A Multimodal Remote Sensing Inversion Method and System for Forest Tree Height Based on Bias Correction

By constructing a deviation correction model and combining data from spaceborne lidar and UAV lidar, the problem of insufficient accuracy in forest tree height inversion under complex terrain was solved, and high-precision and reliable forest tree height distribution maps were generated, supporting forest carbon sink monitoring.

CN122090294APending Publication Date: 2026-05-26AEROSPACE INFORMATION RES INST CAS
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Applications(China)
Current Assignee / Owner
AEROSPACE INFORMATION RES INST CAS
Filing Date
2026-01-12
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing technologies have limitations in the accuracy and robustness of forest tree height inversion in complex terrain areas. They lack clear and quantitative methods for correcting terrain bias in spaceborne lidar data, and multimodal remote sensing data fusion fails to fully model the complex relationship between key terrain factors and system bias.

Method used

By constructing a multimodal remote sensing inversion method for forest tree height based on bias correction, a buffer is constructed using the geometric size of the satellite-borne lidar spot. Combined with the median reference ground value of UAV lidar data, a multivariate nonlinear bias correction model is constructed. Multiple terrain feature parameters are integrated, and a random forest regression model is trained to invert forest tree height.

Benefits of technology

It significantly improves the accuracy and reliability of forest tree height inversion in complex terrain areas, generating spatially continuous forest tree height distribution maps with strong detail representation, supporting precise monitoring and sustainable management of forest carbon sinks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122090294A_ABST
    Figure CN122090294A_ABST
Patent Text Reader

Abstract

This invention discloses a multimodal remote sensing inversion method and system for forest tree height based on bias correction, belonging to the field of forest tree height inversion technology. The method first acquires data from UAV lidar, spaceborne lidar, optical and radar remote sensing, topographic and meteorological data, and performs preprocessing and feature extraction. By constructing a circular buffer based on the geometric dimensions of the spaceborne laser spot and using the median of the UAV canopy height as the ground truth, accurate spatial matching between spaceborne and UAV data is achieved. A machine learning bias correction model integrating multiple terrain features such as slope, aspect, elevation, and mountain shadow is constructed to correct nonlinear systematic errors in the spaceborne lidar data. The corrected spaceborne lidar samples and multimodal remote sensing features are fused and trained using a random forest model to generate a high-resolution, spatially continuous forest tree height distribution map. This invention significantly improves the accuracy and reliability of forest tree height inversion.
Need to check novelty before this filing date? Find Prior Art