A method for joint retrieval of forest structure parameters from full-waveform LiDAR and hyperspectral data

Abstract

The invention discloses a method for jointly inverting forest structure parameters by using a full-waveform lidar and hyperspectral data. Firstly, performing de-noising, smoothing, intensity correction and filtering on airborne full-waveform lidar data, performing interpolation to generate a digital terrain model, and performing high normalization processing on point cloud and waveform data; performing radiation calibration, atmospheric correction and geometric correction preprocessing on a hyperspectral image; then extracting characteristic variables based on normalized point cloud and waveform data and preprocessed hyperspectral data separately; and finally, constructing multivariate regression models combined with ground measured data and the extracted characteristic variables to predict each forest structure parameter. The invention contributes to improving the inversion precision of the forest structure parameters, and effectively suppresses the "saturation" problem of parameter inversion of a stand structure with high forest coverage and high biomass; therefore, the ability and accuracy of forest structure parameter inversion is effectively enhanced; and compared with the stand structure parameter inversion with other similar remote sensing methods, the relative root mean square error is increased by more than 5%.

Description

technical field [0001] The invention belongs to the technical fields of forest resource monitoring and environmental factor investigation, and relates to a method for jointly inverting forest structure parameters with full-waveform laser radar and hyperspectral data. Background technique [0002] Accurate extraction of forest structure parameters is of great significance for forest resource monitoring and environmental factor investigation. At the same time, this information can also be used to grasp the relationship between forest plants and the environment, and to further understand the growth and development of forests, which is of great significance for sustainable forest management, ecological model construction and global carbon cycle research. Conventional forest structure parameter extraction mainly relies on field surveys and interpretation of aerial or satellite images, and its accuracy is often not high, and it is difficult to be practically promoted on the "surfa...

AI Technical Summary

Problems solved by technology

However, the above methods are all based on lidar point cloud data, and do not use full waveform lidar data that can better record the structural characteristics of the canopy

At the same time, there is no method for comprehensively and deeply extracting forest canopy lidar features (waveform features and point cloud features), spectral features and calculating forest structure parameters.

Images