopy height and the true value of the UAV once; Crown coverage, thickness, volume, respectively; Is a constant term; is a crown layer structure parameter coefficient; Is a signal quality parameter coefficient; Is a topographic parameter coefficient; Is a random noise term.
8. The method for correcting forest structure data of spatial lidar according to claim 1, wherein the spatial smoothing method is a thin plate spline method or a kriging interpolation method.
9. The method for correcting forest structure data of spatial lidar according to claim 1, wherein in the hierarchical layout strategy, the layout pitch of UAV control points of a regular grid is 10 to 50km, the layout pitch of UAV control points of an encrypted layout is 5 to 10km, and the complex terrain area is a mountain area with a terrain relief exceeding 300 meters.
10. A spatial lidar forest structure data correction system comprising a processor and a memory, the memory storing a computer program, characterized in that the processor, when executing the computer program, carries out the steps of the method according to any of claims 1 to 9.

Description

Method and system for correcting forest structure data of space laser radar Technical Field The invention relates to the technical field of space laser remote sensing and forest resource monitoring, in particular to a method and a system for correcting forest structure data of a space laser radar. Background The satellite-borne laser radar for global ecological system dynamics investigation (Global Ecosystem Dynamics Investigation, GEDI) is mounted on an international space station, can acquire parameters such as forest canopy height, vertical structure, ground elevation and the like on a global scale, and has important value in carbon reserve accounting, forest resource investigation and ecological system research. However, since the orbit and attitude resolving precision of the observation platform is limited, the influence of the laser waveform on the terrain and the canopy structure is remarkable, and the signal quality is different, GEDI data have various errors in large-scale application, and the reliability and popularization and use of the GEDI data are directly affected. The method is characterized in that firstly, a geometric and positioning system is offset, translation, rotation or scale deviation exists in the whole GEDI footprint due to track and attitude calculation errors and satellite ground reference conversion precision limitation, the position of an observation point is inconsistent with that of an actual ground object, secondly, geometrical distortion is slowly changed, drift or curved surface distortion gradually occurs in a track section due to small accumulated errors of the attitude and geometric parameters in track section observation, so that geometrical relation distortion of a local area is difficult to correct due to integral translation, thirdly, waveforms caused by topography are widened, echo waveforms are elongated along the height direction due to height difference in footprint coverage areas in sloping fields and areas with larger fluctuation, RH height indexes systematically deviate from the actual crown or ground height, fourthly, crown structure influence is caused, structural differences such as forest coverage, thickness, volume and the like of different forest types can change laser penetration and echo energy distribution, ground echo loss can be caused under weak signal conditions, and accordingly, the system deviation of tree height estimation is caused. The existing GEDI data correction method is mostly dependent on a digital elevation model (Digital Elevation Model, DEM) or a small amount of sample to carry out integral deviation correction, or an affine transformation and empirical regression model is adopted, but the methods can not simultaneously solve various errors such as systematic deviation, gradual distortion, terrain widening, canopy structure difference and the like, and a reasonable Unmanned aerial vehicle (Unmanned AERIAL VEHICLE, UAV) control point layout strategy is also lacking, so that GEDI data is insufficient in precision in large-scale popularization and application. Therefore, there is a need to develop a new method that can effectively address the above type of errors under limited UAV cost conditions, enabling high precision applications of GEDI in provincial and even larger areas. Disclosure of Invention The invention mainly aims at the problem of insufficient precision of GEDI data in a large-scale application, provides a method and a system for correcting forest structure data of a space laser radar, and can solve four errors of geometric and positioning system deviation, slow-changing geometric distortion, waveform broadening caused by terrain and influence of canopy structure in GEDI data, and reduce systematic errors in forest height and terrain information. In order to solve the problems, the invention adopts the following technical scheme: The method for correcting the forest structure data of the space laser radar comprises the following steps: Firstly, arranging UAV control points of a regular grid in a target area according to a hierarchical arrangement strategy, arranging the UAV control points in a complex terrain area in an encrypted manner, acquiring UAV data of all the UAV control points, carrying out waveform registration on the UAV data and GEDI footprint points, establishing a homonymy point relation, and calculating corresponding weight values; performing overall geometric correction on plane coordinates of GEDI footprint according to a track section or a period of time based on UAV control points of a regular grid, wherein the overall geometric correction comprises overall translation, rotation and scale adjustment, so as to obtain primary correction coordinates; Thirdly, carrying out local bending and distortion correction on the GEDI footprint after the overall geometric correction based on the UAV control points distributed in an encrypted manner to obtain secondary correction coordinates; step four, based on the t