Pysically Based Data Fusion Between Airborne Lidar and Hyperspectral Data: Geometric and Radiometric Synergies
- Resource Type
- Conference
- Authors
- Brell, Maximilian; Guanter, Luis; Segl, Karl
- Source
- IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium Geoscience and Remote Sensing Symposium, IGARSS 2018 - 2018 IEEE International. :8865-8868 Jul, 2018
- Subject
- Aerospace
Computing and Processing
Fields, Waves and Electromagnetics
Geoscience
Photonics and Electrooptics
Signal Processing and Analysis
Laser radar
Radiometry
Hyperspectral imaging
Lighting
Atmospheric modeling
Data integration
hyperspectral imaging
LiDAR
sensor fusion
in flight
airborne
direct geocoding
radiometric calibration
- Language
- ISSN
- 2153-7003
Combining airborne LiDAR (ALS) and hyperspectral data refers to utilize the LiDAR based Digital Elevation Model (DEM) and the spectral information of the hyperspectral imaging (HSI) sensor. The separation of both discretized data entities leads to a substantial loss of information and does not exhaust the full capabilities of the contrasting sensors. A physically based in-flight fusion of HSI and ALS sensor characteristics is presented. Based on their respective intensity information overlaps, ray tracing and radiative transfer procedures utilize geometric and radiometric synergies. In a first step a rigorous parametric co-alignment procedure is realized using an automated and adjustable tie point detection algorithm. It ensures sub-pixel co-alignment of the contrasting sensors. In a second step we present a rigorous illumination correction of HSI data based on the radiometric cross-calibrated return intensity information of ALS data. This radiometric fusion corrects cloud and cast shadowing effects, across track illumination, partly anisotropy effects and illumination changes over time for the entire HSI wavelength domain. The presented fundamental fusion of the passive and active sensor characteristics is aimed at improving and developing the complete, sensor inherent data density to ensure highest spectral and geometric information content for a variety of applications.