Spectral Imaging From Uavs Under Varying Illumination Conditions

Abstract

ABSTRACT:Rapidly developing unmanned aerial vehicles (UAV) have provided the remote sensing community with a new rapidly deployable tool for small area monitoring. The progress of small payload UAVs has introduced greater demand for light weight aerial payloads. For applications requiring aerial images, a simple consumer camera provides acceptable data. For applications requiring more detailed spectral information about the surface, a new Fabry-Perot interferometer based spectral imaging technology has been developed. This new technology produces tens of successive images of the scene at different wavelength bands in very short time. These images can be assembled in spectral data cubes with stereoscopic overlaps. On field the weather conditions vary and the UAV operator often has to decide between flight in sub optimal conditions and no flight. Our objective was to investigate methods for quantitative radiometric processing of images taken under varying illumination conditions, thus expanding the range of weather conditions during which successful imaging flights can be made. A new method that is based on insitu measurement of irradiance either in UAV platform or in ground was developed. We tested the methods in a precision agriculture application using realistic data collected in difficult illumination conditions. Internal homogeneity of the original image data (average coefficient of variation in overlapping images) was 0.14-0.18. In the corrected data, the homogeneity was 0.10-0.12 with a correction based on broadband irradiance measured in UAV, 0.07-0.09 with a correction based on spectral irradiance measurement on ground, and 0.05-0.08 with a radiometric block adjustment based on image data. Our results were very promising, indicating that quantitative UAV based remote sensing could be operational in diverse conditions, which is prerequisite for many environmental remote sensing applications.

Radiometric Correction of Close-Range Spectral Image Blocks Captured Using an Unmanned Aerial Vehicle with a Radiometric Block Adjustment

“…An irradiance sensor based on the Intersil ISL29004 photodetector with a spectral sensitivity range of 400-1000 nm was integrated into the camera to measure the irradiance during each exposure. Since the sensor is not calibrated, relative, broadband irradiance intensity values can be obtained [31]. Consistent with our previous experience, the flight direction influenced the irradiance recording due to the tilting of the irradiance sensor with respect to the sun's illumination, because the tilt compensation was not applied.…”

“…The image observations of the RCPs are obtained similarly; the reflectance reference is based on a laboratory or in situ measurement of the targets with a spectrometer. Observations for the relative parameters can be based on drone or terrestrial irradiance measurements [31], or else a constant value (e.g., 1.0) can be used. The least squares method is then run to find the optimal values for the unknowns.…”

“…Aasen et al [13] and Yang et al [15] calibrated the Cubert UHD 185-Firefly hyperspectral imagery; they emphasized the sensor calibration and performed reflectance calibration using single or multiple reflectance panels. To cope with varying illumination conditions, methods based on continuous illumination monitoring are highly relevant; in previous studies either ground based [7,31] or onboard-based solutions [8,31] have been used. These studies have reported about challenges related to the irradiance sensor calibration and tilting during the flights.…”

“…In recent studies, the radiometric correction of UAV imagery and spectrometry has mostly emphasized the atmospheric correction using reflectance panels [6,[12][13][14][15][31][32][33][34][35][36][37], atmospheric radiative transfer modeling [2,4] or normalization of UAV based radiance observations using field based radiance observations of a reference panel or irradiance [7,8,31]. Aasen et al [13] and Yang et al [15] calibrated the Cubert UHD 185-Firefly hyperspectral imagery; they emphasized the sensor calibration and performed reflectance calibration using single or multiple reflectance panels.…”

Unmanned Aerial Vehicles for Alien Plant Species Detection and Monitoring

“…Consequently, methods for radiometric corrections inspired by work of (Hakala et al, 2013) are to be implemented. The choice of the best monitoring method always represents a trade-off between accuracy, scale and feasibility.…”