Abstract
Imagery acquired with unmanned aerial vehicles (UAVs) has a great potential for incorporation into natural resource monitoring protocols due to their ability to be deployed quickly and repeatedly. A valid radiometric calibration of radiance measuring instruments is required for physically based analysis of the measured data. In the autumn of 2013, an Unmanned Aerial Vehicle hyperspectral calibration experiment of the HeadWall imaging spectrometer was conducted at a civilian airport in Suizhong, Liaoning province of China. Headwall’s Micro-Hyperspec airborne sensors optimized performance for the most demanding UAV applications which measure radiance in nominally 4nm channels between 380~1000nm. The purpose of the experiment was to validate the radiometric calibration of the spectrometer and to determine whether the image data meet the needs of quantitative scientific analysis. The flight height of the UAV was 800 m, with the ground resolution of this altitude was 0.375 m. The UAV’s image is severely distorted and is hard to geometric corrected very well solely using the low accuracy POS data, we proposed a method to correct the image which combine image line related and POS data. Then, with the help of 2 hyperspectral radiometric targets which were 10.4*10 m wide with reflectivety of 60% and 5%, the HeadWall imaging spectrometer in UAV was calibrated adopting reflectance-based method. The approach used in this work is to measure the surface reflectance and atmospheric properties during the sensor image acquisition. We collected atmospheric data with CE318 and ground reflectance data with an ASD FieldSpec FR portable spectrometer (ASD FR). The barometric pressure and ambient air temperature were measured with a small portable electronic barometer and recorded by hand. These data used as input to 6S radiative transfer code which computes the top of atmosphere spectral radiance. This predicted apparent radiance is compared to the targets of DN value from the image of targets to give the calibration coefficient. At last, the top of atmosphere spectral radiance of concrete runway which was computed by the synchronous spectral reflectance with the mean reflectivety of 28% and 6S was compare with the same area of radiance image which was transformed from calibration coefficient, and the results of the in-flight calibration experiment indicate an average agreement of 4.8% for relative differences between them.
A Method Suitable for In-Flight Calibration of a UAV Hyperspectral Remote Sensor
Imagery acquired with unmanned aerial vehicles (UAVs) has a great potential for incorporation into natural resource monitoring protocols due to their ability to be deployed quickly and repeatedly. A valid radiometric calibration of radiance measuring instruments is required for physically based analysis of the measured data. In the autumn of 2013, an Unmanned Aerial Vehicle hyperspectral calibration experiment of the HeadWall imaging spectrometer was conducted at a civilian airport in Suizhong, Liaoning province of China. Headwall’s Micro-Hyperspec airborne sensors optimized performance for the most demanding UAV applications which measure radiance in nominally 4nm channels between 380~1000nm. The purpose of the experiment was to validate the radiometric calibration of the spectrometer and to determine whether the image data meet the needs of quantitative scientific analysis. The flight height of the UAV was 800 m, with the ground resolution of this altitude was 0.375 m. The UAV’s image is severely distorted and is hard to geometric corrected very well solely using the low accuracy POS data, we proposed a method to correct the image which combine image line related and POS data. Then, with the help of 2 hyperspectral radiometric targets which were 10.4*10 m wide with reflectivety of 60% and 5%, the HeadWall imaging spectrometer in UAV was calibrated adopting reflectance-based method. The approach used in this work is to measure the surface reflectance and atmospheric properties during the sensor image acquisition. We collected atmospheric data with CE318 and ground reflectance data with an ASD FieldSpec FR portable spectrometer (ASD FR). The barometric pressure and ambient air temperature were measured with a small portable electronic barometer and recorded by hand. These data used as input to 6S radiative transfer code which computes the top of atmosphere spectral radiance. This predicted apparent radiance is compared to the targets of DN value from the image of targets to give the calibration coefficient. At last, the top of atmosphere spectral radiance of concrete runway which was computed by the synchronous spectral reflectance with the mean reflectivety of 28% and 6S was compare with the same area of radiance image which was transformed from calibration coefficient, and the results of the in-flight calibration experiment indicate an average agreement of 4.8% for relative differences between them.