An Empirical Line approach for Agrowing Camera Aerial Images of inland waters based on exponential fit and in situ water measurements

Abstract

With the advancement of technology in the area of remote sensing, monitoring the Earth's surface using multispectral cameras attached to unmanned aerial vehicles (UAV) has become promising. However, in many Earth observation applications, it is needed to make compatible the spatial data of images captured by high spatial resolution multispectral sensors with the spectral response of targets on the Earth's surface. This relation is obtained through a radiometric calibration. The empirical line method is commonly used to calibrate the spectral bands of sensors. Thus, applications of this method using linear fit have retrieved negative values in water bodies. So that, attempting different adjustments, as well different reference targets, can solve this issue. In this study, the water quality of small bodies of water was analysed using a Agrowing multispectral camera, which derived negative values when applying linear fit. The aim of this study, therefore, was to fit a radiometric calibration based on empirical line method for Agrowing camera in inland water applications. Besides of standard reference targets, water samples also were attempted because showed a lower radiance response than the darkest (black) calibration target. Two empirical line methods were applied to convert the digital number (DN) from the Agrowing images into remote sensing reflectance (Rrs): linear and exponential. The exponential method showed to be more appropriate, with greater accuracy, unlike the linear method.