Spectral Recalibration for In-Flight Broadband Sensor Using Man-Made Ground Targets.

Huazhong Ren,Guangjian Yan,Rongyuan Liu,Ronghai Hu,Tianxing Wang,Xihan Mu

DOI: https://doi.org/10.1109/tgrs.2012.2226894

IF: 8.2

2013-01-01

IEEE Transactions on Geoscience and Remote Sensing

Abstract:Accurate spectral calibration of the in-flight sensors is crucial for processing and exploration of remotely sensed data. This paper developed a strategy to make spectral recalibration (i.e., spectral response function, central wavelength, and bandwidth) for in-flight broadband sensor using a device-responsivity-decomposition model with a priori knowledge and an optimization algorithm. Sensitivity analysis indicates that an accurate result requires the targets to be observed under a dry and clear atmospheric condition (column water vapor < 2 g/cm(2) and visibility > 23 km) and no more than 5% error is included in the measured data. The new strategy was used to retrieve the spectral parameters along with radiometric calibration coefficients for a multichannel camera onboard an unmanned aerial vehicle from simultaneously remotely sensed and ground measured data sets over 19 (15 color-scaled and four gray-scaled) man-made surface targets, and the retrieved results were validated with a similar data set over another four man-made targets. It demonstrated that the camera's spectral parameters were accurately retrieved and an error less than 3.5 W/m(2)/mu m/sr was brought to the channel radiance.